Can assessment of human assumed central sensitisation improve the predictive accuracy of the STarT Back screening tool in acute low back pain?

Background: The STarT Back Screening Tool (SBT) is recommended to provide risk-stratified care in low back pain (LBP), yet its predictive value is moderate for disability and low for pain severity. Assessment of human assumed central sensitisation (HACS) in conjunction with the SBT may improve its predictive accuracy. Objectives: To examine whether assessment of HACS in acute LBP improves the predictive accuracy of the SBT for LBP recovery at six months in people with acute non-specific LBP. Design: A prospective longitudinal study. Method: Data were drawn from the UPWaRD study. One hundred and twenty people with acute non-specific LBP were recruited from the community. Baseline measures included SBT risk status, nociceptive flexor withdrawal reflex, pressure and heat pain thresholds and conditioned pain modulation. Primary outcome was the presence of LBP (pain numeric rating scale ≥1 and Roland Morris Disability Questionnaire score ≥3) at six-month follow-up. Regression coefficients were penalised using the least absolute shrinkage and selection operator technique to select predictor variables. Internal validation was performed using ten-fold cross-validation. Results/findings: SBT risk status alone did not predict the presence of LBP at six months (area under receiver operating characteristic curve [AUC] = 0.58). Adding measures of HACS to the SBT did not improve discrimination for whether LBP was present at six months (AUC = 0.59). Conclusions: This study confirmed the suboptimal predictive accuracy of the SBT, administered during acute LBP, for LBP recovery at six months. Assessment of HACS in acute LBP does not improve the predictive accuracy of the SBT.