Analysis of individual patient data from clinical trials: epidural morphine for postoperative pain.

BACKGROUND Individual patient information from clinical trials is infrequently available, but can provide insights for clinical trials and practice. METHODS We analysed individual patient information from five randomized trials (913 patients) of i.v. patient-controlled analgesia (IVPCA) plus epidural placebo, morphine sulphate (MS) 5 mg, or extended-release epidural morphine (EREM; DepoDur) at doses of 5-30 mg, to explore effects of a range of epidural morphine doses. Pain and opioid requirement on first and second postoperative days, dose-response, clinically relevant comparisons of IVPCA without epidural morphine, 5 mg MS, and 10 mg EREM, and relationship between patient rating and other measures were described. RESULTS There were three strong findings. Epidural morphine resulted in greater patient satisfaction, despite higher rates of adverse events. Those describing their analgesic medication as 'very good' or 'excellent' used IVPCA opioid less and had pain scores significantly below the global mean, whereas those describing their medication as 'poor' or 'fair' had pain scores significantly above the mean. Epidural morphine meant less need for postoperative IVPCA opioid than epidural placebo. The therapeutic gain with EREM was lower pain scores with less IVPCA opioid. Moderate or severe pruritus was more common with IVPCA plus epidural morphine, whatever the formulation, compared with IVPCA plus placebo. CONCLUSIONS Analysis of individual patient data from high-quality clinical trials provides important insights into characteristics of new agents not immediately apparent from original trials, and also informing clinical practice. Prophylactic epidural morphine provides a better patient experience than IVPCA alone.