The goal of this study was to describe changes in STarT Back Tool (SBT) categorization after 4-weeks of outpatient physical therapy and to evaluate predictive capabilities of SBT categorization when administered at multiple time points. Initial assessment information is commonly used to predict long-term outcomes, however does not account for changes that occur after beginning treatment. Changes in SBT categorization during the course of treatment have potential to provide additional prognostic information that could positively impact management for low back pain (LBP). Patients (n = 123) receiving non-standardized physical therapy care for LBP were administered the SBT at intake and 4-weeks later to evaluate SBT changes that were described as ‘improved’ (SBT categorization changed from medium to low, high to low, or high to medium risk), ‘stable’ (SBT categorization remained low or medium risk), or ‘worsened’ (SBT categorization changed from low to medium, low to high, medium to high, or remained high risk). Clinical outcomes consisted of pain intensity and self-reported disability. Relative contributions of SBT categorization (at intake and 4-weeks) and SBT change patterns as predictors of 6-month clinical outcomes were assessed using separate multiple regression models while controlling for other prognostic variables. The majority of patients (81.8%) initially categorized as SBT high risk were categorized differently by the SBT at 4-weeks. Eleven percent of patients were described as ‘worsened’ based on SBT category changes patterns at 4-weeks. Prediction of 6-month disability scores were improved when considering intake, 4-week, or 4-week change for SBT categorization, with SBT high risk consistently providing unique contributions.
Repeated SBT assessment during the episode of physical therapy has potential to provide additional information for prediction of disability at 6 months. Future studies ought to examine optimal management strategies for patients who have a greater SBT risk following physical therapy intervention.