Forecasting the incidence of acute lymphoid leukaemia in males and females in the Saudi population from 2020 to 2029: application of ARIMA models and public health implications.
Acute lymphoid leukaemia (ALL) is a significant cause of morbidity and mortality globally, with increasing incidence rates observed in Saudi Arabia. Despite advances in treatment, there is a lack of localized, sex-specific forecasts to guide public health interventions.
This study aims to forecast the future incidence of acute lymphoid leukaemia in males and females using ARIMA models.
Saudi national cancer registries data from 1990 to 2019 were used. The dataset was divided into training (80%) and testing (20%) sets. ARIMA models were developed for male and female incidence, with model parameters determined using ACF and PACF plots. Stationarity was assessed using the Augmented Dickey-Fuller test, and model accuracy was validated using MAE, MSE, and MAPE. Forecasts included point estimates and 95% confidence intervals.
For males, the ARIMA (3, 3, 3) model forecasted an increase in ALL incidences from 957 cases in 2020 to 2181 cases in 2029. For females, the ARIMA (4, 3, 0) model projected an increase from 1019 cases in 2020 to 2159 cases in 2029. The models demonstrated high accuracy, with MAE of 1.19895 and 2.749188, MSE of 62.33 and 31.83, and MAPE of 0.6805807 and 1.443453 for males and females, respectively.
Forecasts indicate a substantial rise in ALL incidence among both sexes, highlighting the urgent need for improved surveillance, early detection, and healthcare capacity planning. Incorporating ARIMA modelling into routine monitoring could support proactive resource allocation. Further studies should integrate additional predictive variables to enhance model precision.
This study aims to forecast the future incidence of acute lymphoid leukaemia in males and females using ARIMA models.
Saudi national cancer registries data from 1990 to 2019 were used. The dataset was divided into training (80%) and testing (20%) sets. ARIMA models were developed for male and female incidence, with model parameters determined using ACF and PACF plots. Stationarity was assessed using the Augmented Dickey-Fuller test, and model accuracy was validated using MAE, MSE, and MAPE. Forecasts included point estimates and 95% confidence intervals.
For males, the ARIMA (3, 3, 3) model forecasted an increase in ALL incidences from 957 cases in 2020 to 2181 cases in 2029. For females, the ARIMA (4, 3, 0) model projected an increase from 1019 cases in 2020 to 2159 cases in 2029. The models demonstrated high accuracy, with MAE of 1.19895 and 2.749188, MSE of 62.33 and 31.83, and MAPE of 0.6805807 and 1.443453 for males and females, respectively.
Forecasts indicate a substantial rise in ALL incidence among both sexes, highlighting the urgent need for improved surveillance, early detection, and healthcare capacity planning. Incorporating ARIMA modelling into routine monitoring could support proactive resource allocation. Further studies should integrate additional predictive variables to enhance model precision.
Authors
Kabrah Kabrah, Handoko Handoko, Aljohani Aljohani, Mujalli Mujalli, Alobaidy Alobaidy, Flemban Flemban, Farrash Farrash, Alharbi Alharbi, Alzahrani Alzahrani
View on Pubmed