There were 43 early pregnancy deaths, mostly due to septic abortions (26) and ectopic pregnancies (11). There appears to be significant under-reporting of early pregnancy deaths. Most deaths occurred in women of low parity and between 20 and 34 years of age. The deaths were distributed throughout the levels of care with slightly more ectopic pregnancies occurring in Level 1 hospitals and septic abortions in Level 3 hospitals. Notably the problems occurred in the more advanced pregnancies with the mean duration of pregnancy of women dying of abortion was 17.5 weeks and 4 of the 11 ectopic pregnancy deaths were more than 20 weeks. The most common avoidable factors for abortions lay in the women inducing the termination of pregnancy themselves and delays in seeking help. The lack of termination of pregnancy services, insufficient intensive care beds and availability of theatre were the main administrative problems. Failure to recognise septic abortions and significant delays in management along with poor observations were the major problems related to medical personnel. The main problems for ectopic pregnancies were incorrect diagnoses and delays in performing a laparotomy. The expansion and advertising of the Termination of Pregnancy (TOP) services especially with respect to second trimester terminations is urgently required. Clear guidelines, available at every hospital, on managing incomplete abortions and on the recognition and management of sepsis are also needed.
Key Recommendations
Early pregnancy deaths include all deaths from ectopic pregnancy, abortion (spontaneous or induced) before 24 weeks' gestation. Trophoblastic disease is also included in this chapter.
Early pregnancy deaths contributed 7.6% of all maternal deaths, and 12.0% of direct maternal deaths. There has been probably been considerable under-reporting of early pregnancy deaths (Table 5.1). Early pregnancy deaths are thought to have been well collected in Gauteng Province and here it contributed 11.6% of all maternal deaths and 17.9% of direct maternal deaths. The maternal mortality ratio for abortions in Gauteng Province is 6 per 100 000 live births. This is considerably lower than the estimate of 37 per 100 000 live births made in 1994 for South Africa1. KwaZulu-Natal reported 7.9% of all reported deaths and 14.3% of direct maternal deaths. This in contrast to Northern Province, Eastern Province and North West Province who reported 0%, 3.7% and 2.9% respectively. Deaths occurring in early pregnancy often occur outside of the maternity services and more strenuous efforts will need to be made to collect these. Under reporting of ectopic pregnancies is well recognised with the diagnosis never being made except at autopsy. However, the early pregnancy deaths collected occurred in urban, peri-urban and rural settings and at all levels of care.
As mentioned in chapter 6, septic abortion is a major contributor to pregnancy-related sepsis. This again stresses the importance of sepsis as a major contributor to maternal deaths. Ectopic pregnancies are also an important contributor to deaths due to haemorrhage.
Despite incomplete data collection, the information available allows for analysis and the NCCEMD believes the findings are a close approximation of the truth in South Africa, and the recommendations will result in a substantial decrease in maternal deaths.
The reported number of women who died as a result of an ectopic pregnancy does not allow for detailed demographic analysis. The data is shown below in Table 5.2.
The age and parity of women who died as a result of complications of an abortion are shown in Table 5.3. Surprisingly few teenagers died as a result of abortion.
Four of the 11 women, who died as a result of ectopic pregnancies, had pregnancies more than 20 weeks gestation. The mean gestational age at the time of abortion (where it was recorded) was 17.5 weeks. The distribution is shown in Table 5.4.
More deaths from ectopic pregnancies occurred at Level 1 hospitals, than Level 3 and the opposite happened with deaths due to abortion. See Table 5.5. There were 28.6%, 33.3% and 38.1% of early pregnancy deaths at Levels 1,2 and 3 respectively.
The primary, final and contributory causes of death are shown in Tables 5.6 and 5.7. As expected, pregnancy related sepsis and haemorrhage were the major causes of death.
A global view of the potential for improved care is shown in Table 5.8 and detail in Tables 5.9 for ectopic pregnancies and 5.10 for abortions. The lack of documentation is clearly illustrated by the number of cases that were not assessable. Attention must be paid to making adequate notes. The number of maternal deaths due to ectopic pregnancy is too small to make generalised judgements. However, in the case of abortions, it is clear that problems exist from the patient, administrative and the medical management point of view. In all cases there is room for substantial improvement.A number of problems are clearly illustrated by the case histories given below.
Case 1: A patient presented at a level 1 hospital shocked with the clinical diagnosis of an ectopic pregnancy. Her haemoglobin was 6.4g/dl and blood pressure 90/50 mmHg. The hospital was unable to perform a laparotomy and had only one unit of O negative blood. Resuscitation commenced while arrangements were made for transfer of the patient. "This took sometime because the telephones kept on disconnecting and eventually when we got through the tertiary hospital refused for some reason. So eventually another hospital agreed to accept the patient but unfortunately she died on the way".
Comment: This case clearly illustrates a number of failings in the health system. Firstly this patient should have had a laparotomy as soon as possible after the diagnosis of ectopic pregnancy was made. Level 1 hospitals should have the facilities to perform emergency surgery, and if this had been done the patient would probably have survived. Secondly, the problems with communication and barriers to entry for patients from level 1 hospitals to higher levels are clearly demonstrated. The value of keeping an adequate supply of O negative blood at sites performing basic surgery is again demonstrated.
Case 2: A patient presented to a hospital, requesting TOP. An ultrasound scan was performed and this put her pregnancy at 20 weeks 4 days. TOP was not granted and she was advised to keep the pregnancy. Four days later she was admitted with an incomplete abortion. After initially denying any interference she later stated she drank some mixtures and a rubber catheter was pushed into her "private parts" and some fluid instilled in. After a prolonged hospital stay she died of multiple organ failure.
Comment: Advertising of TOP services could well have saved this woman's life, for if she had presented a week earlier, things might have been different.
Case 3: A patient aborted at 20 weeks and was discharged. She was readmitted the next day "ill with sepsis, cervix open with products felt within. No attempt was made at uterine evacuation. Antibiotics were given. There was no record of any investigations or urinary output until the day of death 4 days later". Comment: This case illustrates the poor initial assessment of this case, followed by lack of recognition of seriousness of the disease. Furthermore, evacuation of the uterus shortly after admission would have been the standard management, but was not performed. Lack of crucial investigations or observations illustrate the lack of recognition of the problem.
Case 4: A patient was admitted at 18h00 with the diagnosis of a urinary tract infection and an incomplete abortion. She was not actively bleeding on admission. Evacuation was scheduled for the next day. Six hours after admission a haemoglobin estimation was performed and recorded as 4g%. The doctor was informed but no measures were taken. Two hours later the patient was "gasping and died". No attempt was made during the whole admission at resuscitation, and there were no documented observations.
Comment: Lack of basic initial assessment, observations and resuscitation clearly led to the demise of this patient.
Case 5: Patient presented with an inevitable abortion. The haemoglobin was measured as 5g% and 2 units of blood were ordered and given the following day. Ampicillin prophylaxis was given. A fever developed the next day. The haemoglobin after transfusion was recorded at 7.5g%. No blood was ordered, but penicillin G and metronidazole was started for sepsis. Patient was not seen the next day and died the day after that.
Comment: What can one say!
A clear illustration of the lack of recognition of the problem with abortion is that fact that only 50% had an anaesthetic of any sort. Many patients who required a hysterectomy never received it because the managing doctor did not appreciate the problem.
Quantification of the major avoidable factors is shown in Table 5.9 and 5.10.
Despite poor reporting, deaths in early pregnancy are an important contributor to maternal deaths. Abortion is the major cause and septic abortion is the main problem and this occurs almost exclusively in the second trimester. Too many deaths are occurring at Level 1 and 2 hospitals. Deaths occur in the lower parity and younger age groups, but not in teenagers.
The major patient-orientated factors relate to women inducing termination of pregnancy themselves or by people outside the recognised health services. There is also considerable delay in seeking help, whether due to ignorance or fear of retribution from the hospital is not clear. However, these delays contribute significantly to the mortality. In a significant proportion of women (44%) the lack of TOP services was directly sited as an avoidable factor. Advertising the whereabouts of TOP services and making every woman aware of their rights under the Choice of Termination of Pregnancy Act of 1996 would considerably decrease mortality due to septic abortion. The effect of this policy has been clearly illustrated by the Report on Confidential Enquiry into Maternal Deaths in England and Wales 1976-1978. Abortion was the most common cause of maternal death in England and Wales from 1952-1969. After introduction of their new Abortion Act in 1968, there was a dramatic decrease in the number of deaths due to abortion. By 1976-8 there were only 19 deaths, dropping from 133 in 1964-66. Health administrators must also pay particular attention to second trimester TOP services. Most deaths occurred in this group, and this procedure is medically the most difficult.
The lack of basic facilities such as blood and theatre availability contributed to deaths in both the ectopic and abortion groups. Health authorities must pay particular attention to defining the level at which a hospital will function and then equip it according, both with personnel and equipment. There were a number of avoidable deaths, which were directly attributable to the availability of facilities.
The medical personnel generally failed to recognise the problem and take appropriate action. This was due to either very superficial assessment of patients or lack of monitoring thereafter. The lack of monitoring may have been due to lack of personnel to perform the monitoring or lack of recognition that there was a problem. There should be clear guidelines in every hospital and community health centre on how to deal with women who present with lower abdominal pain, and/or vaginal bleeding. Basic initial assessment must be specified and standard protocols for the management of suspected or clinically obvious ectopic pregnancies or incomplete abortions should be available in every hospital and health centre. Medical schools should ensure their students are fully conversant with the management of these problems. Ideally regular audit should take place on the management of the problems to keep the health workers aware of the problems.
Most of the early pregnancy deaths were preventable. If particular attention is paid to ensuring women know their rights under the new Choice of Termination of Pregnancy Act, health authorities ensure the basic facilities are available at the appropriate level of care and the health workers follow the basic guidelines of management of women with lower abdominal pain or vaginal bleeding, most of the deaths in early pregnancy would be prevented.