TIBIAL PLATEAU FRACTURES: INTER-RATER RELIABILITY FOR SURGICAL DECISION MAKING

Tibial plateau fractures can result in significant disability and post-traumatic arthritis. They are typically classified using the AO/OTA and Schatzker classifications; however, there is a lack of evidence to define cut-off indications for surgical repair based on joint depression and/or stability (Schatzker I, II, III or AO 41.B1,2,3). The aim of the study was to determine surgeon agreement for surgical indications in a series of real cases deemed ‘borderline’ and to define the most important factors involved in surgical decision making. A survey was designed that included six tibial plateau fracture cases (Schatzker I, II, III, AO Type 41.B1,2,3). These were cases generally deemed borderline for operative management by the senior authors, with depression < 6 mm and condylar widening < 5 mm, and no significant comminution. The survey was sent out through the Canadian Orthopaedic Association (COA), Canadian Orthopaedic Trauma Society (COTS), and Austrian Society of Trauma Surgery (ÖGU). Participants were asked to review x-rays and CT scans as well as a brief clinical history for each patient and then select to offer either surgical or non-surgical management. Questions were then asked to identify and rank factors influencing the decision (fracture pattern, depression, activity level, age, osteoarthritis, comorbidities, malalignment, and instability). Data was collected through Survey Monkey. Data analysis included participant completion rate, demographic data and agreement on surgical indications. Rank sum analysis was performed for most important factors for decision making. The clinical features of cases with high and low treatment agreement were analysed further with a specific emphasis on joint depression. 189 orthopedic surgeons, fellows, and residents responded to the survey. The completion rate was 64% (70% staff, 30% trainees). There were minimal differences between the responses of trainees and staff. Three cases had high agreement (>80%) for either operative or non-operative treatment. The most frequently reported factor in decision making was fracture pattern (86%; rank value 2.0), followed by amount of depression (79%; rank value 2.1) and age (56%). Co-morbidity was the third highest ranked factor. Depression was 5mm and 7mm for the cases offered surgery. Three cases had low agreement (∼50%) for either operative or non-operative treatment. The most frequently reported factors were amount of depression (82%; rank 1.8), followed by fracture pattern (73%; rank 2.0) and age (60%). Absence or presence of knee instability was also deemed important. Depression was ∼4mm for two of the cases and 2mm for the other. The agreement among Orthopaedic Surgeons to operate on a series of ‘borderline’ Type I-III Schatzker tibial plateau fractures ranged between 51 and 99%. Cases with both high and low agreement cited fracture pattern and amount of depression as critically important. This suggests that while surgeons believe those factors are important in their decision-making process they do not uniformly agree on the parameters. The least agreement revolved around cases with depression around 4mm. These results suggest that further research is needed to objectively define the benefits (or lack thereof) of surgical treatment within these common subtypes of fractures.