There were 67 maternal deaths attributable to pregnancy-related sepsis. The 41 cases associated with viable pregnancies are discussed in this chapter, while the 26 septic abortion-related deaths are described in Chapter 5. Pregnancy-related sepsis was the fourth commonest cause of maternal deaths being responsible for 19% of direct and 12% of all maternal deaths. Of the maternal deaths, 70% were <30 years old and 22% was teenagers. Primigravidas accounted for 44 %, and 6 % had a parity of more than 4. The majority (42%) of maternal deaths occurred in level 3 hospitals, however, 24% occurred at level 1 hospitals. Death due to puerperal sepsis followed a caesarean section in 46% of cases. In 27% of the cases, labour had been obstructed. The final cause of death was nearly always septic shock (73%) or multi-organ failure (24%). In 68% of the patients, their HIV status was unknown. Of those who had, 69% were HIV-positive. Patient-related avoidable factors were identified in 42% of the cases. The commonest problem was delay in seeking professional help, largely as a result of the patient failing to appreciate the seriousness of the particular complication of pregnancy. Administrative-related avoidable factors were identified in 32% of the deaths. Medical personnel-related substandard care was identified in 53% of the patients. In 25% of cases the diagnosis was incorrect and in 29% of cases the management protocol had not- or inadequately been followed. Lack of or inadequate monitoring of the patient occurred in 21% of cases.
Key recommendations
There were 67 maternal deaths directly due to pregnancy-related sepsis. Of these, 26 occurred as a result of septic abortion and these have been discussed in Chapter 5. The remaining 41 cases were associated with viable pregnancies and are included in this chapter. Pregnancy-related sepsis was the fourth commonest cause of all maternal deaths, being responsible for 19% of direct, and 12% of all maternal deaths.
The problem may be much larger than reported, because of the under-reporting in a number of Provinces. However, there is a good distribution of cases between rural, peri-urban and urban areas, so the problems detected have a high likelihood of reflecting the actual situation. Another confounding problem has been the lack of HIV testing. In 28 (68%) HIV testing was not performed, and in the remaining 13 cases, 9 were HIV-positive.
The age and parity of women dying as a result of puerperal sepsis reflect those of women dying from the complications of AIDS (see Chapter 9). The age and parity tend to be in the younger age groups (70% less than 30 years) and lower parity (77% having a parity of 2 or less).
The levels of care at which the maternal deaths occurred are shown in Table 6.2.
The patients had a vaginal delivery in 54%, and 46% had a Caesarean section. Obstructed labour was present in 27% of the cases (Table 6.3).
In almost all cases, the final cause of death was septic shock (73%), or multi-organ failure (24%). (See Table 6.4). Contributory causes of death included respiratory failure (17%); immune system failure (15%); cardiac failure (10%); and disseminated intravascular coagulation (7%). It should be noted that patients might have had more than one final and/or contributory cause of death.
Tables 6.5-6.7 illustrate the avoidable factors, missed opportunities and substandard care.
The most common patient-orientated avoidable factor was delay in seeking help (28% of assessable cases). This should be linked to the lack of access to health care facilities, which occurred in 10% of cases.
Assessing women and making the diagnosis of puerperal sepsis was performed poorly in 1 in 5 cases. This led to delays in transferring women (16%) or their being managed at an inappropriate level of care. Furthermore, once the diagnosis had been made, there was non-adherence to accepted protocols for managing puerperal sepsis. In 22% of cases the monitoring of the patients was poor (Table 6.7).
A few case histories below illustrate the problems encountered.
Case 1: A 17-year-old primigravida had an uncomplicated spontaneous vaginal delivery in a Level 1 facility. She and her infant were discharged home in good condition. The patient was readmitted to the Level 1 facility with a history of lower abdominal pain and diarrhoea for 7 days. She had been taking herbal medicines for the past two days. Eleven days post delivery the patient was transferred to a tertiary institution, with an acute abdomen, puerperal sepsis, and renal impairment. Despite emergency hysterectomy and admission to the intensive care unit, the patient died.
Comment: This case illustrates the lack of recognition of puerperal sepsis by the level 1 institution. The diagnosis can be difficult to make, if the doctor does not appreciate that the signs of an acute abdomen will mostly be absent. The abdominal muscles have been stretched during pregnancy and in the immediate postpartum period have lost their tone. The signs of sub-involution of the uterus, a tender lower abdomen, a foul smelling vaginal discharge and an open cervix coupled with the signs of sepsis, (temperature, tachycardia, and tachypnoea) make the diagnosis of puerperal sepsis. The history of herbal medicine ingestion should rather indicate to the doctors that the woman had been feeling ill for some time, than that herbal medicine is causing the symptoms.
Case 2: A 19-year-old primigravida had reached full dilatation of the cervix after 12 hours of labour. Two hours later she had not delivered, and transfer was arranged to a tertiary hospital. Because of ambulance delay, she only arrived there some three hours later. The intern only saw the patient some two hours after admission. Because the contractions were "weak", an oxytocin infusion was commenced. At no stage was the consultant informed of the case. The case notes were "scanty", and there was no record of any examination(s) in labour. The partogram showed grossly inadequate progress of labour, but this was not acted upon. Eventually, after the cervix had been fully dilated for more than 11 hours, a caesarean section was performed. Antibiotic cover was inadequate. On the fourth postoperative day, there was a foul smell from the wound, but no treatment was instituted. The patient collapsed on day 7 and was then seen by a consultant for the first time. At laporotomy, necrotising fasciitis was diagnosed. The wound was debrided. The postoperative course was unsatisfactory and the patient died the next day.
Comment: This tragic case illustrates a number of errors. The lack of appropriate use of the partogram cannot be explained, nor can the lack of post-operative care. This all occurred in a level 3 hospital. Lack of staff, lack of knowledge and lack of facilities cannot explain the happening. Perhaps, the common misconception that once a woman's cervix is fully dilated, she will deliver vaginally and the general attitude that the postpartum period is not important can explain the grossly substandard care. The most common period for the poorest observations is in the puerperium.
The main problems detected are summarised below
In 1998, pregnancy-related sepsis was the cause of 67 maternal deaths, 41 due to puerperal sepsis and 26 due to septic abortion. As a category, it is the fourth commonest cause of maternal deaths, after hypertension, AIDS, and obstetric haemorrhage. The 41 deaths associated with viable pregnancies have been further analysed in this chapter. The septic abortions are dealt with in chapter 5. However, a similar picture exists with respect to the avoidable factors between cases dying of septic abortion and puerperal sepsis.
Of the 41 puerperal sepsis deaths following viable pregnancies, 70% occurred in women less than 30 years old and 24% were teenagers. The corresponding figures for all maternal deaths were 57%, and 13%. Women dying of pregnancy-related sepsis tend to be younger than average. As regards parity, 44% were primigravida and 6% had a parity of more than 4. The age and parity distribution is similar to that of women dying due to AIDS (Chapter 9). 68% of women did not have their HIV status tested. Of those who did have an HIV test, 69% were HIV positive. It is not clear from the reports, to what extent HIV infection had contributed to the severity of the sepsis. Sixty-six percent (6/9) of the HIV-positive patients were less than 26 years. All women with puerperal sepsis should be counselled for an HIV test, regardless of their age or parity.
Delay in seeking help by patients was the most common patient-related avoidable factor. This was coupled to the lack of access to health care facilities, which occurred in 10% of cases. A woman, once discharged from the health institution, theoretically should be seen by the local health care authorities, on days 3 and 7. This is not feasible due to the collapse of the local authorities home-visiting programme. Women are, therefore, asked to go to their local clinic within 7 days of discharge. Some women experience great difficulty in doing this, and early signs of puerperal sepsis, neonatal problems and breast feeding problems are not seen and treated. A recall system whereby women who do not attend their local clinic within 7 days of delivery are identified and visited would be one way of solving this problem. Caesarean sections were performed in 46% of patients. It is important to emphasise that antibiotics should be given to every woman who undergoes a caesarean section. There is good evidence that prophylactic antibiotics significantly reduce postpartum morbidity1. The question that nowadays needs to be asked is whether prophylactic or therapeutic antibiotics need to be given. In cases of prolonged rupture of the membranes or obstructed labour therapeutic antibiotics need to be prescribed.
Obstructed labour occurred in 27% of cases. In many cases, the labour was neglected. Had the partogram been used, earlier intervention could have occurred, perhaps preventing the subsequent events. It must be stressed that the partogram should be used for every woman in labour.
Puerperal sepsis is not a sudden event like obstetric haemorrhage, thus it is surprising that as many as 24% of the maternal deaths occurred in a Level 1 hospital. There was plenty of time to transfer these patients. This aberration is explained by missed diagnosis or lack of appreciation of the severity of the disease in these cases. Continuing Professional Development programmes should focus on this aspect for doctors in level 1 and 2 hospitals. This, coupled with an efficient referral system, should ensure that seriously ill patients are admitted and/or timeously referred to the appropriate level of health care.
It is clear that the absence or inappropriate use of management protocols has been a major factor in a large number of maternal deaths due to pregnancy-related sepsis. Protocols for the management of puerperal sepsis, and indeed for the management of all important obstetric interventions and complications, must be prominently displayed in every health facility. Furthermore, using the partogram for every patient in labour should be stressed.
The prevalence of puerperal sepsis is likely to rise in the coming years, as the AIDS epidemic progresses further. It is important that problems, which have been identified in this chapter, are rectified by implementing the key recommendations.