Academic journal article Bulletin of the World Health Organization

High Mortality despite Good Care-Seeking Behaviour: A Community Study of Childhood Deaths in Guinea-Bissau

Academic journal article Bulletin of the World Health Organization

High Mortality despite Good Care-Seeking Behaviour: A Community Study of Childhood Deaths in Guinea-Bissau

Article excerpt


Many studies have reported a significant decrease in child mortality following general improvements in primary health care (PHC) (1,2). Such decreases are mainly the result of improvements in antenatal care and vaccination coverage; the effect of diarrhoeal disease programmes is less unequivocal. However, as vaccines and antenatal care cannot entirely eradicate the problem of excess childhood mortality in developing countries, better case management of severely ill children is clearly needed.

Improvements in the management of severely ill children are often based on audits of case histories with fatal outcomes (3). However, few studies in developing countries have investigated care-seeking behaviour prior to death in serious childhood illness in order to improve case management (4-6). Mortality surveys have found large variations in the proportion of children seen at a health facility before dying (7-10). It is important to explain why mortality remains high among under-5-year-olds in settings with easy access to health care facilities. For example, in a rural area of the Gambia, where 80% of children were fully immunized and PHC programmes had been active for 10 years, infant mortality was still 120 per 1000 live births in 1990 (11). In the Bandim suburb of Bissau, Guinea-bissau, we previously reported that the infant mortality was 94 per 1000 and under-5-year-old mortality 215 per 1000, despite the presence of two health centres, a mother-and-child health clinic and an outpatient clinic (8).

We conducted the present study on patterns of care-seeking behaviour prior to a child's death to investigate child mortality on the basis of the mother's experience. The aim was to obtain individual case histories that could be used to improve patient management in primary health care programmes and thereby lower childhood mortality in developing countries.

Subjects and methods

Study area

The study was carried out in the suburbs Bandim 1 and Bandim 2 of the capital Bissau, Guinea-Bissau. The population of approximately 25000 persons is served by two local health centres (with senior nurses consulting), one mother-and-child health clinic (with physicians consulting) as well as an outpatient clinic at the paediatric ward of the national hospital (with hospital paediatricians consulting). All inhabitants live within 1 km of a health centre and within 3 km of the mother-and-child health clinic and outpatient clinic. Apart from a small one-time charge levied for a child's vaccination chart, no fees were charged at health facilities in Bissau during the study period. Since 1979. the area has had a demographic and health surveillance system that covers the following: registration of all pregnancies and births; and for children less than 3 years of age, routine collection, by means of 3-monthly visits to all houses. of information on vaccinations, infections, nutritional status, migrations, and deaths. Morbidity and care seeking are monitored by weekly household interviews. Traditional remedies for severe diseases are not generally the first treatment choice.

Study population

The cohort followed in the present study consisted of all 1347 children born in Bandim 1 or 2 between 1 May 1992, and 30 April 1993. Deaths were ascertained by means of the routine surveillance system. Two additional rounds of data collection were carried out in 1993 and 1994; furthermore, a census of the entire population was performed in 1994. Verbal autopsies were conducted by two of the authors (MS & ICA) and a specially trained Guinean midwife. Interviews were carried out from July 1992 to November 1994, by which time the youngest children in the study cohort were 18 months of age and the oldest children 30 months of age. Median time from death to interview was 7 months (25-75th percentile: 6-9 months). The immediate cause of death was determined by combining information from the verbal autopsy, the morbidity survey, and a register of hospital diagnoses. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.