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^st, 2016 and May 31^st, 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:

SOS score and MME were assigned to patients based on data from their pre-operative 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, 4165 patients were included in this study. Pre-operative 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.

Conclusion:

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.