Comparison of the Stopping Opioids after Surgery (SOS) score to preoperative morphine milligram equivalents (MME) for prediction of opioid prescribing after lumbar spine surgery.

BACKGROUND CONTEXT

Reliable estimation of the likelihood for prolonged postoperative opioid use may aid targeted interventions for high-risk patients. Previous studies have recommended differing methodologies for prediction of sustained postoperative opioid use.

PURPOSE

To compare the performance of the Stopping Opioids after Surgery (SOS) score and preoperative morphine milligram equivalents (MME) for postoperative opioid prescription exposure in a contemporary cohort of lumbar surgery patients.

PATIENT SAMPLE

Adult patients undergoing posterior decompression with or without fusion for degenerative lumbar conditions between January 31, 2016 and May 31, 2019.

STUDY DESIGN/SETTING: Retrospective review at two academic medical centers and three community hospitals.

OUTCOME MEASURES

The primary outcome was sustained postoperative prescription opioid exposure at 3 months and 6 months. Reoperations and readmissions were considered secondarily.

METHODS

The Stopping Opioids after Surgery score and MME were assigned to patients based on data from their preoperative surgical evaluation. Performance for both measures was assessed for all outcomes by discrimination, including c-statistic and receiver-operating curve analysis. Calibration of the low, medium and high-risk strata with the observed rates of postoperative adverse events were examined.

RESULTS

Overall, 4,165 patients were included in this study. Preoperative prevalence of prescription opioid use was 31%. Rates of postoperative opioid prescriptions at 3 months and 6 months, were 3.3% (n=136) and 1.5% (n=61). The c-statistics of preoperative oral MME and SOS score for 3-month sustained opioid prescriptions were 0.64 and 0.78, respectively. The c-statistics of preoperative oral MME and SOS score for 6-month sustained opioid prescriptions were 0.64 and 0.82, respectively. C-statistics of preoperative oral MME and SOS score were much lower for reoperation and readmission, although SOS score outperformed MME for both outcomes.

CONCLUSIONS

The SOS score clinically outperformed oral MME as a predictive measure for outcomes following lumbar spine surgery. The SOS score may be valuable for identifying individuals at high-risk for sustained prescription opioid use and associated adverse events following spine surgery.