How well do dog shoulder exams predict arthroscopic instability

Abstract

OBJECTIVE: To determine the diagnostic validity of commonly used physical examination maneuvers for shoulder instability. STUDY DESIGN: Retrospective study. ANIMALS: Dogs (n=24) referred for shoulder arthroscopy. METHODS: Results of physical maneuvers and arthroscopic findings were recorded and sensitivity, specificity, positive likelihood ratios (LR+), and negative likelihood ratios (LR-) were calculated for each of 4 physical examination test findings for arthroscopic changes in the medial, lateral, cranial, or caudal compartments of the shoulder joint viewed in dorsal recumbency by lateral and craniomedial portals. RESULTS: Distribution of compartment changes was: medial (17 dogs), caudal (15), cranial (12), and lateral (5). The biceps test had a moderate effect (LR+=9) on post-test probability of cranial compartment changes and a small effect on post-test probability of lateral and caudal compartment changes (LR+=3 and 2.4, respectively). Hyperabduction had a minimal effect and mediolateral instability test had a small effect (LR+=1.64 and 2.68, respectively) on post-test probability of medial compartment changes. Craniocaudal instability test had little to no effect on post-test probability of changes in any compartment. CONCLUSIONS: Physical examination tests evaluated were limited in their ability to predict the type of arthroscopic pathology in this study population. CLINICAL RELEVANCE: Clinicians should understand that a diagnostic test performs inconsistently based on prevalence of a condition in a given patient population. The use of likelihood ratios can assist clinicians in determining the probability of intraarticular changes from a group with a differing prevalence than the patient population presented.