Rethinking Causal and Time Series Models in Health Forecasting : Prospects and Challenges

Time series and causal models have been longstanding health forecasting technique applied to both clinical and non-clinical decision making. To date the most common approaches of time series forecasting includes exponential smoothing, ARIMA, SARIMA, Time Series Regression etc. However, like most forecasting models that predates the times series and causal approaches, the methods have evolved hence the preponderance of many different types of times series and causal models used to aid forecasting. This review, explores the use of time series models in contemporary clinical and non-clinical decision making. It explores the growing interests, challenges and strengths of the ensemble of techniques developed to augmented and consolidate effective medical forecasting in a constantly changing healthcare environment.