INFANT-MORTALITY RATE INEQUALITIES IN THE WESTERN CAPE PROVINCE OF SOUTH-AFRICA

Background. Cape Town is undergoing rapid urbanization. South African vital statistics have routinely been stratified by racial categories b ut intra-urban and peri-urban geographical variations have been neglec ted. Methods. To examine variations in infant mortality rates (IMR) an d proportional infant mortality between urban, rural and informally se ttled areas, stratified by racial category, birth notifications and in fant death certifications recorded by a large health authority were an alysed. Results. The IMR per 1000 livebirths was as high for coloureds on rural farms (34, 95% confidence interval 29-40) as for blacks in i nformal settlements (35, 95% CI : 32-37) and an exceptionally high IMR (60, 95% CI : 43-82) was found for coloureds in informal settlements. Inequalities between racial categories (11 (95% CI : 9-14) for whites , 19 (95% CI : 18-21) for coloureds and 33 (95% CI : 31-35) for blacks ) were as expected from other South African studies. Of rural farm dea ths, 22% were ascribed to ill-defined causes. Low birthweight was the most common defined cause of death in all areas except rural farm area s (14% ascribed to pneumonia), and gastro-enteritis was important in i nformally settled areas (18%). Conclusions. Routine mortality data are more informative ii stratified by robust and readily available indica tors of socioeconomic status such as residential area and racial categ ory. Place of residence may distinguish risk strata as well as racial category, but the latter is helpful within socioeconomically heterogen eous residential areas.