RESUMO
Implant documentation in total knee arthroplasty (TKA) must be both accurate and comprehensive. A new system enables surgeons to complete a secure, web-based operative report for each surgery. This study evaluated implant documentation using this system and hypothesized that it would be as reliable as the current standard. This was a retrospective study of all primary and revision TKA performed at a single academic institution from January 1, 2015 to March 31, 2015, by eight adult reconstruction surgeons who used the web-based system. Electronic medical records and implant logbooks were chosen as a composite standard reference for implant documentation. Cohen's kappa statistic and Pearson's correlation coefficients were used to determine agreement between the system and the standard reference. Mean kappa value for entire system was 0.916 ± 0.152 (p = 0.015) indicating "almost perfect" agreement (as per Landis and Koch's method) with standard reference. Pearson's correlation coefficient was 0.926 ± 0.147 (p = 0.021) further corroborating this excellent agreement. There was "substantial" or "fair" agreement for items pertaining to cement and augments. Overall, sensitivity was 0.98 (95% confidence interval: 0.71-0.98) and specificity 0.93 (95% confidence interval: 0.53-0.98) indicating that the system was very effective at documenting whether or not an implant was placed. The system exhibited significantly greater sensitivity than specificity (p = 0.027). This study demonstrated excellent performance of this novel system in point-of-care TKA implant documentation. Further research is needed to validate its use in arthroplasty of other joints and its potential for documentation of biomedical device implantation in other fields such as cardiothoracic surgery and gastroenterology.