Abstract
Economically and politically disruptive disease outbreaks are a hallmark of the 21st century. Although pandemics are driven by human behaviours, current surveillance systems for identifying pandemic threats are largely reliant on the monitoring of disease outcomes in clinical settings. Standardized integrated biological-behavioural surveillance could, and should, be used in community settings to complement such clinical monitoring. The usefulness of such an approach has already been demonstrated in studies on human immunodeficiency virus, where integrated surveillance contributed to a biologically based and quantifiable understanding of the behavioural risk factors associated with the transmission dynamics of the virus. When designed according to Strengthening the Reporting of Observational Studies in Epidemiology criteria, integrated surveillance requires that both behavioural risk factors - i.e. exposure variables - and disease-indicator outcome variables be measured in behavioural surveys. In the field of pandemic threats, biological outcome data could address the weaknesses of self-reported data collected in behavioural surveys. Data from serosurveys of viruses with pandemic potential, collected under non-outbreak conditions, indicate that serosurveillance could be used to predict future outbreaks. When conducted together, behavioural surveys and serosurveys could warn of future pandemics, potentially before the disease appears in clinical settings. Traditional disease-outcome surveillance must be frequent and ongoing to remain useful but behavioural surveillance remains informative even if conducted much less often, since behaviour change occurs slowly over time. Only through knowledge of specific behavioural risk factors can interventions and policies that can prevent the next pandemic be developed.