Social contact patterns in Viet Nam and modeling the transmission of respiratory infections
Session I – Horby, Peter, Abstract 2 of 2
Title
Social contact patterns in Viet Nam and modeling the transmission of respiratory infections.
Authors
Peter Horby1,2, Pham Quang Thai3, Annette Fox1,2, Le Quynh Mai3, Nguyen Thi Thu Yen3, N Linh3, Nguyen Tran Hien3, Jeremy Farrar1,2.
Affiliations
1 Oxford University Clinical Research Unit, Vietnam; 2. Centre for Tropical Medicine, Nuffield Department of Clinical Medicine, Oxford University; 3 National Institute of Hygiene and Epidemiology, Vietnam;
Introduction
Prospective community based studies have provided important insights in to the epidemiology of respiratory infections and crucial data for modeling epidemic dynamics. However, no such studies have been undertaken in South East Asia even though this region is believed to be the primary source of influenza A(H3N2) strains causing seasonal influenza outbreaks and is thought to be the most likely source for the next influenza pandemic.
Methods
A household based cohort was established in North Viet Nam to prospectively study the incidence and transmission of influenza in the community. Contact diaries and questionnaires were used to derive Viet Nam specific data on social contact patterns to inform epidemic modeling and intervention policies. The results were compared with similar data from Europe. A smoothed contact surface was created using a negative binomial model with interpolation using a tensor product spline. The contact data were used to undertake preliminary modeling of the early stages of an epidemic in Viet Nam.
Results
Details were obtained of 6,788 contacts for 874 participants aged from one to 90 years. The mean number of contacts per participant per day was 7.8, compared to an average of 13.4 in eight European countries. Of the contacts, 67% occurred at home, 10% at work and 2% during leisure activities; compared to 23%, 21% and 16% for Europe. 52% of all contacts lasted more than four hours. First time contacts were much less common in our cohort than in Europe. Preliminary modeling suggests that if transmission of a new infection to which everyone is susceptible occurs through the social contacts reported in our study, the highest attack rates during the initial phase of an epidemic will be experienced by adults aged 35-49 and children aged 6-15 years.
Discussion
We found notable differences in contact patterns and in the predicted age distribution of a modeled epidemic compared to similar studies conducted in Europe. These findings show
