Abstract

Using proximity sensors to characterize social contact patterns relevant to the transmission of respiratory viruses in schools and households.

Kiti MC
. 2019;Phd

Permenent descriptor

This thesis reports the use of wearable sensors to obtain proximity data from students in schools and household residents in Kenya. Data on who-contacts-whom quantifies behaviour that underpins pathogen transmission usable in mathematical simulations to evaluate interventions such as vaccination. A subset of students from two schools were recruited from rural and urban areas, including residents linked to some of the participating students. Data were collected on different dates for each site, but collection was hindered by low sensor numbers, administrative bottlenecks, student absenteeism, and resident out-migrations. This resulted in 4 study sites enrolling 595 (303 school) and 233 (136 school) participants in rural and urban locations, respectively. Each participant carried a proximity sensor over 7 days. On average, each student was in contact with 38% (rural) and 31% (urban) other participating students. Contact number and duration were assortative by grade in the rural school, with considerable inter-grade mixing in the urban school. Weekends registered low interactions due to school closure, but mixing was reported by older students who attended school on Saturday. In the households, children (5-14 years) in rural and adults (20-49 years) in urban sites interacted most with children under-5 years. The duration of interaction was higher within- versus between-households, suggesting that infections dependent on cumulative exposure time may have higher transmissibility within households. Observed high between-household network density suggests that these interactions may lead to substantial inter-household transmission of endemic infections. Finally, a framework is proposed for an agent-based model that can be used in future to investigate the implications of observed intra- and inter-household interaction patterns on the transmission dynamics of respiratory syncytial virus (RSV). In the absence of a vaccine against RSV, the impact of alternative strategies to prevent infection in infants, such as vaccinating older children, can be safely investigated using computer simulations.