Characterizing social contact patterns in rural and urban Mozambique: the GlobalMix study, 2021-2022
Moses C Kiti,Charfudin Sacoor,Obianuju G Aguolu,Alana Zelaya,Holin Chen,Sara S Kim,Nilzio Cavele,Corssino Tchavana,Americo Jose,Ivalda Macicame,Orvalho Joaquim,Noureen Ahmed,Carol Y Liu,Inci Yildirim,Kristin Nelson,Samuel M Jenness,Herberth Maldonado,Momin Kazi,Rajan Srinivasan,Venkata R Mohan,Alessia Melegaro,Fauzia Malik,Azucena Bardaji,Saad B Omer,Ben Lopman
DOI: https://doi.org/10.1101/2024.06.04.24308064
2024-06-04
Abstract:Introduction:
There are few sources of empirical social contact data from resource-poor settings thus limiting the development of contextual mathematical models of disease transmission and control.
Methods:
We collected and analyzed cross-sectional survey data from rural and urban sites in Mozambique. Participants, including infants, were recruited. They reported retrospectively, in a paper diary, individuals with whom they had a co-located physical or conversation contact, as well as their age, sex, relationship, frequency, and duration of the contact. We compared vaccine effects (VE) by parameterizing transmission models using empirical and synthetic contact rates.
Results:
1363 participants recruited between April 2021 and April 2022 reported a mean of 8.3 (95% CI 8.0-8.6) contacts per person on day 1. Mean contact rates were higher in the rural compared to urban site (9.8 [9.4-10.2 vs 6.8 [6.5-7.1], p<0.01), respectively. Participants aged less than 18 years were the main drivers of higher physical contacts. In the model, we report higher VE in the rural site when comparing empirical to synthetic contact matrices (32% vs 29%, respectively), and lower corresponding VE in urban site (32% vs 35%). These effects were prominent in the younger (0-9 years old) and older (60+ years) individuals.
Conclusion:
Our work suggests differences in contact rates and patterns between rural and urban sites in Mozambique, with corresponding differences in vaccine effects on an infectious pathogen. We also demonstrate the utility of empirical data in infectious disease modelling for high-burden, low-income settings.