A risk-adjusted EWMA chart with dynamic probability control limits for monitoring survival time

Developing control charts for monitoring health-care systems has attracted many researchers attention in recent years. Considering the fact that continuous data have more information rather than discrete data, survival time after cardiac surgery is considered as a continuous quality characteristic in this paper. Accelerated failure time regression model is used to adjust the risk of patients' conditions and surgeon groups on the survival times. After that, a risk-adjusted exponentially weighted moving average control chart is proposed to monitor the standardized residuals of accelerated failure time regression. Since the false alarm rate of each patient change dramatically, dynamic probability control limits have been extended to address the issue. The proposed method is evaluated in Phase II in terms of average run length criterion. The results indicate that the proposed method has acceptable performance in identifying the out-of-control state in the process. Furthermore, considering the effect of surgeon groups in the accelerated failure time regression model leads to improvement in the performance of the proposed method.