BTS guideline for emergency oxygen use in adult patients

Transcription

1 1 Department of Respiratory Mediine, Salford Royal University Hospital, Salford, UK; 2 Hammersmith Hospital, Imperial College Healthare NHS Trust, London, UK; 3 Southend University Hospital, Westliff on Sea, Essex, UK Correspondene to: Dr B R O Drisoll, Department of Respiratory Mediine, Salford Royal University Hospital, Stott Lane, Salford M6 8HD, UK; ronan.o drisoll@srft.nhs.uk Reeived 11 June 2008 Aepted 11 June 2008 BTS guideline for emergeny oxygen use in adult patients B R O Drisoll, 1 L S Howard, 2 A G Davison 3 on behalf of the British Thorai Soiety EXECUTIVE SUMMARY OF THE GUIDELINE Philosophy of the guideline Oxygen is a treatment for hypoxaemia, not breathlessness. (Oxygen has not been shown to have any effet on the sensation of breathlessness in non-hypoxaemi patients.) The essene of this guideline an be summarised simply as a requirement for oxygen to be presribed aording to a target saturation range and for those who administer oxygen therapy to monitor the patient and keep within the target saturation range. The guideline suggests aiming to ahieve normal or near-normal oxygen saturation for all autely ill patients apart from those at risk of hyperapni respiratory failure or those reeiving terminal palliative are. Assessing patients For ritially ill patients, high onentration oxygen should be administered immediately (table 1 and fig 1) and this should be reorded afterwards in the patient s health reord. Oxygen saturation, the fifth vital sign, should be heked by pulse oximetry in all breathless and autely ill patients (supplemented by blood gases when neessary) and the inspired oxygen onentration should be reorded on the observation hart with the oximetry result. (The other vital signs are pulse, blood pressure, temperature and respiratory rate). Pulse oximetry must be available in all loations where emergeny oxygen is used. All ritially ill patients should be assessed and monitored using a reognised physiologial trak and trigger system. Oxygen presription Oxygen should be presribed to ahieve a target saturation of 94 98% for most autely ill patients or 88 92% for those at risk of hyperapni respiratory failure (tables 1 3). The target saturation should be written (or ringed) on the drug hart (guidane in fig 1). Oxygen administration Oxygen should be administered by staff who are trained in oxygen administration. These staff should use appropriate devies and flow rates in order to ahieve the target saturation range (fig 2). BTS guideline Monitoring and maintenane of target saturation Oxygen saturation and delivery system should be reorded on the patient s monitoring hart alongside the oximetry result. Oxygen delivery devies and flow rates should be adjusted to keep the oxygen saturation in the target range. Oxygen should be signed for on the drug hart on eah drug round. Weaning and disontinuation of oxygen therapy Oxygen should be redued in stable patients with satisfatory oxygen saturation. Oxygen should be rossed off the drug hart one oxygen is disontinued. Oxygen is one of the most widely used drugs and is used aross the whole range of speialities. The Guideline Group reognises that many liniians will initially wish to read an abbreviated version of this guideline whih is available to download from the BTS website ( SUMMARY OF KEY RECOMMENDATIONS FOR EMERGENCY OXYGEN USE Ahieving desirable oxygen saturation ranges in aute illness (setions 6.7 and 6.8) 1. This guideline reommends aiming to ahieve a normal or near-normal oxygen saturation for all autely ill patients apart from those at risk of hyperapni respiratory failure. [Grade D] 2. The reommended target saturation range for autely ill patients not at risk of hyperapni respiratory failure is 94 98%. Some normal subjets, espeially people aged.70 years, may have oxygen saturation measurements below 94% and do not require oxygen therapy when linially stable. [Grade D] 3. Most non-hypoxaemi breathless patients do not benefit from oxygen therapy, but a sudden redution of more than 3% in a patient s oxygen saturation within the target saturation range should prompt fuller assessment of the patient (and the oximeter signal) beause this may be the first evidene of an aute illness. [Grade D] 4. For most patients with known hroni obstrutive pulmonary disease (COPD) or other known risk fators for hyperapni respiratory failure (eg, morbid obesity, hest wall deformities or neuromusular disorders), a target saturation range of 88 92% is suggested pending the availability of blood gas results. [Grade C] vi1

2 5. Some patients with COPD and other onditions are vulnerable to repeated episodes of hyperapni respiratory failure. In these ases it is reommended that treatment should be based on the results of previous blood gas estimations during aute exaerbations beause hyperapni respiratory failure an our even if the saturation is below 88%. For patients with prior hyperapni failure (requiring non-invasive ventilation or intermittent positive pressure ventilation) who do not have an alert ard, it is reommended that treatment should be ommened using a 28% Venturi mask at 4 l/min in prehospital are or a 24% Venturi mask at 2 4 l/min in hospital settings with an initial target saturation of 88 92% pending urgent blood gas results. These patients should be treated as a high priority by emergeny servies and the oxygen dose should be redued if the saturation exeeds 92%. [Grade D] 6. Beause oxygenation is redued in the supine position, fully onsious hypoxaemi patients should ideally be allowed to maintain the most upright posture possible (or the most omfortable posture for the patient) unless there are good reasons to immobilise the patient (eg, skeletal or spinal trauma). [Grade C] Clinial and laboratory assessment of hypoxaemia and hyperapnia (setion 7.1) 7. Fully trained liniians should assess all autely ill patients by measuring pulse, blood pressure, respiratory rate and assessing irulating blood volume and anaemia. Expert assistane from speialists in intensive are or from other disiplines should be sought at an early stage if patients are thought to have major life-threatening illnesses and liniians should be prepared to all for assistane when neessary, inluding a all for a 999 ambulane in prehospital are or a all for the resusitation team or ICU outreah team in hospital are. [Grade C D] 8. Initial linial assessment and subsequent monitoring of autely unwell patients should inlude the use of a reognised physiologial trak and trigger system, suh as the Modified Early Warning Soring System (mews), and a hange in this sore should require medial review even if there is no hange in oxygen saturation. [Grade C] 9. Oxygen saturation, the fifth vital sign, should be heked by trained staff using pulse oximetry in all breathless and autely ill patients (supplemented by blood gases when neessary) and the inspired oxygen onentration should be reorded on the observation hart with the oximetry result. [Grade D] 10. The presene of a normal oxygen saturation (arterial oxygen saturation measured by pulse oximetry (SpO 2 ) does not always negate the need for blood gas measurements beause pulse oximetry will be normal in a patient with normal oxygen tension but abnormal blood ph or arbon dioxide tension (PCO 2 ) or with a low blood oxygen ontent due to anaemia). Blood gas measurements and full blood ounts are therefore required as early as possible in all situations where these measurements may affet patient outomes. [Grade D] Arterial and arteriolised blood gases (setions and 8.4) 11. For ritially ill patients or those with shok or hypotension (systoli blood pressure,90 mm Hg), the initial blood gas measurement should be obtained from an arterial speimen. However, for most patients who require blood gas sampling, either arterial blood gases or arteriolised earlobe blood gases may be used to obtain an aurate measure of ph and PCO 2. However, the arterial oxygen tension (PaO 2 )is less aurate in earlobe blood gas samples (it underestimates the oxygen tension by kpa), so oximetry should be monitored arefully if earlobe blood gas speimens are used. [Grade B] 12. Loal anaesthesia should be used for all arterial blood gas speimens exept in emergenies or if the patient is unonsious or anaesthetised. [Grade B] 13. Blood gases should be heked in the following situations: All ritially ill patients. Unexpeted or inappropriate hypoxaemia (SpO 2,94%) or any patient requiring oxygen to ahieve this target range. (Allowane should be made for transient dips in saturation to 90% or less in normal subjets during sleep). [Grade D] Deteriorating oxygen saturation or inreasing breathlessness in a patient with previously stable hypoxaemia (eg, severe COPD). [Grade D] Any previously stable patient who deteriorates and requires a signifiantly inreased fration of inspired oxygen (FIO 2 ) to maintain a onstant oxygen saturation. [Grade D] Any patient with risk fators for hyperapni respiratory failure who develops aute breathlessness, deteriorating oxygen saturation or drowsiness or other symptoms of CO 2 retention. [Grade D] Breathless patients who are thought to be at risk of metaboli onditions suh as diabeti ketoaidosis or metaboli aidosis due to renal failure. [Grade D] Autely breathless or ritially ill patients with poor peripheral irulation in whom a reliable oximetry signal annot be obtained. [Grade D] Any other evidene from the patient s medial ondition that would indiate that blood gas results would be useful in the patient s management (eg, an unexpeted hange in trak and trigger systems suh as a sudden rise of several units in the mews or an unexpeted fall in oxygen saturation of 3% or more, even if within the target range). [Grade D] Oxygen use in speifi illnesses See tables 1 4 and figs 1 and 2 (and setion 8 in main text) Critial illness requiring high levels of supplemental oxygen: see table 1 and setion 8 Serious illness requiring moderate levels of supplemental oxygen if a patient is hypoxaemi: see table 2 and setion 8. COPD and other onditions requiring ontrolled or low-dose oxygen therapy: see table 3 and setion 8. Conditions for whih patients should be monitored losely but oxygen therapy is not required unless the patient is hypoxaemi: see table 4 and setion 8. Oxygen therapy in pregnany (setion ) 14. Women who suffer from major trauma, sepsis or aute illness during pregnany should reeive the same oxygen therapy as any other seriously ill patients, with a target oxygen saturation of 94 98%. The same target range should be applied to women with hypoxaemia due to aute ompliations of pregnany (eg, ollapse related to amnioti fluid embolus, elampsia or antepartum or postpartum haemorrhage). [Grade D] vi2

3 15. Women with underlying hypoxaemi onditions (eg, heart failure) should be given supplemental oxygen during labour to ahieve an oxygen saturation of 94 98%. [Grade D] 16. All women with evidene of hypoxaemia who are more than 20 weeks pregnant should be managed with left lateral tilt to improve ardia output. [Grade B] 17. The use of oxygen during labour is widespread but there is evidene that this may be harmful to the fetus. The use of oxygen during labour is therefore not urrently reommended in situations where the mother is not hypoxaemi (exept as part of a ontrolled trial). [Grade A] Emergeny use of oxygen in prehospital and hospital are (setions 8 and 9) 18. Pulse oximetry must be available in all loations where emergeny oxygen is being used (see also the limitations of using pulse oximetry, setion 7.1.2). [Grade D] 19. Emergeny oxygen should be available in primary are medial entres, preferably using oxygen ylinders with integral high-flow regulators. Alternatively, oxygen ylinders fitted with high-flow regulators (delivering over 6 l/ min) must be used. [Grade D] 20. All douments whih reord oximetry measurements should state whether the patient is breathing air or a speified dose of supplemental oxygen. [Grade C] 21. The oxygen saturation should be monitored ontinuously until the patient is stable or arrives at hospital for a full assessment. The oxygen onentration should be adjusted upwards or downwards to maintain the target saturation range. [Grade D] 22. In most emergeny situations, oxygen is given to patients immediately without a formal presription or drug order. The lak of a presription should never prelude oxygen being given when needed in an emergeny situation. However, a subsequent written reord must be made of what oxygen therapy has been given to every patient (in a similar manner to the reording of all other emergeny treatment). [Grade D] 23. Patients with COPD (and other at-risk onditions) who have had an episode of hyperapni respiratory failure should be issued with an oxygen alert ard and with a 24% or 28% Venturi mask. They should be instruted to show the ard to the ambulane rew and emergeny department staff in the event of an exaerbation. [Grade C] 24. The ontent of the alert ard should be speified by the physiian in harge of the patient s are, based on previous blood gas results. [Grade D] 25. The primary are team and ambulane servie should also be informed by the responsible liniian that the patient has had an episode of hyperapni respiratory failure and arries an oxygen alert ard. The home address and ideal oxygen dose or target saturation ranges of these patients an be flagged in the ambulane ontrol systems and disseminated to ambulane rews when required. [Grade D] 26. Out-of-hours servies providing emergeny primary are servies should be informed by a responsible liniian that the patient has had an episode of hyperapni respiratory failure and arries an oxygen alert ard. Use of oxygen in these patients will be guided by the instrutions on the alert ard. [Grade D] 27. During ambulane journeys oxygen-driven nebulisers should be used for patients with asthma and may be used for patients with COPD in the absene of an air-driven ompressor system. If oxygen is used for patients with known COPD, its use should be limited to 6 min. This will deliver most of the nebulised drug dose but limit the risk of hyperapni respiratory failure (setion ). [Grade D] 28. If a patient is suspeted to have hyperapnia or respiratory aidosis due to exessive oxygen therapy, the oxygen therapy should not be disontinued but should be stepped down to 28% or 24% oxygen from a Venturi mask depending on oxygen saturation and subsequent blood gas results. [Grade C] Equipment used to deliver emergeny oxygen therapy (see setion 10) 29. (a) It is reommended that the following delivery devies should be available in prehospital settings where oxygen is administered: [Grade D] high onentration reservoir mask (non-rebreathe mask) for high-dose oxygen therapy; nasal annulae (preferably) or a simple fae mask for medium-dose oxygen therapy; 28% Venturi mask for patients with definite or likely COPD (patients who have an oxygen alert ard may have their own 24% or 28% Venturi mask); traheostomy masks for patients with traheostomy or previous laryngetomy. (b) Most hospital patients an be managed with the same delivery devie as in 29a, but 24% Venturi masks should also be available. [Grade D] 30. For many patients Venturi masks an be substituted with nasal annulae at low flow rates (1 2 l/min) to ahieve the same target range one patients have stabilised. [Grade D] 31. The flow rate from simple fae masks should be adjusted between 5 and 10 l/min to ahieve the desired target saturation. Flow rates below 5 l/min may ause arbon dioxide rebreathing and inreased resistane to inspiration. [Grade C] 32. Patients with COPD with a respiratory rate of.30 breaths/min should have the flow rate set to 50% above the minimum flow rate speified for the Venturi mask and/or pakaging (inreasing the oxygen flow rate into a Venturi mask inreases the total gas flow from the mask but does not inrease the onentration of oxygen whih is delivered). [Grade C] 33. Trusts should take measures to eliminate the risk of oxygen tubing being onneted to the inorret wall oxygen outlet or to outlets that deliver ompressed air or other gases instead of oxygen. Air flow meters should be removed from the wall sokets or overed with a designated air outlet over when not in use. Speial are should be taken if twin oxygen outlets are in use. [Grade D] 34. Humidifiation is not required for the delivery of low-flow oxygen or for the short-term use of high-flow oxygen. It is not therefore required in prehospital are. Pending the results of linial trials, it is reasonable to use humidified oxygen for patients who require high-flow oxygen systems for more than 24 h or who report upper airway disomfort due to dryness. [Grade B] 35. In the emergeny situation humidified oxygen use an be onfined to patients with traheostomy or an artifiial airway, although these patients an be managed without humidifiation for short periods of time (eg, ambulane journeys). [Grade D] vi3

4 36. Humidifiation may also be of benefit to patients with visous seretions ausing diffiulty with expetoration. This benefit an be ahieved using nebulised normal saline. [Grade C] 37. Bubble bottles should not be used beause there is no evidene of linially signifiant benefit but there is a risk of infetion. [Grade C] 38. When oxygen is required by patients with prior traheostomy or laryngetomy, a traheostomy mask (varying the flow as neessary) should ahieve the desired oxygen saturation (tables 1 4). An alternative delivery devie, usually a two-piee devie fitted diretly to the traheostomy tube, may be neessary if the patient deteriorates. [Grade D] Oxygen therapy during nebulised treatments (see setion 10) 39. For patients with asthma, nebulisers should be driven by piped oxygen or from an oxygen ylinder fitted with a highflow regulator apable of delivering a flow rate of.6 l/min. The patient should be hanged bak to his/her usual mask when nebuliser therapy is omplete. If the ylinder does not produe this flow rate, an air-driven nebuliser (with eletrial ompressor) should be used with supplemental oxygen by nasal annulae at 2 6 l/min to maintain an appropriate oxygen saturation level. [Grade D] 40. When nebulised bronhodilators are given to patients with hyperapni aidosis, they should be driven by ompressed air and, if neessary, supplementary oxygen should be given onurrently by nasal annulae at 2 4 l/min to maintain an oxygen saturation of 88 92%. The same preautions should be applied to patients who are at risk of hyperapni respiratory failure prior to the availability of blood gas results. One the nebulised treatment is ompleted for patients at risk of hyperapnia, ontrolled oxygen therapy with a fixed onentration (Venturi) devie should be reinstituted. [Grade D] During ambulane journeys, oxygen-driven nebulisers should be used for patients with asthma and may be used for patients with COPD in the absene of an air-driven ompressor system. If oxygen is used for patients with known COPD, its use should be limited to 6 min. This will deliver most of the nebulised drug dose but limit the risk of hyperapni respiratory failure (see reommendation 27). Presription, administration, monitoring and disontinuation of oxygen therapy (see setions 11 and 12) Oxygen should always be presribed or ordered on a designated doument. In emergenies, oxygen should be given first and doumented later. See reommendations in setion 11 of the main guideline for presription, administration and monitoring of oxygen therapy and reommendations in setion 12 for guidane on meaning and disontinuation of oxygen therapy. All primary are trusts, ambulane trusts and hospital trusts should take speifi measures to institute safe and effetive administration and doumentation of oxygen as desribed in reommendations in setions 11 and 12 of this guideline. Table 1 Critial illnesses requiring high levels of supplemental oxygen (see setion 8.10) The initial oxygen therapy is a reservoir mask at 15 l/min. One stable, redue the oxygen dose and aim for target saturation range of 94 98% If oximetry is unavailable, ontinue to use a reservoir mask until definitive treatment is available. Patients with COPD and other risk fators for hyperapnia who develop ritial illness should have the same initial target saturations as other ritially ill patients pending the results of blood gas measurements, after whih these patients may need ontrolled oxygen therapy or supported ventilation if there is severe hypoxaemia and/or hyperapnia with respiratory aidosis. Additional omments Grade of reommendation Cardia arrest or resusitation Use bag-valve mask during ative resusitation Grade D Aim for maximum possible oxygen saturation until the patient is stable Shok, sepsis, major trauma, Also give speifi treatment for the underlying ondition Grade D near-drowning, anaphylaxis, major pulmonary haemorrhage Major head injury Early intubation and ventilation if omatose Grade D Carbon monoxide poisoning Give as muh oxygen as possible using a bag-valve mask or reservoir Grade C mask. Chek arboxyhaemoglobin levels A normal or high oximetry reading should be disregarded beause saturation monitors annot differentiate between arboxyhaemoglobin and oxyhaemoglobin owing to their similar absorbanes. The blood gas PaO 2 will also be normal in these ases (despite the presene of tissue hypoxia) COPD, hroni obstrutive pulmonary disease; PaO 2, arterial oxygen tension. vi4

5 Table 2 Serious illnesses requiring moderate levels of supplemental oxygen if the patient is hypoxaemi (setion 8.11) The initial oxygen therapy is nasal annulae at 2 6 l/min (preferably) or simple fae mask at 5 10 l/min unless stated otherwise. For patients not at risk of hyperapni respiratory failure who have saturation,85%, treatment should be ommened with a reservoir mask at l/min. The reommended initial oxygen saturation target range is 94 98%. If oximetry is not available, give oxygen as above until oximetry or blood gas results are available. Change to reservoir mask if the desired saturation range annot be maintained with nasal annulae or simple fae mask (and ensure that the patient is assessed by senior medial staff). If these patients have o-existing COPD or other risk fators for hyperapni respiratory failure, aim at a saturation of 88 92% pending blood gas results but adjust to 94 98% if the PaCO 2 is normal (unless there is a history of previous hyperapni respiratory failure requiring NIV or IPPV) and rehek blood gases after min. Additional omments Grade of reommendation Aute hypoxaemia Reservoir mask at l/min if initial SpO 2,85%, otherwise nasal Grade D (ause not yet diagnosed) annulae or simple fae mask Patients requiring reservoir mask therapy need urgent linial assessment by senior staff Aute asthma Grade C Pneumonia Grade C Lung aner Grade C Postoperative breathlessness Management depends on underlying ause Grade D Aute heart failure Consider CPAP or NIV in ases of pulmonary oedema Grade D Pulmonary embolism Most patients with minor pulmonary embolism are not hypoxaemi and Grade D do not require oxygen therapy Pleural effusions Most patients with pleural effusions are not hypoxaemi. If hypoxaemi, Grade D treat by draining the effusion as well as giving oxygen therapy Pneumothorax Needs aspiration or drainage if the patient is hypoxaemi. Most patients Grades C and D with pneumothorax are not hypoxaemi and do not require oxygen therapy Use a reservoir mask at l/min if admitted for observation. Aim at 100% saturation (oxygen aelerates learane of pneumothorax if drainage is not required) Deterioration of lung fibrosis Reservoir mask at l/min if initial SpO 2,85%, otherwise nasal Grade D or other interstitial lung disease annulae or simple fae mask Severe anaemia The main issue is to orret the anaemia Grades B and D Most anaemi patients do not require oxygen therapy Sikle ell risis Requires oxygen only if hypoxaemi (below the above target ranges or Grade B below what is known to be normal for the individual patient) Low oxygen tension will aggravate sikling COPD, hroni obstrutive pulmonary disease; CPAP, ontinuous positive airway pressure; IPPV, intermittent positive pressure ventilation; NIV, non-invasive ventilation; PaCO 2, arterial arbon dioxide tension; SpO 2, arterial oxygen saturation measured by pulse oximetry. vi5

6 Table 3 COPD and other onditions requiring ontrolled or low-dose oxygen therapy (setion 8.12) Prior to availability of blood gases, use a 28% Venturi mask at 4 l/min and aim for an oxygen saturation of 88 92% for patients with risk fators for hyperapnia but no prior history of respiratory aidosis. [Grade D] Adjust target range to 94 98% if the PaCO 2 is normal (unless there is a history of previous NIV or IPPV) and rehek blood gases after min [Grade D] Aim at a prespeified saturation range (from alert ard) in patients with a history of previous respiratory aidosis. These patients may have their own Venturi mask. In the absene of an oxygen alert ard but with a history of previous respiratory failure (use of NIV or IPPV), treatment should be ommened using a 28% oxygen mask at 4 l/min in prehospital are or a 24% Venturi mask at 2 4 l/min in hospital settings with an initial target saturation of 88 92% pending urgent blood gas results. [Grade D] If the saturation remains below 88% in prehospital are despite a 28% Venturi mask, hange to nasal annulae at 2 6 l/min or a simple mask at 5 l/min with target saturation of 88 92%. All at-risk patients with alert ards, previous NIV or IPPV or with saturation,88% in the ambulane should be treated as a high priority. Alert the A&E department that the patient requires immediate senior assessment on arrival at the hospital. [Grade D] If the diagnosis is unknown, patients aged.50 years who are long-term smokers with a history of hroni breathlessness on minor exertion suh as walking on level ground and no other known ause of breathlessness should be treated as if having COPD for the purposes of this guideline. Patients with COPD may also use terms suh as hroni bronhitis and emphysema to desribe their ondition but may sometimes mistakenly use asthma. FEV 1 should be measured on arrival in hospital if possible and should be measured at least one before disharge from hospital in all ases of suspeted COPD. [Grade D] Patients with a signifiant likelihood of severe COPD or other illness that may ause hyperapni respiratory failure should be triaged as very urgent and blood gases should be measured on arrival in hospital. [Grade D] Blood gases should be reheked after min (or if there is linial deterioration) even if the initial PaCO 2 measurement was normal. [Grade D] If the PaCO 2 is raised but ph is >7.35 ([H + ] (45 nmol/l), the patient has probably got long-standing hyperapnia; maintain target range of 88 92% for these patients. Blood gases should be repeated at min to hek for rising PaCO 2 or falling ph. [Grade D] If the patient is hyperapni (PaCO 2.6 kpa or 45 mm Hg) and aidoti (ph,7.35 or [H + ].45 nmol/l) onsider non-invasive ventilation, espeially if aidosis has persisted for more than 30 min despite appropriate therapy. [Grade A] Additional omments Grade of reommendation COPD May need lower range if aidoti or if known to be very sensitive to oxygen Grade C therapy. Ideally use alert ards to guide treatment based on previous blood gas results. Inrease flow by 50% if respiratory rate is.30 (see reommendation 32) Exaerbation of CF Admit to regional CF entre if possible; if not, disuss with regional entre or Grade D manage aording to protool agreed with regional CF entre Ideally use alert ards to guide therapy. Inrease flow by 50% if respiratory rate is.30 (see reommendation 32) Chroni neuromusular May require ventilatory support. Risk of hyperapni respiratory failure Grade D disorders Chest wall disorders For aute neuromusular disorders and subaute onditions suh as Guillain-Barré Grade D syndrome (see table 4) Morbid obesity Grade D CF, ysti fibrosis; COPD, hroni obstrutive pulmonary disease; CPAP, ontinuous positive airway pressure; IPPV, intermittent positive pressure ventilation; NIV, non-invasive ventilation; PaCO 2, arterial arbon dioxide tension; SpO 2, arterial oxygen saturation measured by pulse oximetry. vi6

7 Table 4 Conditions for whih patients should be monitored losely but oxygen therapy is not required unless the patient is hypoxaemi (setion 8.13) If hypoxaemi, the initial oxygen therapy is nasal annulae at 2 6 l/min or simple fae mask at 5 10 l/min unless saturation is,85% (use reservoir mask) or if at risk from hyperapnia (see below). The reommended initial target saturation range, unless stated otherwise, is 94 98% If oximetry is not available, give oxygen as above until oximetry or blood gas results are available If patients have COPD or other risk fators for hyperapni respiratory failure, aim at a saturation of 88 92% pending blood gas results but adjust to 94 98% if the PaCO 2 is normal (unless there is a history of respiratory failure requiring NIV or IPPV) and rehek blood gases after min Myoardial infartion and aute oronary syndromes Stroke Pregnany and obstetri emergenies Hyperventilation or dysfuntional breathing Most poisonings and drug overdoses (see table 1 for arbon monoxide poisoning) Poisoning with paraquat or bleomyin Metaboli and renal disorders Aute and subaute neurologial and musular onditions produing musle weakness Additional omments Most patients with aute oronary artery syndromes are not hypoxaemi and the benefits/harms of oxygen therapy are unknown in suh ases Most stroke patients are not hypoxaemi. Oxygen therapy may be harmful for non-hypoxaemi patients with mild to moderate strokes. Oxygen therapy may be harmful to the fetus if the mother is not hypoxaemi (see reommendations 14 17) Exlude organi illness. Patients with pure hyperventilation due to anxiety or pani attaks are unlikely to require oxygen therapy Rebreathing from a paper bag may ause hypoxaemia and is not reommended Hypoxaemia is more likely with respiratory depressant drugs, give antidote if available (eg, naloxone for opiate poisoning) Chek blood gases to exlude hyperapnia if a respiratory depressant drug has been taken. Avoid high blood oxygen levels in ases of aid aspiration as there is theoretial evidene that oxygen may be harmful in this ondition Monitor all potentially serious ases of poisoning in a level 2 or level 3 environment (high dependeny unit or ICU) Patients with paraquat poisoning or bleomyin lung injury may be harmed by supplemental oxygen Avoid oxygen unless the patient is hypoxaemi Target saturation is 88 92% Most do not need oxygen (tahypnoea may be due to aidosis in these patients) These patients may require ventilatory support and they need areful monitoring whih inludes spirometry. If the patient s oxygen level falls below the target saturation, they need urgent blood gas measurements and are likely to need ventilatory support Grade of reommendation Grade D Grade B Grades A D Grade C Grade D Grade C Grade D Grade C COPD, hroni obstrutive pulmonary disease; ICU, intensive are unit; IPPV, intermittent positive pressure ventilation; NIV, non-invasive ventilation; PaCO 2, arterial arbon dioxide tension; SpO 2, arterial oxygen saturation measured by pulse oximetry. vi7

8 Thorax: first published as /thx on 6 Otober Downloaded from Figure 1 Chart 1: Oxygen presription for autely hypoxaemi patients in hospital. ABG, arterial blood gas; COPD, hroni obstrutive pulmonary disease; FIO 2, fration of inspired oxygen; ICU, intensive are unit; NIV, non-invasive ventilation; PCO 2, arbon dioxide tension; SpO 2, arterial oxygen saturation measured by pulse oximetry. on 5 July 2018 by guest. Proteted by opyright. vi8

9 Thorax: first published as /thx on 6 Otober Downloaded from Figure 2 Chart 2: Flow hart for oxygen administration on general wards in hospitals. ABG, arterial blood gas; EPR, eletroni patient reord; EWS, Early Warning Sore; SpO 2, arterial oxygen saturation measured by pulse oximetry. on 5 July 2018 by guest. Proteted by opyright. vi9

10 HIERARCHY OF EVIDENCE AND GRADING OF RECOMMENDATIONS Levels of evidene and grades of reommendation Levels of evidene and grades of reommendation are based on the levels of evidene used in the NICE COPD guideline 25 (see tables 5 and 6). For most of the topis overed by the guideline there were either no randomised trials or just a handful of observational studies. Members of the group reviewed the evidene for eah topi and assigned the most appropriate grading whih was usually grade C evidene (ase-ontrol or ohort studies) or grade D evidene (expert opinion or ase reports). Eah reommendation has been alloated a grading whih diretly reflets the hierarhy of evidene upon whih it is based. Please note that the hierarhy of evidene and the reommendation gradings relate to the strength of the literature, not to linial importane. This is espeially important in the field of oxygen therapy where there are very few ontrolled trials. Table 5 Level of evidene Ia Ib IIa IIb III IV Hierarhy of evidene Type of evidene Table 6 Grading of reommendations Level of evidene Type of evidene A B C D Evidene from systemati reviews or meta-analysis of randomised ontrolled trials Evidene from at least one randomised ontrolled trial Evidene from at least one ontrolled study without randomisation Evidene from at least one other type of quasiexperimental study Evidene from non-experimental desriptive studies suh as omparative studies, orrelation studies and ase-ontrol studies Evidene from expert ommittee reports or opinions and/or linial experiene of respeted authorities Based on hierarhy I evidene Based on hierarhy II evidene or extrapolated from hierarhy I evidene Based on hierarhy III evidene or extrapolated from hierarhy I or II evidene Diretly based on hierarhy IV evidene or extrapolated from hierarhy I, II or III evidene SECTION 1: INTRODUCTION 1.1 Clinial ontext Oxygen is probably the ommonest drug to be used in the are of patients who present with medial emergenies. Currently, ambulane teams and emergeny department teams are likely to give oxygen to virtually all breathless patients and to a large number of patients with other onditions suh as ishaemi heart disease, sepsis or trauma. The North West Ambulane Servie serves a population of about 7.25 million people and transports about people to hospital emergeny departments eah year. About 34% of these journeys involve oxygen use at some stage. 1 This translates to about two million instanes of emergeny oxygen use per annum by all UK ambulane servies, with further use in patients homes, GP surgeries and in hospitals. At present, oxygen is administered for three main indiations of whih only one is evidene-based. First, oxygen is given to orret hypoxaemia as there is good evidene that severe hypoxaemia is harmful. Seond, oxygen is administered to ill patients in ase they might beome hypoxaemi. Reent evidene suggests that this pratie may atually plae patients at inreased risk if severe hypoxaemia does atually develop (see setion 6.3.4). Third, a very high proportion of medial oxygen is administered beause most liniians believe that oxygen an alleviate breathlessness. However, there is no evidene that oxygen relieves breathlessness in non-hypoxaemi patients and there is evidene of lak of effetiveness in non-hypoxaemi breathless patients with hroni obstrutive pulmonary disease (COPD) and advaned aner (see setions 6.6 and ). 1.2 Presription of oxygen Most liniians who deal with medial emergenies will enounter adverse inidents and oasional deaths due to underuse and overuse of oxygen. Audits of oxygen use and oxygen presription have shown onsistently poor performane in many ountries. 2 8 One major problem is that healthare professionals reeive onfliting advie about oxygen use from different experts during their training and during their linial areers, and many are onfused about the entire area of oxygen presription and use. 1.3 Need for a guideline for emergeny oxygen therapy and purpose of the guideline There is onsiderable ontroversy onerning the benefits and risks of oxygen treatment in virtually all situations where oxygen is used. Unfortunately, this is an area of mediine where there are many strongly-held beliefs but very few randomised ontrolled trials. The only published UK guideline for emergeny oxygen therapy is the North West Oxygen Guideline published in 2001, based on a systemati literature review by the same authors Against this bakground, the Standards of Care Committee of the British Thorai Soiety (BTS) established a working party in assoiation with 21 other soieties and olleges listed at the front of this doument. The objetive was to produe an evidene-based and up-to-date guideline for emergeny oxygen use in the UK. 1.4 Intended users of the guideline and sope of the guideline This guideline is intended for use by all healthare professionals who may be involved in emergeny oxygen use. This will inlude ambulane staff, paramedis, dotors, nurses, midwives, physiotherapists, pharmaists and all other healthare professionals who may deal with ill or breathless patients. vi10

11 Speifi versions of this guideline will be available on the BTS website for use in the following situations: Hospital use Primary are use Ambulane use Version for use by nursing staff These abbreviated versions of the guideline will ontain the key reommendations and tables and harts that are relevant to the partiular situation. The mini-guidelines an be downloaded by health are trusts for use on trust intranets and to produe paper versions of the guideline for key staff. 1.5 Areas overed by this guideline The guideline will address the use of oxygen in three main ategories of adult patients in the prehospital and hospital setting: Critially ill or hypoxi patients. Patients at risk of hypoxaemia. Non-hypoxi patients who might benefit from oxygen (eg, arbon monoxide poisoning). 1.6 Areas not overed by this guideline Oxygen use in paediatris: the present guideline applies only to subjets aged >16 years. Oxygen use for high altitude ativities. Oxygen use during air travel. Underwater diving and diving aidents. Oxygen use in animal experiments. Oxygen use during surgery and anaesthesia or during endosopy. Oxygen use in high-dependeny units. Oxygen use in intensive are units. Interhospital level 3 transfers. Hyperbari oxygen. Use of heliox mixtures. Use of nitrous oxide/oxygen mixtures (eg, Entonox). Respiratory support tehniques inluding intubation, invasive ventilation, non-invasive ventilation (NIV) and ontinuous positive airway pressure (CPAP). Self-initiated use of oxygen by patients who have home oxygen for any reason (this is overed by the guidelines for home oxygen use). Ongoing are of hypoxaemi patients at home. 1.7 Limitations of the guideline This guideline is based on the best available evidene onerning oxygen therapy. However, a guideline an never be a substitute for linial judgement in individual ases. There may be ases where it is appropriate for liniians to at outwith the advie ontained in this guideline beause of the needs of individual patients. Furthermore, the responsibility for the are of individual patients rests with the liniian in harge of the patient s are and the advie offered in this guideline must, of neessity, be of a general nature and should not be relied upon as the only soure of advie in the treatment of individual patients. In partiular, this guideline gives very little advie about the management of the many medial onditions that may ause hypoxaemia (apart from the speifi issue of managing the patients hypoxaemia). Readers are referred to other guidelines for advie on the management of speifi onditions suh as COPD, pneumonia, heart failure, et. Some of these disease-speifi guidelines suggest slightly different approahes to emergeny oxygen therapy whereas the present guideline aims to provide simple all-embraing advie. All differenes involving oxygen therapy for ommon medial emergenies are disussed in detail in setion 10 of this guideline. SECTION 2: METHODOLOGY OF GUIDELINE PRODUCTION 2.1 Establishment of guideline team The need for a national guideline for emergeny oxygen use was reognised by the BTS Standards of Care Committee in A working party was established with representatives from a wide range of professions involved in oxygen therapy and a lay representative (see full list of guideline group members in setion 16). The original group was expanded in 2006 beause it beame lear that the development and implementation of a national guideline would require input from a very wide range of professional groups. Most development and editing of the guideline took plae subsequent to this expansion of the group. The group agreed the remit of this guideline and a series of key questions as shown below. The group devised a searh strategy for relevant studies. A Medline searh for oxygen yielded over a quarter of a million hits, most of whih were not relevant to this guideline. For this reason, the BTS ommissioned the Centre for Reviews and Dissemination and Centre for Health Eonomis at the University of York to undertake bespoke literature searhes using the searh strategies shown in detail in Appendix 14 on the BTS website ( 2.2 Summary of key questions Key question 1: Physiology and pathophysiology of oxygen What are the dangers of hypoxia/hypoxaemia (ie, what happens to the human body)? What level of hypoxaemia is dangerous to all patients (even healthy adults)?. What level of hypoxaemia is dangerous to vulnerable groups (eg, ishaemi heart disease, stroke, elderly)? Repeat the above searhes with additional key words: elderly, stroke, myoardial infartion, heart failure, hroni obstrutive pulmonary disease (COPD), trauma, renal failure. Same questions for hyperarbia/hyperapnia: Searh for hyperapnia ombined with terms implying a harmful outome (death/tissue injury/brain damage/ oma). What level of hyperapnia is dangerous to all patients? What level of hyperapnia is dangerous to vulnerable groups (as above)? Same questions for respiratory aidosis: Searh for respiratory aidosis ombined with terms implying a harmful outome (death/tissue injury/brain damage/oma). What level of respiratory aidosis is dangerous to all patients? What level of respiratory aidosis is dangerous to vulnerable groups (as above)? Key question 2: Clinial aspets of hypoxaemia and oxygen therapy for ommon medial emergenies How to assess hypoxaemia (linial, early warning sore systems, oximetry, arterial and apillary blood gases). How to assess hyperarbia/hyperapnia. vi11

12 Use of oxygen to relieve symptomati breathlessness. Use of oxygen in aute COPD. Use of oxygen in aute asthma. Use of oxygen in pneumonia. Use of oxygen for pulmonary embolus. Use of oxygen in trauma. Use of oxygen in heart failure. Use of oxygen in myoardial infartion. Use of oxygen in angina. Use of oxygen for other patients with less ommon onditions were searhed individually (eg, ysti fibrosis, musular dystrophy, motor neurone disease, severe kyphosoliosis, anaphylaxis, hyperventilation). Key question 3: Oxygen presription, oxygen delivery systems and oxygen transport Oxygen arriage in transport (pratial issues; safety issues). Oxygen delivery systems in ambulanes. Presription of oxygen. Loal hospital guidelines for oxygen use. Oxygen delivery systems in hospitals. Advantages/disadvantages of eah delivery system (Venturi masks, simple fae masks, nasal annulae, high-flow masks suh as non-rebreathing reservoir masks). Use of oxygen-driven nebulisers. Use of alert ards, alert braelets or similar hazard warning systems for patients who are known to be at risk of hyperapnia. 2.3 How the evidene was assimilated into the guideline The initial searh strategy was devised at two meetings of the group in 2004 and The searhes in Otober 2005 yielded 3306 papers, the abstrats of whih were heked for relevane by group members. One hundred and eighty-four of these abstrats were onsidered to be relevant to the present guideline. Full reprints of all relevant papers were obtained. Further referenes were obtained from the group s personal literature olletions and from the referenes ontained within the papers whih the searh yielded and by foused literature searhes by members of the guideline group. The group ontinued to monitor the literature up to the end of 2007 for important new publiations or very high quality abstrats from international meetings that were thought to be relevant to this guideline. The group was divided into three subgroups to work on speifi areas of oxygen use: (1) emergeny are; (2) hospital are; (3) oxygen physiology and devies. Evidene from the literature searhes was graded aording to the levels of evidene used in the NICE COPD guideline (see tables 5 and 6). The Guideline Development Group orresponded by on a regular basis (usually at least one weekly) for most of 2006 to disuss the evidene and to produe an initial outline of the guideline and its key reommendations. The guideline was onsolidated over the ourse of 2006 and early 2007 with eah setion being led by nominated group members but taking into aount feedbak from the omplete group. Meetings of the full group were held in February 2006, September 2006 and February Between November 2006 and February 2007 the group had an intensive review and disussion of one guideline setion per week with the objetive of ahieving a onsensus on all of the key points before the final meeting of the group in February The draft guideline was first submitted to the BTS Standards of Care Committee in Marh The guideline was further refined by disussion following omments by this ommittee. The resulting draft was sent to 17 peer reviewers (see setion 17) and was posted on the BTS website for 4 weeks in August 2007 and omments were invited. The doument was then sent bak to the Standards of Care Committee and the 21 other Soieties and Colleges for endorsement. 2.4 Piloting the guideline The priniples of the guideline (target saturation ranges, et) have been piloted sine 2004 at Salford Royal University Hospital and Southend University Hospital. The pilot projets have inluded the following elements: Disussion with senior olleagues and management to agree the need for an oxygen guideline (and the ontent). Trust-wide introdution of the agreed hospital poliy. Eduational programme for dotors, nurses and other users of oxygen. Designing presription harts and patient observation harts to failitate the standardisation of oxygen therapy (harts 3 and 4 in figs 17 and 18 in the guideline). Prodution of a detailed implementation doument whih has beome hospital poliy in both hospitals (web appendix 3). The harts whih are neessary to guide the presription and administration of oxygen (harts 1 and 2 in figs 1 and 2) have been piloted suessfully at both hospitals. The eduational materials and leture presentations in web appendix 9 have been piloted in both hospitals. There was a lot of disussion with olleagues about the ideal target saturation range and about how to implement safe oxygen presribing. These issues should not arise with implementation of this national guideline as the key issues are already agreed by all of the relevant speialties and are as evidene-based as is possible. Implementation proeeded smoothly at both hospitals and audit showed improved pratie. However, a lot of effort is required to maintain good quality presribing of oxygen and the role of oxygen hampions has been piloted suessfully in both hospitals (see setion 14.6). 2.5 Planned review and updating of the guideline The guideline will be reviewed by the BTS and by the endorsing organisations within 5 years from publiation (2013). SECTION 3: NORMAL VALUES AND DEFINITIONS Normal blood levels of oxygen and arbon dioxide. Normal oxygen saturation (SaO 2 ) and normal blood ph. Definitions of hypoxaemia, hypoxia, hyperapnia, aidosis, respiratory failure. Oxygen is essential for mammalian life; severe hypoxaemia suh as that seen during ardia arrest, suffoation or drowning will ause loss of onsiousness, rapid organ failure and death. Oxygen is arried in the bloodstream bound to the haemoglobin moleule and delivered to the tissues. Oxygen demand and oxygen delivery inrease during exerise and redue during rest and sleep. 3.1 Blood levels of oxygen and arbon dioxide in health and disease The human lung delivers oxygen to the blood and removes arbon dioxide. Several mehanisms exist to regulate breathing in suh a way that both gases are maintained within quite a narrow range. vi12

13 3.1.1 Normal ranges for oxygen saturation (SaO 2 ) and oxygen tension (PaO 2 ) in the blood at sea level For adults aged,70 years, the two standard deviation (2SD) range for SaO 2 is approximately 94 98% at sea level but this may deline gradually within this age range. 11 The normal range for PaO 2 in the blood in seated adults at sea level is shown in table 7. However, the PaO 2 is 0.8 kpa (6 mm Hg) lower in the supine position than in the upright position 12 and most emergeny measurements are made in the supine position Oxygen saturation in elderly patients The mean SaO 2 may be lower in older people than in young adults. However, it is diffiult to dissoiate the effets of advaning age from the effets of the diseases that beome ommoner in old age. Some papers have reported a fall in the blood PaO 2 in older subjets but others have failed to onfirm this observation The mean SaO 2 in seated adults aged.64 years in one published study was 95.5% ompared with 96.9% in adults aged years, and the standard deviation was wider in the older age group with a 2SD range of % (table 7). 11 The mean (SD) SaO 2 for reumbent healthy men aged >70 years in another study was 95.3 (1.4)% giving a 2SD range of % for men of this age. 13 The mean (SD) SaO 2 was 94.8 (1.7)% for reumbent healthy women aged >70 years with a 2SD range of %. The authors of this study did not observe any agerelated deline in SaO 2 beyond the age of 70 years. The mean SaO 2 in this study of approximately 95.0% for reumbent healthy men and women aged >70 years was below the normal range for seated healthy young adults. The mean PaO 2 in elderly subjets in this study was 10.3 kpa for men and 9.8 kpa for women, whih is lower than two other studies whih reported mean PaO 2 values of 11.2 kpa and 11.1 kpa in healthy elderly subjets Some of these differenes are probably due to different seletion of subjets, but there are also variations in the results obtained by different blood gas analysers. 17 Unfortunately there are no published data whih an provide a normal range for the SaO 2 in the elderly population in the UK. However, an as yet unpublished audit of 320 stable hospital patients in Salford and Southend with no history of lung disease found a mean (SD) SaO 2 of 96.7 (1.77)% (2SD range %) in patients aged >71 years (R O Drisoll, A Davison, L Ward, personal ommuniation). These values were measured by pulse oximetry in UK hospitals in 2008 and are more likely to represent the expeted normal range of pulse oximetry measurements in the elderly UK population than Table 7 mm Hg Mean (SD) PaO 2 Age (kpa and mm Hg) Mean (SD) PaO 2 and SaO 2 values (with range) in kpa and Range 2SD PaO 2 (kpa and mm Hg) Mean (SD) SaO 2 (%) SaO 2 2SD (0.71) (0.4) (5.3) (0.66) (0.7) (4.9) (1.02) (0.6) (7.6) (1.07) (1) (8) (0.60) (0.7) ) (1.43) (1.4) (10.7) PaO 2, arterial oxygen tension; SaO 2, arterial oxygen saturation. Values shown for seated healthy men and women non-smoking volunteers at sea level (adapted from Crapo et al 11 ). previous North Amerian studies based on blood gas estimations. The variation with age, sex and posture makes it diffiult to give a preise target range that will apply to all adults who might require oxygen therapy, but the guideline development ommittee believe that a target range of 94 98% will ahieve normal or near-normal SaO 2 formostadultsintheuk. Normal daytime haemoglobin SaO 2 is 96 98% in young adults in the seated position at sea level but the lower limit falls slightly with age and is about 95% in adults aged.70 years. [Evidene III] Oxygen saturation at altitude The partial pressure of oxygen in the atmosphere is substantially lower at high altitude, even at altitudes where large populations live. The SaO 2 at a given altitude varies with age, sex, ethni group and degree of alimatisation to altitude. For example, a sample of 3812 people of all ages living in Tibet at an altitude of about 4000 m had a mean SaO 2 of only 88.2%, but people native to the Andes had an SaO 2 about 2.6% higher than Tibetans living at the same altitude Millions of people live at these altitudes with SaO 2 values that would ause serious onern at sea level. The ity of La Paz in Bolivia has a mean altitude of 3600 m and a population of approximately 1.5 million people. The SaO 2 of limbers on Mount Everest (8848 m) an fall below 70%. 20 Sudden exposure to altitudes above about 4000 m an ause mountain sikness, high altitude pulmonary oedema and high altitude erebral oedema in unalimatised individuals. Long-term exposure to high altitude (or to hypoxaemia for any other reason) an lead to pulmonary hypertension Oxygen saturation in aute and hroni disease If the blood oxygen level falls to extremely low levels for even a few minutes (eg, during ardia arrest), tissue hypoxia and ell death will our, espeially in the brain. The brain appears to be the most vulnerable organ during profound hypoxaemia; brain malfuntion is the first symptom of hypoxia and brain injury is the most ommon long-term ompliation in survivors of ardia arrests and other episodes of profound hypoxaemia. Sudden exposure to low arterial oxygen saturations below about 80% an ause altered onsiousness even in healthy subjets. It is likely that other organs in patients with ritial illness or hroni organ damage are vulnerable to the risk of hypoxi tissue injury at oxygen levels above 80%. Most experts emphasise the importane of keeping the SaO 2 above 90% for most autely ill patients However, the degree of hypoxia that will ause ellular damage is not well established and probably is not an absolute value. Healthy older adults, for instane, have lower SaO 2 values at rest than younger adults. Patients with hroni lung diseases may tolerate low levels of SaO 2 hronially. However, although hronially hypoxaemi patients may tolerate an abnormally low SaO 2 at rest when in a linially stable ondition, these resting oxygen levels may not be adequate for tissue oxygenation during aute illness when the tissue oxygen demand may inrease (eg, sepsis, trauma, pneumonia, head injury; see setion 8). Aute hypoxaemia with SaO 2,90% and sometimes,80% is seen in many aute illnesses suh as pneumonia and heart failure and it is likely that the linial manifestations of hypoxaemia in illness would be similar to those of experimental hypoxaemia in hypobari hambers (impaired mental funtion followed by loss of onsiousness). However, the linial manifestations of the illness itself make it diffiult to identify vi13

14 whih symptoms and signs are due to hypoxaemia. Pure hypoxaemia, as seen in hypobari hambers and at altitude, does not seem to ause breathlessness in resting subjets. Patients with hroni diseases suh as COPD, lung fibrosis, neuromusular disorders or ongenital heart disease may routinely attend outpatient linis with SaO 2 levels well below 90% even at a time when their disease is stable. In an emergeny a liniian who was not familiar with suh a patient (when stable) might interpret the low saturation as having ourred autely and aim to ahieve an oxygen saturation that was well above the patient s usual oxygen saturation level. Many suh patients would qualify for long-term oxygen therapy. The UK COPD guideline 25 reommends a threshold of 7.3 kpa (55 mm Hg) below whih most patients with COPD will benefit from long-term oxygen therapy (equivalent to a SaO 2 of about 88 89%) and an arterial oxygen tension (PaO 2 ) threshold below 8.0 kpa (60 mm Hg) for patients with established or pulmonale and some other subgroups. Many patients with hroni lung disease, ongenital yanoti heart disease or hroni neuromusular onditions have oxygen saturations substantially below the normal range, even when linially stable. [Evidene III] Variation in oxygen saturation during sleep Healthy subjets in all age groups have greater variation in SaO 2 when sleeping than while awake. A study of 330 people referred to a sleep laboratory with normal results of overnight polysomnography (patients with ranial faial or neurologial abnormalities or previously diagnosed pulmonary disease were exluded) showed that desaturation routinely ourred with a mean (SD) minimum SaO 2 or nadir of 90.4 (3.1)% during the night (2SD range %). 26 The mean (SD) overnight SaO 2 nadir was 89.3 (2.8)% for subjets aged.60 years. 26 In this study subjets aged years spent 10% of the night with SaO 2 levels below 94.8% and half the night below 96.3%, and those aged.60 years spent 10% of the night below 92.8% and half the night below 95.1%. Furthermore, the authors of this study exluded obese patients with any features of sleep apnoea or hypopnoea beause these patients are known to desaturate to very low levels during sleep (often below 70%). The variation in SaO 2 during sleep is exaggerated by alohol and by sedative drugs. This makes it diffiult to evaluate a spot reading of SaO 2 on a sleeping subjet. It is suggested that SaO 2 measurements of sleeping subjets should be interpreted with aution and ideally observed for a few minutes to see if the subjet has got sustained hypoxaemia or just a transient normal noturnal dip. All subjets have transient dips in oxygen saturation at night with a mean nadir of 90.4% (2SD range %) in healthy subjets in all age groups. [Evidene III] Normal range for arbon dioxide tension (PaCO 2 ) in the blood The referene range for arterial arbon dioxide tension (PaCO 2 )is approximately kpa (34 46 mm Hg) for healthy adult men aged years. 27 Although this study was undertaken in 1948, it is onsistent with the linial experiene of the guideline group members and with most modern referene values for PaCO 2. Although different laboratories and textbooks give slightly different referene values, all are within 0.2 kpa of the above referene range. Any value of PaCO 2 of.6.1 kpa (45 mm Hg) should be onsidered abnormal, but values up to 6.7 kpa (50 mm Hg) may be obtained by breath-holding. 3.2 Definitions of hypoxaemia, hypoxia, type 1 respiratory failure and hyperoxia Hypoxaemia Hypoxaemia refers to low oxygen tension or partial pressure of oxygen (PaO 2 ) in the blood. For pratial reasons, hypoxaemia an also be measured in relation to oxyhaemoglobin saturation. In adults the normal range is influened by age and omorbidity and the normal ranges for healthy adults are given in setion The preise level at whih a patient beomes hypoxaemi is debatable. One ould argue that any saturation below the lower limit of normal onstitutes hypoxaemia. Various authors have defined hypoxaemia as SaO 2 of (1),94%; (2),92%; (3),90%; or (4) PaO 2,60 mm Hg or 8 kpa Most authors who have studied this area have defined hypoxaemia as PaO 2,60 mm Hg (8 kpa) or SaO 2,90%. 31 There is no known risk of hypoxi tissue injury above this level and many guidelines on ritial are set 90% as the minimum below whih SaO 2 should not be allowed to fall. Type 1 respiratory failure Type 1 respiratory failure is most widely defined as PaO 2,8 kpa or 60 mm Hg (equivalent to SaO 2 of approximately 90%) with a normal or low PaCO 2 level. 32 Hypoxia The term hypoxia is less speifi and refers to lak of oxygen in a partiular ompartment (eg, alveolar or tissue hypoxia). Tissue hypoxia may result from four main auses (see below). It should be noted that the first two auses hypoxaemia and anaemia do not always result in tissue hypoxia as oxygen delivery to tissues an be augmented in other ways suh as inreasing ardia output. Hypoxaemi hypoxia Hypoxaemi hypoxia (sometimes also referred to as hypoxi hypoxia) is present when the oxygen ontent in the blood is low due to redued partial pressure of oxygen. This ours naturally at altitude and in many diseases suh as emphysema whih impair the effiieny of gas exhange in the lungs. Anaemi hypoxia Anaemi hypoxia results from a redued level of haemoglobin available for oxygen transport. Although the patient may not be hypoxaemi (with a normal PaO 2 and oxygen saturation measured by oximetry (SpO 2 )), the redued oxygen ontent of the blood may lead to tissue hypoxia. Carbon monoxide poisoning may also produe a form of anaemi hypoxia by impairing the ability of haemoglobin to bind oxygen, thereby reduing oxygen-arrying apaity. Stagnant hypoxia Stagnant hypoxia is a low level of oxygen in the tissues due to inadequate blood flow (either globally or regionally). This ondition may our in the extremities if a person is exposed to old temperatures for prolonged periods of time and it is the ause of gangrene in tissue that is deprived of blood in severe peripheral vasular disease. Stagnant hypoxia may our in low ardia output states. Histotoxi hypoxia Histotoxi hypoxia is an inability of the tissues to use oxygen due to interruption of normal ellular metabolism. The best known example of this ours during yanide poisoning whih vi14

15 impairs ytohrome funtion. It is inreasingly thought that mitohondrial dysfuntion may lead to dereased oxygen utilisation in sepsis despite adequate oxygen delivery. This has also been termed ytopathi dysoxia. 33 Hyperoxia and hyperoxaemia Hyperoxia and hyperoxaemia are the ounterparts to the above terms and in this guideline refer to high oxygen ontent in the blood and high oxygen tension in the blood, respetively. As stated above, for pratial purposes the oxygen tension in the blood is often measured as oxyhaemoglobin saturation. Furthermore, this guideline is entred on providing target saturations for various onditions, but it should be noted that above a PaO 2 of approximately 16 kpa (120 mm Hg) the oxyhaemoglobin saturation will obviously not hange from 100%, yet the effets of further inreases in PaO 2 may be important in ertain onditions suh as COPD. This is disussed in further detail in setions 5 and Definition of hyperapnia and type 2 respiratory failure Hyperapnia is present when the PaCO 2 is above the normal range of kpa (34 46 mm Hg) and patients with hyperapnia are said to have type 2 respiratory failure even if the oxygen saturation is in the normal range Definition of aidosis (respiratory aidosis and metaboli aidosis) Aidosis Aidity in any fluid is determined by the onentration of hydrogen ions [H + ], and this is normally regulated between 35 and 45 nmol/l. Aidity is more often expressed in terms of ph where ph = 2log 10 [H + ]. The normal ph range of the blood in humans is between 7.35 and 7.45 units. Aidosis is defined as a ph,7.35 ([H + ].45 nmol/l) and alkalosis is defined as a ph.7.45 ([H + ],35 nmol/l). Aidosis an be aused by respiratory or metaboli disorders. Respiratory aidosis Carbon dioxide (CO 2 ) an ombine with water (H 2 O) to form arboni aid (H 2 CO 3 ) in the blood whih, in turn, dissoiates to biarbonate (HCO 32 ) and a hydrogen ion (H + ). Aute respiratory aidosis ours if the ph of the blood falls below 7.35 ([H + ].45 nmol/l) in the presene of a raised CO 2 level. If respiratory aidosis has been present for more than a few hours the kidney retains biarbonate to buffer the aidity of the blood and, over hours to days, this may be suffiient to produe a normal ph. This situation (high PaCO 2 with high biarbonate and normal ph) is known as ompensated respiratory aidosis. This situation is ommon in patients with hroni severe but stable COPD, but they may have an additional aute rise in PaCO 2 during an aute exaerbation giving rise to aute on hroni respiratory aidosis despite their high biarbonate level. This happens beause the biarbonate level was equilibrated with the previous CO 2 level and is insuffiient to buffer the sudden further inrease in CO 2 level that may our during an exaerbation of COPD. Respiratory aidosis is ommon in linial pratie. Plant and olleagues showed that about 20% of patients with aute exaerbations of COPD requiring hospital admission have respiratory aidosis. 34 Metaboli aidosis This an be aused by failure to exrete aid produed by the body s normal metaboli proesses (eg, during renal failure) or by inreased prodution of aid from abnormal metaboli onditions suh as diabeti ketoaidosis. Alternatively, it may result from diret loss of biarbonate from the kidney or gut (eg, during hroni diarrhoea). In all forms of metaboli aidosis there is a low blood biarbonate level, either due to loss of biarbonate or due to buffering of exess aid by biarbonate whih is exreted as CO 2. A ommon ause of metaboli aidosis is lati aidosis aused by tissue hypoxia. This may result from dereased oxygen delivery suh as ours in hypoxaemia, or low ardia output states or onditions suh as sepsis where oxygen onsumption is impaired in the fae of adequate oxygen delivery. In health, metaboli aidosis will our at peak exerise where oxygen delivery is insuffiient to meet demand. SECTION 4: GENERAL BLOOD GAS PHYSIOLOGY A full understanding of blood gas physiology in the body requires a detailed understanding of the anatomy, physiology and biohemistry of respiration and gas exhange. It is reognised that most readers of this guideline may not have had full training in all of these speialties, so this physiology setion ontains a brief overview of basi priniples for the nonspeialist reader (setion 4) followed by a more detailed overview of the pathophysiology of oxygen for the expert reader (setion 5). The rationale for targeted oxygen therapy is disussed in detail in setion Oxygen physiology Oxygen is the main fuel of the ells in mammalian bodies and it is essential for humans to maintain a safe level of oxygen in the bloodstream. Most of the oxygen arried in the blood is bound to an oxygen-arrying protein in red blood ells alled haemoglobin. Oxygen itself does not dissolve easily in blood so only a small amount is arried dissolved in the bloodstream. As there is a fixed amount of haemoglobin irulating in the blood, the amount of oxygen arried in the blood is often expressed in terms of how saturated with oxygen the irulating haemoglobin is. This is what is meant by oxygen saturation level. If this is measured diretly from an arterial blood sample, it is alled the SaO 2. If the measurement is alulated from a pulse oximeter it is alled the SpO 2. Alternatively, one an measure the oxygen tension of the blood (PaO 2 ), known as the partial pressure of oxygen in the blood. This measurement an be expressed in kilopasals (kpa) (normal range kpa) or in millimetres of merury (normal range mm Hg for young adults). 11 The normal SaO 2 in healthy adults at sea level is maintained within a narrow range of about 95 98% as disussed in setion 3.1 above. This means that almost all of the oxygen-arrying apaity of haemoglobin in the blood is used when the SaO 2 is in the normal range. Therefore, giving supplementary oxygen to a healthy young person will inrease the saturation level only slightly from about 97% to 99% or a maximum of 100%, thus produing only a very small inrease in amount of oxygen made available to the tissues. Sudden exposure to low SaO 2 levels (below about 80%) an ause impaired mental funtioning even in healthy subjets. The brain is the most sensitive organ to the adverse effets of hypoxia, but it is possible that other organs in patients with ritial illness may be vulnerable to the risk of hypoxi tissue injury at oxygen levels above this range. Most experts emphasise the importane of keeping the SaO 2 above 90% for the majority of autely ill patients The present guideline suggests a desirable target saturation range of 94 98%. This vi15

16 range mirrors the normal range for UK adults with a wide margin of safety above the 90% threshold whih is mentioned above. Oxygen passes from inspired air in the lungs into the bloodstream and is delivered to the tissues. If oxygen levels fall in the blood, this is sensed by reeptors in the arotid body (onneted to the arotid artery in the nek) and ventilation is stimulated to inrease the amount of oxygen oming into the lung and therefore the blood. The lung has the ability to divert blood flow away from areas whih are poorly ventilated, so that blood returning from the body an be replenished with oxygen and an also lear arbon dioxide. This ours through a proess alled hypoxi vasoonstrition whereby loalised low oxygen levels in the lung airspaes ause onstrition of feeding blood vessels, therefore diverting blood to areas of the lung with more normal oxygen levels. If the oxygen-arrying apaity of the blood is low as, for example, in anaemia, this is deteted by the kidneys whih produe a hormone, erythropoietin, to stimulate red blood ell prodution. As one of the goals of the irulation is to deliver oxygen to the tissues of the body, the heart also responds to low oxygen levels by inreasing its output, so inreasing oxygen delivery. Hypoxaemia, low PaO 2, an be aused by a number of mehanisms. The most ommon form of hypoxaemia ours when there is suffiient oxygen-arrying apaity (in patients with a normal level of haemoglobin) but insuffiient oxygen taken up in the lungs. This an be the result of poor aeration of areas of lung or due to abnormalities of gas exhange within the lung during serious illnesses suh as pneumonia. This form of hypoxaemia is the easiest to treat with oxygen therapy. Oxygen therapy is less effetive in other auses of hypoxaemia inluding anaemia where there is a low arrying apaity or where the arrying apaity of haemoglobin has been redued by a toxi substane beause oxygen availability is not the limiting feature in these onditions. For example, arbon monoxide bloks oxygen binding to haemoglobin despite having a normal level of oxygen in the lungs and in the blood. 4.2 Carbon dioxide physiology Carbon dioxide is a produt of the body s metabolism. It is leared from the body by being transferred from the bloodstream into the alveoli in the lungs and then exhaled from the lungs. In a similar way to oxygen, arbon dioxide levels in the blood are ontrolled by hemial sensors (both in the arotid body and brainstem). Carbon dioxide is highly soluble in the blood and is arried in three forms: biarbonate (70%), dissolved arbon dioxide (10%) and bound to haemoglobin (20%). As arbon dioxide arriage is not limited by a arrier moleule suh as haemoglobin, it is not expressed as a saturation. Beause its arriage is approximately proportional to the partial pressure (gas tension) of arbon dioxide in the blood within the physiologial range, arbon dioxide arriage is usually expressed in terms of its partial pressure. The normal range is kpa or mm Hg. Inreased levels of arbon dioxide will stimulate ventilation, thus inreasing learane from the lungs and therefore from the bloodstream. However, this mehanism is less effetive in some respiratory diseases suh as COPD where inreased airway resistane and respiratory musle weakness an restrit this response. Hyperapnia will our when there is dereased ventilation for any reason. Safe elimination of arbon dioxide is as important to the body as the intake of oxygen. Too little oxygen an give rise to organ failure but too muh oxygen an also be harmful in some situations, espeially to some vulnerable patients with COPD, hest wall deformities or musle weakness. About a quarter of patients with aute flare-ups of COPD are at risk of arbon dioxide retention if they are given an exessively high dose of oxygen. If high onentrations of oxygen are given to these patients, the oxygen level in the blood will rise but the level of arbon dioxide will also rise and this an ause aidosis with subsequent organ dysfuntion and, when severe, oma. In the past it was thought that the main problem was that these patients were dependent on the stimulus of a low blood oxygen level alled hypoxi drive to stimulate breathing. It was thought that giving oxygen would ause a rise in the arbon dioxide level by simply reduing the stimulus to breathe due to lak of hypoxi drive. It is now known that the mehanisms for arbon dioxide retention in some patients are muh more omplex than this simple model suggested. Muh of the rise in arbon dioxide whih ours during high-dose oxygen therapy is due to deterioration in the mathing of blood flow and gas flow in the lungs. This an be avoided by giving ontrolled lower onentration oxygen therapy to vulnerable patients (see table 3). 4.3 Conept of target oxygen saturation (SaO 2 ) ranges One might ask why one should not aim for an SaO 2 of 100% (hyperoxaemia) in all autely ill patients (and some liniians took this view in the past). This poliy would learly be risky for vulnerable patients with COPD and hest wall problems, but it ould also harm other patients in a variety of ways. The more ontroversial risks of hyperoxaemia inlude oronary and erebral vasoonstrition and dereased ardia output. Although these physiologial effets are well doumented, their signifiane in linial pratie is almost unknown owing to a lak of linial trials of oxygen therapy. High oxygen onentrations lead to an inrease in reative oxygen speies whih may ause tissue damage and may be responsible for some of the detrimental effets observed with high-flow oxygen in myoardial infartion and stroke. It is reognised that very high inhaled oxygen levels an give rise to partial ollapse of some lung units, a ondition known as absorption ateletasis. There is also the potential onern that a high oxygen saturation produed by high onentration oxygen therapy ould mask a major deterioration in the patient s linial ondition ausing dangerous delays in treatment. An example of this is a patient who has taken an opiate overdose whih has produed respiratory depression and the patient is underbreathing. If the patient is given exessive oxygen therapy, high or normal oxygen saturations may be reorded at a time when the arbon dioxide levels are dangerously high. The high oxygen saturation ould mask the real situation and give the health professionals a false sense of onfidene. As alluded to above, autely raised arbon dioxide levels an be dangerous. In aute irumstanes where arbon dioxide levels have risen rapidly, the kidneys are unable to ompensate for the onsequent inreased aid load. There are good data to show that the lower the ph of the blood, the higher the risk of intubation or death in patients with exaerbations of COPD. 34 The purpose of oxygen therapy is to inrease oxygen delivery to tissues, not just to inrease oxygen arried by the blood. It must therefore be remembered that there may be other physiologial disturbanes that need orreting to inrease oxygen delivery suh as low ardia output and severe anaemia. For example, improving these fators will improve oxygen vi16

17 Figure 3 Oxygen dissoiation urve with the Bohr effet. 2, 3 DPG, 2, 3 diphosphoglyerate; PO 2, oxygen tension. delivery muh more than administering oxygen to a patient with a saturation of 90% whih, at most, will produe a 10% rise in delivery. In addition to optimising oxygen delivery from the lungs to the tissues, it is important also to treat problems that might impair delivery of oxygen to the lungs themselves suh as upper airway obstrution, bronhoonstrition and pulmonary oedema (remember the ABC of resusitation airway, breathing, irulation). There is unertainty about defining the ideal target saturation and this is one of the ore debates in oxygen therapy. This unertainty is largely due to a lak of evidene from linial trials. In some speifi disease areas suh as COPD there are good data to inform the ideal target saturation and these will be overed in setions 8 and 9. In the general population without a speifi indiation for running high or low saturations, historially there has been a tendeny to apply oxygen therapy even when saturations are in the normal range. There are no data to support this pratie in ommon onditions suh as ishaemi heart disease or stroke and, indeed, some studies show harm. It should be reassuring that supraphysiologial levels of oxygen delivery are not required in ritial illness unless speifially indiated (eg, arbon monoxide poisoning). The Figure 4 Alveolar-apillary unit. PaCO 2,PaO 2, arterial arbon dioxide and oxygen tensions; PACO 2,PAO 2, alveolar arbon dioxide and oxygen tensions; PICO 2,PIO 2, inspired arbon dioxide and oxygen tensions; PvCO 2, PvO 2, venous arbon dioxide and oxygen tensions. onsensus among the members of the guideline group is that one should aim for a normal or near-normal SaO 2 range of 94 98% for autely ill patients exept those at risk of hyperapni respiratory failure (see reommendations 1 5 in setion 6 of this guideline). SECTION 5: ADVANCED BLOOD GAS PHYSIOLOGY AND PATHOPHYSIOLOGY AND PHYSIOLOGY OF OXYGEN THERAPY Many of the issues disussed in this setion are of a tehnial nature and may not be easily omprehensible to the general reader. However, reommendations 1 5 in setion 6 of the guideline will follow logially from this setion and from the brief overview of oxygen physiology in setion 4. The neuroardiopulmonary axis is designed to optimise global oxygen delivery and arbon dioxide learane and the loal tissue vasular beds are responsible for the distribution of blood flow. Oxygen delivery (DO 2 ) is expressed by the equation: DO 2 = CaO 2 6 Q where CaO 2 is the oxygen ontent of the arterial blood and Q is the ardia output. CaO 2 is the sum of oxygen dissolved in the blood and the amount of oxygen arried by haemoglobin. The solubility of oxygen in the blood is very low and therefore CaO 2 is largely determined by the total amount of haemoglobin and the proportion whih is bound by oxygen, namely saturation. The relationship between haemoglobin SaO 2 and PaO 2 is shown in fig 3 and table 8. In health and disease, haemoglobin saturation is also influened by other fators suh as ph, PCO 2, temperature and 2,3 diphosphoglyerate (Bohr effet; fig 3 and setion 5.1.3). Consequently, there is not an exat relationship between SaO 2 and PaO 2 but table 8 gives approximate equivalents. 5.1 Regulation of blood oxygen ontent (CaO 2 ) Figure 4 shows the level of oxygen and arbon dioxide in the pulmonary artery, in the alveolus and room air and in the pulmonary venous irulation whih leads diretly to the arterial irulation. The PaO 2 of mixed systemi venous blood rises markedly from a low level in the pulmonary artery (about 6 kpa or 45 mm Hg) to about 16 kpa (120 mm Hg) by the end of the pulmonary apillary. However, beause the lung is not homogeneously made up of alveolar apillary units that are mathed for perfusion and ventilation, the PaO 2 in the larger pulmonary veins is lower (13 kpa, 100 mm Hg). This is explained in more detail below. The gradient of arbon dioxide is muh more gradual, falling from about 7 kpa (52 mm Hg) in the venous system and pulmonary artery to about 5 kpa (37 mm Hg) in the pulmonary vein and in the arterial system Arterial oxygen tension (PaO 2 ) The pulmonary vasulature maximises PaO 2 by ensuring that the well ventilated areas of the lung reeive most of the pulmonary blood flow, a proess alled ventilation/perfusion (V/Q) mathing. This is largely ahieved through a proess alled hypoxi pulmonary vasoonstrition (HPV). 35 The pulmonary irulation is unique in this regard ompared with all the other vasular beds in the body whih dilate in response to hypoxia. In poorly ventilated areas of lung the preapillary pulmonary arterioles onstrit in response to sensing low alveolar PO 2 (PAO 2 ). This is a ompensating proess and, despite it, some deoxygenated blood may still leave poorly ventilated alveolar apillary units. Deoxygenated blood leaving poorly ventilated alveolar apillary units annot be ompensated for by vi17

18 Example 1 Before oxygen therapy assume 50% of pulmonary flow is passing through an area of low V/Q and that the pulmonary venous oxyhaemoglobin saturation (SpvO 2 ) from this ompartment is 80% (ie, just above mixed venous SO 2 ). The other 50% is passing through an area of mathed V/Q, resulting in a SpvO 2 of 97%. The final mixed SpvO 2 will be 88.5%. Following maximal oxygen therapy, assuming no hange in flow as a result of release of HPV, SpvO 2 from the low V/Q ompartment rises to 85% and SpvO 2 from the mathed ompartment rises to a maximum of 100%. The resulting mixed SpvO 2 will now only be 92.5%. This has ourred beause fully saturated blood annot inrease its oxygen ontent beyond full saturation despite an inrease in PO 2, apart from the minimal ontribution from dissolved oxygen; ie, the relationship between PO 2 and oxyhaemoglobin saturation/ blood oxygen ontent is not linear. mixing with blood from well ventilated units as the relationship between PaO 2 and CaO 2 is not linear. This physiologial phenomenon is often not fully appreiated and therefore is worth a theoretial worked example (see box). A muh less studied phenomenon that regulates V/Q mathing is hypoxi bronhodilation. This effet inreases ventilation to poorly ventilated areas of the lung. 36 If PaO 2 falls, the peripheral hemoreeptors in the arotid body drive an inrease in ventilation to inrease PaO This will not inrease PaO 2 leaving already well ventilated units, but will inrease PaO 2 leaving less well ventilated alveolar units by inreasing PAO 2 in these units. Although the ventilatory response to SaO 2, and therefore CaO 2, is linear (fig 5), the arotid body senses PaO 2 and not CaO 2. This prevents exessive ventilation in response to anaemia whih would be ineffetive in inreasing CaO 2. The peripheral hemoreeptors are able to do this beause the very high ratio of DO 2 to oxygen onsumption of the arotid body means that the tissue PO 2 in the arotid body ontinues to reflet PaO 2 and will not fall even in the presene of anaemi hypoxia. Figure 5 Ventilatory response to hypoxaemia. The relationship is inversely linear when plotted against oxyhaemoglobin (solid line) saturation but inversely exponential when plotted against arterial oxygen tension (PO 2 ) (dashed line) Haematorit Erythropoiesis is ontrolled by a negative feedbak system involving erythropoietin. By ontrast with the arotid bodies, the peritubular ells in the kidney are well suited to sensing oxygen delivery as oxygen extration is relatively high ompared with oxygen delivery. Although oxygen delivery to the kidneys as a whole organ is high due to high renal blood flow, DO 2 is redued to the renal medulla as oxygen an pass from arterioles to the post-apillary venous system by shunt diffusion due to the parallel organisation of arterial and venous systems. 42 Consequently, the peritubular ellular PO 2 is low. It falls to even lower levels following redutions in DO 2 either as a result of hypoxaemia or low haematorit The Bohr effet The oxygen-arrying apaity of haemoglobin is regulated in response to other metaboli fators to inrease the effiieny of oxygen pik-up and delivery. 43 Aidosis and hyperapnia shift the oxygen dissoiation urve to the right (fig 3), thus favouring the dissoiation of oxygen from haemoglobin in metabolially ative tissues. The onverse would hold true for the lungs where lower arbon dioxide levels favour oxygen loading of haemoglobin. Chroni hypoxaemia inreases 2,3-diphosphoglyerate (2,3-DPG) in erythroytes, shifting the dissoiation urve to the right and therefore inreasing oxygen delivery to the tissues Regulation of DO 2 (oxygen delivery from the lungs to the tissues) Autely, the ardiovasular effets of hypoxaemia will tend to ounter the impat of lower CaO 2 on DO 2 by inreasing ardia output through inreased heart rate and myoardial ontratility and by dereasing afterload by reduing systemi vasular resistane. Anaemi hypoxia is sensed in the aorti body, presumably owing to lower perfusion relative to oxygen onsumption. Consequently, the aorti body an at as a sensor of redued oxygen delivery as a result of either low oxygen tension or low haematorit (unlike the arotid body). 38 At loal tissue level, oxygen delivery an be adjusted to hanges in loal oxygen onsumption. For example, exerising skeletal musle reeives a greater proportion of total ardia output than resting skeletal musle. This relates in part to hypoxaemia reruiting a larger proportion of the apillary bed by the relaxation of periytes, and also through arteriolar vasodilatation Pathophysiology of hypoxia and hyperoxia Hypoxia may result from a number of different diseases disussed in setion 8 of this guideline. In eah ase one or more of the following pathophysiologial mehanisms may apply: Hypoxaemi hypoxia Other mehanisms of hypoxia Hyperoxia Hypoxaemi hypoxia (see definition in setion 3.1.2) Hypoxaemi hypoxia in blood leaving an alveolar apillary unit in the lung may be indued by alveolar hypoxia or inomplete gas exhange. The alveolar gas equation alulates the oxygen level in the alveolus using the following formula: PAO 2 < PIO 2 2 PACO 2 /RER where PAO 2 and PACO 2 represent alveolar levels of oxygen and arbon dioxide, RER is the respiratory exhange ratio or the ratio of arbon dioxide prodution to oxygen onsumption and vi18

19 inspired PO 2 (PIO 2 ) = FIO 2 6 (barometri pressure [100 kpa, 750 mm Hg] water vapour pressure [,6 kpa, 45 mm Hg]). Considering this equation, alveolar hypoxia an be indued by dereased PIO 2 or inreased PACO 2. If an alveolar apillary unit is relatively underventilated for its degree of perfusion (low V/Q ratio), PACO 2 will rise due to inadequate learane and thus PAO 2 will fall. This may happen for a number of reasons suh as inreased dead spae ventilation during the non-fatiguing pattern of shallow respiration in respiratory failure or abnormal lung mehanis in advaned COPD. In diseases that ause global hypoventilation suh as respiratory musle weakness, effetively all areas of lung have low V/Q ratios and this explains the hyperapnia and hypoxaemia assoiated with these onditions. An extreme form of low V/Q pathophysiology ours in intrapulmonary and extrapulmonary shunt where no gas exhange ours at all. An example of intrapulmonary shunt is when the airway to a lung segment is obstruted by muus reating an area of lung tissue that is perfused but not ventilated, thus ating as a right-to-left shunt. An example of extrapulmonary shunt is a ventriular septal defet with rightto-left shunting in Eisenmenger s syndrome. In health and at rest, oxygen has equilibrated aross the alveolar apillary membrane one-third of the way along the length of the apillary. With inreased thikness of this membrane, as in fibroti lung disease, equilibration may take longer and an oxygen gradient may persist between the alveolus and blood at the end of the apillary. The overall effet of this when multiple alveolar apillary units are affeted will lead to an inreased alveolar-to-arterial (A a) gradient. This is exaerbated during exerise, when apillary transit time dereases Other mehanisms of hypoxia (see definitions in setion 3.1.2) Anaemia and arbon monoxide poisoning may result in anaemi hypoxia by reduing oxygen-arrying apaity. A low ardia output state will redue oxygen delivery even in the absene of hypoxaemia. Tissue hypoxia may develop in these irumstanes and this is often termed stagnant hypoxia Hyperoxia Hyperoxia an be aused by hyperoxaemia and polyythaemia. Considering again the alveolar gas equation in the previous setion, hyperoxaemia an only exist in the presene of high inspired PO 2 or low PACO 2 (resulting from hyperventilation). The term hyperoxia ould tehnially be used to desribe a patient with polyythaemia without hyperoxaemia, but most liniians use the term only to desribe situations in whih the PaO 2 is raised. 5.3 Physiology of arbon dioxide Normal arbon dioxide homeostasis Carbon dioxide is prinipally arried in the blood in three forms: arbon dioxide, biarbonate and as a arbamino ompound. 47 In the normal physiologial range of kpa (34 45 mm Hg) the relationship between PaCO 2 and CCO 2 (arbon dioxide ontent) an be onsidered linear (fig 6) Regulation of arbon dioxide PaCO 2 is sensed at the peripheral 37 and entral hemoreeptors (in the medulla oblongata) by its effet on intraellular ph. 48 Consequently, the regulation of PaCO 2 is intimately related to ph homeostasis (fig 7). It is often not appreiated how V/Q mathing relates to PaCO 2. As disussed in setion 5.2.1, alveolar apillary units with a low V/Q ratio have inreased PACO 2. Beause of the high Figure 6 Total arbon dioxide dissoiation urve. PCO 2, arbon dioxide tension; HbCO 2, arbamino haemoglobin. solubility and diffusibility of arbon dioxide, there is little A a gradient for arbon dioxide at the end of the apillary, so blood leaving low V/Q alveolar apillary units has a high PCO 2. As desribed above, areas of low V/Q are usually minimised through hypoxi pulmonary vasoonstrition. It is also thought that a high PCO 2 an ause pulmonary vasoonstrition, adding to the homeostati mehanisms of the lung, mathing perfusion to ventilation As the relationship between PCO 2 and arbon dioxide dissolved in the blood is approximately linear over the physiologial range (unlike oxygen), blood does not beome saturated with arbon dioxide and therefore a high pulmonary venous PCO 2 from low V/Q areas an be partially balaned by a low pulmonary venous PCO 2 from high V/Q areas. Consequently, by inreasing overall alveolar ventilation, the ardiopulmonary system is able to prevent hyperapnia despite signifiant V/Q mismath or shunt, unless respiratory mehanis are limiting. As with the arriage of oxygen (Bohr effet), there is a reiproal relationship between PO 2 and arbon dioxide arriage. This is known as the Haldane effet. 43 Deoxygenated haemoglobin has a higher arbon dioxide buffering apaity than oxygenated haemoglobin. This favours arbon dioxide pik-up in the systemi venous irulation and arbon dioxide offloading in the lungs. Autely, arbon dioxide ats as a sympathomimeti on the heart: it inreases heart rate and stroke volume, inreasing ardia output. Peripherally it auses vasodilation, reduing systemi vasular resistane. Loally, arbon dioxide ats as a Figure 7 Effet of arterial arbon dioxide tension (PaCO 2 ) on ventilation with interation of aidosis and hypoxaemia. vi19

20 vasodilator, thus diverting blood flow to tissues with high metaboli demand. The resulting physial signs of hyperapnia are desribed in setion Pathophysiology of hyperapnia and hypoapnia Mehanisms of hyperapnia and hypoapnia The mehanisms of hyperapnia are simpler than hypoxaemia and there are four possible auses: 51 (1) Inreased onentration of arbon dioxide in the inspired gas. (2) Inreased arbon dioxide prodution. (3) Hypoventilation or ineffetive ventilation. (4) Inreased dead spae. The mehanisms of hyperapnia in COPD (and other onditions predisposing to hyperapni respiratory failure) are disussed in Inreased onentration of arbon dioxide in the inspired gas This iatrogeni ause of hyperapnia is unommon but should be exluded at the outset in any patient unexpetedly found to be hyperapni when breathing from, or being ventilated by, external equipment. The severity of hyperapnia due to rebreathing is limited by the rate at whih the PCO 2 an inrease (no more than kpa/min, 3 6 mm Hg/min). Inreased arbon dioxide prodution This is likely only to ause hyperapnia if the minute ventilation is fixed by artifiial means and if arbon dioxide prodution is inreased (eg, due to sepsis or inreased work of breathing). Hypoventilation or ineffetive ventilation Low alveolar minute ventilation is by far the most ommon ause of hyperapnia. In linial pratie, COPD is the most ommon disease to ause hyperapnia; the problem is seondary to alveolar hypoventilation rather than a redued minute ventilation per se. Patients adopt a rapid shallow pattern of breathing during an aute exaerbation of COPD with the result that the ratio of dead spae to tidal volume is inreased with more ventilation therefore being wasted. A rapid shallow pattern of breathing results in a bigger proportion of eah breath being wasted beause of the need to ventilate the anatomial dead spae. Furthermore, during aute COPD exaerbations, V/ Q mismath may lead to an inrease in physiologial dead spae, exaerbating the problem further. It is important to note that this ommonly ours in the ontext of an apparent overall inrease in minute ventilation. Alveolar hypoventilation due to a redution in minute ventilation is seen following medullary respiratory entre depression by drugs, obstrution of a major airway or restrition of the lungs or hest wall or by respiratory musle weakness, head injury, intraerebral haemorrhage or opioid narosis. Inreased dead spae This would be most ommon in patients breathing through artifiial apparatus whih has been inorretly onfigured. It an also be due to any ause of V/Q mismath in whih the normal response to hypoxaemia (ie, to inrease ventilation) is ompromised beause of lung disease. It is important to note therefore that hyperapnia sometimes may be seen in onditions more usually assoiated with hypoapnia (eg, pulmonary embolus, pneumonia) when it ours in patients with lung disease and an inreased physiologial dead spae. Although alveolar hypoventilation is the most ommon ause of hyperapnia, it is important to onsider the other potential auses, partiularly when patients are reeiving assisted ventilation and an artifiial breathing iruit is used Hypoventilaton and hyperventilation Hypoventilation may be physiologial for example, in the fae of a metaboli alkalosis. Pathologial hypoventilation will our either when the respiratory musles are unable to ventilate the lungs suffiiently beause they are pathologially weak or they are unable to overome abnormal lung mehanis suh as during an exaerbation of COPD. Redued respiratory drive aused by drugs with sedative properties or by neurologial injury will also produe hypoventilation. Using the same physiologial priniples but in reverse, hyperventilation for any reason will produe hypoapnia. This may our during pure hyperventilation during an anxiety attak or during physiologial hyperventilation. 5.5 Physiology of oxygen therapy Oxygen therapy inreases PAO 2 and is therefore only effetive when alveolar apillary units have some funtional ventilation. Oxygen therapy is ineffetive if there is a pure shunt (suh as pulmonary arteriovenous malformations) where mixed venous blood does not pass through an alveolar apillary unit. There will only be a small overall inrease in PaO 2 due to an inrease in dissolved oxygen in the pulmonary venous blood from ventilated alveolar apillary units, whih is small ompared with the ontent of oxygen arried by haemoglobin. Despite this, there is good evidene that breath-hold times an be inreased by breathing oxygen One study found that the breath-hold time of 15 healthy subjets inreased from 56 s after breathing air to 92 s after breathing 4 litres of nasal oxygen for 2 min, and another study found that 31 healthy volunteers had an inrease in breath-hold time from 32 s breathing air to 61 s after breathing oxygen whereas the breath-hold time of 29 patients with hroni pulmonary disease was 9 s ompared with 22 s for a group of 29 similar patients after breathing oxygen. The same priniples are used to preoxygenate patients before intubation during anaesthesia. It is thought that the additional breath-hold time is produed not by the marginal inrease in blood oxygen levels but by the inreased reservoir of oxygen in the lungs after breathing oxygen-enrihed air. In poorly ventilated units (ie, low V/Q ratio), PAO 2 will be low. Inreasing FIO 2 will inrease PAO 2 and therefore PaO 2. Hypoventilation disorders an be onsidered as lungs made up entirely of low V/Q units. When there is diffusion limitation due to inreased alveolar apillary membrane thikness suh as in fibroti lung disease, inreasing PAO 2 will augment the rate of diffusion aross the alveolar apillary membrane by inreasing the onentration gradient. Inreasing dissolved oxygen in plasma by oxygen therapy may also be used to offset the effets of hypoperfusion to some extent (stagnant hypoxia) and may well be important in ertain situations (ardiogeni shok), although the effet is only marginal. Inreased inspired oxygen will only marginally mitigate the effets of anaemi hypoxia but, beause the CaO 2 in patients with anaemia is less than that in patients with normal haemoglobin, the effet of additional oxygen arried in solution may beome more important in these situations. 5.6 Strategies for improving oxygenation and delivery Tissue oxygenation is dependent upon optimal or adequate oxygen delivery to the tissue (DO 2 ). This physiologial proess is omposed of various omponents that independently and interdependently influene and determine DO 2 and therefore vi20

21 Example 2 For a patient with sepsis and anaemia with a haemoglobin level of 50 g/l (5 g/dl) and an oxygen saturation of 90%, inreasing oxygen saturation from 90% to 100% will inrease the total haemoglobin oxygen ontent by about 10%. Transfusing 2 units of paked red ells will inrease the total haemoglobin oxygen ontent by about 40%. tissue oxygenation. These omponents an be onsidered sequentially Optimising PaO 2 The physiology of oxygen therapy has already been disussed in the previous setion. However, inreasing FIO 2 is only one omponent in inreasing oxygen uptake in the lungs. Other key manoeuvres to ensure oxygen delivery to the alveolar apillary bed inlude: Maintaining a satisfatory airway. Ensuring adequate alveolar ventilation. Reversing any respiratory depressants suh as narotis. Invasive or non-invasive ventilation where neessary. Treating airflow obstrution by bronhodilation or sputum learane tehniques. Optimising transfer fator (diffusion apaity). Treatment of pulmonary oedema Optimising oxygen arriage Oxygen is arried in blood mainly by haemoglobin with only a very small amount of oxygen dissolved in the blood itself. Adequate haemoglobin is therefore essential for optimal oxygen ontent (CaO 2 ) of blood. The ideal haemoglobin level for optimal CaO 2 and therefore for optimal DO 2 has long been a subjet of debate. Previous praties have favoured haemoglobin levels lose to 100 g/l (10 g/dl), providing adequate CaO 2 as well as reduing visosity of blood for better perfusion in ritially ill patients. However, studies by Canadian researhers in the late 1990s have shown that haemoglobin levels of 70 g/l (7 g/dl) were as safe as higher levels and may produe fewer ompliations in the ritially ill. 55 However, this study was onduted using non-leuoyte depleted blood and it is possible that some of the infetive ompliations in the group who were given more transfusions might have been avoided by the use of leuoyte-depleted blood. The optimal transfusion target for ritially ill patients therefore remains the subjet of ongoing disussion among experts in ritial are mediine. Although the issue of optimal haemoglobin in patients with unstable or symptomati oronary artery disease is not settled, haemoglobin levels of 100 g/l (10 g/dl) are reommended for adequate DO 2 (see box) Optimising delivery Besides adequate CaO 2 and PaO 2, delivery of oxygen depends upon adequate flow of oxygenated blood. Cardia output in turn depends upon adequate blood (irulating) volume, adequate venous return and adequate and optimal myoardial funtion. To avoid tissue hypoxia, attention must therefore be paid to the volume status of the patient and the adequay of ardia funtion, as well as initiating oxygen therapy. In severely shoked patients (eg, ardiogeni shok, septi shok), invasive monitoring and inotropi/vasopressor therapy will usually be indiated in appropriate higher dependeny environments. It has been shown that deliberately inreasing oxygen delivery in ritially ill patients as well as high-risk surgial patients redues organ failure, redues length of ICU stay and, most importantly, improves mortality Inreased oxygen delivery partly involves oxygen therapy, but these studies did not show any benefit from aiming at supraphysiologial oxygen delivery. The following worked example illustrates how minor abnormalities in eah of the parameters disussed above, when ourring together, an result in dramati falls in oxygen delivery. Oxygen delivery in health an be alulated as follows where CO is ardia output and Hb is haemoglobin (normal values: CO = 5 l/min; SaO 2 = 94 98%, Hb = 15 g/l): DO 2 = CO 6 CaO 2 DO 2 = CO 6 {[SaO 2 /100 6 Hb 6 1.3] + [PaO mm Hg 6 10]} Therefore: DO 2 = 56 {[ ] + [ ]} DO 2 = 970 ml/min or < 1000 ml/min This is well above the normal oxygen onsumption (V O2 )of about 250 ml/min. Now onsider an anaemi patient with a haemoglobin level of 10 g/dl, ardia output 3.5 l/min and SaO 2 of 90%; the oxygen delivery beomes approximately 410 ml/min: DO 2 = {[ ] + [ ]} Although this value is still above the V O2 at resting physiology, in pratie the V O2 would most likely have risen owing to a number of fators suh as inreased work of breathing and inreased ataboli state of sepsis. This example is not rare and ours daily in linial pratie. It is therefore important not to onsider oxygen therapy in isolation. As many patients may not have adequate haemoglobin, ardia output or blood volume, they may suffer from tissue hypoxia when they beome autely ill. All suh patients should have supplemental oxygen therapy until they are evaluated by a responsible healthare professional. SECTION 6: HYPOXIA, HYPEROXIA, HYPERCAPNIA AND THE RATIONALE OF TARGETED OXYGEN THERAPY 6.1 Effets and risks of hypoxia and rationale for target oxygen saturation range As this guideline is addressing emergeny oxygen therapy, this setion will only fous on the effets and risks of aute hypoxia. Setion 8 will disuss the emergeny treatment of aute hypoxia in patients with long-term diseases assoiated with hroni hypoxia. The approximate relationship between PaO 2 and SaO 2 is shown in table 8 and fig 3 (oxygen dissoiation urve). The effets and risks of hypoxia are summarised in table 9. Severe hypoxia may lead to brain damage and death. In general, Table 8 Approximate relationship between arterial blood saturation (SaO 2 ) and arterial oxygen tension (PaO 2 ) 60 PaO 2 (kpa) >17 PaO 2 (mm Hg) >127.5 SaO 2 (%) >99.0 vi21

22 Table 9 Respiratory system Cardiovasular system Metaboli system Neurologial system Renal system Physiologial effets of aute hypoxia and hyperoxia Hypoxia Hyperoxia Effets Risks Effets Risks Inreased ventilation Pulmonary vasoonstrition Coronary vasodilation Dereased systemi vasular resistane (transient) Inreased ardia output Tahyardia Inreased 2,3-DPG Inreased CO 2 arriage (Haldane effet) 2,3-DPG, 2,3-diphosphoglyerate. Inreased erebral blood flow due to vasodilation Renin-angiotensin axis ativation Inreased erythropoietin prodution Pulmonary hypertension Myoardial ishaemia/ infartion Ishaemia/infartion of other ritially perfused organs Hypotension Arrhythmias Lati aidosis Confusion Delirium Coma Aute tubular nerosis many of the physiologial effets of hypoxia are mediated by low PaO 2, irrespetive of oxygen ontent. For example, even when the total blood oxygen ontent is normal in the presene of polyythaemia, hypoxaemia will still exert a physiologial effet suh as stimulation of ventilation. The risks of hypoxia, however, are usually mediated by low tissue PO 2 whih may our as a onsequene of a low PaO 2 and other mehanisms suh as severe anaemia and low ardia output states. These problems an be illustrated in the pathophysiology of myoardial ishaemia whih will develop when there is an imbalane between myoardial DO 2 and oxygen onsumption (V O2 ). DO 2 is not only dependent on PaO 2, but also oronary flow and haematorit. V O2 will also depend on the stroke work of the heart. Defining a lower limit of PaO 2 whih is onsidered safe is therefore impossible given the other variables. Hypoxaemia refers to an abnormally low oxygen tension in the blood (see setion 3.1). However, it is not possible to define a single level of hypoxaemia that is dangerous to all patients. Some patients with hroni lung disease may be austomed to living with SaO 2 as low as 80% (PaO 2 about 6 kpa or 45 mm Hg) while other patients with aute organ failure may be harmed by shortterm exposure to SaO 2,90% (PaO 2,8 kpa or 60 mm Hg). It has been shown that medial patients with sustained desaturation,90% have impaired medium-term survival ompared with medial patients with saturations whih stay.90%. 21 However, muh of this survival disadvantage may be due to the underlying disease whih has aused the low oxygen level (eg, severe COPD or pneumonia) and the degree of hypoxaemia may be a marker of disease severity, therefore the ontribution of modest hypoxaemia to mortality rates is not known. Mental funtioning beomes impaired if the PaO 2 falls rapidly to,6 kpa(45mmhg,sao 2,80%) and onsiousness is lost at,4 kpa (30 mm Hg, SaO 2,56%) in normal subjets Young subjets tolerate aute hypoxaemia for longer than older subjets in terms of time of useful onsiousness. 64 Safe levels of hypoxaemia in COPD have been disussed in detail in a review by Murphy and olleagues. 10 Many patients with COPD have a PaO 2 of,5 kpa (37.5 mm Hg) orresponding to a SaO 2 of,70% during Dereased ventilation (minimal) Dereased 2,3-DPG Dereased CO 2 arriage (Haldane effet) Dereased erebral blood flow Worsened ventilation/ perfusion mathing Absorption ateletasis Myoardial ishaemia (in ontext of dereased haematorit) Redued ardia output Redued oronary blood flow Inreased blood pressure Inreased peripheral resistane Inreased reative oxygen speies Redued renal blood flow an aute exaerbation. 65 Furthermore, sudden hypoxaemia is more dangerous than hypoxaemia of gradual onset both in health and in disease. For example, millions of people live at altitudes above 3000 m despite an average PaO 2 of about 7.3 kpa (55 mm Hg, saturation about 88%) and alimatised limbers on Mount Everest an tolerate short-term exposure to an oxygen saturation of 70% or less with an estimated PaO 2 of about 3.7 kpa (28 mm Hg) Campbell summarised this issue eloquently in 1967 when he said Better a year at a PaO 2 of 50 mm Hg (6.7 kpa) than an hour at a PaO 2 of 20 mm Hg (2.7 kpa). 66 Hypoxi hepatitis has been reported in patients with respiratory failure assoiated with oxygen levels below 4.5 kpa or 34 mm Hg, 67 whereas hypoxi hepatitis in patients with ardia disease is mainly due to dereased hepati blood flow (stagnant hypoxia) and ours at higher blood oxygen levels. 68 An in-flight study of patients with COPD with mean SaO 2 of 96% at sea level showed a fall to 90% in-flight in a ommerial airliner and a further fall to a mean SaO 2 of 87% while walking in the airraft aisles. These patients had no symptoms during these hypoxaemi episodes. 69 A study of 84 healthy airline passengers found that the mean SaO 2 fell from 97% at ground level to 93% (1SD 85 98%) at ruising altitude. 70 A study of healthy airline abin rew has shown that the SaO 2 of flight attendants falls to a mean nadir of 88.6% without ausing breathlessness or any other symptoms. Individual nadirs of SaO 2 ranged from 93% down to 80%. 71 Without any randomised evidene, the guideline prodution team have suggested that the level of saturation whih is tolerated by healthy people without any symptoms (about 85% saturation) should be regarded as the safe lower limit of hypoxaemia. However, other o-morbidities may need to be taken into aount and expert opinion reommends that the SaO 2 should be maintained above 90% for seriously ill patients For this reason, the present guideline reommends a target SaO 2 (and SpO 2 ) above 94% for most hypoxaemi patients to ensure that the atual oxygen level remains above 90% for most of the time with a 4% margin of safety to allow for oximeter error and artefat suh as a weak signal or dark oloured skin. The auray of and pitfalls of vi22

23 oximetry are addressed in setion Speifi targets for oxygen therapy in other diseases will be onsidered theoretially in this setion and pratially in setions 8 and 9. Some patients espeially older people or those with hroni lung disease may have an SaO 2 below 94% when linially stable and oxygen should not be given just to maintain the SaO 2 above 94% if the patient is linially stable. In assessing an ill patient, the SaO 2 level is only one of several physiologial variables that should be monitored. Many patients with sudden aute illness suh as postoperative pulmonary emboli will have a sudden alteration in physiologial variables as assessed by trak and trigger systems suh as the modified Early Warning Soring systems (mews) Suh patients may have only a small fall in SaO 2 owing to physiologial ompensation mehanisms suh as inreased ventilation. Healthare professionals therefore need to be alert for falls in SaO 2 even within the reommended target ranges. Beause of the wide normal range for SaO 2 and the unertainty onerning the possible physiologial onsequenes of minor hypoxaemia, there was more debate among the guideline group about desirable target ranges than about any other aspet of the guideline. If the SaO 2 should fall slightly below 94%, the key issue is to identify and treat the ause of the fall (eg, pulmonary embolism) rather than just orreting the hypoxaemia whih is not of itself dangerous at this level. However, there is a danger that healthare workers might fail to respond appropriately to abnormal hypoxaemia. After muh debate it was onluded that the guideline would reommend a target range of 94 98% for all adult patients. This reflets the approximate normal range of SaO 2 in healthy adults as disussed in setion 3.1. However, a sustained fall in SaO 2 of.3%, even within the normal range, may be an indiator of aute illness and should require linial assessment of a patient while a minority or patients (espeially those aged.70 years) may have an SaO 2 of,94% even when linially stable Desirable oxygen saturation ranges in aute illness Aute hypoxaemia is onsidered dangerous to healthy subjets below a PaO 2 of about 6 kpa (45 mm Hg) or an SaO 2 of about 80% due to impaired mentation and risk of tissue hypoxia, but patients with aute illness or hroni organ disease or ishaemia are likely to be at risk at PaO 2.6 kpa. [Evidene level III] Changes in physiologial trak and trigger systems suh as mews may our in aute illness with either no hange or only a small hange in SaO 2 levels. [Evidene level III] Critial illness may present initially with only a small fall in SaO 2 level beause of ompensating mehanisms. [Evidene level IV] The upper end of the reommended range in this guideline (98%) is the upper limit of SaO 2 in healthy adults. [Evidene level III] The lower end of the suggested target saturation range (94%) is about the lower end of the normal range and ensures that the SaO 2 remains above 90% most of the time. [Evidene level III] 6.2 Potential benefits of hyperoxaemia and supplemental oxygen therapy in non-hypoxaemi patients Supplemental oxygen therapy is most ommonly given to orret hypoxaemia as disussed throughout this guideline. However, there are some irumstanes where supplemental oxygen may be given to non-hypoxaemi patients Benefits of hyperoxaemia in arbon monoxide poisoning and pneumothorax Hyperoxaemia is useful in some linial situations. The best example of this is arbon monoxide (CO) poisoning. CO ombines with haemoglobin and has a higher affinity for the haemoglobin moleule giving rise to arboxyhaemoglobin (COHb). The half-life of COHb is about 4 5 h when breathing air but is redued to about 40 min when breathing 100% oxygen. Hyperoxaemia may also be used to aelerate the resolution of pneumothorax in patients who do not require a hest drain Other potential benefits of oxygen therapy in non-hypoxaemi patients Most guidelines for ardiopulmonary resusitation and the are of patients with ritial illness reommend the use of 100% oxygen in the initial stages of resusitation. Although these reommendations are not evidene-based, it is unlikely that ontrolled trials would ever be undertaken using different levels of oxygen therapy in these emergenies and it seems intuitive to maximise oxygen delivery for ritially ill patients with irulatory ollapse. However, randomised trials have been undertaken of resusitation of neonates breathing room air or oxygen and the unexpeted outome of a Cohrane review was that the outome was possibly better when room air was used. 76 This surprising finding annot be extrapolated to adult patients, but it does emphasise the need for linial trials even in areas where one might intuitively believe that oxygen would be benefiial. Furthermore, there is theoretial evidene that patients who have survived the initial stages of resusitation may be managed more safely with 30% oxygen ompared with 100% oxygen. It has been shown that early intervention to inrease oxygen delivery to the tissues in ritially ill patients as well as highrisk surgial patients redues organ failure, redues length of ICU stay and, most importantly, improves survival Inreased oxygen delivery in part involves oxygen therapy, but these studies did not show any benefit from aiming at supraphysiologial oxygen delivery. Short-term postoperative oxygen therapy (for 2 h) has been shown to redue the risk of surgial wound infetions in double blind trials of patients having bowel surgery but not in general surgery Oxygen therapy has been reported to improve anastomoti integrity in animal models 82 and to have potential benefit in human anastomoti surgery. 83 Reported benefits of oxygen therapy in healing of established wounds and in treatment of wound sepsis are ontroversial. Hyperbari oxygen redued the risk of amputation in patients with hroni diabeti foot ulers and may improve the hane of healing over 1 year, but the Cohrane reviewers had onerns about the size and quality of existing studies and reommended further trials. 84 It is not known if onventional oxygen therapy has any effet on wound healing. 85 Relief from luster headahes has been reported in about 60% of ases but this observation is based on very small studies from the 1980s. Although this ould be onsidered as a form of emergeny oxygen therapy, these patients are not breathless. Through relief of breathlessness and work of breathing, oxygen therapy may derease arbon dioxide prodution and onsequently offset some of the potential inrease in PaO 2 that might otherwise our due to the mehanisms vi23

24 desribed in setion However, there are no ontrolled trials supporting the use of oxygen for this indiation. Oxygen therapy may redue nausea and vomiting in postoperative patients and in ambulanes. Although some reports have suggested that oxygen may have a speifi antiemeti effet during ambulane transfers and in the postoperative state, subsequent studies reported no effet on motion sikness and no anti-emeti effet in postoperative patients Potential adverse effets and risks of supplemental oxygen therapy and hyperoxaemia These are summarised in table 9 and in the review by Downs 93 and in other soures The following paragraphs will summarise the physiology and pathophysiology of supplemental oxygen therapy and hyperoxaemia Respiratory system The most signifiant effet of exess oxygen on the respiratory system is hyperapni respiratory failure in a population of vulnerable patients. This does not our in the absene of signifiant pulmonary disease or musuloskeletal disease affeting the thorax, and it ours while the PaO 2 is still within the normal range or slightly below normal. There are at least five mehanisms responsible for this: V/Q mismath. Ventilatory drive. Haldane effet. Absorption ateletasis. Higher density of oxygen ompared with air. V/Q mismath During air breathing, poorly ventilated alveolar apillary units will be hypoxi and therefore poorly perfused due to hypoxi pulmonary vasoonstrition (HPV). If high onentration oxygen is administered, the PAO 2 will rise, thus reversing the HPV and inreasing blood flow to that unit. However, although the oxygen in the unit has inreased, it remains poorly ventilated with a high PACO 2 and therefore a high pulmonary venous PCO 2. As more blood is now passing through these units, the PaCO 2 will rise. Normally when there is no signifiant lung disease or thorai musuloskeletal disease, the respiratory system is able to ompensate for these hanges by inreasing overall ventilation thereby lowering PaCO 2. However, where respiratory mehanis are suh that inreased ventilation is not possible, PaCO 2 will rise. Several authors have reported that this mehanism is more important than redution in ventilatory drive in produing hyperapnia when supplementary oxygen is administered, but this ontinues to be a ontroversial area of respiratory physiology Ventilatory drive Hypoxaemia drives an inrease in ventilation so it follows that relief of hypoxaemia will ause a derease in ventilation. The onsequent rise in PaCO 2 is inversely proportional to the derease in ventilation suh that a halving of alveolar ventilation will lead to a doubling in PaCO 2, assuming onstant arbon dioxide prodution. As shown in fig 5, any inrease in PaO 2 above 8 kpa (60 mm Hg) will not signifiantly redue ventilation and inreases above 13 kpa (100 mm Hg) will have no impat on ventilaton as the arotid sinus disharge is largely attenuated above 13 kpa. This mehanism is therefore only of importane in explaining inreases in PaCO 2 when PaO 2 rises to 13 kpa, but not inreases thereafter. This phenomenon is also seen in normal individuals. Several linial studies have suggested that hypoxi drive makes only a small ontribution to the rise in PaCO 2 that is seen linially when high-dose oxygen is given to patients with COPD, but one reent study has supported this mehanism Haldane effet The third effet of inreasing FIO 2 is to derease the arbon dioxide buffering apaity of haemoglobin through the Haldane effet. 43 Absorption ateletasis The fourth effet, absorption ateletasis, is thought to our as a result of absorption of oxygen from alveoli with high PAO 2 beyond obstruted airways. This an happen at FIO 2 as low as 30 50% and will result in a shunt (inreased V/Q mismath). 93 Higher density of oxygen ompared with air Johnson and olleagues 106 have shown a redution in fored expiratory volume in 1 s (FEV 1 ) in patients who were breathing pure oxygen ompared with breathing air. They onluded that this effet was probably related to the slightly inreased density and visosity of oxygen relative to air. This would inrease the work of breathing whih ould ontribute to hyperapnia in an exhausted patient. It has been stated for several deades that hyperoxaemia auses hyperapni respiratory failure by produing dereased respiratory drive in patients with intrinsi lung disease, suh as COPD, and many if not most medial textbooks from the 1960s to the present time refer to loss of hypoxi drive as the main ause of hyperapnia and aidosis when high-dose oxygen is given to patients with an aute exaerbation of COPD. This assertion is usually attributed to Campbell who hampioned the onept of ontrolled oxygen therapy in the 1960s. However, Campbell has been widely misquoted. What he atually said in 1967 was as follows: It is usual to attribute the rise in PaCO2 in these patients to removal of the hypoxi drive to ventilation but I share the doubts of Pain and o-workers 103 that this is the whole story; hanges in the pulmonary irulation may also be important. 66 Most but not all subsequent studies have shown that Campbell was orret in this assumption Another urious feature of hyperapnia in aute exaerbations of COPD is that it is not universal. 66 Some patients with COPD are prone to repeated episodes of hyperapni respiratory failure and others may not ever suffer from this ompliation. Even among COPD patients with hroni hyperapnia, not all will develop an inreased arbon dioxide level (and aidosis) during aute exaerbations. The theory of loss of hypoxi drive as the ause of hyperapnia is further onfounded by the observation that PaCO 2 ontinues to rise as PaO 2 is inreased above 13 kpa (100 mm Hg), whih has little impat on dereasing ventilation and most patients with respiratory aidosis during an exaerbation of COPD have PaO 2 above 10 kpa, equivalent to saturation above about 93%. 34 Therefore, while a small redution in ventilation may be a ontributing fator to the rise in arbon dioxide levels during oxygen therapy in COPD, the major fator is the worsening of V/Q mathing. Additional effets of inreasing FIO 2 will relate to ateletasis and perhaps worsening airflow obstrution due to inreased visosity. In diseases where there is little intrinsi lung disease but signifiant respiratory musle weakness, loss of hypoxi vi24

25 respiratory drive will be a greater fator in the development of hyperapnia. However, HPV remains a signifiant regulator of V/Q mathing even in non-diseased lung Rebound hypoxaemia following sudden essation of supplementary oxygen therapy Patients who have developed worsened hyperapni respiratory failure following high onentration oxygen therapy fae a further signifiant danger of rebound hypoxaemia if oxygen is suddenly withdrawn in an attempt to orret the effets of exess oxygen therapy. Rebound hypoxaemia an be explained using the alveolar gas equation and, given its importane, is best illustrated with a working example (box). For the purposes of simpliity, this example makes several assumptions suh as a onstant respiratory exhange ratio and alveolar arterial gradient between stages 1 and 3. It also assumes that ventilation remains unhanged. Although when PaO 2 falls to 3.4 kpa upon removal of oxygen, ventilation will rise, by definition it will not be able to rise suffiiently to meet the need to lear the arbon dioxide stores for the same reason that hyperani respiratory failure developed in the first instane. Rebound hypoxaemia is a major risk and may be more dangerous than the hyperapni respiratory failure itself. Consequently, this guideline will reommend that oxygen therapy be stepped down gradually through sequential Venturi devies while monitoring saturation ontinuously. Sudden essation of supplementary oxygen therapy an ause rebound hypoxaemia with a rapid fall in oxygen tension to below the tension that was present prior to the ommenement of supplementary oxygen therapy. [Evidene level III] Example 3: Rebound hypoxaemia Stage 1: Consider a patient with an exaerbation of COPD whose arterial blood gases are as follows: PaO kpa; PaCO kpa breathing room air. The PAO 2 alulated from the alveolar gas equation (5.2.1) will be 11.6 assuming a respiratory exhange ratio of 0.8, giving an alveolar arterial gradient of 5.1 kpa. Stage 2: Given maximal oxygen through a reservoir bag mask, his blood gases beome: PaO 2 32 kpa; PaCO 2 10 kpa. Beause of the high solubility of arbon dioxide, the total body stores of arbon dioxide will have risen. Stage 3: If oxygen therapy is suddenly withdrawn, PACO 2 and PaCO 2 will remain high initially beause of the high arbon dioxide stores and therefore PAO 2 will fall further than it was initially before oxygen therapy to 8.5 kpa. Assuming the alveolar arterial gradient for oxygen stays at 5.1 kpa for room air breathing, then alulated PaO 2 will beome 3.4 kpa Cardiovasular and erebrovasular system The theoretial risks of hyperoxia have been summarised by Thomson and olleagues in an editorial whih made a strong ase for more trials. 96 Hyperoxaemia auses oronary vasoonstrition and, if the haematorit is suffiiently low, this may theoretially ause paradoxial myoardial hypoxia beause of overall redution in DO 2. One randomised double blind trial of oxygen in unompliated myoardial infartion found higher rates of sinus tahyardia and a signifiantly greater rise in myoardial enzyme in the oxygen group, suggesting a greater infart size. 107 There was a threefold inrease in mortality in patients on oxygen therapy that did not reah statistial signifiane (3 deaths in 77 patients treated with air versus 9 deaths in 80 patients given oxygen at 6 l/min via simple fae mask for 24 h). This trial was published in 1976 and oxygen has been given routinely to millions of normoxaemi patients with myoardial infartion and hest pain for a further 30 years without any evidene to support the pratie. Furthermore, a more reent trial showed inreased mortality in patients with non-hypoxaemi strokes of mild to moderate severity in those randomised to treatment with oxygen. 108 This reates an urgent need for large randomised trials of oxygen therapy for non-hypoxaemi patients with aute ardia and erebral ishaemia. Thompson et al 96 have suggested that oxygen should be presribed, administered and monitored with are in order to ahieve optimal tissue oxygenation, not maximal oxygenation. This view was proposed by other authors suh as Bryan and Jenkinson 109 in the 1980s, but standard medial pratie has not taken note of this advie. Beause there are no published data suggesting benefit from hyperoxaemia for most medial onditions and beause of the theoretial risks, optimal management should aim for physiologial oxygenation. Targets for oxygen therapy in speifi irumstanes, with evidene, are disussed in setion Reative oxygen speies Aside from the potentially detrimental physiologial effets of hyperoxaemia, the toxi effets mediated by reative oxygen speies (ROS) have potential risk. 109 Exess ROS are generated in the presene of high tissue PO 2 in the form of hydrogen peroxide and superoxide, ausing oxidative stress and free radial damage. At physiologial levels ROS at as signalling moleules, but at higher levels they are ytotoxi, notably being released by primed neutrophils as a host defene mehanism. It is thought that ROS are responsible for the development of bronhopulmonary dysplasia in ventilated hyperoxygenated premature infants 110 and reperfusion injury post-myoardial infartion Delay in reognition of physiologial deterioration It was previously believed that a high FIO 2 is protetive and gives patients a margin of safety. However, Downs and Beasley have argued that unstable patients may atually be plaed at risk by the preautionary use of high-dose oxygen therapy. During physiologial deterioration, a patient given high-dose oxygen therapy would have a normal or high pulse oximeter reading masking a progressive deline in the PaO 2 /FIO 2 ratio and therefore not alerting staff to impending deterioration requiring mehanial support. Furthermore, a patient who deteriorated physiologially while at a low FIO 2 would be deteted early by pulse oximetry and ould have the FIO 2 inreased while being transferred to an intensive are unit, whereas a patient who was already reeiving a high FIO 2 would desaturate more slowly but, when the oximeter eventually deteted desaturation, there would be fewer treatment options beause inreasing the FIO 2 further would have little effet Lung injury in patients with aute paraquat poisoning, bleomyin lung injury and aid aspiration Oxygen is known to be hazardous to patients with paraquat poisoning, and oxygen potentiates bleomyin lung injury and vi25

26 may potentiate lung injury from aspiration of aids Further details onerning these onditions are given in setion Summary of risks of hyperoxia and supplemental oxygen therapy Physiologial risks (1) Worsened V/Q mismath. (2) Absorption ateletasis. (3) Coronary and erebral vasoonstrition. (4) Redued ardia output. (5) Damage from oxygen free radials. (6) Inreased systemi vasular resistane. Clinial risks (1) Worsening of hyperapni respiratory failure. (2) Delay in reognition of linial deterioration. (3) Worse outomes in mild to moderate stroke. (4) Speifi risk in patients with previous bleomyin lung damage or with paraquat poisoning or aid aspiration. (5) Unknown risk-benefit balane in aute oronary artery disease with normal oxygen saturation. Unontrolled supplemental oxygen therapy an be harmful to patients who are at risk of hyperapni respiratory failure, espeially if the PaO 2 is raised above 10 kpa. [Evidene level IIa] High-dose oxygen therapy to produe hyperoxaemia (above normal oxygen saturation) an ause absorption ateletasis, myoardial ishaemia and unfavourable outomes in some patient groups (eg, patients with mild and moderate strokes). [Evidene levels Ib III] 6.4 Risks of hyperapnia (and respiratory aidosis) Hyperapnia and respiratory aidosis are inextriably linked and are best onsidered together. If hyperapnia develops slowly (over several days), a patient will have renal ompensation (retention of biarbonate) and aidosis will not our in most suh ases. However, aute elevation of the blood arbon dioxide level produes respiratory aidosis and symptoms of hyperapnia. Some of the onsequenes of an elevated arbon dioxide tension are a onsequene of the resulting aidosis. Sometimes the effet of a raised arbon dioxide tension on a partiular organ system is opposed by an opposite effet of aidosis. Carbon dioxide is a vasodilator and patients with hyperapnia may appear flushed with dilated peripheral veins and a bounding pulse. Cranial vasodilation may ause headahe. Carbon dioxide in high onentrations has hypnoti effets and patients with hyperapnia may progress from drowsiness to onfusion to oma. A link has been shown between severe respiratory aidosis in aute COPD and an inreased risk of death or requirement for mehanial ventilation. 34 However, the problem of respiratory aidosis is not onfined to patients with COPD. Depressed respiration for any reason will give rise to hyperapnia. Examples are opiate overdoses, obesity with hypoventilation and neuromusular disorders affeting the musles of respiration Effets of a raised blood arbon dioxide level Nervous system Carbon dioxide exerts its effet either diretly or as a onsequene of aidosis. Hyperapnia inreases erebral blood flow and thereby may influene the erebrospinal fluid pressure. It is the main fator influening the intraellular ph whih has an important effet on ellular metabolism. It exerts an inert gas naroti effet similar to that of nitrous oxide. It influenes the exitability of neurones partiularly relevant in the retiular ativating system. Carbon dioxide an indue narosis when the PaCO 2 rises above kpa ( mm Hg). 118 Pulmonary irulation An elevated PACO 2 auses vasoonstrition in the pulmonary irulation although the effet is less marked than that of hypoxia. 123 In healthy subjets an end expiratory PCO 2 of 7 kpa (52 mm Hg) inreases pulmonary vasular resistane by 32% whih, along with raised ardia output, inreases mean pulmonary artery pressure by 60%. 124 Changes in ph are thought to be the primary fator responsible for arbon dioxide-mediated hanges in the pulmonary vasulature. Consequently, as with HPV, hanges in PACO 2 help to math perfusion to ventilation. Respiratory system As explained in setion 6.2.1, a raised arbon dioxide level may worsen hypoxia and its effets beause the onentration of arbon dioxide in the alveolar gas redues that of oxygen if the onentration of nitrogen remains onstant. Also an inrease in PaCO 2 shifts the oxygen dissoiation urve to the right. Cardiovasular system In general, both hyperapnia and aidosis have diret depressant effets on ardia myoytes and vasular smooth musle ells. 127 These effets are normally opposed by the inrease in ateholamines aused by the raised PaCO 2. The overall effet of arbon dioxide on the ardiovasular system is therefore unpreditable. In artifiially ventilated hildren a rise in arbon dioxide inreases ardia output and redues total peripheral resistane and blood pressure tends to rise. 128 Although an inrease in arbon dioxide depresses heart rate, tahyardia is more ommon beause of the effets of ateholamine stimulation overriding the depressant effets on the heart. Arrhythmias have been reported but are seldom linially signifiant in normal subjets. Carbon dioxide is a systemi vasodilator. Kidneys Renal blood flow and glomerular filtration rate are redued in the presene of high levels of PaCO 2. If severe, this an lead to anuria. Blood eletrolyte levels The aidosis that aompanies hyperapnia may ause a rise in potassium if the aidosis is severe and sustained. Endorine system Hyperapnia inreases plasma levels of both adrenaline and noradrenaline Clinial signs The linial signs of hyperapnia are produed by the physiologial hanges desribed above and are desribed in detail in setion Risks of aidosis The major effet of aidosis is depression of the entral nervous system with severe aidosis (ph,7.0 or [H + ].100 nmol/l) vi26

27 ausing disorientation and later oma. However, as desribed above, the effets of ph are inextriably linked with both hypoxia and hyperapnia. As a onsequene of opposing effets of aidosis, hypoxia and hyperapnia on different target organs in individual patients together with the fat that derangements of all three omponents may our at the same time, it is very diffiult to predit the effets of aidosis per se in an individual patient. Furthermore, tissue hypoxia will exaerbate aidosis. The onsequenes will depend upon the interplay of the three variables, ompliated by the effets of o-morbid disease states. It is well known that, in patients with COPD, a ph of,7.30 or [H + ].50 nmol/l during an aute exaerbation is assoiated with a muh worse prognosis Rationale of oxygen therapy Oxygen therapy is usually defined as the administration of oxygen at onentrations greater than those found in ambient air. It is usually undertaken to treat or prevent hypoxaemia, thereby preventing tissue hypoxia whih may result in tissue injury or even ell death. In some irumstanes suh as arbon monoxide poisoning or luster headahe, oxygen therapy is used to ahieve hyperoxia. There are no published trials supporting the use of oxygen to relieve breathlessness in non-hypoxaemi patients and there is evidene from randomised studies that oxygen does not relieve breathlessness ompared with air in non-hyoxaemi patients with COPD who are breathless following exertion or in non-hypoxaemi patients with advaned aner At the tissue level, mitohondrial ativity requires oxygen for aerobi ATP synthesis for ellular ativity. PaO 2 of dry air at sea level is 21.2 kpa (159 mm Hg), but at the mitohondrion, PO 2 is in the range of kpa (4 22 mm Hg) depending on tissue type and loal metaboli ativity. This gradient from atmosphere to mitohondrion is known as the oxygen asade. There are many fators in this asade that affet the final mitohondrial PO 2 inluding alveolar gas exhange, oxygen transport in the blood and tissue perfusion. Under pathologial onditions, any hange in one step in this asade may result in hypoxia at the mitohondrial level. Therefore, although not neessarily addressing the underlying ause of tissue hypoxia, inreasing FIO 2 with oxygen therapy is the simplest and quikest way of avoiding hypoxi tissue damage. Besides oxygen therapy, other steps are usually neessary to improve the delivery of oxygen to the tissue (see setion 5.6). 6.7 Target oxygen saturation in aute illness (see also setions 4.3 and 6.1) Many disease states lead to a redued oxygen level and it is standard pratie for breathless patients to be treated with oxygen. However, there have been few ontrolled trials omparing different levels of inspired oxygen for patients with any of the ommon diseases that lead to hypoxaemia. It must also be remembered that oxygen therapy is only one of several strategies that may be used to inrease tissue oxygen delivery for ritially ill patients (setion 5.6). In many linial situations oxygen therapy is applied without a speifi end point in mind. It has been suggested by many studies that hyperoxia an have deleterious physiologial and linial effets (see setion 6.3), albeit suh effets are not widely reported in onditions other than COPD. However, potential for harm may well exist with hyperoxia and good medial pratie should be followed as in all drug presriptions. As the atual PO 2 at the mitohondrial level is so variable and dependent on many variables other than PaO 2, it is often diffiult to set a minimum level of PaO 2 below whih definite ell damage will our or above whih the host is safe from the effets of hypoxi ell damage. In addition, it is not possible to monitor mitohondrial PO 2 linially and the only linially available surrogate of mitohondrial hypoxia is latate prodution. Although blood latate levels are useful and indiate tissue hypoxia, it is a late marker and therefore is an insensitive tool. Thus, targets set for ideal blood gas levels are based on arbitrary goals. Owing to the natural deline in normal arterial oxygen levels with age, it has been suggested that the ideal target PaO 2 an be determined by the following equation: 133 Ideal PaO 2 = 13.3 kpa age (in years) or 100 mm Hg age (in years) In terms of oxygen saturation measured by the bedside, this would translate into an SaO 2 of 94 98% in most situations. This strategy avoids tissue hypoxia in almost all patients and also avoids potential deleterious effets of hyperoxia. Thus, the standard pratie should be to presribe oxygen to a speifi saturation (or PaO 2 ) rather than in terms of FIO 2. Clearly, onsideration will need to be given to patients who have oxygen-sensitive arbon dioxide retention and targets may well have to be set lower for these patients to strike a balane between ahieving a desirable and safe SaO 2 /PaO 2 and arbon dioxide retention. Speifi disease states will be addressed in setion 8. Patients with moderate to severe hypoxaemia are usually breathless and have an inreased respiratory rate. Apart from ausing physial tiredness, this also inreases work of breathing, therefore inreasing both oxygen onsumption and arbon dioxide prodution. In these irumstanes, oxygen therapy may redue the work of breathing and therefore redue arbon dioxide prodution. Therefore, oxygen therapy should theoretially improve breathlessness in hypoxaemi patients. However, this effet has not been demonstrated in linial trials involving patients who were breathless but not hypoxaemi. For example, a reent meta-analysis of all published blinded studies of short burst oxygen therapy for patients with COPD with breathlessness failed to onfirm any linial benefit despite the widespread belief of dotors and patients that oxygen relieves breathlessness in this ondition. 129 A systemati review of oxygen and airflow on the relief of dyspnoea at rest in patients with advaned disease of any ause found only lowgrade sientifi evidene that oxygen and airflow improve dyspnoea in some patients with advaned disease at rest, and almost all of these subjets were hypoxaemi and already using oxygen therapy. 134 Reommendations 1. This guideline reommends aiming to ahieve a normal or near-normal oxygen saturation for all autely ill patients apart from those at risk of hyperapni respiratory failure. [Grade D] 2. The reommended target saturation range for autely ill patients not at risk of hyperapni respiratory failure is 94 98%. Some normal subjets, espeially people aged.70 years, may have oxygen saturation measurements below 94% and do not require oxygen therapy when linially stable. [Grade D] 3. Most non-hypoxaemi breathless patients do not benefit from oxygen therapy, but a sudden redution of more than 3% in a patient s oxygen saturation within the target saturation range should prompt vi27

28 fuller assessment of the patient (and the oximeter signal) beause this may be the first evidene of an aute illness. [Grade D] 4. For most patients with known COPD or other known risk fators for hyperapni respiratory failure (eg, morbid obesity, hest wall deformities or neuromusular disorders), a target saturation range of 88 92% is suggested pending the availability of blood gas results. [Grade C] 5. Some patients with COPD and other onditions are vulnerable to repeated episodes of hyperapni respiratory failure. In these ases it is reommended that treatment should be based on the results of previous blood gas estimations during aute exaerbations beause hyperapni respiratory failure an our even if the saturation is below 88%. For patients with prior hyperapni failure (requiring non-invasive ventilation or intermittent positive pressure ventilation) who do not have an alert ard, it is reommended that treatment should be ommened using a 28% Venturi mask at 4 l/min in prehospital are or a 24% Venturi mask at 2 4 l/min in hospital settings with an initial target saturation of 88 92% pending urgent blood gas results. These patients should be treated as a high priority by emergeny servies and the oxygen dose should be redued if the saturation exeeds 92%. [Grade D] 6.8 Effets of body positioning inluding restraint systems Appropriate positioning of a patient an maximise V/Q mathing. In the healthy self-ventilating adult lung, V/Q mathing improves from non-dependent to dependent areas. In lung disease there is a disruption of this pattern and, in these instanes, appropriate positioning may be advantageous in optimising V/Q mathing, therefore improving gas exhange, oxygenation and arbon dioxide learane. For these reasons, breathless patients usually prefer to sit upright or near upright provided they are able to do so. The relationship between dependeny and V/Q mathing is maintained irrespetive of the position of the subjet. The physiology is then transferable into alternate side lying positions; for example, in left side lying the dependent lung (left) will have the better V/Q mathing. This is important in the presene of asymmetrial lung pathology as the good lung down priniple will maximise V/Q mathing. Many unwell patients are nursed in the semi-reumbent and supine positions. These positions do not failitate V/Q mathing as in the upright and full side lying position due to the hindrane to expansion of the dependent lung by the diaphragm and hest wall. Even in healthy subjets the oxygen tension is 0.7 kpa (5 mm Hg) lower in the supine position than in the upright position. 12 Similarly, 10% of patients with right hemiparesis and onomitant hest disease were more hypoxaemi in the left lateral position. 135 Where there is pathologial lung disease and hene already signifiant V/Q mismath, gas exhange may be further impaired. This is disussed in a review of the effets of position on oxygen saturation in aute stroke. 136 Patients with aute stroke without respiratory o-morbidities may be permitted to adopt any body position that they find most omfortable, while those with respiratory ompromise should be positioned as upright as possible, avoiding slouhed or supine positions to optimise oxygenation. 136 The semi-reumbent/supine position is ommonly adopted in an ambulane. In addition, for safety, the patient is strapped into the strether using abdominal and hest restraints with their arms by their side. While there are a lak of speifi data regarding this, physiologial priniples suggest that the use of suh positioning and restraints would ompromise both respiratory musle funtion and gas exhange. Finally, there are some rare patients with liver disease, ardia shunts or lung fibrosis who have platypnoea and orthodeoxia whih means that they are more hypoxi in the upright position. 137 Other patients with soliosis or with a paralysed hemidiaphragm may feel more omfortable with the good lung up. These patients should be allowed to hoose the position in whih their breathing is most omfortable for them. Reommendation 6. Beause oxygenation is redued in the supine position, fully onsious hypoxaemi patients should ideally be allowed to maintain the most upright posture possible (or the most omfortable posture for the patient) unless there are good reasons to immobilise the patient (eg, skeletal or spinal trauma). [Grade C] SECTION 7: CLINICAL AND LABORATORY ASSESSMENT OF HYPOXAEMIA AND HYPERCAPNIA 7.1 Assessment of hypoxaemia Clinial assessment of breathless patients and assessment of yanosis Cliniians examining a ritially ill patient should remember the ABC of emergeny mediine (airway, breathing, irulation). In the ase of ritially ill patients it may be neessary to seure the airway and resusitate a patient before a detailed history an be obtained and before a full physial examination an be undertaken. In assessing an ill patient the SaO 2 level is only one of several physiologial variables that should be monitored. Many patients with sudden aute illness suh as postoperative pulmonary emboli will have a sudden alteration in physiologial trak and trigger variables as assessed by the modified mews system Suh patients may have only a small fall in SaO 2 due to physiologial ompensation mehanisms suh as inreased ventilation. Cliniians therefore need to be alert for falls in SaO 2 even within the reommended target ranges. Reommendations 7. Fully trained linians should assess all autely ill patients by measuring pulse, blood pressure, respiratory rate and assessing irulating blood volume and anaemia. Expert assistane from speialists in intensive are or from other disiplines should be sought at an early stage if patients are thought to have major life-threatening illnesses and liniians should be prepared to all for assistane when neessary inluding a all for a 999 ambulane in prehospital are or a all for the resusitation team or ICU outreah team in hospital are. [Grade C D] 8. Initial linial assessment and subsequent monitoring of autely unwell patients should inlude the use of a reognised physiologial trak and trigger systems suh as the modified Early Warning Soring system (mews), and a hange in this sore should require medial review even if there is no hange in oxygen saturation. [Grade C] vi28

29 Traditional linial assessment of hypoxaemia involves linial inspetion of the skin and bual muous membranes to deide whether entral yanosis is present or absent. This is a diffiult linial skill, espeially in poor lighting onditions. Clinial assessment of hypoxaemia is made even more unreliable by the presene of anaemia or polyythaemia. Some patients may have peripheral yanosis due to poor peripheral irulation in the presene of normal SaO 2. Several studies have shown that hypoxaemia is often not reognised by emergeny medial servie providers, espeially if the patient does not omplain of respiratory distress A systemati review of the literature in 2005 reported that most hypoxaemi patients had at least one vital sign abnormality but skin olour was a poor indiator of hypoxaemia ompared with pulse oximetry. 31 For these reasons it is reommended that liniians should not rely on visual assessments of yanosis but should instead use pulse oximetry to obtain an aurate assessment of a patient s oxygen saturation. The nature of a patient s presenting illness may make hypoxaemia a likely outome, thus prompting a areful linial searh for evidene of yanosis omplemented by urgent pulse oximetry. This situation applies to many ommon aute illnesses suh as heart failure, COPD exaerbation, pneumonia and pulmonary embolism. A study of 2276 patients with pneumonia showed that hypoxaemia was independently assoiated with six risk fators: age.30 years (odds ratio (OR) 3.2), COPD (OR 1.9), ongestive heart failure (OR 1.5), respiratory rate.24/min (OR 2.3), altered mental status (OR 1.6) and hest radiographi infiltrate involving.1 lobe (OR 2.2). 28 Autely ill patients with signifiant hypoxaemia are likely to have an inreased pulse rate or respiratory rate and, for this reason, usually sore at least 3 points on a mews The respiratory rate is the single best preditor of severe illness. 73 However, many patients with marked hypoxaemia may present with non-speifi findings suh as restlessness and onfusion rather than breathlessness, and oxygen saturation has been shown to be an independent preditor of mortality in multivariate analysis of the outome of emergeny medial admissions. 142 Furthermore, the work of Thrush et al 143 on normal volunteers has shown that heart rate, blood pressure and respiratory rate are not reliable indiators of hypoxaemia down to saturation levels as low as 70%. This would suggest that the hanges in vital signs whih are seen in most hypoxi patients are due to the underlying illness rather than hypoxaemia per se. Hypoxaemia may be assoiated with inreased or dereased ventilation. Although some hypoxaemi patients may have redued levels of ventilation as a ausative fator, the majority of hypoxaemi patients have inreased minute ventilation in an attempt to inrease the blood oxygen level. For example, a patient with an opiate overdose may have redued ventilation ausing hypoxaemia despite having struturally normal lungs, whereas a patient with pneumonia or major pulmonary embolism may have signifiant hypoxaemia due to ventilation-perfusion mismath despite an inreased level of ventilation. The first patient in this example may appear peaeful and non-distressed despite signifiant hypoventilation and hypoxaemia, while the seond patient is likely to have inreased ventilation and tahyardia. The liniian therefore needs to make separate assessments of a patient s oxygen saturation and level of ventilation. Having ompleted the history and rapid assessment of the patient, more detailed physial examination may reveal signs of an illness suh as major pleural effusion, major pneumothorax or unexpeted heart failure that may prompt the liniian to antiipate the presene of hypoxaemia. Advie and reommendations for linial assessment of patients with suspeted hypoxaemia The medial history should be taken when possible in an autely breathless patient and may point to the diagnosis of a partiular aute illness suh as pneumonia or pulmonary embolism or an exaerbation of a hroni ondition suh as COPD, asthma or heart failure. [Evidene level IV] Physial examination may provide evidene of a speifi diagnosis suh as heart failure or a large pleural effusion, but it is ommon for the ause of breathlessness to remain undiagnosed until the results of tests suh as hest radiographs are available. [Evidene level IV] Patients with severe hypoxaemia may present with a nonrespiratory manifestation suh as onfusion or agitation rather than breathlessness and yanosis is a diffiult physial sign to reord onfidently (espeially in poor light or with an anaemi or plethori patient). [Evidene level IV] Tahyardia and tahypnoea are ommoner than a physial finding of yanosis. [Evidene level III] Physiologial trak and trigger systems suh as the Early Warning Soring system (EWS or mews) are extremely valuable in identifying patients with life-threatening illness even if this is not immediately obvious from the patient s history. [Evidene level III] Value and limitations of pulse oximetry Clinial assessment of hypoxaemia has been revolutionised by the advent of pulse oximetry in muh the same manner as the linial assessment of blood pressure was transformed by the invention of the sphygmomanometer. However, it is ommon to see patients with aute respiratory illness who have had multiple measurements of their blood pressure but no reord made of their oxygen saturation, peak expiratory flow or FEV 1. In addition to the linial onsequenes of underassessment, Howes et al 144 and Manab et al 145 have reported that the availability of a pulse oximeter was highly ost-effetive beause the finding of normal oximetry (.94%) in many patients allowed paramedis to use oxygen less frequently with a potential finanial saving of up to $2324 (approximately 1200) per ambulane per annum. Pulse oximetry measures haemoglobin oxygen saturation by deteting the absorption of light at two speifi wavelengths that orrespond to the absorption peaks of oxygenated and deoxygenated haemoglobin. Oximeters are less reliable at low saturation suh as 80%, but modern oximeters reflet the arterial oxygen saturation aurately at saturation above about 88% In almost all linial irumstanes overed by this guideline, patients with a saturation below 88% will be given intensive therapy to bring the saturation up to at least 90%, so the inauray of the instruments at very low saturation levels should not affet patient management. In one study of 123 adult patients who had simultaneous measurements of pulse oximetry and arterial oxygen saturation measured in arterial blood gases, the 95% onfidene interval for the median differene ranged from 20.6 to +0.5%. 148 It has been estimated that an oxygen saturation of 92% or above measured by pulse oximetry has a sensitivity of 100% and speifiity of 86% for exluding hypoxaemia defined as an arterial oxygen saturation below 60 mm Hg (8 kpa). 151 vi29

30 Oximetry may be less aurate in autely ill patients on intensive are units, but there are no diret omparisons of the auray of pulse oximetry in ritially ill patients ompared with stable patients and healthy individuals. The study of Perkins and olleagues showed a mean SpO 2 of 94.6% ompared with a mean SaO 2 of 95.9% from 1132 simultaneous oximeter and arterial blood gas measurements on an intensive are unit. 152 Fortunately, this average differene of 1.3% was lower for pulse oximeter readings, thus allowing a margin of safety in most ases. This study also showed that flutuations in oxygen saturation measured by oximetry tended to be greater than hanges in arterial oxygen saturation measured with samples from an indwelling radial artery atheter. Although oximetry is widely used, there are few linial studies examining its utility. The Cohrane meta-analysis of the use of oximetry in perioperative monitoring of more than patients failed to show any redution in ompliations or deaths where oximetry was used, although oxygen was given more often to patients who were monitored with pulse oximetry. 153 The authors suggested that the orretion of modest hypoxaemia probably does not have muh effet on linial outomes. Pulse oximetry gives no information onerning ph, PCO 2 or haemoglobin level. Blood gases and full blood ount tests are therefore required as early as possible in all situations where these measurements may affet patient outomes. The auray of pulse oximetry is diminished in patients with poor peripheral perfusion whih may our hronially in onditions suh as systemi slerosis or autely in patients with hypotension or hypovolaemia. However, it has been suggested that many types of oximeter may remain aurate at arterial pressures as low as 20 mm Hg so long as the mahine is able to obtain a reading despite the low pulse pressure. 154 Most oximeters give an indiation of the pulse signal strength. It is important to ensure that the oximeter has a good signal if tehnially possible, and the probe may need to be tried on different fingers or toes or on the earlobe to obtain the best available signal for the individual patient. There are some patients with poor perfusion for whom pulse oximetry measurements annot be made. This inludes patients with old peripheries, severe hypotension and peripheral shut down. It must be remembered that oximetry gives a normal reading for oxygen saturation in most patients with anaemia beause the oxygen saturation of the available haemoglobin is normal although the total amount of haemoglobin available for oxygen transport is redued. These patients have normal oxygen saturation levels despite having anaemi hypoxia whih may ause onsiderable redution in the total oxygen ontent of the blood. It is often not reognised that a patient with an SpO 2 of 98% but a haemoglobin of 7 g/dl (( ) = 9.2 ml O 2 /dl) will have a greatly redued blood oxygen ontent ompared with a patient with a haemoglobin of 15 g/dl and a saturation of 85% (( ) = 17 ml O 2 /dl) (eah g/dl haemoglobin when fully saturated arries 1.34 ml oxygen). The auray of oximetry is unreliable in the presene of arbon monoxide or methaemoglobin. Both of these substanes have similar light absorption harateristis to oxyhaemoglobin so an apparently normal SpO 2 in a patient with arbon monoxide poisoning or methaemoglobinaemia may be falsely reassuring. Carboxyhaemoglobin levels above 2% may ause falsely elevated SpO 2 measurements. 155 Many smokers will have arboxyhaemoglobin levels above this level shortly after smoking a igarette, and the arboxyhaemoglobin level may be elevated to 15% in some smokers and up to 50% or more in aute arbon monoxide poisoning. It is not known if the redued blood oxygen ontent in smokers who develop sudden illness within a few hours of smoking igarettes has any effet on linial outomes, or if heavy smokers might benefit from a slightly higher target saturation range than non-smokers during the first few hours of a serious illness in an effort to maintain a similar blood oxygen ontent. Skin pigmentation may also influene the auray of pulse oximetry readings (usually overestimation but sometimes underestimation). In partiular, the auray of pulse oximetry is impaired in dark skinned subjets at saturation levels below 80 85% However, this should rarely be a problem in linial pratie if the saturation is maintained in the range suggested in the present guideline (94 98% for most patients), although the work of Jubran and Tobin on ventilated subjets suggested that an oxygen saturation of 92% was useful in prediting a PaO 2 above 60 mm Hg (8 kpa) in ventilated white subjets but was less reliable in ventilated blak subjets who sometimes had an SpO 2 reading that was more than 4% above the diretly measured PaO In the ase of sikle ell risis, pulse oximetry may underestimate the level of oxygenation. 159 In these irumstanes, under-reading is safer than over-reading beause no truly hypoxaemi patient would be denied oxygen therapy. However, another study found that pulse oximeters did not misdiagnose either hypoxaemia or normoxaemia during a sikle ell risis provided a good wave signal was present. 160 Oximeters an be affeted by motion of the patient s hand, but this is less of a problem with modern oximeters than with older devies. 161 Motion artefat is more of a problem if the patient also has redued perfusion of the measuring site. 162 A malpositioned oximeter sensor an ause artefat whih an overestimate or underestimate the true oxygen saturation; this an be a partiular problem during repositioning of ill patients. 163 The site of oximetry is also important. Finger and earlobe measurements are more aurate than measurements from a probe applied to the toe, and finger probes may be more aurate than ear probes. Finally, linial staff need to remember to remove nail varnish and false nails to avoid artefats in oximetry measurements. Pulse oximeters are aurate to within 1 2% of diretly measured arterial oxygen saturation in most subjets but the error (usually overestimation but sometimes underestimation) is greater in dark skinned subjets, espeially with very low saturation (below 80 85%). [Evidene level IIa] The auray of oximeters in shok, sepsis and hypotension is largely unknown, but most errors are likely to result in falsely low readings whih would result in additional oxygen being given. Most errors in oximetry are therefore not likely to plae patients at risk, but it is important to ensure that the oximeter has a good signal and it is important to avoid artefat due to motion, nail varnish or other potential soures of error. [Evidene level IIa] Oximetry is a valuable linial tool but subjet to artefat and errors of interpretation. All linial staff who use oximeters must therefore be trained in their use and made aware of the limitations of oximetry. [Evidene level IV] It is advised that oximetry measurements on sleeping patients should be reorded over several minutes to avoid the possibility of being misled by a normal transient noturnal dip in oxygen saturation. [Evidene level III] Pulse oximetry an be misleadingly normal in smokers beause of raised blood arboxyhaemoglobin levels whih vi30

31 will ause a redued blood oxygen ontent despite an apparently normal oxygen saturation and a normal oxygen tension. Patients who have smoked igarettes in the previous 10 h may therefore be at inreased risk from hypoxia. [Evidene level III] Reommendations 9. Oxygen saturation, the fifth vital sign, should be heked by trained staff using pulse oximetry in all breathless and autely ill patients (supplemented by blood gases when neessary) and the inspired oxygen onentration should be reorded on the observation hart with the oximetry result. [Grade D] 10. The presene of a normal SpO 2 does not always negate the need for blood gas measurements beause pulse oximetry will be normal in a patient with normal oxygen tension but abnormal blood ph or PCO 2 or with a low blood oxygen ontent due to anaemia. Blood gases and full blood ount tests are therefore required as early as possible in all situations where these measurements may affet patient outomes. [Grade D] Arterial and arteriolised blood gases (indiations for blood gas sampling are given in setion 8.4 and reommendation 13) Arterial blood gases are the gold standard test for assessing respiratory failure. However, reent studies have shown that arteriolised apillary gases from the earlobe (but not from the finger) an provide an assessment of ph and PaCO 2 that is almost idential to that obtained from an arterial sample In both aute and stable situations the earlobe speimen gives a PO 2 measurement whih is kpa ( mm Hg) lower than the simultaneous arterial measurement with most of the divergene ourring at oxygen tensions above 8 10 kpa (60 75 mm Hg) This means that most patients an be managed safely based on the ph and PCO 2 levels measured from earlobe blood gases supplemented by oxygen saturation measured by a pulse oximeter In ritially ill patients the initial speimen should be an arterial speimen to guarantee an aurate initial assessment, but apillary gases are espeially valuable for monitoring progress of the blood gases as a patient stabilises. Patients who have had simultaneous arterial and earlobe samples rated the earlobe punture proedure as being onsiderably less painful than arterial punture. 171 However, the administration of loal anaesthesia before arterial blood gas sampling produed a signifiant redution in pain. 172 There is a very small risk of arterial damage from arterial punture, espeially if the radial site is used. Most reports of hand ishaemia have involved indwelling radial artery annulae, but the vessel ould also be injured by needle punture. 173 The guideline therefore reommends that arteriolised earlobe speimens should be used more widely than at present as a safer and less painful alternative to arterial blood gas sampling and loal anaestheti should be used wherever possible for arterial blood gas sampling, but this is often not pratial in medial emergenies and blood gas sampling should not be delayed in these irumstanes. However, the auray of earlobe samples in shok or hypotension is not known and it is reommended that arterial blood gases should be used in all ases of shok or hypotension (systoli blood pressure,90 mm Hg). The tehnique of patient preparation, sample aquisition and sample proessing for arteriolised apillary gases is omplex and should only be undertaken by fully trained staff. Capillary gases are very vulnerable to errors in tehnique and they should only be implemented for emergeny use in units where staff have been fully trained in their use and where the quality of the tehnique is monitored onstantly. Use of arteriolised apillary blood gas measurements Patients find earlobe speimens less painful than arterial punture without loal anaesthesia. [Level IIa] Arteriolised earlobe blood gases will provide aurate information about PaCO 2 and ph but do not provide an aurate measurement of PaO 2. [Level IIa] The earlobe speimen gives a PO 2 measurement whih is kpa (4 7.5 mm Hg) lower than the simultaneous arterial measurement with greater divergene at oxygen levels above 8 10 kpa (60 75 mm Hg). [Level IIa] However, a ombination of earlobe gases (to monitor ph and PCO 2 ) and oximetry (to measure oxygen levels) will allow safe management of most patients, even in emergeny settings. (The only published evidene is for patients with COPD but this finding is likely to be generalisable to most patients other than those with shok or poor peripheral irulation). [Level IV] The tehnique of patient preparation, sample aquisition and sample proessing for arteriolised apillary gases is omplex and should only be undertaken by fully trained staff. [Level IV] Reommendations 11. For ritially ill patients or those with shok or hypotension (systoli blood pressure,90 mm Hg), the initial blood gas measurement should be obtained from an arterial speimen. However, for most patients who require blood gas sampling, either arterial blood gases or arteriolised earlobe blood gases may be used to obtain an aurate measure of ph and PCO 2.However,the PaO 2 is less aurate in earlobe blood gas samples (it underestimates the oxygen tension by kpa) so oximetry should be monitored arefully if earlobe blood gas speimens are used. [Grade B] 12. Loal anaesthesia should be used for all arterial blood gas speimens exept in emergenies or if the patient is unonsious or anaesthetised. [Grade B] Transutaneous oxygen assessments Transutaneous oxygen devies give different information from pulse oximetry. They are more sensitive to redued perfusion and may be used to monitor tissue oxygenation in trauma patients but their use is beyond the sope of this guideline Assessment of hyperapnia and aidosis Clinial assessment In patients with lung disease hyperapnia may be aompanied by visible respiratory distress, but this will be absent when hyperapnia is a onsequene of a redution in minute ventilation. Patients may have a flushed fae, a full and bounding pulse and musle twithing together with the harateristi flap of the outstrethed hands. In severe ases onsiousness may be depressed and onvulsions may our. Gross hyperapnia usually ours with profound hypoxaemia and it is therefore diffiult to disentangle the diret effet of hyperapnia per se. Coma will usually our when the PaCO 2 is in the range kpa ( mm Hg). vi31

32 Survival has been seen following a PaCO 2 of 67 kpa (500 mm Hg). 175 The presene of hyperapni respiratory failure an be antiipated in patients with severe exaerbations of COPD or other diseases suh as severe neuromusular disorders. Carbon dioxide is a vasodilator so patients with hyperapnia may develop headahe. Carbon dioxide in high onentrations has hypnoti effets and patients with hyperapnia may progress from drowsiness with flapping tremor to onfusion to oma. A study of 127 episodes of aute respiratory aidosis showed that the best linial preditors of respiratory aidosis were drowsiness (OR 7.1), flushing (OR 4.1), the presene of known COPD (OR 3.3) and the presene of interostal retration (OR 2.9). 176 Clinial signs of arbon dioxide retention inlude: Vasodilation produing flushing and warm peripheries with dilated blood vessels (inluding retinal veins). Bounding pulse. Drowsiness. Flapping tremor. Confusion. Coma Blood arterial and arteriolar gases (see setion for further details) Arterial or arteriolised earlobe apillary blood gases will give an aurate estimation of ph and PaCO The blood gases will need to be repeated in min in patients with signifiant hyperapnia or aidosis to monitor the response to treatment. Patients with COPD who remain aidoti despite min of standard treatment (inluding ontrolled low-dose oxygen therapy) are likely to need non-invasive ventilation Venous PCO 2 sampling It has been suggested that the venous PCO 2 level an be used to sreen for hyperapnia in patients with aute respiratory disease. A study of 196 paired samples of arterial and venous blood from patients with aute respiratory disease showed that the PCO 2 in the venous sample was an average of 0.77 kpa (5.8 mm Hg) higher than the simultaneous arterial sample. 177 A venous PCO 2 below 6 kpa (45 mm Hg) had 100% sensitivity for eliminating the risk of hyperapnia (arterial PCO 2 above 6 kpa or 45 mm Hg), although the speifiity was low at 57% and there was more variation in other studies For patients who are not at risk of metaboli aidosis, the presene of a satisfatory oxygen saturation measured by pulse oximetry and a venous PCO 2 below 6 kpa (45 mm Hg) an exlude the possiblility of signifiant arterial hypoxia or hyperapnia and may obviate the need for arterial blood gas measurements. However, venous PCO 2 sampling is not widely used in linial pratie at present and the guideline ommittee have therefore made no reommendations on its use Carbon dioxide monitors and non-invasive assessments of hyperapnia End-tidal arbon dioxide monitors are used primarily to onfirm traheal intubation during anaesthesia, intensive are and for any patients requiring endotraheal intubation. They are onsidered the gold standard by the Royal College of Anaesthetists. The absene of any detetable arbon dioxide output indiates a failed intubation. The management of intubated patients is outside the remit of this guideline. End-tidal arbon dioxide monitors are also useful in the management of ardia arrest and irulatory ollapse. Very low levels of arbon dioxide exretion indiate very low (or absent) ardia output and a low likelihood of survival These devies are also useful in the are of intubated patients in the emergeny department beause, through visualising a typial box wave form, they an onfirm that the tube is in the airway even in the absene of arbon dioxide prodution during a ardia arrest. The appearane of arbon dioxide may be the first sign of spontaneous irulation. 183 End-tidal arbon dioxide measurements orrelate poorly with arterial arbon dioxide levels in patients with COPD, but they may be useful in some researh studies of hyperventilation syndromes. However, these devies are inaurate in patients with airways disease and those with a high respiratory rate, so they should not be used in the management of patients with respiratory failure and they will not be disussed further in this guideline. An exiting new possibility is the development of probes that an assess PCO 2 as well as SpO 2 from a single probe. Early studies indiate that suh devies an be aurate in normal volunteers and there have been some enouraging preliminary studies in patients with respiratory disease. Transutaneous arbon dioxide monitors are also being developed in assoiation with transutaneous oxygen monitors for use in patients with shok and ritial illness. 174 SECTION 8: EMERGENCY OXYGEN USE IN HOSPITAL SETTINGS The hospital management of hypoxaemi patients is presented before the prehospital management beause it represents the ideal management. Some readers may prefer to read setion 9 (prehospital are) first beause most patients reeive prehospital are before hospital are, but the Guideline Development Group preferred to present the ideal management first. 8.1 Assessment and immediate management of breathless patients on arrival in hospital Breathless patients may arrive in hospital diretly (without prior assessment) or in ambulanes where they will usually have been assessed by paramedis who may also have initiated emergeny treatments inluding oxygen therapy. As disussed in setion 7 of this guideline, assessment, triage and resusitation of ritially ill patients must be undertaken in parallel with the initiation of oxygen therapy and speifi treatment must be given for the underlying medial ondition. All ritially ill patients and all patients at risk of hyperapni respiratory failure should be triaged as very urgent and should have blood gases taken on arrival in hospital. Furthermore, all seriously ill patients should be assessed by senior liniians as early as possible. In many ases this may involve liaison with intensive are speialists or with appropriate other speialists who an deal effetively with the patient s major medial or surgial problems. Readers are referred to setion and to disease-speifi guidelines for advie onerning the immediate assessment and management of seriously ill patients. Readers are referred to setion 10 for advie onerning hoie of oxygen delivery devies and systems. Readers are referred to tables 1 4 and harts 1 and 2 (figs 1 and 2) for a summary of the key elements of oxygen therapy in ommon medial emergenies. Remember to ask for senior advie or speialist advie early in the are of profoundly ill patients. vi32

33 8.2 Differenes in management in hospital ompared with a prehospital setting The immediate management of medial emergenies in hospital settings before blood gas results are available is similar in priniple to management in the prehospital setting (setion 9). The main priorities are to avoid harmful levels of hypoxaemia for all patients and to avoid harmful levels of hyperapnia for patients who are at risk of this ompliation. However, the amount of information available to the healthare professionals inreases rapidly in the hospital environment. The hospital management is presented before the prehospital management beause it represents the ideal management. This may also be ahievable in some prehospital settings suh as a well equipped primary are entre. However, in many prehospital settings there will usually be less information available onerning a patient s history and physiology and less equipment available to assess and treat the patient. Differenes between hospital settings and prehospital settings inlude: Pulse oximetry is almost always available in hospital at present. These guidelines also reommend that pulse oximetry must be available in all loations where emergeny oxygen is used (setion 9.1). Blood gas results an be available within minutes of arrival in hospital. Additional diagnosti information may be available from history, linial examination, test results and from the patient s hospital reords. Additional equipment and resoures are available (eg, ability to ventilate). Beause of the universal availability of oximetry in hospitals, it is rare for the hospital medial team to have to administer oxygen on the basis that a patient might be hypoxaemi. However, initial blind management is sometimes neessary for patients with shok or with very poor peripheral irulation where a reliable pulse oximetry trae annot be obtained. Arterial blood gases should be obtained as a matter of urgeny in all suh ases. 8.3 Whih patients need oxygen therapy? Supplementary oxygen therapy is required for all autely hypoxaemi patients and for many other patients who are at risk of hypoxaemia, inluding patients with major trauma and shok. Most autely breathless patients will require supplementary oxygen therapy, but there are some situations suh as aute hyperventilation or diabeti ketoaidosis where an apparently breathless patient will not benefit from oxygen therapy. There are some other linial situations suh as arbon monoxide poisoning where a patient may benefit from oxygen therapy despite a lak of hypoxaemia or breathlessness beause arbon monoxide binds more avidly than oxygen to the haemoglobin moleule. Reommendations Oxygen saturation should be measured in all breathless and autely ill patients (see reommendation 9). Oxygen therapy should be given to hypoxaemi patients (see table 1). Patients do not require oxygen therapy if their oxygen saturation is 94% or above (exeptions are arbon monoxide poisoning and pneumothorax; see setions and ). Patients on oxygen with SpO 2.98% may not require oxygen therapy or may require a lower dose (see reommendatons 1 3 and table 1). All patients with shok, major trauma, sepsis or other ritial illness should be managed initially with high onentration oxygen therapy from a reservoir mask. The dose an be adjusted subsequently one the results of blood gas estimations are known and/or the patient is stable (see table 1). [Grade D] 8.4 Whih patients require blood gas measurements? Blood gases should be measured as soon as possible in most emergeny situations involving hypoxaemi patients 186 and are essential in patients who may develop type 2 respiratory failure (arbon dioxide retention with risk of respiratory aidosis). Blood gases should also be heked (and the linial situation should be reviewed) if the oxygen saturation should fall by more than three perentage points, even if the saturation remains within the target range. For example, a fall from 98% to 93% might be due to a signifiant event suh as a pulmonary embolus. In this situation the saturation of 93% will not harm the patient but the patient will remain at serious risk until the pulmonary embolism is diagnosed and treated. If oximetry shows a patient to be hypoxaemi, the initiation of oxygen therapy should not be delayed while awaiting the results of blood gas measurements. Blood gas measurements are not usually required for patients with no risk fators for hyperapni respiratory failure and an oxygen saturation of 94% or above breathing air unless the patient requires blood gas estimation for other reasons suh as suspeted metaboli aidosis or diabeti ketoaidosis. The BTS asthma guideline reommends that arterial blood gas measurements need not be reorded in patients with aute asthma and an oxygen saturation above 92% and no life-threatening features. 187 Arterial blood gas sampling an be tehnially diffiult, espeially for poorly perfused patients, and junior staff should ask for assistane from more senior staff in diffiult ases. Following initial linial assessment and the availability of a pulse oximetry measurement, a deision an be made regarding the need for blood gas estimation within a few minutes of arrival in the hospital environment or if a previously stable patient develops breathlessness within a hospital environment. Oximetry will give no information onerning arbon dioxide or ph levels and a normal pulse oximetry level may provide false reassurane in patients on oxygen therapy who may have unexpeted hyperapnia and aidosis. However, areful linial assessment supplemented by the use of oximetry will allow the setting of an appropriate oxygen saturation target for different groups of patients until blood gas results are available. If repeated blood gas estimations are required, the timing will depend on the indiation. In general, the oxygen saturation (and PaO 2 ) stabilises at a new higher level within a few minutes of inreasing the dose of oxygen but the PaCO 2 an take min to equilibrate. The rise in blood oxygen level an be monitored with oximetry, so repeat blood gas tests are done mostly to assess ritial illness (immediate sampling required) or to monitor ph and PCO 2 levels (best done min after inreasing the dose of oxygen). Reommendation 13. Blood gases should be heked in the following situations: All ritially ill patients. Unexpeted or inappropriate hypoxaemia (SpO 2,94% in patients breathing room air or oxygen) or any patient requiring oxygen to ahieve the above vi33

34 target range. (Allowane should be made for transient dips in saturation to 90% or less in normal subjets during sleep). [Grade D] Deteriorating oxygen saturation or inreasing breathlessness in a patient with previously stable hypoxaemia (eg severe COPD). [Grade D] Any previously stable patient who deteriorates and requires a signifiantly inreased FIO 2 to maintain a onstant oxygen saturation. [Grade D] Any patient with risk fators for hyperapni respiratory failure who develops aute breathlessness, deteriorating oxygen saturation, drowsiness or other symptoms of arbon dioxide retention. [Grade D] Breathless patients who are thought to be at risk of metaboli onditions suh as diabeti ketoaidosis or metaboli aidosis due to renal failure. [Grade D] Autely breathless or ritially ill patients with poor peripheral irulation in whom a reliable oximetry signal annot be obtained. [Grade D] Any other evidene from the patient s medial ondition that would indiate that blood gas results would be useful in the patient s management (eg, an unexpeted hange in trak and trigger systems suh as a sudden rise of several units in the mews sore or an unexpeted fall in oxygen saturation of 3% or more, even if within the target range). [Grade D] 8.5 Can arteriolised earlobe gases be used as a substitute for arterial blood gases? Readers are referred to setion for advie onerning when to use arterial blood gases and when to use arteriolised earlobe blood gases. 8.6 Should oxygen be presribed at a fixed dose or to ahieve a target saturation? In the past, oxygen was presribed at a fixed FIO 2 or at a fixed flow rate via nasal annulae or variable performane fae masks. However, several audits have shown that many (or most) patients do not reeive the presribed dose of oxygen. 2 8 Furthermore, a patient s oxygen requirement may vary over time so the presribed oxygen dose may be too high or too low even a short time after the presription was written. For this reason it is reommended that oxygen should be presribed to a target saturation range rather than presribing a fixed dose of oxygen or fration of inspired oxygen. This is analogous to an insulin sliding sale where the presriber speifies a variable dose of insulin to ahieve a target blood gluose range rather than presribing a fixed dose of insulin. This will allow the appropriate healthare professional usually a dotor, nurse or physiotherapist to adjust eah patient s dose of oxygen to ahieve the safest oxygen saturation range for eah patient. The presriber may indiate a starting dose, devie or flow rate, but there needs to be an agreed system for adjusting the oxygen dose upwards or downwards aording to a patient s needs (see harts 1 and 2 (figs 1 and 2); setions and and harts 3 and 4 (figs 17 and 18)). As a patient improves, he or she is likely to require a lower FIO 2 over a time period that will vary between patients. Most reovering patients will eventually require no supplemental oxygen. On the other hand, a deteriorating patient may need an inreased dose of oxygen. This inrease an be initiated by nursing staff or physiotherapists, but the requirement for an inreased dose of oxygen is an indiation for urgent linial reassessment of the patient (and repeat blood gas measurements in most instanes). It is reommended that oxygen should be presribed to a target saturation range rather than presribing a fixed dose of oxygen or fration of inspired oxygen (see reommendations 1, 2, 4 and 5). 8.7 What should be the target oxygen saturation range for patients reeiving supplementary oxygen? Oxygen saturation target range for most patients As disussed in setions 4 6 of this guideline, there is no evidene of benefit from above normal oxygen saturation in most medial emergenies and there is evidene that exessive doses of oxygen an have adverse effets, even in some patients who are not at risk of hyperapni respiratory failure. A target oxygen saturation range of 94 98% will ahieve normal or near normal oxygen saturation for most patients who are not at risk of hyperapni respiratory failure. Furthermore, the suggested lower limit of 94% allows a wide margin of error in the oximeter measurement, thus minimising the risk of any patient being allowed to desaturate below 90% due to inaurate oximetry Oxygen requirements for speifi groups of patients Patients with ritial illness requiring high dose oxygen therapy are disussed in setion Patients with medial emergenies whih frequently ause breathlessness and hypoxaemia are disussed in setion Patients with COPD and other onditions that may predispose to type 2 respiratory failure are disussed in setion Medial emergenies for whih oxygen is ommonly given at present but is not atually indiated unless the patient is hypoxaemi are disussed in setion Importane of blood gas measurements in guiding oxygen therapy As soon as blood gas measurements are available, a patient s further treatment an be guided by the results of this test. For patients with a normal or low PaCO 2 and no risk fators for hyperapni respiratory failure, it is safe to aim at an oxygen saturation in the normal range (94 98%). For patients with a raised PaCO 2, a lower oxygen saturation is indiated (88 92%), espeially if the patient is aidoti. Non-invasive ventilation is reommended for patients with COPD who have hyperapnia and a ph,7.35 ([H + ].45 nmol/l) despite 1 h of standard medial treatment inluding ontrolled oxygen therapy. 8.9 What should be the initial hoie of oxygen delivery system in hospital settings? The tehnial and pratial aspets of different oxygen delivery systems are disussed in setion 10. For major trauma ases and for severely hypoxaemi patients without risk fators for hyperapni respiratory failure, a non-rebreathing mask (reservoir mask) at l/min is the suggested first hoie. The delivery system and FIO 2 may be adjusted later to a lower dose of oxygen as a patient improves or towards supported ventilation if the patient deteriorates. The majority of patients with modest hypoxaemia an be treated with nasal annulae or a simple fae mask at a flow rate whih is adjusted to maintain the oxygen saturation in the target range for their speifi linial presentation. Chart 2 (fig 2) shows a suggested sheme that allows the oxygen level to be adjusted upwards or downwards in gradual inrements depending on a patient s vi34

35 linial progress (see also setions and ). Venturi masks are reommended for low-dose oxygen therapy beause they deliver a more reliable oxygen onentration than nasal annulae or variable flow masks. 188 They an also be ombined with a humidifier system when neessary (see setion ). The mask and/or flow should be rapidly hanged if the initial hoie does not ahieve the target saturation Devies used in emergeny oxygen therapy in hospitals (see setion 10 for further details) High onentration oxygen from reservoir mask (10 15 l/ min) or bag-valve mask for ritial illness or severe hypoxaemia or during resusitation. Nasal annulae (2 6 l/min) or simple fae masks (5 10 l/ min) for medium-dose oxygen therapy. 24% Venturi mask at 2 l/min or 28% Venturi masks at 4 l/ min for patients at risk of hyperapni respiratory failure (hange to nasal annulae at 1 2 l/min when the patient has stabilised). Traheostomy masks for patients with prior traheostomy (adjust flow to ahieve desired saturation) Reommended oxygen therapy for major medial emergenies and ritial illness (see also table 1) There are a number of major medial emergenies where patients are very likely to suffer from hypoxaemia. High-dose oxygen therapy from a reservoir mask at l/min is reommended in the initial management of all suh patients prior to stabilisation in a ritial are area or high dependeny unit. Following stabilisation, the dose of oxygen an be titrated downwards to maintain a target saturation of 94 98%. It is reommended that patients with COPD or other risk fators for hyperapnia who develop a ritial illness should be treated by emergeny servies in the same manner as other ritially ill patients until urgent blood gas results beome available beause the primary issue is the ritial illness. Critially ill patients with hyperapnia, hypoxaemia and aidosis will require immediate assessment by intensive are teams and will usually require intubation and mehanial ventilation Cardia arrest and other onditions requiring ardiopulmonary resusitation (CPR) The 2005 guideline for Adult Advaned Life Support issued by Resusitation Counil UK reommends the use of non rebreathing reservoir masks (or 100% oxygen via a self-inflating bag mask system) to deliver the highest possible inspired oxygen level to patients requiring resusitation. 189 The present guideline endorses these proposals during the period of resusitation. Subsequent management will depend on the underlying ondition and the patient s degree of reovery. There is theoretial evidene that patients who have survived the initial stages of resusitation may be managed more safely with 30% oxygen than with 100% oxygen. Some patients will require invasive ventilation following CPR, but others will reover rapidly and an oxygen saturation target of 94 98% is reommended during the onvalesent period. Reommendation (see table 1) Use high-dose oxygen from a reservoir mask at 15 l/min or bag-valve mask during resusitation. [Grade D] Critially ill patients inluding major trauma, shok and major sepsis There is evidene that early intervention to normalise oxygen delivery to the tissues using volume expansion and vasoative agents is benefiial in the management of ritially ill patients with shok or sepsis, but there is no evidene of benefit from attempts to ahieve supranormal oxygen delivery. In fat, there is evidene that hyperoxia an ause a paradoxial derease in whole body oxygen onsumption in ritially ill patients, 195 and it has been demonstrated reently that hyperoxia an impair oxygen delivery in septi patients. 196 Most suh patients are at risk of multiorgan failure and therefore require intensive are assessment as a matter of urgeny. Critial are onsensus guidelines set 90% saturation as the minimum level below whih oxygen saturation should not be allowed to fall, and the Surviving Sepsis Campaign guideline reommends a target arterial oxygen saturation of 88 95% for patients with sepsis. However, these reommendations are based on diretly measured arterial oxygen saturations in ritial are settings with intensive levels of nursing and monitoring. The present guideline reommends a slightly higher target saturation range prior to the transfer of these seriously ill patients to ritial are failities. For most ritially ill or severely hypoxaemi patients, initial oxygen therapy should involve the use of a reservoir mask, aiming at an oxygen saturation of 94 98%. If the patient has onomitant COPD or other risk fators for hyperapni respiratory failure, the initial saturation target should also be 94 98% pending the results of blood gas estimations and assessment by intensive are speialists. If ritially ill COPD patients have hyperapnia and aidosis, the orretion of hypoxaemia must be balaned against the risks of respiratory aidosis and ventilatory support using non-invasive or invasive ventilation should be onsidered. It is also reognised that many patients with long bone fratures may develop hypoxaemia even in the absene of injury to the airway or hest (possibly due to opiate treatment and fat embolism) and they should be monitored with oximetry and given oxygen if neessary These patients, if not ritially ill, should have a target oxygen saturation of 94 98% or 88 92% if they have o-existing COPD or other risk fators for hyperapni respiratory failure. Reommendation (see table 1) In ritial illness, inluding major trauma and sepsis, initiate treatment with a reservoir mask at l/min and aim at a saturation range of 94 98%. [Grade D] Near-drowning Survivors of near-drowning may have suffered inhalation of fresh or sea water into the lungs and may beome hypoxaemi. Supplemental oxygen should be given to all patients with saturation below 94%, aiming at a target saturation of 94 98%. Reommendation (see table 1) In ases of near-drowning, aim at an oxygen saturation of 94 98%. [Grade D] Anaphylaxis Patients with anaphylaxis are likely to suffer from tissue hypoxia due to a ombination of upper and/or lower airway obstrution together with hypotension. In addition to speifi vi35

36 treatment of these problems, the Resusitation Counil UK reommends high onentration oxygen (10 15 l/min, presumably by reservoir mask if available) for patients with anaphylaxis. 189 The present guideline would endorse this pratie in the immediate management of anaphylaxis followed by a target saturation of 94 98% one the patient s ondition has stabilised. Reommendation (see table 1) In anaphylaxis, initiate treatment with a reservoir mask at l/min and aim at a saturation range of 94 98%. [Grade D] Major pulmonary haemorrhage or massive haemoptysis Major pulmonary haemorrhage and massive haemoptysis an our for a large number of reasons ranging from aute pulmonary vasulitis to erosion of a blood vessel by a lung tumour. In addition to speifi treatment of the ausative ondition, most suh patients require supplementary oxygen treatment. A target saturation range of 94 98% is reommended. Treatment should be initiated with high onentration oxygen via a reservoir mask and subsequently adjusted aording to Chart 2 (fig 2) to maintain a saturation of 94 98% pending the results of blood gas measurements. Reommendation (see table 1) In pulmonary haemorrhage, aim at an oxygen saturation of 94 98%. [Grade D] Major head injury Patients with major head injury are at risk of hypoxaemia and hyperapnia. They require urgent assessment and maintenane of airway pateny, either through positioning, simple adjunts or early intubation and ventilation to avoid further brain injury due to brain oedema whih may be aggravated by hyperapnia. These patients should be referred immediately to appropriately trained speialists, even if this requires an interhospital transfer. Initial treatment should inlude high onentration oxygen via a reservoir mask pending availability of satisfatory blood gas measurements or until the airway is seured by intubation. Although hypoxaemia is ommon in patients with head injury, the relative ontribution of hypoxaemia to outome is not yet established All authors agree that hypoxaemia should be orreted, but a reent review of the literature onluded that there is no evidene of linial benefit from hyperoxia in braininjured patients and a subsequent linial study showed that normobari hyperoxia did not improve brain metabolism in five patients with aute severe brain injury There are no UK guidelines for oxygen therapy in the immediate phase after head injury, but US guidelines reommend maintaining an oxygen saturation above 90% for patients with aute brain injury. 22 The present guideline advises giving supplementary oxygen if required to maintain an oxygen saturation in the range of 94 98%. Reommendation (see table 1) In ases of major head injury, aim at an oxygen saturation of 94 98%. Initial treatment should involve high onentration oxygen from a reservoir mask at l/min pending availability of satisfatory blood gas measurements or until the airway is seured by intubation. [Grade D] Carbon monoxide poisoning Patients with arbon monoxide poisoning have a normal level of PaO 2 but a greatly redued level of oxygen bound to haemoglobin beause this has been displaed by arbon monoxide. 207 Pulse oximetry annot sreen for arbon monoxide exposure as it does not differentiate arboxyhaemoglobin from oxyhaemoglobin and blood gas measurements will show a normal PaO 2 in these patients. The blood arboxyhaemoglobin level must be measured to assess the degree of arbon monoxide poisoning. The half-life of arboxyhaemoglobin in a patient breathing room air is approximately 300 min; this dereases to 90 min with high onentration oxygen via a reservoir mask. The most important treatment for a patient with arbon monoxide poisoning is therefore to give high-dose oxygen via a reservoir mask. Comatose patients or those with severe mental impairment should be intubated and ventilated with 100% oxygen. The role of hyperbari oxygen remains ontroversial. A 2005 Cohrane review onluded that existing randomised trials did not establish whether the administration of hyperbari oxygen to patients with arbon monoxide poisoning redued the inidene of adverse neurologial outomes. 208 However, a randomised trial published in 2007 has suggested that patients with loss of onsiousness or high arboxyhaemoglobin levels may have less ognitive sequelae if given hyperbari oxygen. 209 Reommendation (see table 1) In ases of arbon monoxide poisoning, an apparently normal oximetry reading may be produed by arboxyhaemoglobin, so aim at an oxygen saturation of 100% and use a reservoir mask at 15 l/min irrespetive of the oximeter reading and PaO 2. [Grade C] 8.11 Serious illnesses requiring moderate levels of supplemental oxygen if the patient is hypoxaemi (see also table 2) Patients who present with aute medial emergenies who are not ritially ill or grossly hypoxi an be treated with medium-dose oxygen therapy from nasal annulae or a simple fae mask with a target saturation range of 94 98%. Some of these patients (eg, patients with pneumonia) may subsequently deteriorate, requiring high onentration oxygen from a reservoir mask or requiring respiratory support suh as invasive ventilation. Others may turn out to have an additional diagnosis of COPD or neuromusular disease with a risk of hyperapni respiratory failure and they should be managed with a Venturi mask or 2 litres of oxygen via nasal annulae, aiming at a target saturation of 88 92%. There are no published trials supporting the use of oxygen to relieve breathlessness in non-hypoxaemi patients, and there is evidene from randomised studies that oxygen does not relieve breathlessness ompared with air in non-hypoxaemi patients with COPD who are breathless following exertion Patients with aute onset of hypoxaemia of unknown ause with no pre-existing respiratory disorders or risk fators It is ommon for breathless and hypoxaemi patients to have no firm diagnosis at the time of presentation. For most autely hypoxaemi patients whose medial problem is not yet diagnosed, an oxygen saturation range of 94 98% will avoid the potential hazards assoiated with hypoxaemia or hyperoxia (see setions 4 6 and table 1). Aiming for an oxygen saturation in the normal range (rather than an abnormally high oxygen vi36

37 level) will also have the effet of allowing the lowest effetive FIO 2 to be used, thus avoiding risks suh as absorption ateletasis and V/Q mismath that may be assoiated with the use of very high frations of inspired oxygen (see setions 5 and 6). The priority for suh patients is to make a speifi diagnosis as early as possible and to institute speifi treatment for the underlying ondition. Early blood gas measurement is mandatory in the management of patients with sudden unexplained hypoxaemia. Reommendations (see table 2) For autely breathless patients not at risk of hyperapni respiratory failure who have saturations below 85%, treatment should be ommened with a reservoir mask at l/min in the first instane. The oxygen dose an be adjusted downwards (using nasal annulae or a simple fae mask) to maintain a target saturation of 94 98% one the patient has stabilised. [Grade D] In all other ases without risk fators for hyperapni respiratory failure, treatment should be ommened with nasal annulae (or a simple fae mask if annulae are not tolerated or not effetive) with the flow rate adjusted to ahieve a saturation of 94 98%. [Grade D] If medium-dose therapy with nasal annulae or a simple fae mask does not ahieve the desired saturation, hange to a reservoir mask and seek senior or speialist advie. [Grade D] Aute asthma The BTS/SIGN guideline for the management of aute asthma reommends that the oxygen saturation should be maintained above 92%. 187 The present guideline suggests a target saturation of 94 98% for most disease onditions, inluding asthma. The lower limit of 94% in this guideline is reommended in order to maintain onsisteny throughout the guideline. The rationale for this approah is explained in setion 6.7 and in reommendation 3. Although there is no danger of tissue hypoxia at any saturation above 90%, a drop of oxygen saturation below 94% may indiate deterioration and should prompt a further assessment. Supplementary oxygen should be started using nasal annulae at 2 4 l/min or a simple fae mask at 5 l/min or 35 40% Venturi mask and adjusted as neessary to maintain a saturation of 94 98%. 210 The BTS asthma guideline reommends giving high-flow oxygen to all patients with aute severe asthma. However, this aspet of the guideline was written in the early 1990s, before oximetry was in routine use. A study whih was published in 2003 showed that the administration of 100% oxygen to patients with aute severe asthma produed an inreased PaCO 2 and a dereased peak expiratory flow ompared with patients treated with 28% oxygen. 211 The authors of that study reommended the use of targeted oxygen therapy rather than giving high onentration oxygen to all patients with aute severe asthma. It remains appropriate to give oxygen to patients with aute severe asthma in the absene of oximetry or blood gas results, but there is no evidene of benefit from giving oxygen to patients who are not hypoxaemi. Oxygen should not be withheld from hypoxaemi patients with severe asthma beause of onerns about possible hyperapnia, although there is some evidene that this phenomenon does our. Hyperapnia in aute asthma indiates a near-fatal attak and indiates the need for onsideration of intensive are admission and ventilation. 187 Reommendation (see table 2) In aute asthma, aim at an oxygen saturation of 94 98%. [Grade C] Pneumonia The BTS guideline for pneumonia reommends aiming at an oxygen saturation above 92% and PaO 2.8 kpa (60 mm Hg) in unompliated pneumonia with appropriate adjustments for patients with COPD, guided by blood gas measurements. 213 The present guideline endorses these priniples. For internal onsisteny, a saturation range of 94 98% is reommended for most adults and it is reommended that patients with COPD ompliated by pneumonia should be managed in aordane with the COPD setion of the present guideline. Reommendation (see table 2) In ases of pneumonia, aim at an oxygen saturation of 94 98%. [Grade D] Lung aner and other aners with pulmonary involvement Most patients with aner who present with aute breathlessness have a speifi ausative fator suh as a pleural effusion, pneumonia, COPD, anaemia or ollapse of a lobe or of the left or right lung. One small double blind trial reported that hypoxaemi patients with advaned aner (SpO 2,90%) had redued dyspnoea breathing oxygen ompared with air, but a larger and more reent study failed to show benefit from oxygen ompared with air, even when hypoxaemia was present A single blind study involving 38 hospie patients with dyspnoea at rest showed a redution in breathlessness when oxygen or air was given. 217 Other studies have shown improvements in breathlessness in patients with aner given opiates or benzodiazepines but not with oxygen Asystematireviewofoxygen and air flow on the relief of dyspnoea at rest in patients with advaned disease of any ause found low-grade sientifi evidene that oxygen and airflow improve dyspnoea in some patients with advaned disease at rest. 134 This systemati review ould only find evidene involving a total of 83 patients and most were hypoxaemi and already reeiving oxygen therapy. Based on the existing evidene, it is likely that aner patients with signifiant hypoxaemia may have some relief from breathlessness if given oxygen, but there is no evidene for any benefit in patients who are breathless but not hypoxaemi,andthereisevidenethatopiatesareeffetive in palliating breathlessness in this group of patients. In addition to speifi management of the ausative fator, oxygen should be given to maintain a saturation of 94 98% exept for patients with o-existing COPD who should be treated in aordane with the COPD setion of this guideline. Monitoring of oxygen saturation is not neessary when the patientisinthelastfewdaysoflife. Reommendations (see table 2) In breathlessness due to lung aner, oxygen therapy may be benefiial and a trial of oxygen therapy is reommended. Aim at an oxygen saturation of 94 98% unless there is oexisting COPD. However, monitoring of oxygen saturation is not neessary when the patient is in the last few days of life. [Grade D] vi37

38 Deterioration of fibroti lung onditions and other onditions involving parenhymal lung disease or alveolitis It is reognised that patients with fibrosing lung onditions suh as idiopathi pulmonary fibrosis may have aute deteriorations or exaerbations, often during interurrent hest infetions. Other patients may present autely with breathlessness due to extrinsi allergi alveolitis, saroidosis or other types of parenhymal lung disorders. These patients often have a high degree of V/Q mismath and a requirement for high oxygen onentrations to ahieve satisfatory blood gases and they are not at risk of hyperapnia. It is reommended that treatment is started with 60% oxygen from a Venturi mask or 6 l/min via nasal annulae if the patient an tolerate a high nasal flow rate. The oxygen level should be adjusted upwards or downwards to maintain an oxygen saturation in the range of 94 98%, but this level may not be ahievable or only ahievable with a reservoir mask. Patients with end-stage pulmonary fibrosis are rarely suitable for invasive or non-invasive ventilation beause of the progressive nature of the ondition. Reommendation (see table 2) In aute deterioration of pulmonary fibrosis or other parenhymal lung diseases, aim at an oxygen saturation of 94 98% or the highest possible if these targets annot be ahieved. [Grade D] Pneumothorax As with pleural effusions, patients with a large pneumothorax may be breathless and hypoxaemi and may require supplementary oxygen for symptom relief pending definitive treatment by aspiration or drainage. However, high onentration inhaled oxygen an also inrease the rate of reabsorption of air from a pneumothorax up to fourfold. 74 For this reason, the BTS guideline on the management of pneumothorax reommends the use of high onentration oxygen (reservoir mask) in all non-copd patients who require hospital admission for observation due to a moderate-sized pneumothorax that does not require drainage. 75 One a pneumothorax is drained or aspirated suessfully, the patient should not require oxygen therapy unless there is additional pathology suh as pneumonia, asthma or COPD requiring speifi treatment. Reommendations (see table 2) In most ases of pneumothorax, aim at an oxygen saturation of 94 98% if the patient is at risk of hyperapni respiratory failure. [Grade D] In patients having hospital observation without drainage, the use of high onentration oxygen (15 l/min flow rate via reservoir mask) is reommended. [Grade C] Pleural effusion If a pleural effusion is ausing signifiant breathlessness, the most effetive treatment is to drain the effusion (but not too quikly in view of the risk of re-expansion pulmonary oedema). Hypoxaemi patients with pleural effusions are likely to benefit from supplementary oxygen therapy. The BTS guidelines for management of pleural effusions do not give any speifi advie onerning oxygen therapy, but it seems reasonable to give supplementary oxygen to hypoxaemi patients to maintain a saturation of 94 98%. Reommendation (see table 2) In pleural effusion, aim at an oxygen saturation of 94 98% (or 88 92% if the patient is at risk of hyperapni respiratory failure). [Grade D] Pulmonary embolism Most patients with suspeted pulmonary embolism have normal oxygen saturation and the main fous of treatment is to reah a speifi diagnosis and to ommene antioagulant treatment. These patients do not require oxygen therapy unless there is hypoxaemia. In these ases, the lowest dose of oxygen that will ahieve a target saturation of 94 98% is reommended. However, patients with massive or multiple pulmonary embolism may be profoundly hypoxaemi and should initially be given high onentration oxygen via a reservoir mask to ahieve an oxygen saturation of 94 98% pending definitive treatment suh as thrombolysis. It has been suggested that the blood oxygen saturation may underestimate the severity of pulmonary artery obstrution in aute pulmonary embolism if shok is present. 220 Reommendation (see table 2) In pulmonary embolism, aim at an oxygen saturation of 94 98% or 88 92% if the patient is at risk of hyperapni respiratory failure. [Grade D] Aute heart failure Most patients with aute heart failure are breathless, usually due to pulmonary oedema or low ardia output, espeially if ardiogeni shok is present. The pathophysiology of oxygen transport in ardiogeni shok has been disussed in detail by Creamer and olleagues. 221 It has been shown in an animal model that the ventilatory failure of ardiogeni shok may be due to an impairment of the ontratile proess of the respiratory musles. 222 In addition to speifi treatment for heart failure, patients should be given supplementary oxygen to maintain a saturation of 94 98%. This is onsistent with the European Soiety of Cardiology Task Fore and European Soiety of Intensive Care reommendation that patients with aute heart failure should reeive oxygen to maintain SpO 2 of 92 96%. 223 It is reasonable to initiate treatment with 40% or 60% oxygen for hypoxaemi patients with heart failure, followed by upward or downward adjustment to maintain saturation in the desired range. Patients with marked hypoxaemia (saturation,85%) should be treated with a reservoir mask initially and patients with o-existing COPD will require a lower target saturation of 88 92% pending the availability of blood gas results. In hospital settings, patients with aute pulmonary oedema may benefit from ontinuous positive airway pressure and from non-invasive ventilatory support Reommendations (see table 2) In aute heart failure, aim at an oxygen saturation of 94 98% or 88 92% if the patient is at risk of hyperapni respiratory failure. [Grade D] Consider treatment with ontinuous positive airway pressure if there is hypoxaemia and treatment with non-invasive ventilation (BiPAP) if there is o-existent hyperapnia. [Grade C] vi38

39 Postoperative breathlessness or hypoxaemia on general surgial wards These guidelines do not over immediate postoperative are in post-anaestheti reovery units, high dependeny units or intensive are units (ICUs). Some reent trials have shown a redued inidene of wound infetion when high-dose oxygen was given perioperatively to patients having bowel surgery but not general surgery This planned use of oxygen postoperatively is also outside the sope of this guideline. There is some ontroversy about the use of routine supplemental oxygen postoperatively and no good evidene supporting suh a poliy. The SIGN guideline on postoperative are reommends supplemental oxygen therapy for ertain high-risk groups suh as those with oronary artery disease, obesity, thorai and upper abdominal surgery, but aknowledges lak of evidene to support these suggestions and does not speify an oxygen dose or target saturation for suh patients. 229 This SIGN guideline reommends maintaining an oxygen saturation above 92% for postoperative patients, whih fits well with the suggested target saturation in the present guideline of 94 98% for most patients who require supplementary oxygen therapy. Patients on general surgial wards an develop sudden breathlessness or hypoxaemia due to a variety of postoperative ompliations suh as pneumonia, pulmonary embolism, opiate analgesia and ateletasis. The use of oxygen for speifi postoperative ompliations suh as pneumonia should follow the guidane for eah ondition (for most patients the target will be 94 98%). Speial are must be taken in ases of COPD and other risk fators for hyperapni respiratory failure. Management of these ases an be enhaned by early speialist referral or the input of expert assistane from ICU Outreah Teams. These ases should be identified as being at risk during preoperative assessment and a target saturation of 88 92% is suggested pending the availability of blood gas results. Reommendations (see table 2) For postoperative surgial patients, aim at a saturation of 94 98% or 88 92% if at risk of hyperapni respiratory failure. [Grade D] For postoperative surgial patients with COPD or other risk fators for hyperapni respiratory failure, aim at a saturation of 88 92% pending results of blood gas analysis. If the PaCO 2 is normal, adjust target range to 94 98% and repeat blood gas measurements after min (see table 3 and hart 1 (fig 1)) Breathlessness due to severe anaemia If breathlessness is due to severe anaemia, the speifi treatment is blood transfusion. Studies by Canadian researhers in the late 1990s have shown that haemoglobin levels of 70 g/l (7 g/dl) were as safe as higher levels and may produe fewer ompliations in the ritially ill. 55 However, this study was onduted using non-leuoyte-depleted blood and it is possible that some of the infetive ompliations in the group who were given more transfusions might have been avoided by the use of leuoyte-depleted blood. The optimal transfusion target for ritially ill patients therefore remains the subjet of ongoing disussion among experts in ritial are mediine (setion 5.6.2). Giving oxygen to inrease an already normal oxygen saturation will have very little effet on the oxygen-arrying power of the blood, but it is reasonable to administer supplemental oxygen to maintain a saturation of 94 98% (if the saturation is below these levels breathing air of if breathlessness is a very prominent symptom). Reommendations (see table 2) In anaemia, aim at an oxygen saturation of 94 98% or 88 92% if the patient is at risk of hyperapni respiratory failure. [Grade D] Give paked red ells if the haemoglobin level falls below g/l (7 8 g/dl) in most ases or 100 g/l (10 g/dl) if the patient has unstable or symptomati ishaemi heart disease. [Grade B] Sikle ell risis Patients with sikle ell disease frequently present with an aute painful risis and less frequently with an aute hest syndrome omprising breathlessness, hest pain and fever with pulmonary infiltrates on the hest radiograph. The exat auses and mehanisms are not well understood, but oxygen should be given to all hypoxaemi patients with sikle ell risis to avoid further intravasular sikling. There are no randomised studies of oxygen therapy in aute hest syndrome and no randomised studies of aute painful risis in adults, but two small randomised trials showed no linial benefit in non hypoxaemi hildren with aute painful risis. Patients with sikle ell disease may have a redued oxygen saturation even when linially stable. Homi and olleagues reported a mean saturaton of only 92.5% (95% CI 92.0% to 93.0%) in a group of hildren and young adults (age 9 18 years) with stable sikle ell disease ompared with an average saturation of 97.1% (95% CI 98.8% to 97.3%) in a loal ontrol group. 233 The British Committee for Standards in Haematology have reommended that oxygen should be given if the oxygen saturation falls below what is normal for the individual patient or a default target of 95% if the usual saturation is unknown. 234 This is onsistent with the advie in the present guideline to aim at a normal or near-normal oxygen saturation for non-hypoxaemi patients with a target saturation of 94 98%. Readers are referred to the guideline on sikle ell disease for disease-speifi management of this ondition. 234 Reommendation (see table 2) In sikle ell risis and aute hest syndrome, aim for an oxygen saturation of 94 98% or aim at the saturation level that is usual for the individual patient. [Grade B] 8.12 Reommended oxygen therapy for patients who may be vulnerable to medium or high doses of oxygen (see also table 3) COPD is the best known ondition that an predispose to hyperapni (type 2) respiratory failure with aidosis, espeially if the blood oxygen level is inreased above 10 kpa (75 mm Hg). However, there are a number of other onditions whih an render patients vulnerable to hyperapni respiratory failure. The emphasis for suh patients is to avoid linially harmful levels of hypoxaemia or hyperapnia by giving arefully titrated oxygen therapy or, if neessary, by supporting the patient with the use of non-invasive or invasive mehanial ventilation. Non-COPD patients at risk of hyperapni respiratory failure inlude the following: Cysti fibrosis. Non-CF bronhietasis (often in assoiation with COPD or severe asthma) Severe kyphosoliosis or severe ankylosing spondylitis. vi39

40 Severe lung sarring from old tuberulosis (espeially with thoraoplasty). Morbid obesity (body mass index.40 kg/m 2 ). Musuloskeletal disorders with respiratory musle weakness, espeially if on home ventilation. Overdose of opiates, benzodiazepines or other respiratory depressant drugs COPD exaerbations There is an extensive literature doumenting the effets of highdose oxygen therapy in aute COPD These reports show that the administration of supplemental oxygen to patients with exaerbated COPD often auses a rise in PaCO 2 with subsequent respiratory aidosis for reasons summarised in setions 5.3, 5.4 and The literature is summarised in detail in the review by Murphy et al. 10 Some patients with COPD are prone to repeated episodes of hyperapni respiratory failure and others may not ever suffer from this ompliation. Even among patients with COPD with hroni hyperapnia, not all will develop an inreased arbon dioxide level (and aidosis) during aute exaerbations. Apart from patients with reurrent hyperapni respiratory failure, it is not possible to predit if individual patients with COPD will develop hyperapnia during an aute exaerbation, so all patients with moderate or severe COPD should be onsidered to be at risk of this ompliation until the results of blood gas measurements are available. It is therefore essential that patients who are at risk of having COPD should be diagnosed aurately, and this an only be done by measurement of FEV Patients with aute severe exaerbations of COPD may be too breathless to undertake spirometry on arrival in hospital, but many patients are able to perform spirometry on arrival in hospital and all patients should have the test performed before disharge from hospital to onfirm the diagnosis of COPD and to assess the severity of the ondition. There is very little literature desribing the effets of oxygen therapy in the other onditions listed above, but they are reognised to be at risk of hyperapni respiratory failure and should be treated in a manner analogous to patients with COPD. It has been shown that patients with COPD with a ph reading,7.35 ([H + ].45 nmol/l) despite ontrolled oxygen therapy are more likely to die and more likely to meet riteria for intubation and ventilation One of these reports also showed that patients with a high PaO 2 on arrival in hospital (.10.0 kpa or 75 mm Hg) were more likely to meet riteria for ventilation and the severity of aidosis was related to high PaO 2 values. 34 Based on these results, Plant and olleagues reommended an upper limit of about 92% saturation for patients with exaerbations of COPD to prevent the PaO 2 rising above 10 kpa. 34 This report was supported by the reent work of Joosten et al whih showed that a PaO 2 of.74.5 mm Hg (10 kpa) in aute COPD was assoiated with an inreased likelihood of admission to a high dependeny unit, inreased need for non-invasive ventilation and a longer stay in hospital. 235 Consequently, the guideline group has reommended a maximum saturation of 92% while awaiting blood gas results in aute exaerbations of COPD and other onditions that may predispose to type 2 respiratory failure. Although the rise in PaCO 2 (and fall in ph) is greatest in patients who are given suffiient oxygen therapy to elevate the PaO 2 above 10 kpa, it is important to note that hyperapnia an our in aute COPD even if the oxygen saturation is,88%. 249 The best management strategy for persistently aidoti COPD patients is a trial of non-invasive ventilation with supplementary oxygen therapy. Some patients with previous hyperapni respiratory failure will have alert ards or an entry in their eletroni reord to alert the emergeny team to the optimal dose of oxygen required during the patient s previous hospital admissions (see setion 9.7). In the absene of suh information, it is suggested that a target of 88 92% should be set initially for patients with a history of previous non-invasive or invasive ventilation and, if neessary, modified later based on blood gas results. These patients should be ategorised as very urgent by ambulane teams and emergeny servies, requiring immediate blood gas measurement and senior assessment on arrival at the hospital emergeny department. Unfortunately, many linial studies have shown that patients with COPD are frequently given very high doses of oxygen, either beause of misdiagnosis or beause the risks of hyperoxia in patients with COPD have been overlooked. Many patients with COPD are unaware of the diagnosis or are mislabelled as having asthma (see setion 9.5). The onsensus from the literature is that patients with aute exaerbations of COPD should be treated with Venturi masks to minimise the risks of hyperapni respiratory failure and to ahieve a high gas flow from the mask in patients with a high inspiratory flow rate. 10 It is not yet known if it is better to start with a 28% Venturi mask or a 24% Venturi mask. Management with a 28% Venturi mask appears to be safe. 252 The urrent guideline reommends starting with a 28% Venturi mask in ases of COPD with no known history of hyperapni respiratory failure, with downward adjustment to a 24% mask (in hospital) if the saturation rises above 92%. In ases of prior hyperapni failure who do not have an oxygen alert ard, it is reommended that prehospital treatment should be ommened using a 28% Venturi mask at 4 l/min or a 24% Venturi mask in hospitals with a target saturation of 88 92%. Observational studies in the 1960s suggested that a PaO 2 of 50 mm Hg or 6.7 kpa (saturation about 84%) will prevent death from hypoxaemia in aute COPD exaerbations If the saturation should fall below 88% despite treatment with a 24% or 28% Venturi mask, the patient should be treated with nasal annulae or a simple fae mask with the flow adjusted to maintain a saturation of 88 92% pending the availability of blood gas results. This small subgroup of patients is at very high risk of death and should be treated as a high priority on arrival in emergeny departments, requiring immediate senior assessment and arterial blood gas measurements. Measurement of FEV 1 may onfirm (or exlude) a diagnosis of airflow obstrution and the FEV 1 level is a useful indiator of disease severity in COPD. 25 [Evidene level III] Patients with exaerbations of COPD are at risk of hyperapni (type 2) respiratory failure with respiratory aidosis. [Evidene level IIa] The risk of respiratory aidosis in patients with hyperapni respiratory failure is inreased if the arterial oxygen tension is above 10.0 kpa due to previous exessive oxygen use. [Evidene level IIa] These patients with hroni lung disease are usually alimatised to living with an oxygen saturation whih may be in the high 80s or low 90s and there is not likely to be any benefit from inreasing the saturation above these levels during aute illness. [Evidene level III] vi40

41 Reommendations (see table 3) If the diagnosis is unknown, patients over 50 years of age who are long-term smokers with a history of hroni breathlessness on minor exertion suh as walking on level ground and no other known ause of breathlessness should be treated as if having COPD for the purposes of this guideline. Patients with COPD may also use terms suh as hroni bronhitis and emphysema to desribe their ondition but may sometimes mistakenly use asthma. FEV 1 should be measured on arrival in hospital if possible and should be measured at least one before disharge from hospital in all ases of suspeted COPD. [Grade D] Patients with a signifiant likelihood of severe COPD or other illness that may ause hyperapni respiratory failure should be triaged as very urgent on arrival in hospital emergeny departments and blood gases should be measured on arrival in hospital. [Grade D] Prior to availability of blood gas measurements, use a 28% Venturi mask at 4 l/min or 24% Venturi mask at 2 l/min and aim for an oxygen saturation of 88 92% for patients with risk fators for hyperapnia but no prior history of type 2 respiratory failure. [Grade D] For patients with known previous hyperapi respiratory failure but no oxygen alert ard, aim at a saturation of 88 92% until the results of blood gas measurements are available (see reommendation 5). If the saturation remains below 88% in prehospital are despite a 28% Venturi mask, hange to nasal annulae at 2 6 l/min or a simple fae mask at 5 l/min with target saturation of 88 92% and alert the A&E department that the patient is to be treated as a high priority. [Grade D] Patients with a respiratory rate.30 breaths/min should have the flow rate set to 50% above the minimum flow rate speified for the Venturi mask and/or pakaging. Inreasing the oxygen flow rate into a Venturi mask does not inrease the onentration of oxygen whih is delivered (see reommendation 32). Aim at a prespeified target saturation range (if available) in patients with a history of previous respiratory aidosis. In many ases the ideal target saturation will be speified on the patient s alert ard. If no information is available, aim at a saturation level of 88 92% pending blood gas results. [Grade D] Patients with prevous hyperapni respiratory failure should have a personalised oxygen alert ard and this information should be available to primary are staff, ambulane staff and hospital staff (see reommendations 23 25). If following blood gas measurements the ph and PCO 2 are normal, aim for an oxygen saturation of 94 98% unless there is a history of previous hyperapni respiratory failure requiring non-invasive ventilation or intermittent positive pressure ventilation. [Grade D] Rehek blood gases after min (or if there is evidene of linial deterioration) for all patients with COPD or other risk fators for hyperapni respiratory failure even if the initial PaCO 2 measurement was normal. [Grade D] If the PaCO 2 is raised but ph is >7.35 ([H + ] (45 nmol/l), the patient has probably got long-standing hyperapnia; maintain target range of 88 92% for these patients. Blood gas measurements should be repeated at min to hek for rising PaCO 2 or falling ph. [Grade D] If the patient is hyperapni (PaCO 2.6 kpa or 45 mm Hg) and aidoti (ph,7.35 or [H + ].45 nmol/l), onsider noninvasive ventilation, espeially if the aidosis has persisted for more than 30 min despite appropriate therapy. [Grade A] One patients have stabilised, onsider hanging from Venturi mask to nasal annulae at 1 2 l/min (see reommendation 31) Exaerbation of ysti fibrosis Patients with breathlessness due to ysti fibrosis should be managed in a Cysti Fibrosis Centre unless this is not possible for geographial reasons. If not possible, all ases should be disussed with the Cysti Fibrosis Centre or managed aording to a protool that has been agreed with the regional entre. Patients with advaned ysti fibrosis may suffer from exaerbations whih are similar to exaerbations of advaned COPD with assoiated hypoxaemia and hyperapnia. The priniples of management are similar to those in aute exaerbations of COPD, inluding a need to maintain adequate oxygen saturation and avoiding exessive hyperapnia and aidosis. As in COPD, non-invasive ventilation may be of value in severe ases. 253 Non-invasive ventilation in ysti fibrosis may also be helpful to redue symptoms (eg, work of breathing and dyspnoea) and assist in airway learane. It is reommended that patients with aute exaerbations of ysti fibrosis should be managed on similar lines to patients with aute exaerbations of COPD with a target oxygen saturation of 88 92% for most patients, but reognition that individual patients may need to be managed differently on the basis of previous and urrent blood gas measurements. One study has shown that patients with a respiratory rate above 30 breaths/min often have an inspiratory flow rate above the minimum flow rate speified on the mask pakaging. 254 However, there is no diret experimental evidene of the linial effetiveness of inreased flow rates from Venturi devies. It is possible that patients with very high inspiratory flow rates might benefit from a 28% Venturi mask with the flow rate set at 6 8 l/min to minimise the risk of the inspiratory flow rate exeeding the gas flow rate (see table 10 in setion 10). Patients with ysti fibrosis who have had previous episodes of hyperapni respiratory failure should be issued with an oxygen alert ard with reommendations based on previous blood gas measurements (see reommendations 23 25). Reommendation (see table 3) Initial treatment of ysti fibrosis exaerbations should be similar to the initial treatment of COPD exaerbations (see setion ). [Grade D] Chroni musuloskeletal and neurologial disorders Hypoxaemia due to musuloskeletal and neurologial disorders is usually assoiated with aute illness (suh as a hest infetion) superimposed on a hroni neuromusular ondition. However, musle weakness an be aute or subaute (eg, Guillain-Barré syndrome, see setion ). For most patients with inadequate ventilation due to neuromusular weakness, non-invasive or invasive ventilatory support is more useful than supplementary oxygen and these patients are at risk of hyperapni respiratory failure whih may be aggravated by high doses of oxygen. For this reason it is reommended that spirometry should be monitored arefully and blood gases should be obtained as early as possible in all suh ases. Pending the availability of blood gas results, a saturation target of 88 92% will avoid the risks of severe hypoxaemia or severe hyperapnia. vi41

42 Reommendation (see table 3) In the initial management of musuloskeletal and neurologial disorders with aute respiratory failure, aim at an oxygen saturation of 88 92%. Many suh patients will be suitable for non-invasive ventilation. [Grade D] Obesity-hypoventilation syndrome Patients with the obesity-hypoventilation syndrome often develop hroni hyperapni respiratory failure and they may deompensate autely to produe hyperapni respiratory failure with aidosis. 255 For purposes of oxygen therapy, these patients should be treated in a similar manner to patients with hyperapni respiratory failure due to an aute exaerbation of COPD (but they learly do not require bronhodilator and steroid therapy). The initial target saturation will usually be 88 92% but, as with COPD, a lower target range may be appropriate for individual patients based on blood gas measurements during a previous exaerbation or due to aute aidosis. Assessment of patients with inreasing shortness of breath or worsening oxygen saturation must inlude blood gases. As in COPD, patients with respiratory aidosis may benefit from noninvasive ventilation. Reommendations (see table 3) In the initial management of the obesity-hypoventilation syndrome with aute exaerbation, aim at an oxygen saturation of 88 92%. [Grade D] Non-invasive ventilation should be onsidered for all of the above groups of patients if the ph is,7.35 or [H + ].45 nmol/l. [Grade C] 8.13 Common medial emergenies for whih oxygen therapy is indiated only if hypoxaemia is present (see also table 4) There are a number of onditions suh as myoardial infartion, angina and stroke for whih oxygen was traditionally given to all patients in an attempt to inrease oxygen delivery to the heart or brain. However, the administration of supplemental oxygen to normoxaemi patients has very little effet on blood oxygen ontent but may redue myoardial and erebral blood flow due to vasoonstrition whih is a physiologial response to hypoxia in most organs. There is no evidene of benefit from the administration of supplemental oxygen to non-hypoxaemi patients with these onditions and there is some evidene of possible harm, so it is reommended that oxygen should only be given to patients with these onditions if hypoxaemia is present, usually due to ompliations suh as heart failure or pneumonia. There are no published trials supporting the use of oxygen to relieve breathlessness in non-hypoxaemi patients, and there is evidene from randomised studies that oxygen does not relieve breathlessness ompared with air in non-hyoxaemi COPD patients who are breathless following exertion Aute myoardial infartion, suspeted myoardial infartion and aute oronary syndromes Some patients with aute myoardial infartion have heart failure and should be treated aordingly (see setion ). Most patients with suspeted or onfirmed myoardial infartion are not hypoxaemi and most are not breathless. In the ase of non-hypoxaemi patients, it is not known if supplementary oxygen may be benefiial by inreasing the amount of oxygen delivered to the hypoxaemi area of myoardium or whether it may atually ause vasoonstrition with inreased systemi vasular resistane and redued myoardial oxygen supply with worsened systoli myoardial performane A reent study of patients having oronary arteriography found that breathing 100% oxygen redued oronary blood flow veloity by 20% and inreased oronary resistane by 23%. 265 There is also a theoretial possibility that high oxygen levels might exaerbate reperfusion injury to the heart. 263 Despite a multitude of large studies of intervention in myoardial infartion, there has been only one randomised study of oxygen therapy (in 1976) and this study did not identify any benefit from suh therapy but found some evidene of potential harm. 107 This trial reported a signifiantly greater rise in myoardial enzyme in the oxygen group, suggesting a greater infart size. There was a threefold inrease in mortality on oxygen therapy that did not reah statistial signifiane (3 deaths in 77 patients treated with air versus 9 deaths in 80 patients given oxygen at 6 l/min via a simple fae mask for 24 h). A systemati review and a historial review of oxygen therapy in aute myoardial ishaemia have both onluded that there was no evidene to support this pratie in non-hypoxaemi patients and some evidene of possible harm One study from 1969 showed that hypoxia did not affet the availability of oxygen for myoardial metabolism in normal subjets until the oxygen saturation fell to about 50%, but evidene of myoardial ishaemia was seen at saturations of 70 85% in subjets with oronary artery disease. 268 In these irumstanes it is advised that patients with myoardial infartion or hest pain suspiious of myoardial infartion should be given supplementary oxygen if required to maintain a saturation of 94 98%. The study by Lal and olleagues 269 in the 1960s showed that hypoxaemia was present in a high proportion of patients diagnosed with myoardial infartion and ould usually be reversed by medium-dose oxygen, but sometimes required treatment with a reservoir mask to ahieve a PaO 2 oxygen tension.60 mm Hg (8 kpa). The study by Wilson and Channer 2 in 1997 showed that desaturation below 90% was ommon in patients with myoardial infartion within the first 24 h of admission to a oronary are unit, but these authors may not have been aware that noturnal desaturation to this level is very ommon in healthy individuals. 26 Wilson and Channer did not demonstrate any orrelation between hypoxaemi events and adverse ardia events. 2 They did, however, show that monitoring by oximetry was inadequate in UK oronary are units in the mid 1990s. There are no UK guidelines for oxygen therapy in aute myoardial infartion. The 1998 European Soiety for Cardiology/European Resusitation Counil Task Fore reommended the use of 3 5 l/min oxygen via fae mask to all patients with hest pain of presumed ardia origin, but no evidene was presented to support this advie. 270 However, most of the papers that have raised onerns about the effets of oxygen on myoardial blood flow have been published sine that date (see preeding paragraph). The European Soiety of Cardiology published subsequent guidane on the management of ST elevation myoardial infartion in This revised guidane reommended the use of oxygen at 2 4 l/min by mask or nasal annulae for patients with heart attaks assoiated with breathlessness or heart failure. The 2007 SIGN guideline for aute oronary syndromes states that there is no evidene that routine administration of oxygen to all patients with aute oronary syndromes improves linial outome or redues infartion size. 272 The SIGN guideline gives a grade D reommendation that oxygen should be administered to patients with hypoxaemia, pulmonary oedema or ontinuing myoardial ishaemia. vi42

43 The European Resusitation Counil Guidelines for the management of aute oronary syndromes in 2005 reommended the use of supplementary oxygen at 4 8 l/min (devie not speified) for patients with arterial oxygen saturation,90% and/or pulmonary ongestion. 273 The guideline aknowledged the lak of evidene of benefit for non-hypoxaemi patients but reommended supplementary oxygen in ase of unreognised hypoxaemia. This situation might apply in the prehospital setting but not in the hospital setting. The limited available evidene therefore supports the suggestion that liniians should aim at normal or near-normal oxygen saturation in patients with myoardial infartion, aute oronary syndrome and hest pain suspiious of oronary artery disease. A target saturation range of 94 98% will meet all of these goals, and further researh of this topi should be prioritised beause this is suh a ommon medial problem and there is so little existing evidene. Most 999 alls to ambulane servies beause of hest pain are urrently treated with high onentration oxygen in aordane with the Joint Royal Colleges Ambulane Liaison Committee (JRCALC) guidane. 274 However, most suh patients have a final diagnosis of undifferentiated hest pain rather than aute oronary artery syndrome and most patients with undifferentiated hest pain are normoxaemi. The linial management of a very large number of patients will therefore be hanged following the introdution of this guideline. Reommendation (see table 4) In myoardial infartion and aute oronary syndromes, aim at an oxygen saturation of 94 98% or 88 92% if the patient is at risk of hyperapni respiratory failure. [Grade D] Stroke In the past it was ustomary to give supplementary oxygen to all patients with stroke to try to improve erebral oxygenation. However, there has been only one randomised trial of oxygen therapy in stroke. 108 This trial found no differene in 1-year survival for the entire ohort of patients with stroke and no differene in survival for patients with more severe strokes. However, for patients with minor or moderate strokes, 1-year mortality was 18% in the group given oxygen and 9% in the group given air (OR 0.45; 95% CI 0.23 to 0.90, p = 0.023). Based largely on the results of this trial, the Royal College of Physiians stroke guideline reommends that oxygen saturation should be maintained in the normal range in patients with stroke. 275 It is reommended that patients with stroke should reeive supplementary oxygen only if this treatment is required to ahieve an oxygen saturation of 94 98% (88 92% for patients with o-existing risk of COPD or other risk of respiratory aidosis). There has also been some disussion onerning the optimal body position for the management of patients with stroke and potential hypoxaemia. A systemati review onluded that there was limited evidene that sitting in a hair had a benefiial effet and lying positions had a deleterious effet on oxygen saturation in patients with aute stroke with respiratory omorbidities, but patients with aute stroke without respiratory o-morbidities an adopt any body position. 136 The authors of this review reommended that people with aute stroke and respiratory o-morbidities should be positioned as upright as possible. Reommendation (see table 4) In stroke, aim at an oxygen saturation of 94 98% or 88 92% if the patient is at risk of hyperapni respiratory failure. [Grade B] Obstetri emergenies and labour The use of oxygen has been reommended during many obstetri emergenies and, in partiular, for ollapse related to haemorrhage, pulmonary embolism, elampsia or amnioti fluid embolism. Severe pre-elampsia and elampsia may oasionally present with pulmonary oedema and this an our in the antenatal or postnatal periods. Medial problems suh as pneumonia or aute exaerbations of asthma are not unommon during pregnany. Peripartum ardiomyopathy is rare but may present with heart failure in the postnatal period. Major trauma is inreasingly ommon, partiularly related to road traffi aidents. The use of oxygen during pregnany should follow the same general priniples as the use of oxygen for other patients. Pregnant women suffering major trauma or severe hypoxaemia should be started on high onentration oxygen via a nonrebreathing reservoir mask and those with milder hypoxaemia an use nasal annulae or a simple fae mask or Venturi mask to ahieve an oxygen saturation of 94 98% in most ases. If an undelivered woman is hypoxaemi, she should be managed with left lateral tilt applied. This will improve ardia output 276 and may also failitate breathing for mehanial reasons. Oxygen is ommonly given as part of the treatment for many obstetri emergenies. However, it is reommended that, when oxygen is administered during pregnany or labour, liniians should aim to ahieve normoxaemia (saturation 94 98%). There is no randomised trial evidene to suggest that maternal hyperoxaemia is benefiial to mother or fetus. Oxygen is often given when aute fetal ompromise is suspeted in labour in the hope of inreasing oxygen delivery to the fetus. A Cohrane review found no trials addressing the use of oxygen for fetal ompromise. However, two trials of prophylati oxygen in labour found a signifiant inrease in the inidene of ord blood aidosis (ph,7.20 or [H + ].63 nmol/l) in the oxygenation group (RR 3.5 (95% CI 1.34 to 9.19)). 277 It is reommended that pregnant women with evidene of hypoxaemia should have their blood oxygen saturation maintained in the normal range (94 98%) using supplemental oxygen as neessary to ahieve this effet. This applies before or during labour as well as in the postnatal period. The auses of maternal hypoxaemia may inlude trauma, pre-existing or de novo medial onditions as well as pregnany-speifi ompliations. In all of these situations the aim should be normoxaemia (saturation 94 98%). Reommendations 14. Women who suffer from major trauma, sepsis or aute illness during pregnany should reeive the same oxygen therapy as any other seriously ill patients, with a target oxygen saturation of 94 98%. The same target range should be applied to women with hypoxaemia due to aute ompliations of pregnany (eg, ollapse related to amnioti fluid embolus, elampsia or antepartum or postpartum haemorrhage). [Grade D] 15. Women with underlying hypoxaemi onditions (eg, heart failure) should be given supplemental vi43

44 oxygen during labour to ahieve an oxygen saturation of 94 98%. [Grade D] 16. All women with evidene of hypoxaemia who are more than 20 weeks pregnant should be managed with left lateral tilt to improve ardia output. [Grade B] 17. The use of oxygen during labour is widespread but there is evidene that this may be harmful to the fetus. The use of oxygen during labour is therefore not urrently reommended in situations where the mother is not hypoxaemi (exept as part of a ontrolled trial). [Grade A] Anxiety and hyperventilation or dysfuntional breathing Many patients who present to hospital with breathlessness are found to have no ardiopulmonary problems and many suh patients have a speifi diagnosis of hyperventilation, dysfuntional breathing, upper airway dysfuntion or pani attaks, sometimes in addition to asthma or some other underlying respiratory disorder. 278 Many suh patients will have an abnormally high oxygen saturation of 99% or 100% and learly do not require supplemental oxygen therapy. Many other nonhypoxaemi patients will present to hospital with aute breathlessness of unknown ause, and the majority of patients with an elevated respiratory rate are likely to have an organi illness. In some ases simple investigations will reveal a speifi diagnosis suh as pneumothorax or pneumonia or pulmonary embolism, but many ases remain undiagnosed. A poliy of giving supplementary oxygen if the saturation falls below 94% will avoid exposing patients with undiagnosed medial illnesses to the risk of hypoxaemia while avoiding the unneessary use of oxygen in patients with behavioural or dysfuntional breathlessness. Studies in normal volunteers have demonstrated that ompensatory desaturation may our shortly after voluntary hyperventilation. 279 The mean PaO 2 of 10 male volunteers inreased from 13.7 kpa (103 mm Hg) to 18.6 kpa (140 mm Hg) during hyperventilation but fell to a nadir of 7.8 kpa (58 mm Hg) about 7 min after essation of hyperventilation and did not normalise until after a total of 17 min of observation. It is not known whether or not this ours after pathologial hyperventilation, but this phenomenon ould ause onsiderable onfusion if it should our in an emergeny department. A traditional treatment for hyperventilation was to ask the subjet to rebreathe from a paper bag to allow the arbon dioxide level in the blood to normalise. However, it has been shown that this pratie an ause hypoxaemia with potentially fatal onsequenes. 280 The average fall in oxygen tension during rebreathing was 26 mm Hg (3.5 kpa) and the maximum fall was 42 mm Hg (5.6 kpa). This guideline does not reommend rebreathing from a paper bag in ases of hyperventilation unless the patient has been shown to have hyperoxia and a low arbon dioxide level, and any suh treatment should be monitored with ontinuous oximetry and disontinued if the patient should desaturate. Reommendations (see table 4) Organi illness must be exluded before making a diagnosis of hyperventilation. [Grade C] Patients with a definite diagnosis of hyperventilation should have their oxygen saturation monitored. Those with normal or high SpO 2 do not require oxygen therapy. [Grade B] Rebreathing from a paper bag an be dangerous and is NOT reommended as a treatment for hyperventilation. [Grade C] Poisoning with substanes other than arbon monoxide Many poisons and drugs an ause respiratory or ardia depression or diret toxi effets on the lungs. The treatment of individual toxi agents is beyond the sope of this guideline. Speifi antidotes suh as naloxone should be given if available and oxygen saturation should be monitored losely. Supplementary oxygen should be given to ahieve a target saturation of 94 98% pending the results of blood gas analysis (88 92% if at risk of hyperapni respiratory failure). All potentially serious ases of poisoning should be monitored in a level 2 or level 3 environment (high dependeny unit or intensive are unit). Three speifi types of lung injury deserve speial mention. Oxygen is known to be hazardous to patients with paraquat poisoning, and oxygen potentiates bleomyin lung injury. Beause of these risks, oxygen should be given to patients with these onditions only if the oxygen saturation falls below 90%. Some authors have suggested the use of hypoxi ventilation with 14% oxygen as a speifi treatment for paraquat poisoning. 281 Bleomyin lung injury an be potentiated by high-dose oxygen therapy, even if given several years after the initial lung injury. 115 It is therefore reommended that high doses of oxygen should be avoided in patients with possible bleomyin-indued lung injury and a lower oxygen saturation target range should be aepted (eg, 88 92%). There is evidene from animal experiments that oxygen may potentiate lung injury from aspiration of aids The effet in humans is not known so patients with aid inhalaton should have the usual adult target saturation range of 94 98%, but it would appear prudent to aim in the lower half of the target range for these patients and linial trials in humans are learly required. Reommendations (see table 4) In most poisonings, aim at an oxygen saturation of 94 98%. [Grade D] In poisoning by paraquat and bleomyin, aim at a saturation of 88 92%. [Grade D] Metaboli, endorine and renal disorders Many metaboli and renal disorders an ause metaboli aidosis whih inreases respiratory drive as the body tries to orret the aidosis by inreased exretion of arbon dioxide via the lungs. Although these patients have tahypnoea, they do not usually omplain of breathlessness and most have a high oxygen saturation (unless there is a o-existing pulmonary or ardia problem). Supplementary oxygen is not required for suh patients unless the oxygen saturation is redued. In suh ases, oxygen should be given to maintain a saturation of 94 98%. Reommendation (see table 4) In most metaboli and renal disorders, aim at an oxygen saturation of 94 98%. [Grade D] vi44

45 Aute and subaute neuromusular disorders produing respiratory musle weakness Patients with aute or subaute onditons affeting the respiratory musles are at risk of sudden onset of respiratory failure with hypoxaemia and hyperapnia and may require noninvasive or invasive ventilatory support. This applies espeially to patients with Guillain-Barré syndrome for whom spirometry should be monitored arefully as this should detet the onset of severe respiratory failure prior to the development of hypoxaemia. If the oxygen level falls below the target saturation, urgent blood gas measurements should be undertaken and the patient is likely to need ventilatory support. SECTION 9: EMERGENCY USE OF OXYGEN IN AMBULANCES, COMMUNITY AND PREHOSPITAL SETTINGS This setion applies to a range of linial settings to inlude emergeny oxygen use in patients homes, GP praties or health entres and during emergeny ambulane journeys to hospital. Management in some prehospital settings suh as a Primary Care Centre or in a paramedi ambulane may be almost idential to hospital management. Readers are referred to setion 10 for advie onerning hoie of oxygen delivery devies and systems. Readers are referred to tables 1 4 and harts 1 and 2 (figs 1 and 2) for a summary of the key elements of oxygen therapy in ommon medial emergenies. A brief summary of this setion an be downloaded from Ongoing are at home of hronially hypoxaemi patients is not overed by this guideline. There is little literature on whih to base any reommendations when suh patients have an aute exaerbation of their ondition, but patient safety should be the priority. The NICE guidelines on COPD reommend that patients reeiving long-term oxygen and those with an arterial PO 2 of,7 kpa should be onsidered for treatment in hospital during exaerbations (reommendation 135 in NICE guideline 25 ). The BTS Emergeny Oxygen Guideline Group would add that hronially hypoxaemi patients with a linial exaerbation assoiated with a 3% or greater fall in oxygen saturation on their usual oxygen therapy should be assessed in hospital with blood gas estimations. Arterial PO 2 of,7 kpa equates to SpO 2 below approximately 85%. 9.1 Pulse oximetry and availability of oxygen It is essential to provide optimal oxygen therapy at the earliest possible opportunity while the autely breathless patient is being assessed and treated in the ommunity and during transfer to hospital. For most suh patients the main onern is to give suffiient oxygen to support their needs. Hypoxaemia an lead to ardia arrhythmias, renal damage and, ultimately, erebral damage. However, exessive oxygen therapy an also be dangerous for some patients, espeially those with advaned COPD. Target saturation should be used; pulse oximetry is neessary to ahieve this. Setion provides advie onerning the hoie of oxygen ylinders in primary are praties. Emergeny ambulanes and emergeny/fast response type vehiles and ambulane servie motorbikes and yles should be equipped with oxygen and oximeters germane to the mode of transport. Thus, fast response ars/motorbikes and yles will require handheld finger oximeter-type devies and staff initiating oxygen in the home will need a portable or finger oximeter. Community First Responder (CFR) shemes are enouraged to seek the opinion of the ambulane servie to whih they are affiliated to disuss the purhase and use of pulse oximeters. Likewise Voluntary Aid Soieties (VAS) medial diretors are enouraged to disuss the purhase and use of pulse oximieters. Reommendations 18. Pulse oximetry must be available in all loations where emergeny oxygen is being used (see also the limitations of using pulse oximetry setion 7.1.2). [Grade D] 19. Emergeny oxygen should be available in primary are medial entres, preferably using oxygen ylinders with integral high-flow regulators. Alternatively, oxygen ylinders fitted with highflow regulators (delivering over 6 l/min) must be used. [Grade D] 20. All douments whih reord oximetry measurements should state whether the patient is breathing air or a speified dose of supplemental oxygen. [Grade C] 9.2 Clinial assessment by initial responder(s) (GP, nurse or ambulane team) It is suggested that the first healthare professional(s) to enounter an autely breathless patient should perform an initial ABC assessment (airway, breathing, irulation), followed by obtaining a quik history from the patient and/or family or friends. Immediate assessment should inlude a reording of pulse rate and respiratory rate and pulse oximetry should be reorded. Clinial assessment of a breathless patient starts with ABC (airway, breathing, irulation) (see reommendation 7). A brief history should be taken from the patient or other informant. Initial assessment should inlude pulse and respiratory rate in all ases (see reommendation 7). Pulse oximetry should always be measured in patients with breathlessness or suspeted hypoxaemia (see reommendation 9). Disease-speifi measurements should also be reorded (eg, peak expiratory flow in asthma, blood pressure in ardia disease). 9.3 Immediate management of hypoxaemi patients Having asertained that the airway is lear, the emergeny responders should ommene oxygen treatment if the oxygen saturation is below the target. The initial oxygen therapy should follow the general priniples given in tables 1 4 and harts 1 and 2 (figs 1 and 2). There is some evidene that bronhodilator therapy, however given, an ause inreased V/ Q mismath and redued blood oxygen levels in autely ill patients shortly after treatment (see setion ). Reommendations The initial oxygen therapy to be used in the various linial situations is given in tables 1 4. If there is a lear history of asthma or heart failure or other treatable illness, appropriate treatment should be instituted in aordane with guidelines or standard management plans for eah disease. [Grade D] 21. The oxygen saturation should be monitored ontinuously until the patient is stable or arrives at hospital for a full assessment. The oxygen onen- vi45

46 tration should be adjusted upwards or downwards to maintain the target saturation range. [Grade D] 22. In most emergeny situations oxygen is given to patients immediately without a formal presription or drug order. The lak of a presription should never prelude oxygen being given when needed in an emergeny situation. However, a subsequent written reord must be made of what oxygen therapy has been given to every patient (in a similar manner to the reording of all other emergeny treatment). [Grade D] 9.4 Patients with known COPD A proportion of breathless patients will have COPD (hroni bronhitis and emphysema). Unfortunately, a reent Cohrane review of oxygen therapy for COPD in the prehospital setting found no relevant studies. 282 Audits of emergeny admissions in UK hospitals have shown that about 25% of breathless medial patients who require hospital admission have COPD as a main diagnosis. Many of these patients will require arefully ontrolled oxygen therapy beause they are at risk of arbon dioxide retention or respiratory aidosis. In a large UK study, 34 47% of patients with exaerbated COPD had PaCO kpa (45 mm Hg), 20% had respiratory aidosis (ph,7.35 or [H + ].45 nmol/l) and 4.6% had severe aidosis (ph,7.25 or [H + ].56 nmol/l). Aidosis was more ommon if the blood oxygen was.10 kpa (75 mm Hg). Plant and olleagues 34 reommended that patients with aute COPD should be maintained within a PaO 2 range of kpa (55 75 mm Hg) to avoid the dangers of hypoxaemia and aidosis. Reommendation (see table 3) Patients with COPD should initially be given oxygen via a Venturi 28% mask at a flow rate of 4 l/min or a 24% Venturi mask at a flow rate of 2 l/min. Some patients may benefit from higher flow rates via the Venturi mask (see reommendation 32). The target oxygen saturation should be 88 92% in most ases or an individualised saturation range based on the patient s blood gas measurements during previous exaerbations. [Grade C] 9.5 Patients who should be assumed to have COPD One of the hallenges faed by the initial linial response staff is that the diagnosis may be unlear and the patient s medial reords or detailed history may not be available. It has been shown that ambulane teams may be aware of a diagnosis of COPD in only 58% of ases. 283 The guidelines group onsider that an initial diagnosis of COPD should be assumed if there is no lear history of asthma and the patient is.50 years of age and a long-term smoker or ex-smoker with a history of longstanding breathlessness on minor exertion. The diagnosis should be reassessed on arrival at hospital where more information will probably beome available, and the FEV 1 should be measured unless the patient is too breathless to undertake spirometry. Reommendation If the diagnosis is unknown, patients.50 years of age who are long-term smokers with a history of hroni breathlessness on minor exertion suh as walking on level ground and no other known ause of breathlessness should be treated as if having COPD for the purposes of this guideline. Patients with COPD may also use terms suh as hroni bronhitis and emphysema to desribe their ondition but may sometimes mistakenly use asthma (see table 3). 9.6 Other patients at risk of hyperapni respiratory failure with respiratory aidosis Any patient with severe kyphosoliosis or severe ankylosing spondylitis. Severe lung sarring from old tuberulosis (espeially with thoraoplasty). Morbid obesity (body mass index.40 kg/m 2 ). Patients with neuromusular disorders (espeially if musle weakness has led to wheelhair use). Any patient on home mehanial ventilation. Use of home mehanial ventilation. Overdose of opiates, benzodiazepines or other drugs ausing respiratory depression. 9.7 Oxygen alert ards and 24% or 28% Venturi masks in patients with COPD (and others at risk of respiratory aidosis) who have had an episode of hyperapni respiratory failure The administration of high oxygen onentrations in aute COPD and other onditions (see setion 8.12) leads to worsening of hyperapni respiratory failure and respiratory aidosis. 34 Patients with COPD and a PaO 2.10 kpa (75 mm Hg) and a PaCO kpa (45 mm Hg) may be assumed to have had exessive oxygen therapy. If a patient is found to have respiratory aidosis due to exessive oxygen therapy, the oxygen therapy should not be disontinued immediately beause the oxygen level will fall signifiantly over 1 2 min by virtue of the alveolar gas equation (see setion 5.2.1) whereas the arbon dioxide level will take muh longer to orret itself (see setion 6.3.2). In this situation the oxygen treatment should be stepped down to 28% or 24% oxygen from a Venturi mask depending on oxygen saturation and blood gas results. A saturation target of 88 92% is reommended for aidoti patients in type 2 respiratory failure and non-invasive ventilation is required if the aidosis does not resolve quikly. This avoidable problem has ourred historially during the transfer to hospital, prior to measurement of arterial blood gases or before a definitive diagnosis is known. Furthermore, ambulane teams are often not informed at present of a diagnosis of COPD 283 and may not be aware of the presene of other high-risk onditions suh as kyphosoliosis or respiratory failure due to neuromusular onditions. These patients an be issued with an oxygen alert ard and a 24% or 28%Venturi mask based on previous blood gas results. The reommended oxygen saturation will be based on the linial senario for eah individual patient but will usually be 88 92%, oasionally 85 88% or 85 90% based on previous blood gas results. Patients should be instruted to show this ard to the ambulane rew and emergeny department staff in order to avoid the use of high oxygen onentrations. This sheme an be suessful. 284 The ambulane servie an also be informed about whih patients are issued with oxygen alert ards. 285 The urrent Joint Royal Colleges Ambulane Liaison Committee (JRCALC) guideline for the use of oxygen in COPD are being revised to aommodate these hanges. 274 An example of an oxygen alert ard is shown in fig 8. Reommendations 23. Patients with COPD (and other at-risk onditions) who have had an episode of hyperapni respiratory failure should be issued with an oxygen alert ard vi46

47 Figure 8 Example of oxygen alert ard. and with a 24% or 28% Venturi mask. They should be instruted to show the ard to the ambulane rew and emergeny department staff in the event of an exaerbation. [Grade C] 24. The ontent of the alert ard should be speified by the physiian in harge of the patient s are, based on previous blood gas results. [Grade D] 25. The primary are team and ambulane servie should also be informed by the responsible liniian that the patient has had an episode of hyperapni respiratory failure and arries an oxygen alert ard. The home address and ideal oxygen dose or target saturation ranges of these patients an be flagged in the ambulane ontrol systems and disseminated to ambulane rews when required. [Grade D] 26. Out-of-hours servies providing emergeny primary are servies should be informed by a responsible liniian that the patient has had an episode of hyperapni respiratory failure and arries an oxygen alert ard. Use of oxygen in these patients will be guided by the instrutions on the alert ard. [Grade D] 27. During ambulane journeys oxygen-driven nebulisers should be used for patients with asthma and may be used for patients with COPD in the absene of an air-driven ompressor system. If oxygen is used for patients with known COPD, its use should be limited to 6 min. This will deliver most of the nebulised drug dose but limit the risk of hyperapni respiratory failure (setion ). [Grade D] 28. If a patient is suspeted to have hyperapnia or respiratory aidosis due to exessive oxygen therapy, the oxygen therapy should not be disontinued but should be stepped down to 28% or 24% oxygen from a Venturi mask depending on oxygen saturation and subsequent blood gas results. [Grade C] 9.8 Choie of devies in prehospital are The range of oxygen delivery devies is very wide as disussed in setion 10. However, most patients an be managed with one of five types of oxygen delivery devie. Table 10 Examples of ommon oxygen ylinder sizes and apaities Oxygen ylinder sizes Reommendation 29a. It is reommended that the following delivery devies should be available in prehospital settings where oxygen is administered (see setion 10): [Grade D] high onentration reservoir mask (non-rebreathe mask) for high-dose oxygen therapy; nasal annulae (preferably) or a simple fae mask for medium-dose oxygen therapy; 28% Venturi mask for patients with definite or likely COPD (patients who have an oxygen alert ard may have their own 24% or 28% Venturi mask); traheostomy masks for patients with traheostomy or previous laryngetomy. SECTION 10: PRACTICAL ASPECTS OF OXYGEN THERAPY Oxygen delivery systems an be onsidered as two omponents: the method of storage and provision of oxygen (eg, ylinders); the method of delivery to the patient (eg, Venturi mask). The options available for both will depend on the environment in whih it is being used and the needs of the patient Oxygen storage and provision Cylinders (ompressed gas) Cylinders ontain ompressed gas held under a very high pressure. They ome in an array of sizes and hene apaity, ranging from small portable ylinders for individual patient use to large ylinders suitable for hospital use (table 10). These an be used for bedside administration where piped oxygen is not available or an be the supply for a piped system. With reent hanges in tehnology, high pressure ylinders are now available (ie, filled to 200 bar rather than 137 bar whih an ontain 54% more gas for the same size ylinder). It is important for all users of oxygen to be aware that most oxygen ylinders are olour-oded (blak ylinder with white shoulder). Small lightweight ylinders are also available for ambulatory use (eg, some weigh 3.2 kg when full). All systems ontaining ompressed gases in the UK are subjet to the Pressure Systems Safety Regulations 2000 (SI 2000 No 128). These regulations are intended to prevent the risk of injury from pressurised systems. Trusts must ensure that they have a poliy in plae whih ensures the safety of patients, staff and ontrators in the provision, storage, use and maintenane of ompressed gas systems as required by the Health and Safety at Work et At Cliniians using oxygen ylinders should hek the labelling of the ylinder to ensure that it is an oxygen ylinder and heks should be made to ensure that the ylinder is not empty or near-empty Liquid oxygen Liquid oxygen is ontained in pressure tanks and is obtained from atmospheri oxygen by frational distillation. It has to be evaporated into a gas before use. Large tanks are often used by Size C CD D E F G HX J Height (m) Capaity (l) vi47

48 hospitals and small tanks an be used domestially. Portable liquid oxygen is also available in small portable ontainers whih an be filled from the larger tanks Oxygen onentrators Oxygen onentrators are largely used in the domiiliary setting for the provision of long-term oxygen therapy and are therefore not used in the aute setting so will not be overed further Patient delivery methods/interfaes High onentration reservoir mask (non-rebreathing mask) (fig 9) This type of mask delivers oxygen at onentrations between 60% and 90% when used at a flow rate of l/min. 286 The onentration is not aurate and will depend on the flow of oxygen and the patient s breathing pattern. These masks are most suitable for trauma and emergeny use where arbon dioxide retention is unlikely (table 1) Simple fae mask (fig 10) This type of mask delivers oxygen onentrations between 40% and 60%. It is sometimes referred to as an MC Mask, Medium Conentration Mask, Mary Catterall Mask or as a Hudson Mask, but the latter desription is disouraged beause the Hudson Company make many types of mask (inluding high onentration reservoir masks). The guideline group favours the term simple fae mask. The oxygen supplied to the patient will be of variable onentration depending on the flow of oxygen and the patient s breathing pattern. The onentration an be hanged by inreasing or dereasing the oxygen flows between 5 and 10 l/min. However, different brands of simple fae mask an deliver a different oxygen onentration at a given flow rate. Flows of,5 l/min an ause inreased resistane to breathing, and there is a possibility of a build-up of arbon dioxide within the mask and rebreathing may our. 287 Figure 9 High onentration reservoir mask (non-rebreathing mask). Figure 10 Simple fae mask. This mask is suitable for patients with hypoxaemi respiratory failure (type 1) but is not suitable for patients with hyperapni (type 2) respiratory failure. The mask may deliver a high onentration of oxygen (.50%) and is therefore not reommended for patients who require low-dose oxygen therapy beause of the risk of arbon dioxide retention. Patients using a simple fae mask may have an inspiratory flow rate greater than the gas flow rate from the mask, so the simple fae mask should not be used at flow rates below 5 l/min. 287 Several publiations have shown that patients who require medium-dose oxygen therapy tend to prefer nasal annulae to simple fae masks and the annulae are more likely to be left in position by the patient and less likely to fall off Venturi mask (fig 11) A Venturi mask will give an aurate onentration of oxygen to the patient regardless of oxygen flow rate (the minimum suggested flow rate is written on eah Venturi devie and the available options are shown in table 11). The oxygen onentration remains onstant beause of the Venturi priniple. The gas flow into the mask is diluted with air whih is entrained via the age on the Venturi adaptor. The amount of air suked into the age is related to the flow of oxygen into the Venturi system. The higher the flow the more air is suked in. The proportions remain the same and therefore the Venturi mask delivers the same onentration of oxygen as the flow rate is inreased. Venturi masks are available in the following onentrations: 24%, 28%, 35%, 40% and 60%. They are suitable for all patients needing a known onentration of oxygen, but 24% and 28% Venturi masks are partiularly suited to those at risk of arbon dioxide retention (eg, patients with COPD). A further benefit of Venturi masks is that the flow rate of gas from the mask will usually exeed the inspiratory flow rate of the patient. One study has shown that patients with a respiratory rate.30 breaths/min often have an inspiratory flow rate above vi48

49 Figure 11 (A) Venturi mask. (B) Range of onentrations available. (C) Operation of Venturi valve. For 24% Venturi mask the typial oxygen flow of 2 l/min gives a total gas flow of 51 l/min. For 28% Venturi mask, 4 l/min oxygen flow gives a total gas flow of 44 l/min (table 11). the minimum flow rate speified on the mask pakaging. 254 Therefore, for patients with a high respiratory rate, it is suggested that the flow rate for Venturi masks should be set above the minimum flow rate listed on the pakaging (inreasing the oxygen flow rate into a Venturi mask does not inrease the onentration of oxygen whih is delivered). The auray of oxygen delivery from a Venturi mask is greatly redued if the mask is not aurately plaed on the patient s fae. 292 Patients with a respiratory rate.30 breaths/min often have a flow rate whih is above the minimum delivered by the Venturi system as speified by the flow rate reommended for the mask. [Evidene III] Venturi masks deliver a onstant perentage of oxygen but the effet on the patient will depend on the ondition being treated and on the breathing pattern and baseline oxygen saturation of the patient. As might be expeted from the oxygen dissoiation urve, patients with an oxygen saturation that is already in the normal range will have a very small rise in oxygen saturation (although the arterial oxygen tension is likely to rise substantially). However, patients with very low oxygen saturation will have a marked rise if given even a small dose of oxygen. This is beause the oxygen dissoiation urve is atually a rapid esalator rather than a slippery slope. This is illustrated in fig 12 whih uses atual oxygen saturations from King et al 246 and Warrel et al 65 together with alulated saturations from Bone et al 293 (two different groups), DeGaute et al 294 and Shiff and Massaro Nasal annulae (fig 13) Nasal annulae an be used to deliver low- and medium-dose oxygen onentrations. However, there is wide variation in patients breathing patterns so the same flow rate of nasal oxygen may have widely different effets on the blood oxygen and arbon dioxide levels of different patients. Nasal annulae at 1 4 l/min an have effets on oxygen saturation approximately equivalent to those seen with 24 40% oxygen from Venturi masks. The oxygen dose ontinues to rise up to flows above 6 l/min, but some patients may experiene disomfort and nasal dryness at flows above 4 l/min, espeially if maintained for several hours. Although one might expet mouth breathing to redue the effiieny of nasal annulae, the majority of studies have shown that mouth breathing results in either the same inspired oxygen onentraton or a higher onentration, espeially when the respiratory rate is inreased. 296 This is important beause patients with aute breathlessness are likely to breathe quikly and via the mouth rather than the nose. As there is marked individual variation in breathing pattern, the flow rate must be adjusted based on oximetry measurements and, where neessary, blood gas measurements. A rossover omparison of nasal annulae versus a Venturi mask (both adjusted to give satisfatory initial oxygen saturation) showed that the oxygen saturation of patients with exaerbated COPD fell below 90% for 5.4 h/day during treatment with a Venturi mask ompared with only 3.7 h/day during treatment with nasal annulae. 297 The upper range of oxygen delivery from nasal annulae is a little lower than the output of a simple fae mask, but the lower range goes a lot lower than a simple fae mask whih should not be used below a flow rate of 5 l/min (about 40% oxygen). 287 The performane and variation of nasal annulae for medium onentration oxygen therapy is broadly similar to that of the simple fae mask, both in laboratory experiments and in linial pratie One study 299 suggested that the saturation was lower with nasal annulae than with simple fae masks in a subgroup of men following abdominal surgery. Further studies are required to see if this was a hane finding or Table 11 Total gas flow rate (l/min) from Venturi masks at different oxygen flow rates Oxygen flow (l/min) Venturi values 24% oxygen 28% oxygen 35% oxygen 40% oxygen 60% oxygen vi49

50 Figure 12 Oxygen saturation response to treatment with 24%, 28% and 35% oxygen in patients with COPD. a genuine linial differene between the devies. Three patient preferene studies omparing nasal annulae with simple fae masks in postoperative are found that patient preferene was strongly in favour of nasal annulae with up to 88% of patients preferring annulae to masks Another advantage of annulae over simple fae masks is that they are less likely to be removed aidentally and they allow the patient to speak and eat There are no omparisons of these devies in aute are, but there is no reason to believe that the results would be any different for patients requiring medium-dose oxygen therapy. Advantages of nasal annulae ompared with simple fae masks for medium-dose oxygen therapy (Evidene level III): Comfort (but a minority of patients dislike the flow of oxygen into the nose, espeially above 4 l/min). Adjustable flow gives wide oxygen dose range (flow rate of 1 6 l/min gives FIO 2 from approximately 24% to approximately 50%). Patient preferene. No laustrophobi sensation. Figure 13 Nasal annulae. Not taken off to eat or speak and less likely to fall off. Less affeted by movement of fae. Less inspiratory resistane than simple fae masks. No risk of rebreathing of arbon dioxide. Cheaper. Disadvantages of nasal annulae: May ause nasal irritation or soreness. Will not work if nose is severely ongested or bloked Traheostomy mask (fig 14) These devies are designed to allow oxygen to be given via a traheostomy tube or to patients with previous laryngetomy (ie, nek breathing patients ). The oxygen flow rate should be adjusted to ahieve saturation in aordane with tables 1 4 and hart 1 (fig 1). Oxygen given in this way for prolonged periods needs onstant humidifiation and patients may need sution to remove muus from the airway. vi50

51 Figure 14 Traheostomy mask Continuous positive airways pressure and non-invasive ventilation These treatment options are beyond the sope of the present guideline. Readers are referred to the BTS guideline onerning the use of non-invasive ventilation in patients with exaerbations of COPD Oxygen arriage and delivery during patient transport in ambulanes Transport of oxygen ylinders in vehiles omes under the Transport of Dangerous Substanes At or the Carriage Regulations only if 1000 litres or more (measured by the water apaity of the ylinder) is arried at any one time. Ambulanes are therefore exempt from this. Normal health and safety requirements will still apply Health and Safety Exeutive guidane for safe use of oxygen ylinders All ylinders must be seured appropriately so they annot move in transit (inludes portable ylinders). No smoking in the viinity of ylinders. Cylinders must be heked regularly for obvious signs of leakage. Cylinders must be kept out of diret sunlight. Green warning triangle ompressed gas should be displayed on the vehile. Cylinders should never be lifted by the nek. They should only be hanged by suitably trained personnel. Apart from portable ylinders, all ylinders should be moved using a ylinder trolley Oxygen use by UK ambulane servies Currently within the UK the Ambulane Servie whether NHS or private has a range of vehiles and oxygen delivery systems at their disposal. There is an inreasing use of yle response units whih tend to use the lightweight AZ or C sized ylinder with a apaity of 170 litres. Motoryle response units are generally equipped with the same AZ or C sized ylinders. Fast response units based on ars tend to be equipped with at least two of the lightweight CD sized ylinders whih hold 460 litres. The CD ylinder is also the size favoured by mountain, ave and mines resue teams. Front-line ambulanes are usually equipped with piped oxygen fittings (Shraeder type) and supplied from two HX sized (2300 litres) ylinders, as well as arrying at least two CD sized ylinders to power a portable oxygen-powered resusitator. The piped supply has several outlet points plaed in strategi positions to whih are attahed standard Shrader flow meters (0 15 l/min). This enables oxygen to be given throughout the patient s journey. The ambulane is also equipped with a portable supply whih an be used at the site of an aident, taken into a patient s home or an be used when transferring a patient. They arry a range of patient interfaes for delivering the oxygen under the different irumstanes enountered. Portable resusitators are always apable of supplying freeflow oxygen therapy as well as their resusitator faility. Again, there are a variety of portable oxygen-powered resusitators and it is beyond the sope of these guidelines to desribe eah one available for use in prehospital are. It is strongly suggested that those pratitioners who need to work losely with the Ambulane Servie should beome familiar with the equipment used by their loal Ambulane Servie provider. With the possible exeption of the yle response units, all types of Ambulane Servie response will have portable resusitators, bag-valve-mask devies and portable sution as a minimum. Front-line emergeny department ambulanes will also have vehile-powered sution available. It is also very ommon now for patient transport servie ambulanes to be equipped with an oxygen supply, normally an HX ylinder (2300 litres) delivering the oxygen via a flow meter attahed diretly to the ylinder. Suh vehiles also tend to arry basi hand-held sution devies. The masks available are generally high onentration reservoir masks and are provided speifially for emergeny use for patients who might beome ill on the vehile. Vehiles that are equipped with an oxygen supply should also arry oximeters to ensure appropriate use of oxygen (see setion 9.8 for advie on whih oxygen delivery devies should be arried in ambulanes) Oxygen arriage in other vehiles and in primary are settings and patients homes Oxygen arriage in private ars (Health and Safety Exeutive Guidane) When travelling by ar, patients have the freedom to arry their own portable oxygen ylinder. Some general pratitioners in rural areas also arry oxygen in their ars. However, it is advised that ertain safety preautions should be followed: It is good pratie for the ar to display a green warning triangle for ompressed gas. The ylinder should be seure within the ar and annot move during transport or in the event of an aident Medial entres and primary are praties The majority of medial entres and praties should have a supply of oxygen for emergeny use. Generally, ylinders with integral high-flow regulators should be ordered. Otherwise, the ylinder must be fitted with a high-flow regulator apable of delivering a flow of.6 l/min in order to deliver medium and high-dose oxygen therapy. A reommended list of oxygen delivery devies for use in prehospital are is given in setion 9.8. Emergeny oxygen should be available in primary are medial entres, preferably using oxygen ylinders with integral high-flow regulators. Alternatively, oxygen ylinders fitted with high-flow regulators (delivering.6 l/min) must be used (see reommendation 19). vi51

52 Patients homes In patients homes oxygen is either provided for long-term therapy where an oxygen onentrator is provided (with or without a lightweight ylinder for ambulatory needs) or for short-term/short-burst therapy. Long-term oxygen therapy is overed in other guidelines. This existing home oxygen supply may be used by a patient or general pratitioner in an emergeny situation before the arrival of an ambulane. The patient/arers should be made aware of the following Health and Safety reommendations: 301 All ylinders should be stored on a ylinder trolley or suitably seured so they annot be knoked over. There should be no trailing oxygen tubing. A green warning triangle for ompressed gas should be displayed by the front door (warns emergeny servies in the event of a fire). The minimum number of ylinders should be stored in the house. There should be no smoking in the viinity of oxygen ylinders. Cylinders must be heked regularly for obvious signs of leakage. Cylinders must be kept out of diret sunlight. Oxygen must not be used near a naked flame Oxygen delivery systems in hospitals Most hospitals have piped oxygen systems as desribed previously, although some wards an still be found where piped oxygen is not available and large ompressed gas ylinders are used to supply the oxygen. Aute hospitals an spend up to per annum on liquid oxygen, so any devie that uses lower oxygen flow rates ould have signifiant eonomi savings for hospitals (eg, nasal annulae instead of a simple fae mask for medium-dose oxygen) Postoperative are on general surgial wards Medium onentration masks and nasal annulae are usually suffiient (target saturation 94 98%) exept for patients with known signifiant COPD who should reeive oxygen from a 24% or 28% Venturi mask or 1 2 l/min from nasal annulae aiming at a saturation range of 88 92% Emergeny departments Medium or high onentration oxygen is normally used (via nasal annulae, simple fae mask or reservoir mask), but partiular attention should be given to patients who have type 2 respiratory failure when a 24% or 28% Venturi mask or nasal annulae at a flow rate of 1 2 l/min would be appropriate General wards and respiratory wards The method of oxygen delivery will depend on the following irumstanes: Expeted duration of treatment. Type of respiratory illness. Pattern of breathing (high or low respiratory rate and drive). Need for humidifiation. Risk of arbon dioxide retention. Presene of onfusion and its effet on potential ompliane. Nasal annulae, simple fae masks, reservoir masks and Venturi masks should be used where appropriate (see tables 1 4 and harts 1 and 2 (figs 1 and 2)). Nasal annulae at flow rates or l/min are sometimes used as a substitute for Venturi masks in aute or post-aute patients with COPD on respiratory wards (adjusting flow as neessary to ahieve the desired arterial blood gas tensions). This pratie requires the use of paediatri flow meters to ensure onsistent and finely alibrated oxygen delivery and is not reommended outside speialist units Devies used in emergeny oxygen therapy Based on the advantages of eah delivery system disussed above, the following reommendations are made for delivery of oxygen in medial emergenies. It is likely that additional equipment will be maintained in speialist units, but speialist treatment is outside the sope of the present guideline. Reommendations (see tables 1 4) 29b. Most hospital patients an be managed with the same delivery devie as in reommendation 29a, but a 24% Venturi mask should also be available. The high onentration reservoir mask at l/min is the preferred means for delivering high-dose oxygen to ritially ill patients. [Grade D] Nasal annulae should be used rather than simple fae masks in most situations requiring medium-dose oxygen therapy. Nasal annulae are preferred by patients for reasons of omfort and they are less likely to be removed during meals (see setion ). There is also a ost saving. [Grade C] The flow rate from nasal annulae for medium-dose oxygen therapy should be adjusted between 2 and 6 l/ min to ahieve the desired target saturation. [Grade C] Venturi masks are reommended for patients requiring preise ontrol of FIO 2. Venturi masks an deliver a onstant FIO 2 of 24%, 28%, 35%, 40% and 60% oxygen with a greater gas flow than a simple fae mask. In those at risk of developing hyperapni respiratory failure with oxygen therapy, the use of Venturi masks may redue this risk. Furthermore, there is less likelihood of dilution of the oxygen stream by room air if the patient s inspiratory flow rate exeeds the flow rate delivered by the fae mask. [Grade D] 30. For many patients Venturi masks an be substituted with nasal annulae at low flow rates (1 2 l/min) to ahieve the same target range one patients have stabilised. [Grade D] 31. The flow rate from simple fae masks should be adjusted between 5 and 10 l/min to ahieve the desired target saturation. Flow rates below 5 l/min may ause arbon dioxide rebreathing and inreased resistane to inspiration. [Grade C] 32. Patients with COPD with a respiratory rate of.30 breaths/min should have the flow rate set to 50% above the minimum flow rate speified for the Venturi mask and/or pakaging (inreasing the oxygen flow rate into a Venturi mask inreases the total gas flow from the mask but does not inrease the onentration of oxygen whih is delivered). [Grade C] Flow meters All oxygen delivery systems must have a method of taking the high pressure/flow of gas and reduing it so it an be administered to the patient at a speifi flow depending on the individual s needs and mask being used. vi52

53 Piped oxygen points have Shrader flow meters and ylinders have pressure and flow regulators. Most oxygen flow meters use a floating ball to indiate the flow rate. The entre of the ball should be aligned with the appropriate flow rate marking. The example shown in fig 15 indiates the orret setting to deliver 2 l/min Oxygen tubing and oxygen wall outlets Oxygen tubing is needed to onnet flow meters and regulators to the patient delivery devie. It is important to ensure that all tubing is onneted orretly at both ends. The National Patient Safety Ageny has reported frequent adverse events related to oxygen use, inluding four reports of instanes where an oxygen mask was onneted in error to a ompressed air outlet instead of an oxygen outlet. Compressed air outlets are often used to drive nebulisers on in hospitals beause they are quieter than eletrial ompressors. However, the flow meter looks very similar to an oxygen flow meter and is often mounted beside an oxygen flow meter so it is very important to ensure that air flow meters are learly labelled. There is a similar risk with other piped gas outlets suh as those delivering nitrous oxide in some hospitals. Air flow meters are never required in an emergeny and should be removed from wall sokets or overed by a designated hood when not in use. The guideline authors are also aware of some ases where twin oxygen outlets were in use and the wrong one had been turned on or off. For example, one patient tried to turn off the oxygen flow after finishing a nebulised treatment but aidentally turned off the oxygen flow to a neighbouring patient with serious onsequenes. It is reommended that patients should not be allowed to adjust oxygen flow, espeially if there are dual outlets. Reommendation 33. Trusts should take measures to eliminate the risk of oxygen tubing being onneted to the inorret wall oxygen outlet or to outlets that deliver ompressed air or other gases instead of oxygen. Air flow meters should be removed from the wall sokets or overed with a designated air outlet over when not in use. Speial are should be taken if twin oxygen outlets are in use. [Grade D] 10.6 Use of humidified oxygen Rationale for use of humidified oxygen The upper airway normally warms, moistens and filters inspired gases. When these funtions are impaired by a pathologial proess or when they are bypassed by an artifiial airway, it is ommon pratie to provide humidifiation. The main reason for using humidifiation, espeially with high-flow oxygen, is that it may redue the sensation of dryness in the upper airways that oxygen an ause. However, in the non-intubated population there appears to be little sientifi evidene of any benefit from humidified oxygen exept that single doses of nebulised isotoni saline have been shown to assist sputum learane and redue breathlessness in patients with COPD. There is also evidene that humidifiation, when ombined with physiotherapy, an inrease sputum learane in bronhietasis. 305 Randomised ontrolled trials of the effets of humidified high-flow oxygen on patient omfort are required. Reommendations 34. Humidifiation is not required for the delivery of low-flow oxygen or for the short-term use of highflow oxygen. It is not therefore required in prehospital are. Pending the results of linial trials, it is reasonable to use humidified oxygen for patients who require high-flow oxygen systems for more than 24 h or who report upper airway disomfort due to dryness. [Grade B] 35. In the emergeny situation, humidified oxygen use an be onfined to patients with traheostomy or an artifiial airway although these patients an be managed without humidifiation for short periods of time (eg, ambulane journeys). [Grade D] 36. Humidifiation may also be of benefit to patients with visous seretions ausing diffiulty with expetoration. This benefit an be ahieved using nebulised normal saline. [Grade C] Thorax: first published as /thx on 6 Otober Downloaded from on 5 July 2018 by guest. Proteted by opyright. Figure 15 Flow meter showing orret setting to deliver oxygen at a rate of 2 l/min. vi53

54 Use of bubble humidifiation systems Humidified oxygen is widely administered in hospitals aross the UK and this is presumed to alleviate nasal and oral disomfort in the non-intubated patient. Humidifiation of supplemental oxygen is ommonly delivered by bubbling oxygen through either old or warm sterile water before it reahes the patient. However, the effet on patient omfort is negligible Bubble humidifiers may, however, represent an infetion hazard and should not be used. 308 There is no evidene of a linially signifiant benefit from bubble bottle systems but there is an infetion risk. [Evidene level III] Reommendation 37. Bubble bottles should not be used beause there is no evidene of a linially signifiant benefit but there is a risk of infetion. [Grade C] Large volume nebulisation-based humidifiers If humidifiation is required, it should ideally deliver the inspired gas at a of 32 36uC. Cold water humidifiers are simple and inexpensive but less effiient than a warm water system (about 50% relative humidity at ambient temperatures). The warm water option is more effetive, targeting a relative humidity of 100%, but both systems are thought to be a potential infetion ontrol risk. Warm water humidifiers are expensive and mostly onfined to intensive are units and high dependeny units and thus outside the sope of this guideline. Newer humidifying systems are really giant nebulisers with a 1-litre reservoir of saline or sterile water and an adjustable Venturi devie (fig 16). These systems are attahed diretly to the oxygen flow meter and onneted to an aerosol mask via Flex tube. They allow delivery of preise oxygen onentrations of 28%, 35%, 40% and 60% oxygen via their Venturi devie. This requires a speifi oxygen flow rate as well as adjusting the Venturi nozzle on the devie. It is possible to deliver 24% oxygen using a speial adaptor. These large volume humidifiers have a high humidifiation output. The main indiation for use is to assist with expetoration of visous sputum. There are no published randomised studies involving these devies, but it has been shown that single doses of nebulised saline an assist sputum prodution and relieve breathlessness in patients with COPD. Patients requiring high flow rates or longer term oxygen might benefit from a large volume oxygen humidifier devie, espeially if sputum retention is a linial problem. [Evidene level III] In the absene of an artifiial airway the deision to humidify supplemental oxygen needs to be made on an individual basis but this pratie is not evidene-based. [Evidene level IV] 10.7 Use of oxygen in patients with traheostomy or laryngetomy The number of patients with a traheostomy being ared for in a ward setting is inreasing as ritial are personnel use this as a method of failitating weaning from mehanial ventilation. In the absene of a pressurised iruit, oxygen is predominantly delivered via traheostomy mask. This is a variable performane devie and delivers onentrations up to 60 70%. If the patient deteriorates and requires an inreased oxygen onentration exeeding the onentration that a variable performane Figure 16 Large volume nebulisation-based humidifier. interfae an deliver (60 70%), it will be neessary to seek an alternative delivery system, usually a T-piee devie fitted diretly to the traheostomy tube. With a mask system the interfae will be onneted to a humidifiation system via elephant tubing. As inserting a traheostomy tube bypasses the patient s natural mehanisms to warm and moisturise inspired gases, it is essential to humidify any supplemental oxygen being delivered to the traheostomised patient. This will help maintain a patent traheostomy tube, reduing the build-up of seretions within the inner tube or the traheostomy itself and minimising any subjetive disomfort that the patient may experiene. Reommendation 38. When oxygen is required by patients with prior traheostomy or laryngetomy, a traheostomy mask (varying the flow as neessary) should ahieve the desired oxygen saturation (tables 1 4). An alternative delivery devie, usually a two-piee devie fitted diretly to the traheostomy tube, may be neessary if the patient deteriorates. [Grade D] vi54