A statistical calibration framework for subnational health surveillance from routine data and survey audits

A calibration framework for subnational HIV surveillance in four African countries estimates undiagnosed HIV positivity by district while auditing and correcting selection bias using intermittent surveys.

• Why it matters: Routine health records and satellite data can improve precision, but survey-based estimates are often too infrequent and underpowered for frequent subnational reporting.
• What they did: The method inverts data roles: routine registers form a biased baseline over time, and probability surveys intermittently audit and adjust selection bias, with models and diagnostics for transport across out-of-survey groups and time points.
• The result: Applied to undiagnosed HIV positivity by district, it quantifies selection bias by age and sex and provides a transferable approach for other surveillance settings with similar data constraints.

AI summary of the abstract · may contain errors · not medical advice · read the paper.