Morocco's population contact matrices: A crowd dynamics-based approach using aggregated literature data.

Area of Science:

• Epidemiology
• Computational Social Science
• Public Health

Background:

• Contact pattern estimation traditionally relies on questionnaires and time-use data.
• These methods inform epidemiological models for predicting disease spread and evaluating non-pharmaceutical interventions (NPIs) like social distancing.
• Existing methods have limitations, particularly for dynamic scenarios and specific populations like children.

Purpose of the Study:

• To introduce a novel method for estimating contact patterns using a modified social force model.
• To integrate social distancing and panic scenarios into epidemiological models.
• To assess the impact of these scenarios on contact patterns in diverse social contexts.

Main Methods:

• Developed a modified social force model calibrated with individual movement data (walking velocity, interpersonal distance) and demographics.
• Applied the model to assess contact patterns in various social contexts in Morocco.
• Quantified the effects of social distancing and panic on contact matrices.

Main Results:

• Estimated contact matrices were highly assortative, aligning with previous studies (e.g., POLYMOD).
• Simulations indicated social distancing could reduce contacts by 95%.
• Panic scenarios showed an 11% increase in contacts.

Conclusions:

• The modified social force model provides a viable alternative to questionnaire-based methods for studying NPIs and specific scenarios (e.g., rush hours).
• This approach offers a substitute for estimating children's contact patterns, overcoming limitations of proxy reporting.
• The findings highlight the significant impact of social distancing and panic on contact dynamics.