Health Forecasting Using Time Series Analysis in Biomedicine and Public Health: A Case Study of Salah al-Din Governorate

Abstract

     This study analyzes temporal trends in selected health and demographic indicators in Salah al-Din Governorate during 2010–2022, examines their associations with population growth and mortality, and develops a model explaining variation in mortality using the available disease indicators. The analysis used 13 annual observations for viral hepatitis, malaria, food poisoning, diabetes, cardiovascular disease, population growth, and mortality. Descriptive statistics, Pearson correlation, multiple linear regression, and predictive-fit accuracy measures were applied. Cardiovascular disease had the highest annual mean (6265.692), followed by viral hepatitis (3397.846). Population growth correlated positively with diabetes (r=0.55) and cardiovascular disease (r=0.69), while mortality correlated with malaria (r=0.56) and diabetes (r=0.55). The regression model showed relatively high explanatory power (R²=0.895; adjusted R²=0.819; F=11.868; p≈0.003). Malaria, food poisoning, diabetes, and cardiovascular disease had statistically significant coefficients at the 0.05 level, whereas viral hepatitis was not significant. In-sample fit measures were RMSE=22.069, MAE=12.812, and MAPE=9.016%. The findings reveal substantial temporal variability in disease burden and support integrating health and demographic indicators in public-health planning, while interpreting the observed relationships as statistical associations rather than causal effects.