Time series analysis of the admission to the emergency department due to respiratory and cardiovascular diseases between 2010 and 2014 in Kirklareli, Turkey

Mercan Y., Hapcioglu S. B. , Issever H.

JOURNAL OF CLINICAL AND ANALYTICAL MEDICINE, cilt.10, ss.301-306, 2019 (ESCI İndekslerine Giren Dergi) identifier

  • Cilt numarası: 10 Konu: 3
  • Basım Tarihi: 2019
  • Doi Numarası: 10.4328/jcam.6113
  • Sayfa Sayıları: ss.301-306


Aim: The aim of this study was to estimate the admissions to the emergency department due to cardiovascular and/or respiratory diseases for the next twelve months. Material and Method: This research was characterized as an ecological study. The data were obtained from the hospital database between years 2010 and 2014. Emergency department admissions (N=148.169) from >= 15 years due to cardiovascular and/or respiratory diseases were evaluated according to the monthly average. Multiplicative Seasonal Auto-Regressive Integrated Moving Average (SARIMA) modeling method was used for the research. Results: It is observed that the emergency department admissions display seasonal changes. ARIMA(1,1,2)(1,0,1)12 model (MAPE-98,039) was ascertained to be the most suitable model with the success of 99.6% in the predictions. It was predicted that the admissions would be higher in the winter period. Model success for admissions according to disease groups vary between 752% and 89.2% and was estimated the highest level of admissions in January and February. The most suitable models used to estimate the number of admissions were the ARIMA(2,1,3)(1,0,0)12 for respiratory diseases, the ARIMA(2,1,2)(1.0,0)12 for cardiovascular diseases and the ARIMA(1,1,1)(1,0,0)12 for both for cardiovascular and respiratory diseases. It was estimated that the admissions due to cardiovascular diseases which had a conjuncture structure would increase mostly in April and the admissions due to respiratory diseases and both of the diseases would be higher mostly in the winter period. Discussion: SARIMA models are a good prediction model that can be used to estimate emergency department admissions due to cardiovascular and/or respiratory diseases. The estimations derived comprise a good evidence-based source for policymakers and health service providers.