Lumbar fusion surgery is usually prompted by chronic back pain, and many patients receive long-term preoperative opioid analgesics. Many expect surgery to eliminate the need for opioids. We sought to determine what fraction of long-term preoperative opioid users discontinue or reduce dosage postoperatively; what fraction of patients with little preoperative use initiate long-term use; and what predicts long-term postoperative use. This retrospective cohort study included 2491 adults undergoing lumbar fusion surgery for degenerative conditions, using Oregon's prescription drug monitoring program to quantify opioid use before and after hospitalization. We defined long-term postoperative use as ≥4 prescriptions filled in the 7 months after hospitalization, with at least 3 occurring >30 days after hospitalization. Overall, 1045 patients received long-term opioids preoperatively, and 1094 postoperatively. Among long-term preoperative users, 77.1% continued long-term postoperative use, and 13.8% had episodic use. Only 9.1% discontinued or had short-term postoperative use. Among preoperative users, 34.4% received a lower dose postoperatively, but 44.8% received a higher long-term dose. Among patients with no preoperative opioids, 12.8% became long-term users. In multivariable models, the strongest predictor of long-term postoperative use was cumulative preoperative opioid dose (odds ratio of 15.47 [95% confidence interval 8.53-28.06] in the highest quartile). Cumulative dose and number of opioid prescribers in the 30-day postoperative period were also associated with long-term use. Thus, lumbar fusion surgery infrequently eliminated long-term opioid use. Opioid-naive patients had a substantial risk of initiating long-term use. Patients should have realistic expectations regarding opioid use after lumbar fusion surgery.
Lumbar fusion surgery infrequently eliminated the use of long-term opioids, despite patient expectations. Both preoperative and perioperative doses were predictors of long-term postoperative use.
aDepartment of Family Medicine, Oregon Health and Science University, Portland, OR, USA
bDepartment of Medicine, The Oregon Institute for Occupational Health Sciences, Oregon Health and Science University, Portland, OR, USA
cHealthInsight Oregon, Portland, OR, USA
dBiostatistics Group, OHSU-PSU School of Public Health, Oregon Health and Science University, Portland, OR, USA
eInjury and Violence Prevention Program for the State of Oregon, Oregon Health Authority, Portland, OR, USA
fSystems Science Program, Portland State University, Portland, OR, USA
Corresponding author. Address: Department of Family Medicine, Oregon Health and Science University, 3181 SW. Sam Jackson Park Rd, Mail code FM, Portland, OR 97239, USA. Tel.: 503-494-1694; fax: 503-494-2746. E-mail address: email@example.com (R.A. Deyo).
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
Received November 30, 2017
Received in revised form February 08, 2018
Accepted March 02, 2018