PREDICTING FUTURE OUTLIER STATUS FOR TOTAL KNEE ARTHROPLASTY SURGEONS

Introduction The New Zealand Joint Registry monitors surgeon-level arthroplasty outcomes using funnel plots. If future outlier status could be predicted, this would provide an opportunity to intervene and potentially improve performance prior to exceeding outlier thresholds. This study evaluated the predictive value of revision rate trend and exceeding a lower “alert” threshold for subsequent identification as an outlier. Methods Three separate analyses were undertaken. The first used the revision within five years timeframe and evaluated trend in revision rate for 2021 outliers over the previous ten years. The second summarized the predictive value of exceeding the 80% control limit for being identified as an outlier (exceeding the 95% control limit) within the subsequent three years. The third identified outliers and summarized the percentage of these that exceeded the 80% control limit in the three years preceding identification as an outlier. These two latter analyses identified outliers from 2017 through 2021. Results Six of eight outliers with a tracked longitudinal performance showed an upward trend in revision rate, relative to the control limit, prior to exceeding the outlier threshold. Thirty-six surgeons exceeded the 80% control limit in the five reporting years (2014 to 2018). Ten of these surgeons became outliers within three years of first exceeding the lower threshold; a significantly higher proportion than those remaining below the 80% control limit (p<0.001). All seven surgeons identified as outliers from 2017 to 2021 were outside the 80% control limit in at least one of the three years prior to exceeding the outlier threshold. Conclusion Trend in revision rate and exceeding the 80% control limit have predictive value for future outlier status. Including this information in reports could facilitate earlier changes to practice and avoidance of outlier status.