Better Forecasts, Fairer Access: Modelling family planning where data is scarce

A/Prof Ole Maneesoonthorn, 2026

Purpose

Reliable health planning depends on good data – but data is often scarce at the subnational level. Ole Maneesoonthorn set out to build a robust statistical approach for projecting contraceptive method supply shares even where data are limited.

Practice

The proposed Bayesian probabilistic model combines information across contraceptive methods, regions and time, and was tested and benchmarked against existing methods to assess its predictive performance.

Output

The research delivers a reproducible modelling framework generating more accurate forecasts than existing approaches, with methodology and code released as a practical, open-access tool for researchers and public health agencies.

Outcome and impact

Governments, international organisations and health programme managers can now make more informed decisions on contraceptive procurement, resource allocation and service planning – strengthening health system resilience and progress towards Sustainable Development Goals 3 and 5, with an adaptable methodology offering wider benefits wherever reliable forecasts are needed despite limited data.