Opioid use following cervical spine surgery: trends and factors associated with long-term use

Background Context: Limited or no data exist evaluating risk factors associated with prolonged opioid use following cervical arthrodesis. Purpose: The objectives of this study were to assess trends in postoperative narcotic use among preoperative opioid users (OUs) versus non-opioid users (NOUs) and to identify factors associated with postoperative narcotic use at 1 year following cervical arthrodesis. Study Design/Setting: This is a retrospective observational study. Patient Sample: The patient sample included 17,391 patients (OU: 52.4%) registered in the Humana Inc claims dataset who underwent anterior cervical fusion (ACF) or posterior cervical fusion (PCF) between 2007 and 2015. Outcome Measures: Prolonged opioid usage was defined as narcotic prescription filling at 1 year following cervical arthrodesis. Methods: Based on preoperative opioid use, patients were identified as an OU (history of narcotic prescription filled within 3 months before surgery) or a NOU (no preoperative prescription). Rates of opioid use were evaluated preoperatively for OU and trended for 1 year postoperatively for both OU and NOU. Multivariable regression techniques investigated factors associated with the use of narcotics at 1 year following ACF and PCF. Based on the model findings, a web-based interactive app was developed to estimate 1-year postoperative risk of using narcotics following cervical arthrodesis (http://neuro-risk.com/opiod-use/ or https://www.neurosurgerycost.com/opioid/opioid_use). Results: Overall, 87.4% of the patients (n=15,204) underwent ACF, whereas 12.6% (n=2187) underwent PCF. At 1 month following surgery, 47.7% of NOUs and 82% of OUs had a filled opioid prescription. Rates of prescription opioids declined significantly to 7.8% in NOUs versus 50.5% in OUs at 3 months, but plateaued at the 6- to 12-month postoperative period (NOU: 5.7%–6.7%, OU: 44.9%–46.9%). At 1 year, significantly higher narcotic prescription filling rates were observed in OUs compared with NOUs (45.3% vs. 6.3%, p<.001). Preoperative opioid use was a significant driver of 1-year narcotic use following ACF (odds ratio [OR]: 7.02, p<.001) and PCF (OR: 6.98, p<.001), along with younger age (≤50 years), history of drug dependence, and lower back pain. Conclusions: Over 50% of the patients used opioids before cervical arthrodesis. Postoperative opioid use fell dramatically during the first 3 months in NOU, but nearly half of the preoperative OUs will remain on narcotics at 1 year postoperatively. Our findings serve as a baseline in identifying patients at risk of chronic use and encourage discontinuation of opioids before cervical spine surgery.