Wean Earlier and Automatically with New Technology (The WEAN Study): A Multicentre, Pilot Randomized Controlled Trial

Transcription

1 Wean Earlier and Automatically with New Technology (The WEAN Study): A Multicentre, Pilot Randomized Controlled Trial Karen E. A. Burns MD, MSc, Maureen O. Meade MD, MSc, Martin R. Lessard MD, Lori Hand RRT, Qi Zhou PhD, Sean P. Keenan MD, MSc and Francois Lellouche MD, PhD Online supplement content This online supplement provides: Additional methods Appendix E1: Detailed Methods Appendix E2: Study Inclusion Process Appendix E3: Automated Weaning Protocol Appendix E4: Protocolized Weaning Appendix E5: Other Important Considerations (criteria for NIV, criteria for reintubation) Appendix E6: SAS Sedation Protocol Appendix E7: PEEP/FiO 2 Titration Chart (Both Study Groups) Additional results Appendix E8: Detailed Sedation Scores (SAS and RASS equivalents in both groups) Appendix E9: Adverse Events Appendix E10: Median Cumulative and Per Diem of Mechanical Ventilation Use of Sedative, Analgesic and Hypnotic Agents Appendix E11: Subgroup Analyses Based on Humidification Type Appendix E12: Subgroup Analyses Based on COPD 26

2 Appendix E1: Detailed Methods Background Patients were considered to be in a state of normal ventilation when they breathed with a RR between15 and 30 b/min (up to 34 b/min with central neurologic disease), a V T above a minimum threshold (based on weight) and an ETCO 2 level below a maximum threshold of 55 and 65 mmhg in non-copd and COPD patients, respectively. When values fall outside of these ranges, the system adjusts the level of PS provided in increments or decrements of 2 to 4 cm H 2 O to restore the patient to the comfort zone. Methods We held training sessions at each of the participating centres to familiarize personnel with the ventilators and software. Patient Identification and Consent We defined severe heart failure as grade 3 or 4 left ventricular function or New York Heart Association class 4 dyspnea. Patients or their legal representatives were approached for consent to undergo a PST and a SBT, if criteria were present and subsequent randomization. Run In Phase We conducted a run-in phase including two patients at each centre with the first patient randomized to either AW or PW and the second patient assigned to the alternate strategy. These patients are not included in pilot study analyses. Upon completion of the two patients, a clinical evaluation committee reviewed the completed case report forms to assess the ability of participating centres to screen and enroll patients, comply with study procedures and manage data in a timely fashion. Approval from the clinical evaluation committee enabled sites to proceed to enrolling patients in the pilot RCT. Pre-randomization Pressure Support Trials Patients who met the following criteria underwent a PST: (i) intubated on mechanical ventilation, (ii) SpO 2 > 90% with FiO 2 50% and PEEP 10 cm H 2 O, (iii) no requirement for high dose vasopressors (i.e., no epinephrine or norepinephrine > 15 µg/min (or 0.2 µg/kg/min) or equivalent dose vasopressin or phenylephrine), (iv) motor component of the Glasgow Coma Scale score > 4, (v) stable neurological status with no deterioration in the previous 24 hrs, intact respiratory drive, and intracranial pressure < 20 mm Hg, and (vi) were not expected to be extubated on the day of study randomization with (vii) no surgery or procedure requiring sedation planned for the next 48 hrs. The minimum and maximum acceptable levels of PS (above PEEP) were 10 and 22 cm H 2 O, respectively, during PSTs. Patients who failed the PST were returned to their previous ventilator settings or those that restored comfort. 27

3 A PST was successful if the patient remained clinically stable with a RR > 10 and < 35 b/minute with no decrease in pulse oximetry saturations [(SpO 2 ) > 90% on an FiO 2 < 50% with PEEP < 10 cm H 2 O] for 60 minutes without a PS change. PSTs were terminated for sustained hemodynamic or respiratory distress (heart rate < 50 or > 140 beats/minute or new significant dysrhythmias, systolic blood pressure < 80 mm Hg or > 180 mm Hg or RR > 40 b/min). Criteria for SBT Failure Patients were considered for a SBT if they met the following criteria: (i) partial or complete reversal of the cause of respiratory failure with SpO 2 90% on an FiO (or at baseline level in chronically hypoxemic patients) and PEEP 5 cm H 2 O, (ii) hemodynamic stability [off vasopressors or on low levels of vasopressors (i.e., norepinephrine 7 µg/min or 0.1 µg/kg/min) or equivalent], (iii) absence of uncontrolled sepsis, and (iv) stable haemoglobin > 70 g/l. We consider the presence of any one of the following: (i) RR > 35 breaths/min, (ii) clinical signs of respiratory distress (i.e., abdominal paradox), (iii) SpO 2 < 90% (or below baseline in chronically hypoxemic patients) with FiO 2 > 50%, (iv) systolic blood pressure 80 mmhg or 180 mmhg, (v) heart rate 50 or 140 beats/min or new significant dysrhythmias, (vi) severe agitation or diaphoresis, or (vii) Increased somnolence with elevated arterial pressure of carbon dioxide (PaCO 2 ) and ph < 7.30 to signify SBT failure. Randomization Patients underwent randomization if a study ventilator was available. If both diagnoses (COPD and a central neurologic disorder) were present, we prioritized the central neurologic disorder during stratification. We clustered humidification strategies. Reusable circuits were not permitted. Weaning Strategies All participating centres had at least 2 AW capable ventilators. When available, we preferentially used Evita ventilators, for patients randomized to PW. In both groups, PS could increase or decrease during weaning in response to patient s needs or events. Maximum inspiratory pressure was set at 35 cm H 2 O in the AW group. We used comparable ventilators (Evita 2 dura, Evita 4, Evita XL, Servo i, Servo 300, Puritan Bennett 840, Puritan Bennett 760, Avea or Galileo) in the PW group. PS mode (manually or automatically set) was used for weaning. The initial PS setting was similar to that used during the PST. 28

4 Criteria to return to or remaining on an alternate mode of ventilation included:: (i) surgery or invasive procedures that required sedation, (ii) respiratory distress defined by a) sustained hypoxemia (SpO 2 < 90%) with FiO 2 > 60% and PEEP > 10 cm H 2 O or hypercapnia with ph < 7.30 or clinical respiratory distress, b) repeated episodes ( 3 episodes within 1 hour wherein an inspiratory pressure (pressure support + PEEP) of 35 cmh 2 O was attained (despite suctioning, bronchodilation etc.), (iii) hemodynamic instability despite fluid boluses with a requirement for high dose vasopressors: norepinephrine > 15µg/min (0.2µg/kg/min) or equivalent, (iv) suspected myocardial ischemia based on electrocardiogram and/or elevated Troponin i, (v) neurologic deterioration with need to control PaCO 2 (both groups) or an alarm indicating central hypoventilation (AW group) or (vi) increased sedation resulting in RR < 10 breaths/min. Patients who returned to or remained on an alternate mode of ventilation were reassessed at least daily with a PST. An SBT was not required to resume the assigned weaning protocol. A PEEP/FiO 2 chart guided titration of oxygen and PEEP during weaning in both groups. For patients randomized to AW, a SmartCare ventilator was kept in the patient s room following extubation until the patient was deemed successfully extubated. Ventilator start up data were reentered following a tracheostomy or humidification change. We evaluated RT and physician acceptance of the weaning protocols and critical care nurse acceptance of the sedation protocol daily using Likert [21] scales [ranging from 0 highly unacceptable (extremely difficult to use) to 10 highly acceptable (extremely easy to use)]. During study implementation, we made two modifications to the study protocol to enable inclusion of patients with a do not resuscitate status and to permit patients to undergo a PST with a PEEP of < 10 cm H 2 O. Data Collection and Analysis We collected data at ICU admission, study enrollment and daily thereafter. We considered p- values < 0.05 to be statistically significant. All analyses adhered to the intention-to-treat principle. 29

5 Lickert scales: The Likert scale is a categorical linear scale that allows quantification of subjective information [21]. It can be compared to visual analogue scale (VAS) but the VAS uses a continuous linear scale. The respondents of the Likert scale specify their level of agreement or disagreement on a symmetric agree-disagree scale for a series of statements as described on the following example: DAILY RT ACCEPTANCE OF THE WEANING PROTOCOL Please circle the number that corresponds to your evaluation of how easy the assigned weaning protocol was to implement today using the following scale: =Highly unacceptable (difficult to use) 10=Highly acceptable (extremely easy to use) Results Primary Outcomes: Compliance with the Weaning and Sedation Protocols The overall time-off assigned study protocols was 16.0 (± 8.7) hours and ranged from 1-24 hours across centres. 30

6 Appendix E2: Study Inclusion Process 31

7 Appendix E3: Automated Weaning Protocol Initiation: STEPS FOR EVITA XL SET UP: Applies to Automated Weaning patients following: (1) randomization, (2) reinitiation after interruption, (3) reintubation within 48 hrs (4) reinitiation after tracheostomy. Step 1: Switch to Pressure Support (CPAP/ASB) Step 2: Select SmartCare tab Step 3: Enter patient weight (measured body weight) Step 4: Type of humidification at randomization: HH or HME based on clusters. Step 5: Airway access: endotracheal tube (oro or nasal) at randomization or tracheostomy (after randomization) Step 6: Known or suspected COPD yes/no Neurologic disorder (only central neurologic disorder) yes/no Step 7: Night rest select no Step 8: Set alarms: Maximum Inspiratory Pressure: 35 cm H 2 O Max RR 40 b/min ETCO 2 limits: 15 mm Hg (minimum) and 70 mm Hg (maximum) Step 9: You must ensure the following prior to initiating a session a) ETCO 2 monitoring is activated b) Check that the ETCO 2 cuvette (connector) is clean on the interior aspect c) Flow Monitoring is activated and d) ETCO 2 is calibrated Step 10: Apnea ventilation must be activated and set Step 11: ATC (Automatic Tube Compensation) must be deactivated Step 12: Leak compensation must be activated Before initiating mechanical ventilation with SmartCare, baseline information including [patient measured body weight, presence or absence of COPD/central neurologic disorder, type of airway prosthesis (tracheostomy or endotracheal tube) and humidification used [heat and moisture exchanger (HME) or heated humidification (HH)] is entered using the touch pad ventilator interface. These items are required to determine limits for V T, PETCO 2 and the minimal level of pressure support to initiate an SBT (range: 5-12 cm H 2 O; with HH 5-7 cm H 2 O and HME 9-12 cm H 2 O). Pressure support, PEEP and FiO 2 are initiated as per the levels used in the PST. 32

8 Titration: Pressure support be continuously adapted (by averaging RR, V T, PETCO 2 every 2 to 5 minutes) with RRs [ranging from 15 to 34 breaths/min, minimum V T of at least 300 cc (250 cc if patient s weight < 55kg)] and maximum P ET CO 2 of 55 mm Hg (or 65 mm Hg for COPD patients). When ventilatory parameters fall within these constraints, a normal ventilatory state is diagnosed. Conversely, the system adjusts the level of pressure support to attain these targets when ventilatory parameters fall outside of these constraints. Specifically, the automated system increases pressure support by 2 cm H 2 O in response to tachypnea (RR> 30 breaths/min or 34 breaths/min with a central neurologic disease) and by 4 cm H 2 O (RR > 36 breaths/min) and lowers the pressure support delivered by 2 cm H 2 O when bradypnea (RR< 12 breaths/min) is diagnosed with no increase in PETCO 2. When V T or PETCO 2 are out of the defined ranges, insufficient ventilation is diagnosed and pressure support is increased by 2 cm H 2 O. Patients are weaned at night and during the day. If ventilation becomes unstable the mechanical assistance provided will be augmented. Specific management strategies depend on the duration of instability and the level of pressure support required previously. PEEP and FiO 2 are titrated to maintain PaO mm Hg (or at baseline levels in hypercarbic patients) or SpO %%. A PEEP/FiO 2 chart is used to titrate FiO 2 and PEEP during weaning (see Appendix E6). We assess and document ventilator parameters at least every 4-6 hrs. Sedation is titrated to either a SAS score of 3-4 or a RASS score of -3 to 0 during weaning. Discontinuation: SBTs are automatically conducted when a minimum level of pressure support is reached [5 cm H 2 O (HH) with a tracheostomy or 7 cm H 2 O (HH) with an oro/nasotracheal airway] and PEEP has been reduced to 5 cm H 2 O. Alternatively, SBTs are automatically conducted when a minimum level of pressure support is reached [9 cm H 2 O (HME) with a tracheostomy or 12 cm H 2 O (HME) with an oro/nasotracheal airway] and PEEP has been reduced to 5 cm H 2 O. SBT duration (60 or 120 minutes) is contingent upon the level of pressure support initially used and the magnitude of the decrease in pressure support (ranging from 2 to 4 cm H 2 O) tolerated. If the initial pressure support required was <20 cm H 2 O, the SBT duration is 60 minutes and if the initial pressure support required was > 20 cm H 2 O, the SBT duration is 120 minutes. The level of pressure support used during the SBT is the minimal level of pressure support delivered during automated ventilation. This level will remain steady as long as the patients breathing pattern remains stable. If more assistance is required, the SBT will be terminated and pressure support will be increased. Alternatively, if less support is required (ventilatory diagnosis of hyperventilation) pressure support will decrease to 5 cm H 2 O. When a successful SBT is completed, a directive is issued by the ventilator informing staff to consider separation from the ventilator. In the event of a successful SBT in patients with an endotracheal tube patients must be assessed for extubation. If the message appears and extubation criteria are fulfilled then patients must be extubated immediately (must be performed within 15 hours of meeting the 33

9 extubation criteria). After an unsuccessful SBT, ventilator settings are adjusted to restore respiratory comfort until criteria are met to undergo a subsequent SBT. Extubation Criteria: Patients are assessed for extubation criteria after successful completion of a SBT. To be extubated patients must have: 1. An SpO 2 > 90% on an FiO 2 < 40% and PEEP< 5 cm H 2 O (or at baseline level in chronically hypoxemic patients) 2. A cough of sufficient strength to clear secretions and must not require suctioning more than every 2 hours. 3. No procedures requiring sedation or surgery should be planned within 48 hours after extubation 4. Patients should be hemodynamically stable (off vasopressors or on minimal norepinephrine i.e. < 7 ug/min or equivalent) and 5. A level of consciousness sufficient to ensure airway protection. Disconnection criteria (tracheostomized patients): All of the above criteria, except #5 above, apply to patients with a tracheostomy who successfully complete a SBT. Disconnection from the ventilator must be performed immediately after meeting criteria # 1-4 above. Peri-Extubation Ventilator Management: *If a patient passes an SBT and meets Extubation Criteria: (s)he should be immediately extubated (and must be extubated within 15 hrs). *If a patient passes an SBT and does not meet Extubation Criteria: (s)he should continue on pressure support 5-7 cm H 2 O PEEP < 5 cm H 2 O (HH) or 9-12 cm H 2 O PEEP < 5 cm H 2 O (HME). Patients should be continued on SmartCare pressure support. Pressure support will be titrated automatically. *If a patient fails an SBT the assigned weaning protocol should be resumed. Ventilator Management After Extubation: Protocols continue until successful extubation (at least 48 hrs of unassisted spontaneous breathing), ICU death, ICU discharge or until day 90 after randomization. The AW ventilator must remain in the patient s room until successful extubation or ICU discharge (which ever comes first). 34

10 Appendix E4: Protocolized Weaning Initiation: Pressure support is the recommended mode of ventilation. Pressure support, FiO 2 and PEEP are initiated as per the levels used in the Pressure Support Trial. Titration: The level of pressure support is reevaluated at least every 4-6 hrs and titrated to avoid respiratory distress (use of accessory muscles or RR > 30 breaths/min (alternatively, > 34 breaths/min in patients with COPD or a central neurologic disease) or overassistance. Patients are weaned during the day and at night. PEEP and FiO 2 are titrated to maintain PaO mm Hg (or at baseline levels in hypercarbic patients) or SpO %%. A PEEP/FiO 2 chart is used to titrate FiO 2 and PEEP during weaning. We assess and document ventilator parameters at least every 4 6 hrs. Sedation is titrated to either a SAS score of 3-4 or a RASS score of -3 to 0 during weaning. Discontinuation: SBTs are performed when the following criteria are met on pressure support: a) The cause of respiratory failure is partially or completely reversed with SpO 2 90% on an FiO (or at baseline in chronically hypoxemic patients) and PEEP 5 cm H 2 O, b) The patient is hemodynamic stable [off vasopressors or on low levels of vasopressors (i.e. norepinephrine 7µg/min or 0.1µg/kg/min) or equivalent], c) Absence of uncontrolled sepsis, d) Hemoglobin > 70 g/l and stable, The above criteria are evaluated at least once per day. SBTs are performed using either a T- piece (or Trach Mask) with oxygen or CPAP (< 5 cm H 2 O) or pressure support 5-7 cm H 2 O PEEP < 5 cm H 2 O (HH group) or cm H 2 O PEEP < 5 cm H 2 O (HME group). All SBTs are minutes in duration. SBT failure/termination criteria is defined by the presence of any ONE of (i) RR > 35 breaths/min, (ii) Clinical signs of respiratory distress (i.e. abdominal paradox), (iii) SpO 2 < 90% (or below baseline in chronically hypoxemic patients) with FiO 2 > 50%, (iv) Systolic blood pressure 80 mmhg or 180 mmhg, (v) Heart rate 50 bpm or 140 beats/min or new significant dysrhythmias, (vi) Severe agitation or diaphoresis, (vii) Increased somnolence with elevated PaCO 2 and ph < After an unsuccessful SBT, ventilator settings are adjusted to restore respiratory comfort until criteria are met to undergo a subsequent SBT. For patients who fail an SBT, at least one SBT is attempted 35

11 daily. In the event of a successful SBT in patients with an endotracheal tube, patients must be must be extubated immediately (must be performed within 15 hrs of meeting the extubation criteria). Alternatively, in the event of a successful SBT in a tracheostomized patient the patient must be disconnected. Extubation Criteria: Patients are assessed for extubation criteria after successful completion of an SBT. To be extubated patients must have: 1. An SpO 2 > 90% on an FiO 2 < 40% and PEEP < 5 cm H 2 O (or at baseline level in chronically hypoxemic patients) 2. A cough sufficiently strong to clear secretions and not requiring suctioning more than every 2 hours. 3. No procedures requiring sedation or surgery planned within 48 hours after extubation 4. Patients should be hemodynamically stable [off vasopressors or on minimal norepinephrine i.e. < 7 ug/min (or 0.1 ug/kg/min) or equivalent] and 5. A level of consciousness sufficient to ensure airway protection (not required for tracheostomized patients). Disconnection criteria (tracheostomized patients): All of the above criteria, except #5 above, apply to patients with a tracheostomy who successfully complete a SBT. Disconnection from the ventilator must be performed immediately after meeting criteria # 1-4 above. Peri-Extubation Ventilator Management: *If a patient passes an SBT and meets Extubation Criteria: (s)he should be immediately extubated (and must be extubated within 15 hrs). *If a patient passes an SBT and does not meet Extubation Criteria: (s)he should continue on pressure support 5-7 cm H 2 O PEEP < 5 cm H 2 O (HH) or cm H 2 O PEEP < 5 cm H 2 O (HME). Patients should be continued on these settings. Pressure support and PEEP should be titrated to avoid respiratory distress or over assistance at least Q4-6 hr. *If a patient fails an SBT the assigned weaning protocol should be resumed. Ventilator Management After Extubation: Protocols continue until successful extubation (at least 48 hrs of unassisted spontaneous breathing), ICU death, ICU discharge or until day 90 after randomization. 36

12 Appendix E5: Other Important Considerations A) Use of Noninvasive Ventilation After Extubation: In the event that respiratory distress develops after extubation, patients may be treated with NIV if the criteria for initiation are achieved. Criteria for Initiation of Noninvasive Positive Pressure Ventilation Initiation Criteria: The presence of any two of the following criteria are used to initiate NIV [29,30] (i) clinical signs of respiratory distress/muscle fatigue and RR > 30 breaths/min or a 50% increase in RR from baseline, (ii) respiratory acidosis (ph < 7.35 with PaCO 2 > 45 mmhg), (iii) hypoxemia (SpO 2 < 90% or PaO 2 < 60mmHg with FiO 2 > 50%) For patients randomized to AW, the AW system is not be used to manage NIV. Initiation Technique: NIV is delivered with the patient in a seated position and the upper body elevated at a minimum of 30. The patient-ventilator interface is fitted to the patient with oronasal masks preferred over nasal masks. Patients are encouraged to hold the mask up to their face before securing it with head straps. Excessive tension on the straps is avoided. A nasal hydrocolloid protection dressing is used to minimize discomfort related to mask application. Initial breaths are made with inspiratory pressure of 8 cm H 2 O and no expiratory pressure. In the first 2-5 minutes, progressive increases in inspiratory and expiratory pressure are made in increments of 2 cm H 2 O. PEEP (or expiratory positive airway pressure) is increased to a maximum of 3 cm H 2 O in COPD patients and 10 cm H 2 O in hypoxemic patients and for treatment of atelectasis. Heated humidification is preferentially used in patients, especially those with persistent hypercapnia. A variety of interfaces (different shapes/sizes) are available to minimize leaks and ensure patient comfort. Leaks and asynchronies are minimized. Regardless of treatment assignment, the total duration of mechanical ventilation includes the time spent on NIV. B) Reintubation and Reinitation of Invasive Weaning: We use objective criteria to guide reintubation during the pilot RCT. Reintubated patients are ventilated according to initial treatment assignment once the criteria for a PST are re-attained until successful extubation, ICU death, ICU discharge or until day 90 after randomization. Reintubation Criteria 37

13 Criteria for Reintubation/Reconnection in Tracheostomized Patients: Intubation is indicated if at least one major criterion or two minor criteria are met [31]: Major criteria: (i) cardiac arrest, (ii) respiratory pauses with loss of consciousness or gasping for air, (iii) psychomotor agitation interfering with nursing care, (iv) heart rate < 50 bpm with loss of alterness or (v) hemodynamic instability with systolic pressure < 80 mmhg for > 30 minutes despite adequate volume challenge, use of vasopressors or both. For patients on NIV: (vi) a change in mental status (decrease in level of consciousness, severe agitation) rendering the patient unable to tolerate NIV, (vii) persistent or worsening signs of respiratory distress/muscle fatigue despite NIV, (viii) abundant secretions that can not be effectively cleared or are associated with acidosis, hypoxemia or a change in mental status with NIV, (ix) failure to improve ph or PaCO 2 with NIV. Minor criteria: (i) RR >35 breaths/min and exceeding RR after extubation, (ii) ph < 7.30 and less than the value after extubation, (iii) sustained decrease in SpO 2 < 85% despite high FiO 2 (> 0.80) or (iv) decreasing level of consciousness (deterioration from baseline). C) Tracheostomy: Tracheostomy may be performed in either group when clinically indicated. Notwithstanding, we encourage clinicians to delay decisions regarding elective tracheostomy until day 10 after randomization. 38

14 Appendix E6: SAS Sedation Protocol (** A RASS sedation protocol was also developed using comparable sedation scores) Sedation and/or analgesia are available to all participants in whom it is required. We administer sedation and analgesia using a nurse-implemented sedation protocol. The goals of sedation administration for all study participants are to (i) minimize agitation, (ii) promote ventilator synchrony, (iii) relieve anxiety and pain and (iv) avoid inappropriate oversedation. Critical Care nurses titrate sedatives to achieve a calm, cooperative patient [Sedation Agitation Scale of 3 (acceptable) or 4 (preferred)] [32,33] who is capable of following simple commands (i.e. hand squeezing to command). We do not assess strict adherence to the protocol, but rather, whether the goals of the protocol are achieved. Nurses evaluate SAS scores at least every 4 hrs or at least every 2 hours if dose changes were made on the prior measurement. We record SAS scores on the case report forms every 4 hours. Sedation titration is not mandatory if the patient is on an alternate mode of MV Administration of sedation and analgesia during weaning is according to the following principles: General Principles: 1. The goal is to attain a SAS scale score of 3 or 4; that is, to avoid oversedation or undersedation. 2. Attempts are made to distinguish the etiology of agitation during weaning and to administer appropriate treatment with either sedatives (anxiolytics) or analgesics or both. 3. Regardless of the medication(s) administered, the amount administered (per dose and total daily dose) is individualized and is expected to vary from patient to patient and within patients over time. 4. If possible, attempts are made to minimize sedation/analgesia over time and to minimize the use of continuous IV sedation/analgesia. Oversedation (SAS scores of 1 or 2) 1. In the presence of either intermittent or continuous sedation/analgesia and a SAS score of 1 or 2, sedation/analgesia except in the presence of ventilator dyssynchrony or signs of respiratory distress refractory to supportive measures (e.g., suctioning, bronchodilation, diuresis). 2. If a SAS score > 3 is obtained within 8 hrs after the sedation/analgesia was stopped, a physician should judge whether the patient requires ongoing sedation/analgesia. 3. If the patient is judged to require additional sedation/analgesia, intermittent or continuous sedation/analgesia should be resumed at half the previous dose and adjusted accordingly thereafter to attain a SAS score of 3 to 4 at the discretion of the attending physician. 39

15 Optimal Sedation (SAS scores of 3 to 4) 1. Once SAS scores of 3 to 4 have been obtained, a judgment should be made as to whether the patient requires ongoing sedation/analgesia. 2. If ongoing sedation/analgesia is required, routes of administration, doses administered and dosing intervals can be maintained and preferably decreased. The goal should be to minimize the amount and duration of sedation/analgesia administered over time. Undersedation (SAS scores of 5, 6 and 7) without continuous IV sedation 1. Preferential and sparing use of small doses of intravenous short-acting sedatives (i.e. Midazolam; range: 1 to 5 mg) and analgesics (i.e., Fentanyl; range: 25 to 75 g). 2. When short-acting agents are required more than every 4 hrs, small doses of longer-acting sedatives, (i.e., Lorazepam 1 to 4 mg or or Clonazapem 0.5 to 2.0 mg q 8h) and analgesics (i.e. Morphine 2 to 4 mg q 4-6h), administered IV or orally is encouraged. 3. Infusions of sedative or analgesic agents are reserved for patients refractory to long acting agents or requiring frequent boluses (> every 2 hours) of short-acting agents or persistently attaining SAS scores > When required, continuous infusions of sedative/analgesic agents are titrated to achieve SAS scores of 3 to For SAS scores persistently > 5, additional adjuvant therapies including Olanzepine, Haldol, Respirodone are permitted at the discretion of the attending physician. Undersedation (SAS scores of 5, 6 and 7) with continuous IV sedation/analgesia 1. With SAS scores of 5 to 7, a bolus of short acting agent (sedative/analgesic) should be administered and consideration should be given to increasing the infusion (sedative/analgesic) by approximately 50% (i.e. from Midazolam 2 mg/hr to 3 mg/hr) to achieve a SAS score of 3 to 4 at the discretion of the attending physician. 2. If the SAS score remains persistently > 5 despite at least 2 boluses and at least 2 attempts to increase infusion rates, consideration should be given to adding adjuvant antipsychotic therapy. 3. For SAS scores persistently > 5, additional adjuvant therapies including Olanzepine, Haldol, Respirodone are permitted at the discretion of the attending physician. 40

16 Appendix E7: PEEP/FiO 2 Titration Chart (Both Study Groups) This scale is followed with the objective of obtaining PaO 2 between 60 and 80 mm Hg or SpO 2 between 90 and 95% (or at baseline saturation in chronically hypoxemic and hypercarbic patients) in both groups. SpO 2 is assessed at least every 4-6 hours and arterial blood gases are performed at the discretion of the RRT/physician at least once per day. The PEEP/FiO 2 chart is not mandatory if the patient is on an alternate mode of MV. Acceptable FiO 2 and PEEP pairs are: FiO 2 PEEP (in H 2 O) cm H 2 O cm H 2 O , 6, 7, or 8 cm H 2 O , 9 or 10 cm H 2 O 41

17 Appendix E8: Detailed sedation scores displayed as equivalent SAS in both groups SEDATION equivalent SAS Automated Weaning (n=762) Protocolized Weaning (n=1502) P value* P value** 1 3 (0.39%) 42 (2.80%) < < (16.27%) 225 (14.97%) (32.02%) 377(25.10%) (37.28%) 737 (49.07%) < (12.86%) 119 (7.92%) (1.05%) 1 (0.07%) (0.13) 1 (0.07%) 1 * p-value in binomial distribution ** p-value in multinomial distribution Automated Weaning (n=762) Protocolized Weaning (n=1502) p value OVERSEDATION 127 (16.7%) 267(17.8%) < TARGET (SAS 3 4) 528 (69.3%) 1114 (74.2%) UNDERSEDATION 107(14.0%) 121(8.1%) Equivalent RASS/SAS from Mehta et al. JAMA Nov 21;308(19):

18 Appendix E9: Adverse Events Adverse Events Automated Weaning Protocolized Weaning p-value Nosocomial Pneumonia, n (%) 5 (10.2) 6 (14.0) 0.6 Myocardial Infarction, n (%) 0 3 (7.0) 0.1 Pneumothorax, n (%) 2 (4.1) 2 (4.7) 1.0 Unplanned extubation, n (%), N 1 (2.4), 42 0, Self-extubation, n (%), N 0, 42 1 (2.6), N reflects the number of patients used for the outcome analysis 43

19 Appendix E10: Median Cumulative and Per Diem of Mechanical Ventilation Use of Sedative, Analgesic and Hypnotic Agents Medication Automated Weaning (median, IQR, n) Protocolized Weaning (median, IQR, n) p-value Lorazepam per patient per patient/per day of MV 6.5 (3.0, 9.5), (1.0, 3.0), (3.0, 36.0), (0.2, 2.9), 16 Midazolam per patient 33.0 (16.0, 121.0), (10.0, 255.0), 38 per patient/per day of MV 8.0 (4.5, 31.0), (1.5, 32.0), 33 Propofol per patient (30.0, 2068), (180.0, ), 21 per patient/per day of MV 34.1 ( ), (13.4, 123.3), 19 Morphine per patient 29 (3.0, 78.0), (15.0, 259.0), 20 per patient/per day of MV 6.0 (2.0, 30.4), (3.1, 35.7), 19 Fentanyl or patch (mcg) per patient (300.0, ), (150.0, ), 27 per patient/per day of MV (150.0, ), (68.8, ), 24 Hydromorphone per patient 16.0 (3.0, 50.0), (9.0, 223.0), 12 per patient/per day of MV 1.5 (0.7, 6.3), (2.3, 22.7), 11 Haloperidol per patient 10.0 (2.0, 22.0), (17.5, 112.5), 16 per patient/per day of MV 5.5 (0.4, 10.0), (2.5, 10.1), 15 The unit of the drugs dosage is mg except for one as indicated (mcg)

20 Appendix E11: Subgroup Analyses Based on Humidification Type HH p-value HME p-value AW n=20 PW n=17 AW n=29 PW n=26 RT acceptance, median (IQR), # forms 7 (5, 8), (8, 10), 88 < (7, 9), (8, 9), Physician acceptance, median (IQR), # forms 8 (5, 9), (7, 9), (7, 10), 9 (8, 10), Time to first extubation, median (IQR), n 4 (2, 10), 15 4 (2, 6), (2, 5), 27 9 (3,16,,, Time to successful extubation, median (IQR), n 4 (3, 10), 15 4 (2, 4), (2, 7), (3, 22), Time to successful completion of an SBT median 0 (0, 2), (1.5, 4), (0, 3), 27 4 (2, 10), 25 < (IQR), # forms Total duration of mechanical ventilation, median 10 (8, 20), 19 9 (6, 17), (8, 17), 18.5 (11, 29), (IQR), n ICU length of stay, median (IQR), n 10 (5, 16), 15 6 (5, 9), (5, 12), (6, 28), Nosocomial pneumonia, count (%), n 2 (10.0), 20 1 (5.9), (10.3), 29 5 (19.2), Tracheostomy, count (%), n 4 (20.0), 20 1 (5.9), (13.8), (53.9), Prolonged mechanical ventilation at 21 days, n (%) 1 (6.7), 15 0, , 26 6 (28.6),

21 Appendix E12: Subgroup Analyses Based on COPD AW n=10 COPD PW n=13 p-value AW N=39 Non-COPD PW n=30 p-value RT acceptance, median (IQR), # forms 8 (7, 9), 59 8 (7, 10), (5, 9), (8, 9), 279 < Physician acceptance, median (IQR), # forms 8 (7, 9), 52 9 (8, 10), (5, 9), (7, 10), Time to first extubation, median (IQR), n 1 (1,3), 9 6 (4, 12), (2,6), (2.5, 11.5), Time to successful extubation, median (IQR), n 3 (2,4), (4,19), (2,7.5), 32 5 (2.5, 17), Time to successful completion of an SBT median (IQR), # forms Total duration of mechanical ventilation, median (IQR), n 0 (0,0), 6 1 (0,3), (0, 3), 33 1(0,2), (6, 10), 10 9 (6, 25), (9, 18), (8, 27), ICU length of stay, median (IQR), n 5 (4, 8), (5, 22), (5, 15), 31 9 (5, 25), Nosocomial pneumonia, count (%), n 1 (10.0), 10 1 (7.7%), (10.3), 39 5 (16.7), Tracheostomy, count (%), n 2 (20.0), 10 5 (38.5%), (15.4), (33.3), Prolonged mechanical ventilation at 21 days, n (%) 1 (11.1), 9 2 (20.0), (0), 32 4 (16.7%)