Postoperative sleep disturbance: influences of opioids and pain in humans.

STUDY OBJECTIVES To test the hypothesis that opioids and pain contribute independently to postoperative sleep disturbance, 10 women undergoing surgery requiring a low abdominal incision for treatment of benign gynecologic conditions were randomized to receive either epidural opioid (fentanyl) (n=6) or epidural local anesthetic (bupivacaine) (n=4) for intraoperative and postoperative analgesia. DESIGN N/A. SETTING N/A. PATIENTS OR PARTICIPANTS N/A. INTERVENTIONS N/A. MEASUREMENTS Polysomnography was performed in a standard patient room on the preoperative and first three postoperative nights. Pain at rest and with coughing was evaluated using a visual-analogue pain scale each evening and morning. RESULTS On the first postoperative night, rapid eye movement (REM) sleep was abolished in all patients. On the third postoperative night, the mean +/- SE REM sleep time increased significantly (p=.003) to 9.8% +/- 3.1% in the fentanyl group, and 12.9% +/- 3.8% in the bupivacaine group. Conversely, light non-REM (NREM) sleep (%stage 1 + %stage 2) was higher on the first postoperative night and significantly lower on the third postoperative night (p=0.011). Between group comparison revealed only that the mean % slow-wave sleep (SWS) in the fentanyl group (6.0%, 2.0%, and 14.7%) was different from the bupivacaine group (7.8%, 9.1%, and 10.6%) in the postoperative period after adjusting for the preoperative night % SWS (p=0.021). Pain was well controlled in all patients, but was slightly better controlled in the fentanyl group than in the bupivacaine group on postoperative night 2 (p=0.024). There was no statistically significant association between pain score and any polysomnographically defined stage. CONCLUSION Postoperative patients suffer a profound sleep disturbance even when opioids are avoided and pain is well controlled.