Electronic Thesis and Dissertation Repository


Master of Science




Dr. Dianne Bryant

2nd Supervisor

Dr. J Robert Giffin



We asked 217 sport surgery and 135 total knee arthroplasty (TKA) patients to complete a questionnaire (e-Visit) before attending their two and six-week post-operative appointment. Our primary objective was to use the questions asked of patients prior to their appointment to develop a model that could be used as web-based e-Visit to predict early post-operative adverse events. Gold standard comparison was the surgeon’s opinion as to the presence or absence of an event at follow-up. Secondary objective was evaluation of a simplified model. We found good area under the curve (AUC) statistics (0.76 (95% CI 0.69 - 0.84) and 0.80 (95% CI .74 - .85)) and good sensitivity (0.70, 0.83) and specificity (0.70, 0.80) for the two- week model and for the six-week model respectively. The simplified models and raw-data models were similar. Future work should improve the web-based interface, include educational content, and be validated using a large multicenter RCT.