Context: Establishing reliable palpatory tests continues to be a critical, yet elusive, step in osteopathic medical research and evidence-based clinical practice.
Objective: The authors investigated the interobserver reliability of common osteopathic palpatory tests used to evaluate the lumbar spine.
Design and methods: Subjects (N=119) were recruited from the faculty, staff, and students of Kirksville (MO) College of Osteopathic Medicine (KCOM) of A.T. Still University of Health Sciences. Three osteopathic medical examiners residency trained in neuromusculoskeletal medicine initially evaluated lumbar segments on subjects from one subgroup (n=42) in a blinded assessment. The examiners performed palpatory tests of tenderness and tissue texture changes, as well as--in three planes--vertebral positional asymmetry and motion asymmetry. Kappa statistics (kappa) were used to evaluate interobserver reliability. Following a period of consensus training, subjects from another subgroup (n=77) were evaluated in a blinded assessment for those palpatory tests that seemed most likely to produce reliable findings. The interobserver reliability was then re-evaluated.
Results: During the initial evaluation of interobserver reliability, kappa ranged from -0.02 to 0.34, within the poor-to-fair reliability range. Following consensus training, reliability improved, rising into the moderate range for tissue texture changes (kappa=0.45) and into the substantial range for tenderness assessments (kappa=0.68). Reliability for positional asymmetry in the transverse plane (kappa=0.34) and rotational motion asymmetry (kappa=0.20) were improved but remained in the fair range.
Conclusion: The authors concluded that consensus training improved the interobserver reliability of common osteopathic palpatory tests of the lumbar spine. In two of the four tests that were studied--tissue texture and tenderness--acceptable kappa values for clinical tests were achieved after consensus training.