Quick Detection Updating Variable Life-Adjusted Display

The variable life-adjusted display was the earliest risk-adjusted charting procedure developed for monitoring the performance of surgeons. The risk adjustment is done using a simple monitoring statistic which is the difference between the predicted probability of death of a patient from an operation and the binary surgical outcome, zero for death within a 30-day period and one for survival. The risk adjustment is simple to understand, and the procedure has since been widely used for monitoring cardiac operations and in many other applications. The original procedure did not come with any signaling rule and various signaling rules have since been developed. One critical problem remains is the 30-day wait before a surgical outcome can be used by the procedure even though if death has occurred earlier. The 30-day wait results in unnecessary delay in detection when a deterioration occurs. We study a method that updates a surgical outcome as soon as there is any change in order to allow for a quicker detection of deterioration. The traditional and updating procedures are compared in terms of run length performance. Both procedures are illustrated and compared using the case study data of two surgeons.