Exponentially weighted moving average scheme with curtailment for detecting COVID-19 cases: a case study

The exponentially weighted moving average (EWMA) scheme has been commonly utilised to monitor shifts in infection rate p. In this research, an enhanced EWMA chart employing the curtailment technique (called Curt-EWMA chart) is used for detecting COVID-19 cases in airports. This chart can also be used for detecting infections associated with other respiratory viruses. The Curt-EWMA chart is able to improve detection effectiveness by minimising an overall performance measure, being the average number of infections (ANI), while maintaining a satisfactory false alarm rate. The utilisation of the ANI as an objective function provides the chart with a commendable detection effectiveness across a broad spectrum of shift sizes. A comparison between the Curt-EWMA and the traditional EWMA scheme is performed across multiple cases. The Curt-EWMA scheme is found to be 42% more effective than the traditional EWMA scheme in terms of ANI. The Curt-EWMA design is validated using a real-life case study in a major airport, where the results show a superiority of the Curt-EWMA scheme over the traditional EWMA by 20% in terms of ANI.