The feasibility and validity of studies comparing orthopedists and non‐orthopedists caring for musculoskeletal injuries: Results of a pilot study

Objective . To review care provided by orthopedists and non‐orthopedists for a variety of common acute injuries to the shoulder, knee, and ankle; and to evaluate the difficulties in assessing results from diagnostic data collected for administrative reporting by comparing these data with data collected by medical record abstraction for the same patients. Methods . The Rochester Epidemiology Project database was used to identify new cases of Colles fracture, knee sprain, ankle sprain, and shoulder sprain/rotator cuff injury among Olmsted county residents >18 years of age, who were evaluated between January 1, 1990 and December 31, 1992. Data were collected from an administrative database and medical records. Results . Although the diagnoses recorded in the administrative database and those obtained from medical record review were the same in 94% of the 500 cases reviewed, only 270 (67.7%) actually represented acute injuries. Medical record review showed that (for the shoulder and ankle) injury severity was highly correlated with treatment specialty: 13 of 15 severe injuries were cared for by orthopedists, while 64 of 68 mild injuries were treated by non‐orthopedists. Diagnostic uncertainty was more common among non‐orthopedists; in 21 of the 149 patients treated by non‐orthopedists (15.1%), the final (i.e., recorded for coding) diagnosis was different from the initial diagnosis. In 14 of these cases, an orthopedist made the final diagnosis. None of the diagnoses made by orthopedists were revised. Conclusion . Current diagnostic coding does not capture injury acuity, severity, or diagnostic uncertainty. Studies that limit themselves to such data for diagnostic information are likely to suffer from biases which can skew results in unpredictable ways. Conclusions drawn from such studies are likely to be flawed.