INTRODUCTION: Psychosocial risk factors are important determinants of postoperative spine outcomes. Psychosocial risk is often determined preoperatively; however, early postoperative status has shown stronger relationships with 1-year outcomes and may have greater relevance for postoperative decision making. The STarT Back Screening Tool (SBT) is a 9-item measure developed for efficient screening in primary care to determine risk of low back pain-related disability and direct rehabilitation treatment. There is limited evidence on the utility of postoperative psychosocial risk stratification using the SBT in patients after lumbar spine surgery. The aim of this study was to examine the association between 3-month psychosocial risk based on the SBT and patient-reported disability, physical function, and pain at 1-year after surgery.
METHODS: Data from 251 patients (mean age = 61.0 years, 115 (46%) females, 120 (48%) fusion surgery, 52 (21%) revision surgery) enrolled in the Vanderbilt Comprehensive Spine Registry and with preoperative, 3-month, and 1-year patient-reported outcomes and 3-month SBT were analyzed. Outcomes included disability (Oswestry Disability Index), physical function (PROMIS), back and leg pain (Numeric Rating Scale), and pain interference (PROMIS). Three-month SBT scores were used to classify patients as low, medium, and high risk. Separate hierarchical multivariable regressions were conducted predicting each 1-year outcome controlling for age, race, insurance type, smoking, opioid use, fusion, revision, ambulation status, comorbidities, and preoperative and 3-month outcome score and depression in the first block and adding 3-month SBT risk category in the second block. The association of 3-month SBT risk category with 1-year outcomes was assessed with change in r-squared, beta coefficient and 95% confidence interval, and standardized beta effect size.
RESULTS: Based on 3-month SBT scores, 182 (73%) patients were low risk, 47 (19%) were medium risk, and 22 (9%) were high risk. The increase in adjusted r-squared for SBT was significant for all outcomes (p < 0.05). After controlling for confounders, patients classified as high risk had greater 1-year disability (beta = 7.0 [95% CI = 0.4 to 13.6]), back pain (beta = 1.4 [95% CI = 0.3 to 2.5]), and leg pain (beta = 1.5 [95% CI = 0.3 to 2.7]) compared to patients at low risk. Patients classified as high and medium risk category had lower 1-year physical function (beta for high risk = -3.2 [95% CI = -6.2 to -0.2], beta for medium risk = -2.4 [95% CI = -4.4 to -0.4]) and higher pain interference (beta for high risk = 6.2 [95% CI = 2.3 to 10.2], beta for medium risk = 3.2 [95% CI = 0.5 to 5.9]) compared to patients at low risk. Standardized beta effect sizes for SBT were small, ranging from 0.10 to 0.18.
DISCUSSION: Postoperative risk stratification using the SBT developed for primary care appears to have utility for postoperative rehabilitation decision making in patients who have undergone lumbar spine surgery. Future work is needed to determine whether targeted treatment strategies for patients at medium and high risk based on the SBT can improve disability, physical function, and pain.