Multivariate exponentially weighted moving average chart for monitoring patient's progress after cardiac surgery

Statistical process control has emerged in the medical literature after wide expansion in the industry. In clinical monitoring, there are always more than one quality characteristics of interest which are usually correlated. In such cases, multivariate control charts would be deployed to monitor the medical process. In this paper, Multivariate Exponentially Weighted Moving Average control chart (MEWMA) is applied to monitor the patient's progress in the Intensive Care Unit, which is characterised by nine quality characteristics. One difficulty encountered with multivariate control charts is the interpretation of out-of-control signals. The univariate control charts are employed to obtain a rough estimate of the sources of multivariate out-of-control signals. Issues of non-normality in the data are addressed, and suitable transformations are offered. A comparison is made between the performance of EWMA and MEWMA methods in monitoring of patient recovery process. The results clearly show the superiority of MEWMA over univariate EWMA chart.