Sex difference in outcomes and risk factors for lumbar decompressive surgery

Evaluate the sex difference in perioperative patient-reported outcome measures (PROMs), 12-month PROM risk-factors, and 12-month PROM prediction models for lumbar decompressive surgery.

The study included 2938 cases. The primary 12-month PROM was Oswestry Disability Index (ODI), and secondary PROMs were back and leg pain. Other outcomes included 12-month EQ-5D, and 3-month opioid / non-opioid analgesic intake and employment. Twelve-month PROM predictive models were developed with 2008-2022 data and validated on 2023 data. Sex-specific PROM model predictive performance was compared to non-specific models (both sexes in model). Model calibration was by R², and discrimination by area-under-operating-curve (AUC) to predict an acceptable symptom state (ODI ≤ 22%).

Across seven metrics, an absolute 4-15% more females than males had a poor outcome. The risk-factor effect on 12-month PROMs was equal for each sex; however, females have more PROM risk-factors. Females have higher prevalence of depression/anxiety, rheumatoid arthritis, osteoarthritis, mobility aided, extreme BMI, scoliosis, vertebral body fracture, and more severe and longer duration of preoperative symptoms linked to poorer PROM. In contrast, males have higher prevalence of diabetes and current smokers. Model validation found sex-specific 12-month ODI model predictive performance (R² 40%, 95%CI 0.31-0.48; AUC = 0.815, 95%CI 0.765-0.864) was equivalent to the non-specific model (R² 42%, 95%CI 0.33-0.50; AUC = 0.819, 95%CI 0.770-0.868).

Compared to males, females had poorer outcomes, more PROM risk-factors, and accelerated age-related thoracolumbar degenerative change. However, a similar proportion of each sex achieves improvement in each outcome metric after surgery. Sex-specific 12-month PROM models are not indicated.