Respiratory and haemodynamic effects of acute postoperative pain management: evidence from published data.

BACKGROUND This study examines the evidence from published data concerning the adverse respiratory and haemodynamic effects of three analgesic techniques after major surgery; i.m. analgesia, patient-controlled analgesia (PCA), and epidural analgesia. METHODS A MEDLINE search of the literature was conducted for publications concerned with the management of postoperative pain. Information relating to variables indicative of respiratory depression and of hypotension was extracted from these studies. Over 800 original papers and reviews were identified. Of these papers, 212 fulfilled the inclusion criteria but only 165 provided usable data on adverse effects. Pooled data obtained from these studies, which represent the experience of a total of nearly 20,000 patients, form the basis of this study. RESULTS There was considerable variability between studies in the criteria used for defining respiratory depression and hypotension. The overall mean (95% CI) incidence of respiratory depression of the three analgesic techniques was: 0.3 (0.1-1.3)% using requirement for naloxone as an indicator; 1.1 (0.7-1.7)% using hypoventilation as an indicator; 3.3 (1.4-7.6)% using hypercarbia as an indicator; and 17.0 (10.2-26.9)% using oxygen desaturation as an indicator. For i.m. analgesia, the mean (95% CI) reported incidence of respiratory depression varied between 0.8 (0.2-2.5) and 37.0 (22.6-45.9)% using hypoventilation and oxygen desaturation, respectively, as indicators. For PCA, the mean (95% CI) reported incidence of respiratory depression varied between 1.2 (0.7-1.9) and 11.5 (5.6-22.0)%, using hypoventilation and oxygen desaturation, respectively, as indicators. For epidural analgesia, the mean (95% CI) reported incidence of respiratory depression varied between 1.1 (0.6-1.9) and 15.1 (5.6-34.8)%, using hypoventilation and oxygen desaturation, respectively, as indicators. The mean (95% CI) reported incidence of hypotension for i.m. analgesia was 3.8 (1.9-7.5)%, for PCA 0.4 (0.1-1.9)%, and for epidural analgesia 5.6 (3.0-10.2)%. Whereas the incidence of respiratory depression decreased over the period 1980-99, the incidence of hypotension did not. CONCLUSIONS Assuming a mixture of analgesic techniques, Acute Pain Services should expect an incidence of respiratory depression, as defined by a low ventilatory frequency, of less than 1%, and an incidence of hypotension related to analgesic technique of less than 5%.