Using time series analysis to predict affecting factors of thallium myocardial perfusion scan service usage frequency

Heat diseases have become the second main cause of death in recent years. Cardiac examinations play important roles to offer related medical diagnosis and treatments for reducing the mortality rate. Tl-201 myocardial perfusion scan is a non-invasive cardiac examination. It has easier procedure for diagnosis and extremely lower risk of complications. The major aim of this study is to find the distribution trend of the usage number of thallium scan services according to the weather, medical department, and socioeconomic variables. We try to explore the influence of these variables toward the number of thallium scan services, and find out the best model to predict exam usage in a hospital database. This is a longitudinal retrospective study. By using Box-Jenkins method, we selected the most suitable model of the number of total thallium scan examinations and construct a predictive model. Our data came from a teaching regional hospital in southern Taiwan from 2008 to 2013. Independent variables including weather, numbers of outpatient visiting medical divisions, the number of patients receiving cardiac catheterization, and a socioeconomic Issue Weighted Stock Price Index was selected to predict the number of thallium scan service usage. ARIMA(0,1,1) was identified to be the most suitable model of numbers of Tl-201 myocardial perfusion scan usage per month. The lowest temperature, average moisture and the number of visiting outpatients at cardiovascular division are identified as three significantly predict factors of thallium scan service usage. These information can provide strategic references in management of equipment and human resources for hospitals and healthcare organizations.