British Journal of Anaesthesia 99 (4): (2007) doi: /bja/aem206 Advance Access publication July 25, 2007 Observational study of perioperat

Transcription

1 British Journal of Anaesthesia 99 (4): (2007) doi: /bja/aem206 Advance Access publication July 25, 2007 Observational study of perioperative Ptc CO2 and Sp O2 in nonventilated patients receiving epidural infusion or patient-controlled analgesia using a single earlobe monitor (TOSCA) A. Kopka 1 *, E. Wallace 2, G. Reilly 2 and A. Binning 2 1 Department of Anaesthesia, Greater Glasgow University Hospitals, Southern General Hospital, Glasgow, UK. 2 Department of Anaesthesia, Greater Glasgow University Hospitals, Gartnavel General Hospital, Glasgow, UK *Corresponding author: Department of Anaesthesia, Greater Glasgow University Hospitals, Southern General Hospital, 1345 Govan Road, Glasgow G51 4TF, UK. a.kopka@doctors.org.uk Background. TOSCA, a non-invasive monitor with a single earlobe probe incorporating a Stow Severinghaus electrode and optical sensor (Linde Medical Sensors AG, Basel, Switzerland), has previously been used with ventilated patients and in sleep laboratories. We recorded transcutaneous carbon dioxide pressures (Ptc CO2 ) and oxygen saturations (Sp O2 ) in non-ventilated patients to investigate opioid-induced respiratory depression. Methods. This observational cohort study included 28 ASA I and II patients, monitored between 10 p.m. and 6 a.m., before and after elective major laparotomy. After operation, patients were kept on oxygen, 4 litre min 21, and received either bupivacaine (0.1%) containing fentanyl (2 mg ml 21 ) via epidural catheter (epidural analgesia group, EPI; n¼14) or morphine via patientcontrolled analgesia infusion pump (PCA-morphine group, PCA; n¼14). Results. The preoperative median (lower/upper quartile) Ptc CO2 was similar in both groups at around 5.5 kpa, but significantly higher after operation in PCA with 6.9 kpa (5.6/7.3) (P¼0.02), accompanied by a longer hypercarbia time.6 kpa of 6.6 h (0.1/8.0) (P¼0.04), and lower respiratory rates of 13.9 breaths min 21 (13.3/15.4) (P¼0.04). In EPI, the corresponding results were 5.8 kpa (5.5/6.0), 1.2 h (0.1/4.3), and 16.2 breaths min 21 (14.8/16.7). The perioperative median Sp O2 in both groups was comparable within the normal range, although generally higher when on supplemental oxygen (P¼0.26). The Sp O2 time,94% was similar in both groups (P¼0.33) as were pain scores (P¼0.25). Conclusions. Ptc CO2 recording in patients on PCA-morphine and supplemental oxygen revealed hypercapnia in the presence of normal respiratory rates and Sp O2 values. This is recommended as an easy and sensitive monitor of respiratory depression and may have a role in the safe administration of opioid-analgesia. Br J Anaesth 2007; 99: Keywords: capnometry; carbon dioxide, hypercapnia; carbon dioxide, hypercarbia; oxygenation, tissue, cutaneous; oxygenation, tissue, subcutaneous Accepted for publication: June 6, 2007 The TOSCA monitor (Linde Medical Sensors AG, Basel, Switzerland) generates estimations of arterial oxygen saturations and carbon dioxide partial pressures. Processed transcutaneous readings correlate closely with directly obtained arterial blood gas results. 12 Hypoventilation leading to hypoxaemia, hypercarbia, and respiratory acidosis is an unwanted central effect of general anaesthesia. It is also commonly associated with opioid use, irrespective of the route of administration. Increasing blood carbon dioxide partial pressures may lead to arrhythmias, hypoxaemia, and myocardial depression. Part of this work was presented at the 2005 annual meeting of the European Society of Anaesthesiology in Vienna, Austria, and published in the European Journal of Anaesthesiology 2005; 22 (Suppl 34): 16. # The Board of Management and Trustees of the British Journal of Anaesthesia All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Kopka et al. The work of breathing, sympathetic stimulation, stress levels, oxygen consumption, and intracranial pressures can all be increased. The latter will cause headaches and confusion and may even result in CO 2 -narcosis 3 or coma. Patients on opioid-containing patient-controlled analgesia (PCA) or epidural infusion pumps are routinely prescribed supplemental oxygen to avoid hypoxaemia. Respiratory monitoring on general wards is commonly based on the assessment of respiratory rates and, less frequently, pulse oximetry. However, respiratory rates may be influenced by various factors, and the diagnostic value of pulse oximetry is limited when on supplemental oxygen. 4 Importantly, outside the operating theatre or intensive care unit, carbon dioxide pressures are not routinely monitored for as the real measure of respiratory function. This prospective, observational clinical pilot study was primarily designed to generate baseline data for continuous Ptc CO2 and Sp O2 in patients breathing spontaneously without evident cardio-pulmonary disease. A secondary purpose was to carry out a comparison of those patients managed with either epidural or patient-controlled infusion analgesia after elective major laparotomy. Finally, we hoped to evaluate transcutaneous carbon dioxide measurement with TOSCA as a method of non-invasive respiratory monitoring. Methods After institutional ethics committee approval and after obtaining written informed consent, we included 32 ASA I and II patients. Excluded was anyone with a BMI.30 (kg m 22 ), or if diagnosed with acute or chronic cardiopulmonary, renal and vascular disease, skin anomalies, oedema, or with perioperative oxygen requirements.4 litre min 21. Protocol Patients were allocated to receive either bupivacaine (0.1%) including fentanyl (2 mg ml 21 ) via continuous epidural infusion (epidural analgesia group¼epi) or morphine via PCA infusion pump (PCA-morphine group¼pca) to provide postoperative analgesia. The general anaesthetic technique was comparable for both groups. In EPI, a thoracic epidural catheter was inserted before induction of anaesthesia, either at T6 for an anticipated rooftop incision or at T10 for an anticipated midline incision. A test dose of 2 ml of bupivacaine (0.5%) was administered. The epidural infusion was started intraoperatively at 10 ml h 21 and adjusted after operation to maintain adequate analgesia. The rate was limited to 18 ml h 21, thus allowing for a maximum hourly total of 36 mg fentanyl. Analgesia was supplemented by i.v. remifentanil, starting at 0.2 mg kg 21 min 21 and adapted during the operation. In PCA, morphine was given intraoperatively and included in the total requirements recorded until 6 a.m. on the first postoperative day. The morphine PCA bolus dose was set at 1 mg with a 5 min lockout, without a background infusion. After operation, all patients were kept on oxygen, 4 litre min 21, via Hudson mask and transferred to the high dependency unit when pain free and maintaining oxygen saturations of.94%. Pain management in both groups followed established protocols and pain was classified according to the institutional assessment score, that is 0, no pain at rest and no pain on movement; 1, no pain at rest but slight pain on movement; 2, intermittent pain at rest and moderate pain on movement; 3, continuous pain at rest and severe pain on movement. Measurements Data were collected continuously from 10 p.m. until 6 a.m. on both the pre- and postoperative night using a non-invasive single earlobe sensor (TOSCA monitor; Linde Medical Sensors AG, Basel, Switzerland). 5 The sensor temperature is selectable between 378C and 458C, in steps of 0.58C, utilizing two independent circuits to provide a safe and reliable control with an accuracy of +0.28C. For this study, we used the default temperature setting of 428C. Oxygen saturations were measured optically with a light-emitting diode in the red (660 nm) and infrared (880 nm) spectrum. The accuracy for Sp O2 is +2 digits in the range % with a resolution of 1%. Carbon dioxide tensions were calculated by determining the ph of an electrolyte solution, deducted from the potential difference between a miniaturized ph glass measurement and an Ag/AgCl reference electrode (Stow Severinghaus type electrode). The in vitro response time for Ptc CO2 is,50 s and the in vitro drift,0.5% h 21. Hypercarbia was defined as transcutaneous carbon dioxide pressures 6.0 kpa, and termed hypercapnia if caused by hypoventilation. Desaturation was any reading of Sp O2,94%. Both definitions were applied in accordance with the definitions used in the institutions biochemistry department based on international standards. Data analysis Specifically designed computer software (Download 2001 for TOSCA on CD-ROM) was purchased in conjunction with the monitor. The software enables both transfer and analysis of stored data. A summary of descriptive statistics, including a variety of graphs, tables, and times spent above or below defined Ptc CO2 or Sp O2 values can be obtained. Statistical comparison was performed with SAS software, version 9.1. The descriptive statistics for continuous variables were mean, standard deviation, median, and lower and upper quartiles (Table 1; Table 2). The post- vs preoperative comparison between the two groups was carried out using analysis of covariance (ANCOVA; 568

3 Perioperative Ptc CO2 and Sp O2 in non-ventilated patients monitored with TOSCA Table 3). Postoperative pain scores and respiratory rates were compared with Mann Whitney Wilcoxon test. Statistical significance was defined as P,0.05. Results Thirty-two patients were enrolled. Two patients decided not to participate, after initial consent. A further two were excluded after completion of preoperative measurements. One patient was confused, most probably morphine-induced. Another patient withdrew from the study due to sleep disruption from the alarm tone. Thus, a total of 28 patients (ASA I and II) completed this study, equally distributed between the two groups, EPI (n¼14) and PCA (n¼14). Patient characteristics and opioid intake are presented in Table 1. Both groups were well matched for age and BMI. Twelve female patients took part, four in EPI and eight in PCA. Of the 16 male patients, 10 were in EPI and six in PCA. The median (lower/upper quartile) opioidconsumption, recorded until 6 a.m. on the first postoperative day, was 152 mg (144/180) for fentanyl in EPI and 52.5 mg (32/64) for morphine in PCA, respectively. The surgical procedures included: gastrectomy (n¼10) carried out via abdominal rooftop incision; anterior resection (n¼5), ileal conduit (n¼2), pan-proctocolectomy (n¼5), reversal of Hartmann s (n¼2), and right hemicolectomy (n¼4), all facilitated by abdominal midline incision. Table 2 summarizes pre- and postoperative findings. Before operation, the median Ptc CO2 was similar in both groups at around 5.5 kpa. After operation, this result was found to be 1.3 kpa higher in PCA with 6.9 kpa (5.6/7.3), but increased only slightly in EPI to 5.8 kpa (5.5/6.0). A significantly longer postoperative hypercarbia time.6kpa of 6.6 h (0.1/8.0) and lower respiratory rates of 13.9 breaths min 21 (13.3/15.4) were also observed in PCA. In comparison, the postoperative hypercarbia time.6 kpain EPI was insignificantly reduced to 1.2 h (0.1/4.3) from 1.5 h (0.1/3.8), and respiratory rates were higher by an Table 1 Patient characteristics and opioid intake. EPI, epidural infusion analgesia group; PCA, patient-controlled analgesia group; SD, standard deviation; Med, median; LQ, lower quartile; UQ, upper quartile; BMI, body mass index (kg m 22 ); S-Opioid, total fentanyl (Fent) (mg) or morphine (Mo) (mg) intake until 6 a.m. on the first postoperative day Table 2 Pre- and postoperative results for both groups. MPtc CO2, median CO 2 (kpa); Mt.6 kpa, median Ptc CO2 time.6 kpa (h); MSp O2, median Sp O2 (%); Mt,94%, median Sp O2 time,94% (h); Resp. rates, breaths min 21 Variable Med (pre) LQ UQ Med (post) LQ UQ EPI MPtc CO Mt.6 kpa MSp O Mt,94% Resp. rates Pain score PCA MPtc CO Mt.6 kpa MSp O Mt,94% Resp. rates Pain score average of more than 2 breaths min 21. The perioperative median Sp O2 in both groups was comparable within normal values, but generally higher on the first postoperative night. Significant oxygen desaturations,94% were not noted. Post- vs preoperative statistical comparison (Table 3) failed to demonstrate differences of median Sp O2 times,94% (P¼0.33) or median Sp O2 readings (P¼0.26). However, the postoperative results for hypercarbia, that is the median Ptc CO2.6kPa(P¼0.02), and hypercarbia time, that is Ptc CO2 time.6 kpa, were significantly higher in PCA (P¼0.04), and accompanied by significantly lower respiratory rates (P¼0.04). Postoperative pain scores were similar in both groups (P¼0.25). Discussion Considerable and prolonged hypercapnia was observed in the group receiving PCA-morphine. Importantly, this was accompanied by normal respiratory rates and oxygen saturations. All opioids, irrespective of the route of administration but dependent on effect-site concentrations, may cause significant or even life-threatening respiratory depression. The true incidence, however, is unknown. 6 The absence of a consensus definition for hypoventilation or respiratory depression, aggravated by insufficient Group Variable Mean SD Med LQ UQ Age All EPI PCA BMI All EPI PCA S-Opioid EPI Fent PCA Mo Table 3 Post- vs preoperative comparison between both groups. Significant changes at P,0.05 can only be demonstrated for carbon dioxide pressures (DMPtc CO2 ) and Ptc CO2 times spent.6 kpa(dmt.6 kpa). The PCA vs EPI difference is presented as mean difference (95% confidence interval) Variable P-value PCA EPI difference DMPtc CO (0.14 to 1.34) kpa DMt.6 kpa (0.16 to 5.09) h DMSp O (20.41 to 1.47)% DMt,94% (20.52 to 1.48) h 569

4 Kopka et al. monitoring and underreporting, can be blamed. 7 Evidence of opioid-induced respiratory depression is generally defined by low respiratory rates, oxygen saturations, or elevated carbon dioxide pressures. 8 Some regard the administration of opioid-receptor antagonists such as naloxone as evidence; a misleading concept since its application is commonly triggered by the presence of any of the former three parameters, usually low respiratory rates. However, respiratory rates can vary 910 and may be depressed due to many reasons. 11 Thus, rate counts are perhaps particularly useful as a trend indicator in conjunction with other assessment tools. The value of pulse oximetry to detect hypoventilation is limited in the presence of supplemental oxygen. 4 Carbon dioxide monitoring, however, as the real measure of respiratory function appears to be widely underestimated and its importance is confirmed by the findings in this study. A clinically useful, graded definition of respiratory depression should be based on carbon dioxide partial pressures, which can be easily and safely measured transcutaneously. Limitations of this method are caused in principle by any condition that will result in skin disease, hypo-perfusion, oedema, or very high carbon dioxide pressures. 12 Considering the level of respiratory monitoring on general wards mainly based on rate counts and, to a lesser degree, pulse oximetry, one must question the practice of admitting patients with opioid-containing PCA-infusions on supplemental oxygen. Our own observations suggest that this level of monitoring is all but inadequate. Evolving respiratory depression, that is opioid-induced hypercapnia, may not be detected in time to prevent its potentially very serious side-effects. 3 Furthermore, the routine administration of supplemental oxygen has been blamed for masking and therefore delaying the detection of hypoventilation and appropriate respiratory care. 413 However, based on currently available evidence, this practice must be regarded as an accepted clinical standard. On the other hand, patients receiving epidural opioids are often initially admitted to postoperative anaesthesia care or high dependency units. The argument for choosing this higher level of postoperative surveillance, apart from obvious medical or surgical reasons, is safe epidural management. However, epidural analgesia can also be safely managed on general wards The fact is that opioid-induced respiratory depression is often not appropriately monitored for a frequent omission, also seen in these studies investigating pain relief and safety after major surgery. The sequelae of hypercapnia are therefore commonly unnoticed, ignored, or quietly accepted. Furthermore, it is unknown what impact these prolonged elevated carbon dioxide levels may have on patient outcome. Epidural fentanyl did not cause any relevant hypoventilation at a dosage regime commonly employed in the UK. Interestingly, opioid-induced hypercapnia is more often associated with the epidural drug administration. The incidence of respiratory depression, defined by arterial carbon dioxide pressures above a predetermined value, was stated to be 6%, but only 1.3% when opioids are given i.v. or i.m. 8 This relatively high incidence may vary and is most probably influenced by choice of opioid. Nevertheless, we would have required a higher number of participants in the epidural analgesia group to draw a firm conclusion. Another weakness of our study is its design as a prospective cohort. Suitable patients were recruited from different surgical lists and received either thoracic epidural or patient-controlled analgesia, depending on the preference of the anaesthetist involved. This method does not fulfil statistical randomization criteria and could also lead to selection bias. However, our primary goal to generate observational baseline data did not necessitate prospective randomization. In conclusion, considerable and prolonged respiratory depression observed as transcutaneous hypercapnia in the presence of normal respiratory rates and oxygen saturations was observed in the group receiving PCA-morphine and supplemental oxygen. The impact on postoperative outcome and recovery from surgery remains to be evaluated. Transcutaneous carbon dioxide monitoring is recommended as an easy and sensitive monitor of respiratory depression and may have a role in the safe administration of opioid-analgesia. Acknowledgements The authors are indebted to Mr Stephen Henderson, Academic Research Assistant, for his help in conducting this study, and Dr Chris Weir, Robertson Centre for Biostatistics, University of Glasgow, UK, for his valuable statistical advice and data analysis. References 1 Rohling R, Biro P. Clinical investigation of a new combined pulse oximetry and carbon dioxide tension sensor in adult anaesthesia. J Clin Monit Comput 1999; 15: Gisiger PA, Palma JP, Eberhard P. OxiCarbo, a single sensor for the non-invasive measurement of arterial oxygen saturation and CO 2 partial pressure at the ear lobe. Sens Actuators B Chem 2001; 76: Smyth E, Egan TD. Apneic oxygenation associated with patientcontrolled analgesia. J Clin Anesth 1998; 10: Fu ES, Downs JB, Schweiger JW, Miguel RV, Smith RA. Supplemental oxygen impairs detection of hypoventilation by pulse oximetry. Chest 2004; 126: Linde Medical Sensors AG. TOSCA Operating Manual. Basel, Switzerland 6 Rathmell JP, Lair TR, Nauman B. The role of intrathecal drugs in the treatment of acute pain. Anesth Analg 2005; 101: S Ko S, Goldstein DH, VanDenKerkhof EG. Definitions of respiratory depression with intrathecal morphine postoperative analgesia: a review of the literature. Can J Anesth 2003; 50: Cashman JN, Dolin SJ. Repiratory and haemodynamic effects of acute postoperative pain management: evidence of published data. Br J Anaesth 2004; 93: Gagliardi L, Rusconi F. Respiratory rate and body mass in the first three years of life. Arch Dis Child 1997; 76:

5 Perioperative Ptc CO2 and Sp O2 in non-ventilated patients monitored with TOSCA 10 Hooker EA, O Brien DJ, Danzl DF, Barefoot JAC, Brown JE. Respiratory rates in emergency department patients. J Emerg Med 1989; 7: Loadsman JA, Hillman DR. Anaesthesia and sleep apnoea. Br J Anaesth 2001; 86: Cuvelier A, Grigoriu B, Molano LC, Muir J-F. Limitations of transcutaneous carbon dioxide measurements for assessing long-term ventilation. Chest 2005; 127: Downs JB. Has oxygen administration delayed appropriate respiratory care? Fallacies regarding oxygen therapy. Respir Care 2003; 48: Scott DA, Beilby DSN, McClymont C. Postoperative analgesia using epidural infusions of fentanyl with bupivacain: a prospective analysis of 1,014 patients. Anesthesiology 1995; 83: Liu SS, Allen HW, Olsson GL. Patient-controlled epidural analgesia with bupivacaine and fentanyl on hospital wards: prospective experience with 1,030 surgical patients. Anesthesiology 1998; 88: Flisberg P, Rudin A, Linner R, Lundberg CJ. Pain relief and safety after major surgery. A prospective study of epidural and intravenous analgesia in 2696 patients. Acta Anaesthesiol Scand 2003; 47: