Tracking progress towards Sustainable Development Goal 3.2 in Kenya using time series models

Area of Science:

• Public Health
• Biostatistics
• Demography

Background:

• Sustainable Development Goal (SDG) 3.2 aims to reduce the under-five mortality rate (UFMR) to below 25 deaths per 1000 live births by 2030.
• Sub-Saharan Africa faces challenges in child survival, with high UFMRs and stalled progress.
• Kenya's UFMR trends are critical for evaluating child survival strategies.

Purpose of the Study:

• Model the likelihood of Kenya achieving the SDG 3.2 target by 2030.
• Analyze historical trends in Kenya's under-five mortality.
• Forecast future UFMR trajectories.

Main Methods:

• Fitted autoregressive integrated moving average (ARIMA), autoregressive fractionally integrated moving average (ARFIMA), and hybrid models to national UFMR data (1995-2022).
• Selected ARIMA (0,2,1) as the best-fitting model based on information criteria, predictive accuracy, and residual diagnostics.
• Validated the model using mean absolute error, root mean square error, mean absolute percentage error, and an 80/20 train-test split.

Main Results:

• Kenya's UFMR shows a slight decline, but the rate of decrease is slowing.
• Projected UFMR for 2030 is 27.8 deaths per 1000 live births (95% PI: 25.2-30.3), exceeding the SDG 3.2 goal.
• Achieving SDG 3.2 requires an accelerated annual decline of approximately 2.43 fatalities per 1000 from 2023 onwards.

Conclusions:

• Kenya's UFMR has decreased significantly, but meeting the SDG 3.2 target by 2030 is unlikely without enhanced interventions.
• Accelerating progress necessitates improvements in maternal and child health services, community interventions, and addressing social determinants.
• Future child mortality monitoring and prediction can be improved with higher-quality data and advanced modeling techniques.