Population-based seroprevalence studies provide important data on susceptible groups and the potential for future outbreaks. However, the invasive nature of serum collection has limited studies. This paper describes the first postal population-based survey using noninvasive oral fluid technology to collect antibody prevalence data in conjunction with extensive risk factor data to assess the distribution of immunity to common viral infections in England and Wales. These results pertain to hepatitis A virus (HAV). Approximately 5,500 oral fluid samples were collected between August 2001 and May 2002, as well as individual risk factor data through a questionnaire, from persons aged less than 45 years randomly sampled from general practices countrywide. Samples were tested for immunoglobulin G-specific antibody marking a past infection or immunity to HAV using an antibody-capture enzyme-linked immunosorbent assay. The age-specific HAV seroprevalences indicated a low incidence of infection (overall seroprevalence of 18.9% (95% confidence interval: 17.0, 20.9) and of 9.2% (95% confidence interval: 7.1, 11.3) after the exclusion of vaccinees). Vaccination proved the most important determinant of seropositivity. Ethnic minority groups were underrepresented, and adjustment increased the overall prevalence to 20.1% and to 12.1% in unvaccinated individuals. The availability of comprehensive risk factor data allowed the description of two risk profiles related to natural infection and vaccination.

SUMMARYSerological surveys among representative population samples have proved rare given their reliance on invasive sample collection. We therefore completed the first population-based postal survey of immunity in England and Wales using new oral fluid technology. This paper examines the feasibility of this new methodological approach. Nearly 5500 oral fluid samples were collected, with individual demographic and social data via a questionnaire, from persons under 45 years of age recruited through general practices. Instructions were accurately followed with only 1 % of samples returned without risk-factor data. The overall response rate was 40%. Response was independently associated with age, sex and location. Response was highest in children aged 5-14 years, adult females and in rural locations. This approach allowed the successful collection of comprehensive individual risk data, but response rates in adults must be improved if oral fluid surveys are to routinely complement serological surveillance.

After adjustment for age, sex, and geographic location, there was no evidence of a change in seroprevalence between 1991 and 2000.