Abstract
Objective The purpose of this article is to determine whether patient-reported history items
are predictive of shoulder pathology and have the potential for use in triaging patients
with shoulder pathology to orthopaedic outpatient clinics.
Setting It is set at two tertiary orthopaedic clinics.
Patients All new patients reporting pain and/or disability of the shoulder joint were prospectively
recruited. A total of 193 patients were enrolled, 15 of whom withdrew, leaving 178
patients composing the study sample.
Design Patients completed a questionnaire on the history of their pathology, then the surgeon
took a thorough history indicating the most likely diagnosis. The clinician then performed
appropriate physical examination. Arthroscopy was the reference standard for those
undergoing surgery and magnetic resonance imaging (MRI) with arthrogram for all others.
We calculated the sensitivity, specificity, and likelihood ratios (LRs) of history
items alone and in combination. We used the LRs to generate a clinical decision algorithm.
Main Outcome Measures Diagnosis was determined through arthroscopy or MRI arthrogram. Reporting was standardized
to ensure review of all structures.
Results The physical examination and history agreed in 75% of cases. Of those that did not
agree, the physical examination misdirected the diagnosis in 47% of our cases. In
particular, history items were strong predictors of anterior and posterior instability
and subscapularis tears and were combined in a tool to be utilized for screening patients.
Conclusion The patient-reported history items were effective for diagnosing shoulder pathology
and should be considered for use in a triaging instrument.
Keywords
shoulder - diagnosis - history
