STATISTICAL MODELLING AND FORECASTING OF DIABETIC AND RETINOPATHY IN KAURA NAMODA METROPOLIS

This study employed statistical modeling and forecasting to analyze the prevalence of diabetic retinopathy in Kaura Namoda Metropolis, with data sourced from General Hospital Kaura Namoda covering a five-year period. The dataset comprised annual and monthly case records, coded and analyzed using descriptive and inferential statistical techniques. Measures of central tendency (mean, median, mode) and dispersion (variance, standard deviation, and coefficient of variation) were computed to understand data distribution. Time series decomposition isolated the trend, seasonal, and irregular components, while the Augmented Dickey-Fuller (ADF) test confirmed stationarity after first differencing. Autoregressive Integrated Moving Average (ARIMA) modeling was applied, and model selection was based on the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), with ARIMA(1,1,1) emerging as the best fit. Diagnostic checks via the Ljung-Box Q-test indicated no significant autocorrelation in residuals (p > 0.05). The fitted model forecasted a 14.7% increase in cases over the next three years, projecting a mean monthly case load rising from 26.4 (±4.5) to approximately 30.3 (±4.8). Seasonal peaks were observed in March–May and September–November, possibly linked to dietary habits, agricultural cycles, and healthcare access variations. These results suggest a statistically significant upward trend (t = 3.42, p < 0.01) in diabetic retinopathy incidence. The study recommends targeted community awareness campaigns, early screening at primary healthcare centers, expanded ophthalmic training, and investment in diagnostic equipment to curb the projected rise. Findings provide a statistically validated evidence base for policymakers to implement data-driven public health interventions.