Practice variation may occur when there is no standardised approach to specific clinical problems and there is a lack of scientific evidence for alternative treatments. Practice variation suggests that a segment of the patient population may be managed sub-optimally, and indicates a need for further research in order to establish stronger evidence-based practice guidelines. We surveyed Australian orthopaedic surgeons to examine practice variation in common orthopaedic presentations. In February 2009, members of the Australian Orthopaedic Association were emailed an online survey, which collected information regarding experience level (number of years as a consultant), sub-specialty interests, state where the surgeon works, on- call participation, as well as five common (anecdotally controversial) orthopaedic trauma cases with a number of management options. Surgeons were asked to choose their one most likely management choice from the list provided, which was either surgical or non-surgical in nature. A reminder was sent two weeks later. Exploratory regression was modeled to examine the predictors of choosing surgical management for each case and overall.

Of 760 surgeons, 358 (47%) provided responses. For undisplaced scaphoid fractures, respondents selected short-arm cast (53%), ORIF (22%), percutaneous screw (22%) and long-arm cast (3%). Less experienced (0 to 5 years) (p=0.006) and hand surgeons (p=0.008) were more likely to operate. For a displaced mid-shaft clavicle fracture, respondents selected non-operative (62%), plating (31%) and intramedullary fixation (7%). Shoulder surgeons were more likely to operate (p<0.001). For an undisplaced Weber B lateral malleolus fracture, respondents selected plaster cast or boot (59%), lateral plating (31%), posterior plating (9%) and no splinting (2%). For a displaced Colles fracture in an older patient, respondents selected plating (47%), Kirschner wires (28%), cast/splint (23%) and external fixation (1%). Less experienced (p<0.001) and hand surgeons (p=0.024) were more likely to operate. For a two-part neck of humerus fracture in an older patient, respondents selected non-operative (74%), locking plate (14%), and hemiarthroplasty (7%). Shoulder surgeons were more likely to operate (p<0.001). Accounting for all answers in multiple regression modeling, it was found that more experienced surgeons (>15 years) were 25% less likely to operate (p=0.001). Overall, there was no difference among sub-specialties, or whether a surgeon participated in an on-call roster.

Considerable practice variation exists among orthopaedic surgeons in the approach to common orthopaedic problems. Surgeons who identify with a sub-specialty are more likely to manage conditions in their area of interest operatively, and more experienced surgeons are less likely to recommend surgical management.