Diarrhoea, the third major cause of morbidity in children aged less than five years in Ghana (1), accounts for 15% of outpatient visits to health centres. The proportional rate of mortality due to diarrhoea is 10% in children aged less than one year and 8% in children aged 1- 5 year(s) (1). Mortality from diarrhoea is often due to dehydration, which needs to be properly managed to reduce high mortality rates. Education on the use of oral rehydration solution (ORS) for the treatment of diarrhoea in the home is encouraged. For severe diarrhoea, however, the healthcare system plays an important role in case management since this is a potentially lifethreatening condition, which requires immediate treatment with intravenous and/or oral fluids (2).
Rotavirus infection is one of the important causes of severe diarrhoea in children throughout the world and is a leading cause of death in early childhood (3,4). The incidence of diarrhoea due to rotavirus is 0.035 per childyear in northern Ghana (5), occurring more commonly in the dry season (5-7). Improvements in sanitation and in food and water quality have failed to alter the incidence of rotavirus infection in many countries (8). Vaccines are considered to be the intervention most likely to prevent the disease, and a number of candidate rotavirus vaccines are undergoing clinical testing (8).
Care-seeking behaviours and treatments for children with diarrhoeal illnesses vary considerably between countries. Recent surveys suggest an increasing tendency to seek care for diarrhoea outside the home (9). Hospitalbased surveillance of severe diarrhoea in children aged less than five years has been proposed to address the burden of rotavirus infection in communities. Likewise, passive hospital-based surveillance has been proposed for clinical trials to assess the protective efficacy of candidate rotavirus vaccines against severe diarrhoea.
Before initiating any hospital surveillance, it is important to determine hospital-use patterns among carers of children with severe diarrhoea in catchment areas. It is also important to understand the attitudes and practices of the population towards paediatric gastroenteritis and treatment preferences for diarrhoea in general. For this reason, the World Health Organization (WHO) has developed a new protocol for a community-based survey to assess the use of health services for children with gastroenteritis (10). To field-test this protocol, we conducted two surveys in Ghana__one in a rural district and one in an urban district. In each district, the purpose of the survey was to assess whether the district hospital would be a suitable site for passive surveillance of severe diarrhoea in children aged less than five years.
MATERIALS AND METHODS
The healthcare system
The healthcare system in Ghana has operated on primary healthcare concepts and strategies since the early 1980s. At Level A, the community level, trained community members and community health nurses screen patients and offer basic treatment. At Level B, the health centre, trained nurses and, sometimes, doctors offer treatment and preventive health services. At Level C, the district level, there is a hospital where doctors and nurses manage referred cases and offer tertiary care.
Financing of health services in Ghana has moved from the previously heavily-subsidised care to a situation where all patients pay for all services and drugs. Exemptions from such fees exist for special cases, the elderly, children aged less than five years, and antenatal care. However, many centres often ignore the exemption policy (11,12). The current trend is to shift to insurance systems for healthcare. The Ministry of Health has supported the diarrhoeal disease-treatment programme in Ghana for about 15 years, with ORS 'corners' in many health institutions where special attention was paid to children with diarrhoea. This programme is now fully integrated into the national services at all levels. …