Controlled Exponentially Weighted Moving Average Chart in Cardiac Surgery: A Simulation Study Across 9 Italian Cardiac Centers

Abstract

Application of statistical process charts has led to consistent quality production improvement in the industrial sector. Aim of this simulation study is to assess if the use of exponentially weighted moving average chart with control limits (CL) could help to identify mortality trends in a cardiac surgery scenario. Mortality rate of 9 cardiac centers has been continuously monitored by a central clinical governance unit since 2010; prospectively collected monthly mortality rate of calendar year 2013-2014 from each center was used to retrospectively build an exponentially weighted moving average chart; mortality level was set at 4% as per threshold defined by the Italian Ministry of Health recommendation; upper CLs were set as 1.5 standard deviation from the specified level; lowest mortality rate (2.6%) was observed during calendar year 2012-2013, hence that was considered the center of the chart. All centers were considered as 1 entity and consecutively plotted in the chart following a geographic distribution, from North to South. A total number of 4049 operations were performed; 108 patients died while in hospital (2.6%). Different mortality trends that consisted of minor and major out-of-control process defined as a point of the chart outside the upper CLs were demonstrated. In conclusion, mortality trends could have been potentially identified at earlier time points before reaching the 4% limits of mortality; exponentially weighted and controlled chart may facilitate clinical governance units to their monitoring role.