Background Multimodal analgesia protocols have shortened hospitalizations after total joint arthroplasty. It is unclear whether individual components of these protocols are associated with respiratory depression during phase I postanesthesia recovery.

Objectives To test the hypothesis that sedating analgesics used in a multimodal protocol are associated with an increased rate of phase I postanesthesia respiratory depression.

Methods Our Department of Anesthesiology records were searched to identify patients undergoing total joint arthroplasty with a multimodal analgesia protocol, including peripheral nerve blockade, from 2008 through 2012. Patient records were reviewed for episodes of postanesthesia respiratory depression, and potential causative factors were abstracted and analyzed for potential associations. Respiratory depression was defined as apnea, hypopnea, oxyhemoglobin desaturations, or episodes of severe pain despite moderate to profound sedation.

Results Of 11,970 patients who underwent joint arthroplasty, 2836 (23.7%; 237 per 1000 cases; 95% confidence interval [95% CI], 214–262) had episodes of respiratory depression. A higher rate of respiratory depression was observed among patients who underwent general anesthesia (312 per 1000 cases; 95% CI, 301–323) than neuraxial anesthesia (144 per 1000 cases; 95% CI, 135–153) (P < 0.001). With both anesthetic techniques, respiratory depression was associated with preoperative use of gabapentin (>300 mg) (P < 0.001 for both anesthesia groups) and sustained-release oxycodone (>10 mg) (P = 0.01 for general anesthesia; P = 0.008 for neuraxial anesthesia).

Conclusions Use of medications with long-acting sedative potential was associated with increased risk of respiratory depression during phase I anesthesia recovery. These effects were more pronounced when used in conjunction with general anesthesia than with neuraxial anesthesia.