OBJECTIVES: Individual-based models (IBMs) are increasingly utilized to simulate disease transmission and guide public health policies. In theory, these models rely on extensive societal data concerning human movement and interactions to simulate behavior at spatial and temporal levels. However, it remains unclear what types and quality of data are employed in practice. This systematic review aims to provide an overview of the data used for calibrating IBMs that model infectious disease spread.

METHODS: A systematic literature search was conducted on Embase and Medline databases. Articles describing the data used to calibrate IBMs used in cost-effectiveness analyses of vaccination assessed in a general population (not specific one) were included. Screening and extraction of information regarding population data used in the calibration were performed.

RESULTS: The search yielded 914 candidate articles, out of which 63 met the inclusion criteria for the review. These articles described 64 IBMs calibrated for distinct populations. Population data were explicitly described for 89% of the models, while for the remaining 11%, the data used either could not be identified or was inadequately described. Among the 57 models with explicit data descriptions, there was significant variation in the type and quality of sources used to simulate population mobility and interaction. Specifically, only 5% of models incorporated observational data, 82% utilized socioeconomic data (mostly social contact matrices), and 12% did not consider human mixing.

CONCLUSIONS: The reporting of data used for calibrating IBMs falls short of optimal standards. Enhanced detail and comprehensive documentation of this process would improve the reliability, reproducibility, and quality of inferences based on IBM-predicted epidemiology, consequently strengthening resulting public health policies. Further research is needed to evaluate the representativeness of available mobility data for general populations.