Patient Factors Associated with High Opioid Consumption after Common Surgical Procedures Following State-Mandated Opioid Prescription Regulations
Autumn Sacklow,Ajai K Malhotra,Ethan Jones,Peter Callas,Mayo Fujii,Christos Colovos
DOI: https://doi.org/10.1097/XCS.0000000000000185
2022-06-01
Abstract:Background: State regulations have decreased prescribed opioids with more than 25% of patients abstaining from opioids. Despite this, 2 distinct populations of patients exist who consume "high" or "low" amounts of opioids. The aim of this study was to identify factors associated with postoperative opioid use after common surgical procedures and develop an opioid risk score. Study design: Patients undergoing 35 surgical procedures from 7 surgical specialties were identified at a 620-bed tertiary care academic center and surveyed 1 week after discharge regarding opioid use and adequacy of analgesia. Electronic medical record data were used to characterize postdischarge opioids, complications, demographics, medical history, and social factors. High opioid use was defined as >75th percentile morphine milligram equivalents for each procedure. An opioid risk score was calculated from factors associated with opioid use identified by backward multivariate logistic regression analysis. Results: A total of 1,185 patients were enrolled between September 2017 and February 2019. Bivariate analyses revealed patient factors associated with opioid use including earlier substance use (p < 0.001), depression (p = 0.003), anxiety (p < 0.001), asthma (p = 0.006), obesity (p = 0.03), migraine (p = 0.004), opioid use in the 7 days before surgery (p < 0.001), and 31 Clinical Classifications Software Refined classifications (p < 0.05). Significant multivariates included: insurance (p = 0.005), employment status (p = 0.005), earlier opioid use (odds ratio [OR] 2.38 [95% CI 1.21 to 4.68], p = 0.01), coronary artery disease (OR 0.38 [95% CI 0.16 to 0.86], p = 0.02), acute pulmonary embolism (OR 9.81 [95% CI 3.01 to 32.04], p < 0.001), benign breast conditions (OR 3.42 [95% CI 1.76 to 6.64], p < 0.001), opioid-related disorders (OR 6.67 [95% CI 1.87 to 23.75], p = 0.003), mental and substance use disorders (OR 3.80 [95% CI 1.47 to 9.83], p = 0.006), headache (OR 1.82 [95% CI 1.24 to 2.67], p = 0.002), and previous cesarean section (OR 5.10 [95% CI 1.33 to 19.56], p = 0.02). An opioid risk score base was developed with an area under the curve of 0.696 for the prediction of high opioid use. Conclusions: Preoperative patient characteristics associated with high opioid use postoperatively were identified and an opioid risk score was derived. Identification of patients with a higher need for opioids presents an opportunity for improved preoperative interventions, the use of nonopioid analgesic therapies, and alternative therapies.