STUDY DESIGN: Prospective cohort/predictive validity study. OBJECTIVE: To determine the diagnostic accuracy of common clinical examination items and to construct a preliminary clinical prediction rule for diagnosing hip osteoarthritis (OA) in individuals with unilateral hip pain. BACKGROUND: The current gold standard for the diagnosis of hip OA is a standing antero-posterior (AP) radiograph of the pelvis. Other than for Altman's criteria, little research has been done to determine the accuracy of clinical examination findings for diagnosing hip OA. METHODS: Seventy-two subjects completed the study. Each subject received a standardized history, physical examination, and standing AP radiograph of the pelvis. Subjects with a Kellgren and Lawrence score of 2 or higher based on the radiographs were considered to have definitive hip OA. Likelihood ratios (LRs) were computed to determine which clinical examination findings were most diagnostic of hip OA. Potential predictor variables were entered into a logistic regression model to determine the most accurate set of clinical examination items for diagnosing hip OA. RESULTS: Twenty-one (29%) of the 72 subjects had radiographic evidence of hip OA. A clinical prediction rule consisting of 5 examination variables was identified.  If at least 4 of 5 variables were present, the positive LR was equal to 24.3 (95% confidence interval=4.4-142.1), increasing the probability of hip OA to 91%. CONCLUSION: The preliminary clinical prediction rule provides the ability to a priori identify patients with hip pain who are likely to have hip OA. A validation study should be done before the rule can be implemented in routine clinical practice.

J Orthop Sports Phys Ther., Epub 14 July 2008. doi:10.2519/jospt.2008.2753

KEYWORDS: arthritis, diagnosis, OA, predictive validity

Study Design: Nonexperimental, retrospective design. Objectives: This study was designed to compare clinical diagnostic accuracy (CDA) between physical therapists (PTs), orthopaedic surgeons (OSs), and nonorthopaedic providers (NOPs) at Keller Army Community Hospital on patients with musculoskeletal injuries (MSI) referred for magnetic resonance imaging (MRI). Background: US Army PTs are frequently the first credentialed providers privileged to examine and diagnose patients with musculoskeletal injuries. Physical therapists assigned at Keller Army Community Hospital have also been credentialed with privileges to order MRI studies for several years. Methods and Measures: To reduce provider bias, a retrospective analysis was performed on 560 patients referred for MRI over an 18-month period. An electronic review of each patient’s radiological profile was performed to assess agreement between clinical diagnosis and MRI findings. Data analyses were performed through descriptive statistics and contingency tables. Results:Analysis on agreement between clinical diagnosis and MRI findings produced a CDA of 74.5% (108/145) for PTs, 80.8% (139/172) for OSs, and 35.4% (86/243) for NOPs. There was a significant difference in CDA between PTs and NOPs (P<.001), and between OSs and NOPs (P<.001). There was no difference in CDA between PTs and OSs (P>.05). Conclusions: Clinical diagnostic accuracy by PTs and OSs on patients with musculoskeletal injuries was significantly greater than for NOPs, with no difference noted between PTs and OSs.

J Orthop Sports Phys Ther. 2005;35(2):67-71. doi: 10.2519/jospt.2005.1344

Key Words: diagnostic agreement, direct access, primary care