Postoperative opioid cessation rates based on preoperative opioid use: analysis of common orthopaedic procedures.

AIMS

The aim of this study was to characterize the relationship between pre- and postoperative opioid use among patients undergoing common elective orthopaedic procedures.

PATIENTS AND METHODS

Pre- and postoperative opioid use were studied among patients from a national insurance database undergoing seven common orthopaedic procedures using univariate log-rank tests and multivariate Cox proportional hazards analyses.

RESULTS

A total of 98 769 patients were included; 35 701 patients were opioid-naïve, 11 621 used opioids continuously for six months before surgery, and 4558 used opioids continuously for at least six months but did not obtain any prescriptions in the three months before surgery. Among opioid-naïve patients, between 0.76% and 4.53% used opioids chronically postoperatively. Among chronic preoperative users, between 42% and 62% ceased chronic opioids postoperatively. A three-month opioid-free period preoperatively led to a rate of cessation of chronic opioid use between 82% and 93%, as compared with between 31% and 50% with continuous preoperative use (p < 0.001 for significant changes in opioid use before and after surgery in each procedure). Between 5.6 and 20.0 preoperative chronic users ceased chronic use for every new chronic opioid user. Risk factors for chronic postoperative use included chronic preoperative opioid use (odds ratio (OR) 4.84 to 39.75; p < 0.0001) and depression (OR 1.14 to 1.55; p < 0.05 except total hip arthroplasty). With a three-month opioid-free period before surgery, chronic preoperative opioids elevated the risk of chronic opioid use only mildly, if at all (OR 0.47 to 1.75; p < 0.05 for total shoulder arthroplasty, rotator cuff repair, and carpal tunnel release).

CONCLUSION

Chronic preoperative opioid use increases the risk of chronic postoperative use, but an opioid-free period before surgery decreases this risk compared with continuous preoperative use Cite this article: 2019;101-B:1570-1577.