Infectious Disease
SickMix: How Do People Change Social Contact Patterns When They Are Sick? Grissel Lopes* Grissel Lopes Aarushi Tuli Judy Donald Mark Schmidt Samuel Jenness Benjamin Lopman Kayoko Shioda
Background: Sick individuals are the major drivers of infectious disease transmission, but there is limited empirical data on how they interact with others while infectious. We aimed to understand how acute gastroenteritis (AGE) or acute respiratory infection (ARI) cases and their household members (close contacts) modify their social behaviors over the course of illness.
Methods: We recruited cases of all ages who sought any type of healthcare service (e.g., telehealth, outpatient/inpatient care) at Kaiser Permanente Northwest clinics for AGE or ARI in 2024. Cases were asked to complete three longitudinal online social contact diaries: the first regarding their social contacts and behaviors on the day they felt most ill, and follow-up surveys at 1 and 2 weeks. Their household members were also invited to participate in the surveys.
Results: A total of 526 AGE/ARI cases completed all three surveys. Additionally, 834 household members of these cases consented to participate. The mean number of social contacts among AGE/ARI cases on the sick day was 6.4 (95% confidence interval [CI]: 5.2-7.7), which increased to 10.4 (8.5-12.3) at the 1-week follow-up and 11.9 (9.8-13.9) at the 2-week follow-up. On average, contacts at the 2-week follow-up were 1.84 times (95% CI: 1.79–1.88) higher than on the sick day, with the largest difference observed among those aged 5–17 years, who had a ratio of 2.73 (95% CI: 2.50–2.98). Stratifying by setting, contacts with household members remained stable over time, whereas contacts with non-household members increased as cases recovered.
Discussions: Our data collection is ongoing and is expected to conclude in May 2025. This study found that AGE/ARI cases reduced social contacts during the acute phase of illness and gradually increased interactions, especially with non-household members, as they recover. These data can help generate reliable estimates of transmission parameters and evaluate the impact of interventions.
