Call detail record aggregation methodology impacts infectious disease models informed by human mobility.

Area of Science:

• Epidemiology
• Computational Social Science
• Network Science

Background:

• Population mobility data derived from Call Detail Records (CDR) is crucial for understanding infectious disease transmission.
• Different data aggregation methodologies for CDR can lead to variations in mobility patterns.
• Metapopulation models are widely used to simulate disease spread based on human movement.

Purpose of the Study:

• To investigate how two distinct CDR data aggregation methods affect predictions of epidemic spread.
• To assess the influence of these methodologies on metapopulation SEIR models of disease transmission.
• To identify conditions under which aggregation methodology choices impact disease spread predictions.

Main Methods:

• Utilized a Call Detail Record (CDR) dataset from Ghana (2021) detailing inter-district movement.
• Applied two aggregation methodologies: 'all pairs' (retaining long-distance connections) and 'sequential' (reflecting travel volume).
• Integrated mobility data into a metapopulation SEIR model to simulate disease transmission.

Main Results:

• The choice of CDR aggregation methodology significantly influences epidemic spread predictions.
• Impact of methodology varies based on pathogen introduction location and disease transmissibility.
• For less transmissible diseases or remote introductions, methodology affects spatial spread speed and peak infection numbers.

Conclusions:

• Methodological choices in processing CDR data can substantially alter epidemiological model outputs.
• Understanding these methodological impacts is vital for accurate infectious disease modeling.
• The significance of aggregation methodology is context-dependent, varying with disease and introduction characteristics.