An integrated framework for modelling respiratory disease transmission and designing surveillance networks using a sentinel index.

Defining epidemiologically relevant placements for sentinel units is critical for establishing effective health surveillance systems. We propose a novel methodology to identify optimal sentinel unit locations using network approaches and metapopulation modelling. Disease transmission dynamics were modelled using syndromic data on respiratory diseases, integrated with road mobility data. A generalizable sentinel index is introduced as a metric that evaluates the suitability of a site to host a sentinel unit, based on topological metrics and metapopulation dynamics. A case study using syndromic data from primary health care attendances in Bahia, Brazil, validated the relevance of existing sentinel units while identifying opportunities for local re-designs to improve disease surveillance coverage.