Most studies on maternal mortality have looked at the direct clinical
causes and the distribution of actual rates. Much less attention has b
een given to prevailing health care systems or community factors assoc
iated with such deaths. A case-control study design using incident cas
es was used to identify the magnitude of maternal deaths and community
and health care operational factors in both an urban and a rural sett
ing in Zimbabwe. The maternal mortality ratio for the rural setting wa
s 168 per 100 000 live births and that for the urban setting was 85 pe
r 100 000 live births. For the rural setting, the major direct causes
of death were haemorrhage (24.8%), abortion complications (15.2%), pue
rperal sepsis (13.3%), and eclampsia (4.8%). For the urban setting the
y were eclampsia (26.2%), abortion complications (23.0%), puerperal se
psis (14.8%) and haemorrhage (9.8%). Whereas rural-urban variations in
maternal mortality were observed, inter-rural district variations wer
e also apparent, especially with poor medical resources, poor communic
ation and delayed interventions. Risk factors for maternal mortality w
ere present at each of the various levels of care. Lack of antenatal c
are (ANC) had a significant Odds Ratio (OR 10.7 rural and 4.6 urban) c
ontribution to maternal mortality. When abortions and ectopics were ex
cluded the OR for absent ANC was 4.1 (rural) and 2.6 (urban). Lack of
timely transport to nearest clinic or hospital adversely affected preg
nancy outcome in both rural and urban settings. Despite delivery place
planning, predisposing health conditions and some danger signals, few
of the women utilized the venue originally planned for delivery. Heal
th education, community sensitization and teaching on risk signal awar
eness as well as health care delivery system strengthening are recomme
nded for reducing the high maternal mortality rates.