There is an important need for epidemiological studies on enteric diseases for which vaccines are available or under development, including shigellosis, typhoid fever, cholera, and rotavirus diarrhoea. Traditionally, the epidemiology of diarrhoeal diseases has been explored in longitudinal, prospective community-based surveys. As long as families can be reached at home and interviews are carefully conducted, prospective community-based studies provide a picture of the spectrum of disease ranging from mild to severe; however, this type of study is increasingly seen as costly and logistically difficult, especially in large populations. For this reason, hospital-based or clinic-based surveillance has been advocated as a more affordable method of monitoring the epidemiology of diarrhoeal disease. It is thought that health facility-based surveillance is likely to provide information on more severe disease, but it is not known to what extent families with diarrhoea make use of local healthcare providers and how this varies in different locations. These limitations of health facility-based surveillance have led to the development of different types of studies that will allow investigators to understand local patterns of healthcare use for diarrhoeal disease, including the proportion of individuals with diarrhoea who will seek care at a health facility. Six papers in this issue of the Journal report on a total of eight studies that explore patterns of use of traditional and non-traditional healthcare services when a family member has diarrhoea. The target populations for these studies were in the range of 50,000 to 500,000. Altogether, the papers report on three types of studies: a one-time census survey of all households in the study area; a rapid cluster survey of a limited number of representative households; and an anthropological interview of selected families and healthcare practitioners. In the census survey, investigators visit all households in the study population once. The census survey is designed to enumerate the population, to record episodes of diarrhoea in the previous month, and to learn where families sought treatment for diarrhoea. Two census surveys are reported in this issue, one from a slum area of Kolkata, India (1) and one from a slum area of Jakarta, Indonesia (2). Both of these studies were conducted in planning for pilot projects for vaccination against either cholera or typhoid. Healthcare-use patterns were analyzed for those individuals with diarrhoea in the month preceding the census survey--428 (0.7%) of 57,099 persons in Kolkata, India and 8,074 (5%) of 160,261 persons in Jakarta, Indonesia. The proportion of individuals reporting diarrhoea in Kolkata was lower because the census survey was conducted in the low-prevalence season, whereas the Jakarta survey was carried out during the peak diarrhoea season. The rapid cluster survey uses a simple sampling method (3) to select a small but reasonably representative sample of households for interview to assess healthcare-use patterns for diarrhoea. A rapid cluster survey questionnaire developed by the International Vaccine Institute (IVI) for proposed Shigella disease-burden study sites is published in this issue as an annex to the study from China (4). The questionnaire asks about healthcare use for family members who had diarrhoea in the previous month and, where no family member has experienced diarrrhoea in the past month, the respondent is asked how the family would respond to a hypothetical vignette concerning diarrhoea. Three rapid cluster surveys using the IVI questionnaire are reported in this issue, and these were conducted in China (4), Thailand (5), and Viet Nam (6). Another rapid cluster survey questionnaire has been developed by the World Health Organization (WHO) for proposed rotavirus disease-burden study sites (7). Two rapid cluster surveys using the WHO questionnaire are reported in this issue, one from a rural district and one from an urban district in Ghana (8). …
Read full abstract