Responsiveness of the Bournemouth questionnaire in determining minimal clinically important change in subgroups of low back pain patients
Newell, David; Bolton, Jennifer E.
STUDY DESIGN: Prospective single cohort. OBJECTIVE: To determine the ability of the Bournemouth Questionnaire (BQ) to distinguish between improved and nonimproved patients who present with either short (acute) or long (subacute/chronic) duration low back pain (LBP), and with either high or low baseline scores (severity). SUMMARY OF BACKGROUND DATA: Recent evidence suggests that the responsiveness of outcome measures used to determine clinical change is dependent on the chronicity and severity of the condition. METHODS: Data from 437 back patients undergoing chiropractic treatment were used for analysis. Patients completed the BQ before treatment and 4 weeks later. Patients also completed the Patient Global Impression of Change scale at follow-up. Responsiveness was determined by calculating Standardized Response Means (SRM) and by the area under the receiver operator curve (ROC) with best cut-point analysis. The minimal clinically important change (MCIC) was calculated by the change score with the best balanced sensitivity and specificity. RESULTS: The responsiveness of the BQ at 4 weeks was dependent on both duration and severity of the condition. As expected, the responsiveness of the total BQ was greater in improved compared to nonimproved patients in the acute (SRM [95% confidence interval], 1.9 [1.7-2.0] and 1.2 [0.9-1.5], respectively), as well as in the subacute/chronic group (SRM, 1.7 [1.5-1.8] and 0.5 [0.3-0.7]), respectively. For the psychological domains, SRMs in the acute patients failed to distinguish improved from nonimproved patients (SRM [95% confidence interval], 1.3 [1.1-1.4] and 0.9 [0.5-1.2] for anxiety, and 0.9 [0.8-1.0] and 0.8 [0.5-1.2] for depression). In acute and subacute/chronic patients, the MCIC for the total BQ was 26 and 18 points, respectively. In patients with lower and higher BQ scores at baseline, the MCIC was 10 and 31 points, respectively. CONCLUSION: The BQ can distinguish between improved and nonimproved LBP patients but the amount of change needed to achieve this is lower in more chronic patients and in individuals with less severe presentation at baseline.