Low interrater reliability of examiners performing the prone instability test: a clinical test for lumbar shear instability.

Objective: To test the interrater reliability of examiners performing the prone instability test (PIT), a clinical test proposed to identify lumbar shear instability.

Design: Cross-sectional test-retest design examining individuals with mechanical low back pain (LBP).

Setting: University-based musculoskeletal analysis laboratory.

Participants: Individuals (N=30) with mechanical LBP recruited from community sources in a metropolitan region.

Interventions: Not applicable.

Main Outcome Measures: Repeated measures of a clinical examination test proposed to identify lumbar shear instability.

Results: Interrater reliability of examiners' judgments of PIT results were indexed with percentage of agreement and κ statistic. Examiners obtained 63% agreement and κ of .10 (95% confidence interval, -.27 to .47). Adjusted κ values based on prevalence and bias indexes were calculated to evaluate the effect on κ. The prevalence index associated with examiner judgments of the PIT was .43, and bias index was .03. The prevalence-adjusted bias-adjusted κ value was slightly higher than the unadjusted κ value (κ=.27; 95% confidence interval, -.08 to .61).

Conclusions: Results of our study are not consistent with those of previous studies examining the reliability of therapists performing the PIT. We conclude that examiners do not attain acceptable interrater reliability when performing procedures for the PIT based on the information currently provided in the literature. Based on our experience, we suggest further exploration, standardization, and clarification of procedural details to improve therapists' ability to conduct the PIT on individuals with LBP.