Reliability of the Buttock Applied Strain test to uncover radicular pain in Low Back Pain patients

BACKGROUND: Low-back pain (LBP) pathophysiological conditions include nociceptive back pain, somatic referred pain, radicular pain (RP), and radiculopathy. Differential diagnosis is challenging; guidance may come from patients' thorough clinical history and physical examination, and particularly, for lumbar RP, from the evaluation of subjective responses of injured lumbar nerves to a strain applied at the buttock (Buttock Applied Strain, BUAS-test). METHODS: In a sample of n=395 consecutive LBP patients, sensitivity, specificity, and prior probability (positive and negative predictive values, PPV and NPV, respectively) of the BUAS-test were evaluated against two Reference tests: the Straight Leg Raising Test (SLRT) and the painDETECT (PD) questionnaire. Multinomial Logistic Regression (MLR) and χ2 analyses were used to evaluate the BUAS-test outcomes dependence upon independent variables (gender, age group, pain localization, SLRT, and PD outcomes). Cohen's Kappa statistic was used to assess inter-rater agreement. RESULTS: Against PD, the BUAS-test showed a sensitivity of 92%, specificity, and PPV of 100%, respectively, and NPV of 82%; against the SLRT, sensitivity and NPV of 82%, respectively, and specificity and PPV of 40%, respectively. Interrater agreement Cohen's Kappa was 0.911. Significant associations were found between BUAS-test outcomes and pain localization, SLRT, and PD outcomes, but not with gender or age group predictors. MLR showed significant congruent relationship between BUAS-test and the PD outcomes. CONCLUSION: Among LBP patients, the BUAS-test showed satisfactory sensitivity, specificity, Prior Probability, and interrater reliability, and thus, it may be considered a useful adjunctive tool to uncover RP in LBP patients. For results generalization, more research, in different clinical settings other than pain clinics, is needed.