Saving Babies 2003

Executive summary

Key findings

The PPIP users have performed a mammoth task and coupled with the superb PPIP v2 software, a clear picture of perinatal care in South Africa has emerged. There have been some incorrect classifications and there were differences in interpretation of avoidable factors, missed opportunities and substandard care. However, the strength of the data lies in its size of the sample (462348 births and 12773 perinatal deaths) and the distribution of health institutions participating (102 institutions throughout all areas in South Africa). The picture drawn by the data is the best scenario for South Africa, as usually only the most concerned hospitals perform audit and are prepared to share their problems with other institutions.

The most immediate and important problem identified is that of managing the pregnant woman in labour. Intrapartum asphyxia and birth trauma were responsible for about one in five of all deaths recorded in the database. This varied between one in four in rural areas to one in seven in metropolitan areas. More than three quarters presented with a live baby at the institution and more than three quarters of the infants weighed more than 2.5 kg. More than three quarters of the neonatal deaths due to hypoxia could be attributed to this primary cause. These babies should not have died. Analysis of avoidable factors and further in-depth analysis of another 100 cases indicated that the majority occur in low-risk women with apparently uncomplicated labour. The major avoidable factors were the failure to detect or respond to evidence of fetal distress. In other words, basic standard management of labour is not being provided to all women in labour. Labour is unpredictable and all women in labour must have at least hourly monitoring and this requires staffing of labour wards at the equivalent level of a high care ward. Resuscitation of the hypoxic infant must be an essential skill of all health workers involved in delivering babies.

The next clearly identified problem is the high proportion of unexplained intrauterine deaths, almost one in four. More than 80% of the babies were macerated, they constituted the majority of deaths in the 1.5-2.5 kg weight category, and comprised 23% of the deaths in babies over 2.5 kg. This must be seen in the context of idiopathic intrauterine growth restriction, post-maturity and congenital abnormalities being rarely diagnosed outside of institutions associated with medical schools, and the high proportion (39% of cases) where the syphilis serology was unknown. The higher proportion of neonatal deaths resulting from congenital abnormalities than intrauterine deaths is further evidence of the under diagnosis of congenital abnormalities. If good antenatal care is provided, all four of these conditions can be easily detected, and with intrauterine growth restriction, post maturity and infections the deaths can be prevented. Although poor or no attendance at antenatal care was recorded as an avoidable factor in one in five cases of unexplained intrauterine deaths, the majority of women did attend antenatal care. Furthermore, surveys have indicated that 95% of women countrywide attend antenatal care when pregnant. Hence, the high proportion of unexplained intrauterine deaths is probably a good indication that the quality of antenatal care is poor.

The finding that more than half of the deaths due to complications of hypertension in pregnancy were macerated stillbirths and the most common avoidable factor was lack of referral to the appropriate level of care provides further evidence of poor quality of antenatal care. Hypertension was a common cause of perinatal deaths (10%), with two thirds occurring in the 1-2 kg weight categories, disturbing is the fact that one third occurred over 2 kg, clearly where intervention could have prevented the death.

Spontaneous preterm birth is a major cause of perinatal death with approximately one in six deaths being due to this primary cause. Almost 80% of the deaths due to immaturity could be attributed to this primary cause. Prevention of premature births is not possible at present, leaving care of the immature infant as the only possible mechanism to reduce this neonatal mortality rate. The finding that the metropolitan areas have less than half the neonatal morality rates experienced for infants born between 1 and 2 kg in the cities and towns indicates it can be done.

Abruptio placenta is a major cause of death (12%) and is the only major primary cause where there is no clear solution or strategy to reduce the deaths. Perhaps the most important finding is the much higher prevalence in urban areas than rural areas. Lifestyle issues such as smoking might be an important factor where preventative measures might be directed.


  1. Perinatal mortality audits must occur in each institution conducting births
  2. Ensure adherence to standard protocols in:a. Monitoring the mother, progress in labour and the fetus during labour and ensure appropriate action is taken when abnormalities occur;b. Neonatal resuscitation;c. Basic care of all neonates post resuscitation;

    d. Antenatal care especially with respect to hypertension, detecting and managing intrauterine growth restriction and post term pregnancies and syphilis;

    e. Voluntary counselling and HIV testing, prevention of mother to child transfer and antiretroviral treatment for those that meet the criteria;

    f. Evaluation of a stillbirth.

  3. Kangaroo mother care should be the primary way to manage stable low birth weight infants.
  4. All level 2 and 3 hospitals involved in the care of neonates should be able to provide respiratory support using at least nasal CPAP.
  5. A system of outreach programmes should be established to cover the whole country that includes support for, education and audit of the sites visited.
  6. Early confirmation of pregnancy and immediate initiation of antenatal care on confirmation of pregnancy should be the norm.
  7. Health promotion to the public should include messages on early confirmation of pregnancy and initiation of antenatal care, attention to fetal movements, appropriate action to danger signs, and plans for getting to the institution where the birth is planned.
  8. Establishing staffing and equipment norms per level of care must be performed for every health institution concerned with the care of pregnant women.
  9. Criteria for referral and referral routes must be established and utilized appropriately in all provinces.
  10. Provincial Maternal Child and Women’s Health units to primarily ensure the minimal perinatal data set is completed and analysed for each institution in their province.

Implementation strategy

  1. Quality assurance programmes to be introduced in each institution providing care for pregnant women for:
    a. Antenatal care;
    b. Intrapartum care;
    c. Neonatal resuscitation;
    d. Basic post resuscitation neonatal care.
  2. In-service training on:
    a. Antenatal card;
    b. Partogram;
    c. Fetal Heart Rate monitoring including electronic monitoring;
    d. Neonatal resuscitation;
    e. Basic neonatal care;
    f. Kangaroo mother care;
    g. Nasal CPAP use;
    h. Conducting audits.
  3. Incorporate in the job description, change the job description or create dedicated posts for implementing the outreach strategy such that the outreach programmes can support, educate and audit the institutions in the above fields.
  4. Provide the infrastructure to establish pregnancy confirmation and initiation of antenatal care at a single visit at the same site. Distribution of antenatal cards to general practitioners should be part of this. This could be by promoting “antenatal friendly clinics” and allowing for easy access and incorporating general practitioners in the antenatal care system.
  5. Use the different media to promote specific clear messages about pregnancy.
  6. Insist all institutions conducting births complete the minimal perinatal data set monthly and submit it to their provincial MCWH unit.


Analysing the data clearly suggests that the factor underlying most of the deaths is the poor quality of care, whether, antenatal, intrapartum or in the neonatal period. There are many reasons for this; probably most important are lack of personnel, facilities, knowledge and poor morale. The recommendations and strategies dealing with these issues must be implemented.

Full report

Download the complete report here