Validity of the new Backache Index (BAI) in patients with low back pain

Abstract

Background context: The Backache Index (BAI) is applied to patients with low back pain (LBP) in order to help therapists, doctors, and surgeons perform physical examinations easily. It is carried out within a short space of time (<2 min) without using inclinometric instruments.

Purpose: To explore the reliability, validity, and responsiveness of this new Backache Index in patients with LBP, which can fulfil the existing need for a reliable routine examination in the clinical environment.

Study design/setting: Patients with LBP filled in disability questionnaires; pain rating scales and physical impairment tests were completed in function of construct validity and correlation studies. A subgroup was evaluated for interobserver and test-retest reliability, and a second group was reassessed after two active treatment sessions to verify the responsiveness compared with other examined variables.

Patient sample: In total, 75 patients with subacute LBP (3-12 weeks) participated in a randomized controlled study.

Outcome measures: The validity of the BAI was explored through a correlation with the standard Oswestry LBP Disability Index (ODI), the McGill LBP Questionnaire Index (MPQ), and the Visual Analogue Scale (VAS).

Methods: The BAI consisted of a scoring system that includes pain factors and stiffness estimation at the end of a series of five different lumbar movements of a patient standing in an erect position.

Results: The correlations between the separate outcomes and the BAI ranged from 0.61 to 0.76 (p<.001). The interobserver reliability between two experienced observers for the five outcome scores was good (intraclass correlation coefficient [ICC]>0.86) and even perfect for the BAI (ICC=0.96). A BAI change of one unit is able to exclude a measurement error. A significantly good correlation (p<.001) was found between the BAI at baseline, the ODI (R=0.62), and the total degree of pain rating index (MPQ-PRI-T) (R=0.57), a moderate correlation with the total number of chosen adjectives from the whole list of adjectives (MPQ-NWC-T) (R=0.48), and the VAS (R=0.47), but a lower correlation was found with the MPQ-Quality of life index (R=0.43). The effect size and discriminative ability of the measures were explored after two treatment sessions of deep transverse friction myotherapy by means of study of the receiver operating characteristics curve (ROC) and the greatest area under the curve (AUC). The greatest level of distinction was found for the MPQ-PRI-T and the BAI (AUC>0.93), followed by the ODI (AUC=0.92). A lower level of distinction was found for the MPQ-NWC-T and the VAS (AUC>0.82).

Conclusions: The BAI appears to be a reliable and valid assessment of overall restricted spinal movements in case of LBP and discriminates between successful and unsuccessful treatment outcome.