BACKGROUND : Septic incomplete miscarriages remain a cause of maternal deaths in South Africa. There was an
initial decline in mortality when a strict protocol based approach and the Choice of Termination of Pregnancy Act
in South Africa were implemented in this country. However, a recent unpublished audit at the Pretoria Academic
Complex (Kalafong and Steve Biko Academic Hospitals) suggested that maternal mortality due to this condition is
increasing. The objective of this investigation is to do a retrospective audit with the purpose of identifying the
reasons for the deteriorating mortality index attributed to septic incomplete miscarriage at Steve Biko Academic
METHODS : A retrospective audit was performed on all patients who presented to Steve Biko Academic Hospital with
a septic incomplete miscarriage from 1st January 2008 to 31st December 2010. Data regarding patient
demographics, initial presentation, resuscitation and disease severity was collected from the “maternal near-miss”/
SAMM database and the patient’s medical record. The shock index was calculated for each patient retrospectively.
RESULTS : There were 38 SAMM and 9 maternal deaths during the study period. In the SAMM group 86.8% and in
the maternal death group 77.8% had 2 intravenous lines for resuscitation. There was no significant improvement in
the mean blood pressure following resuscitation in the SAMM group (p 0.67), nor in the maternal death group
(p 0.883). The shock index before resuscitation was similar in the two groups but improved significantly following
resuscitation in the SAMM group (p 0.002). Only 31.6% in the SAMM group and 11.1% in the maternal death group
had a complete clinical examination, including a speculum examination of the cervix on admission. No antibiotics
were administered to 21.1% in the SAMM group and to 33.3% in the maternal death group.
CONCLUSION : The strict protocol management for patients with septic incomplete miscarriage was not adhered to.
Physicians should be trained to recognise and react to the seriously ill patient. The use of the shock index in the
identification and management of the critically ill pregnant patient needs to be investigated.