Pre-existing maternal disease remains an important cause of indirect maternal mortality in countries with poor socio-economic status. There were 59 maternal deaths due to pre-existing maternal disease. This comprises 10.4% of the total mortality. There were 28 deaths due to maternal cardiac disease. The non-cardiac deaths occurred mostly in insulin dependent diabetics and neurological conditions of stroke and epilepsy. Valvular heart disease (54%) and peripartum cardiomyopathy (25%) were the most frequent cardiac conditions. Fifty-four percent of the 35 women died within 24 hours of admission and 60% of the deaths occurring in the postpartum period. The majority (77%) have demonstrated some deficiency in their care. The major shortcomings recognised were a failure to recognise symptoms and/or signs of mild to moderate cardiac failure. There was failure to consider peripartum cardiomyopathy when faced with unexplained cardiac failure especially in cases associated with pre-eclampsia. A failure to perform ancillary electrocardiogram and chest radiography together with "reluctance" to seek specialist help and advice was also noted.
The following recommendations are made:
Key Recommendations
Pre-existing maternal disease remains an important cause of indirect maternal mortality in countries with poor socio-economic status. Maternal cardiac disease, in particular rheumatic valvular heart disease and peripartum cardiomyopathy, are the important treatable causes identified in this report.
There were 59 maternal deaths due to pre-existing maternal disease. This comprises 10.4% of the total mortality. There were 28 (47%) deaths due to cardiac disease. There was a trend to a greater mortality in the higher age groups and those with greater parity for women with cardiac disease than the general pregnant population (Table 10.1). The effect was more marked for women with pre-existing medical conditions other than cardiac disease (Table 10.2).
The various causes of maternal cardiac deaths are listed in Table 10.3. It is to be noted is that rheumatic valvular heart disease and peripartum cardiomyopathy are the most frequent.
The mean gestation at presentation was 33 weeks. 53% women died within 24 hours of admission, 31% died with hospitalisation duration of 2-7 days and the remaining 14% died at duration beyond 7 days. 39% deaths occurred antepartum with the remaining 61% deaths occurring postpartum at mean 11 days and range 1-35 days. These women were managed at level 1 (5, 18%); level 2 (11, 39%) and level 3 (12, 43%).
The 31 (53%) non-cardiac deaths are shown in Table 10.4.
Tables 10.5-10.7 summarises the avoidable factors, missed opportunities and substandard care encountered in 1998. The management of cardiac disease stands out as a major problem. A large proportion of patient orientated factors is noted in maternal deaths due to heart disease. This possibly is a reflection of lack of understanding on the part of the women about the nature of their condition, or alternatively avoidance behaviour because they receive a hostile reception from doctors if they declare they want to have a baby or present when they are pregnant. Patients' wishes must be respected, and a sympathetic attitude should prevail in dealing with women with heart disease. A co-operative relationship will achieve more than a confrontational one.
The majority of maternal deaths due to cardiac disease demonstrated some deficiency in their medical care. All managed at level 1 had errors, 87% had errors at level 2 and 53% had errors or deficiencies at level 3. Appropriate case histories are used to highlight the shortcomings and are discussed hereafter.
The majority of these were mitral stenosis. The errors in management occurred mainly as a result of inaccurate diagnosis; inability to appreciate severity of the cardiac disease and inappropriate management at level 1 and 2 care.
Case 1: She was 33 years and unbooked. She presented with pulmonary oedema from her mitral stenosis. She was inappropriately managed at level 1 care for 2 days before transfer to level 2 and then to a level 3 hospital. At level 3, her confused and restless behaviour from hypoxia was treated with sedatives. She died after a 12-hour stay at the level 3 hospital.
Case 2: She presented to the antenatal clinic at 20 weeks with symptoms of heart failure. She was not recognised as a cardiac and discharged with haematinics. Her symptoms persisted and she was admitted with a diagnosis of bronchopneumonia to level 2 care. Nursing notes confirm symptoms pointing to cardiac failure. A tachycardia was noted and digoxin was given. Level 3 help was not sought. She demised after 5 days from seizure and cardio-respiratory arrest.
Case 3 and 4: Two women presented in cardiac failure at 20 and 23 weeks gestation to level 3 care. The correct diagnoses of severe mitral stenosis were made. They were stabilised on medical therapy and asked to return for review in 4 weeks. There was failure to map out a plan for pregnancy and delivery. Both presented 3 weeks thereafter in florid pulmonary oedema and demised thereafter.
Case 5: She was 36 years old, P4 with prosthetic mitral valve 2 years prior. She had made 3 antenatal visits. She had declined tubal ligation. Her warfarin anticoagulation was changed to aspirin/persantin at 25 weeks gestation by level 2 care. Level 3 care was not sought. She next presented to level 1 care in extremis at 32 weeks gestation and demised within 20 hours of admission. She probably developed prosthetic valve thrombosis. She demonstrates inappropriate care regarding her anticoagulation therapy.
Case 6: She was 18 years, P0 who had had a mitral valve replacement at age 10. She did not present for antenatal care and was not on any contraception. She presented late at night in pulmonary oedema, to level 3 care. Physicians and cardiologists failed to respond to obstetrician request for help. She was seen the following morning, noted to be in extremis and died soon thereafter. She probably developed thrombosis of her mitral valve prosthesis with resultant pulmonary oedema. She demonstrates failure of appropriate level 3 doctors to respond to an obstetric emergency. This case also highlights the need for contraceptive advice in young girls who have a prosthetic valve as pregnancy can bring on disastrous consequences.
Comment: Prosthetic valve disease is particularly difficult to manage. The preceding 2 case history's highlights some of the difficulties and shortcoming in management.
There were 7 deaths due to dilated peripartum cardiomyopathy. In a further 3, classified as undiagnosed, the probable diagnosis was peripartum cardiomyopathy. The mean age was 25 years with a range of 15-33 years. Six were P0, 2 were P1 and the remaining 2 were P3 and P5. The majority presented late in the third trimester. Most were managed at level 2 care (n=6) and 2 each were managed at level 1 and 3 care respectively. Nine of the 10 deaths occurred in the postpartum period. Errors in management were noted in 9 of these 10 cases, the major feature being a failure to diagnose mild cardiac failure with a resultant failure to make a diagnosis of a dilated cardiomyopathy.
The following case histories highlight shortcomings and difficulties that are often encountered.
Case 1: A 15 year old, P0, attended level 1 care at 37 weeks with symptoms of cardiac failure. An incorrect diagnosis of pneumonia was made and she was managed as an outpatient with antibiotics. She presented a week later to level 2 care in severe cardiac failure. Fetal distress was noted. Delivery was effected by emergency caesarean section under spinal anaesthetic during which she developed cardiovascular collapse. Autopsy confirmed a severe dilated cardiomyopathy with pulmonary emboli and bronchopneumonia. This case demonstrates shortcoming at level 1 antenatal care and a failure at level 2 care to appreciate the severity of maternal disease and to act primarily in maternal interest.
Case 2: A 17 years old, P0 who made 5 antenatal visits to level 1 care. She delivered a 3,2kg FSB at 38 weeks. She was noted to be ill and referred to level 2 care. A diagnosis of puerperal sepsis/pneumonia was made. She deteriorated and a chest X-ray done 2 days later showed a dilated heart and pulmonary oedema. She was referred in extremis to level 3 care where she demised the following day. This case demonstrates failure to recognise cardiac failure at level 1 and 2 care.
Case 3: She was 32 years old, P1, unbooked who presented with twin pregnancy to level 1 care. She developed heart failure post delivery and was referred to level 3 care. A diagnosis of a dilated peripartum cardiomyopathy was made but she was discharged home one week thereafter. She represented 2 weeks later in extremis to level 1 care and demised.
Comment: This case demonstrates a failure to appreciate the severity and nature of peripartum cardiomyopathy at level 3 care. Bed rest is an important aspect of therapy and such patients often need many weeks of hospitalisation and medical therapy.
Case 4: She was 26 years old, parity not documented with 4 antenatal visits at level 2 care. She had a caesarean section for pre-eclampsia and was discharged on day 3 with methyldopa (aldomet). She was noted to be coughing at discharge, a symptom of cardiac failure that was not appreciated. She returned 3 days later with in florid pulmonary oedema and died within 9 hours.
Comment: This case demonstrates a failure to recognise mild to moderate cardiac failure early postpartum resulting in inappropriate discharge and lack of specific therapy for cardiac failure from peripartum cardiomyopathy. Peripartum cardiomyopathy is not uncommonly seen in the setting of pre-eclampsia often resulting in a missed diagnosis.
Case 5: She was 18 years old, P0. She presented at term to level 2 care in labour and pulmonary oedema. A diagnosis of pre-eclampsia was also made. She had a caesarean section delivery. Level 3 requested stabilisation prior to transfer. A specific diagnosis of peripartum cardiomyopathy was not made. She deteriorated and was only transferred 10 days thereafter to level 3 care in extremis where she died.
Comment: This case demonstrates a failure to make a specific diagnosis of peripartum cardiomyopathy and thereby institute early specific therapy. It also demonstrates the need for level 1 and 2 doctors to maintain close liaison with level 3 so that when stabilisation is not achieved, earlier transfer can be arranged. Level 1 and 2 doctors must liase closely with level 3 institution doctors. Failure of such liaison contributes to suboptimal care.
Case 6: She was 33 years old, P5 known to be HIV positive with a low CD4 count. She had an uneventful caesarean section delivery at level 2 care. She developed wound sepsis, anaemia and cardiac failure post delivery. A diagnosis of peripartum cardiomyopathy was confirmed. She remained in intractable cardiac failure and died thereafter.
Comment: Retroviral disease can co-exist or cause a dilated cardiomyopathy. In women with retroviral disease and breathlessness, the diagnosis of a dilated cardiomyopathy must be considered once a pulmonary cause has been excluded.
Medical conditions other than cardiac disease These conditions are relatively rare in themselves, but as a group contribute a significant proportion of maternal deaths. Some case histories are given below.
Case 1: A woman aged 37 years old, 24 weeks gestation, died suddenly at home. Autopsy confirmed a ruptured berry aneurysm.
Comment: This case illustrates the importance of obtaining autopsies on women who die suddenly.
Case 2: A woman with severe kyphoscolisis, 29 years old, P0 and presented at 22 weeks gestation. She declined termination of pregnancy. She became progressively breathless with wheezing and was in respiratory failure at 28 weeks. Ventilation was not offered as her prognosis was assessed as being poor.
Comment: The dangers of respiratory complications with kyphoscolisis are illustrated here.
Case 3: A woman with dwarfism, 21 years old, P0 and presented to level 2 hospital at 38 weeks. There was no medical "work-up". An epidural caesarean section was performed 4 days following admission. On day 1 post delivery, she developed unexplained cardiac arrest. Pulmonary oedema was noted at autopsy. However, no structural heart disease was detected.
Comment: Fluid supplementation must be adjusted to the size of the patient.
The following 4 cases illustrate problems encountered with women with diabetes.
Case 4: A woman, 25 years old, P2 presented at 17 weeks gestation. She had 2 previous intrauterine deaths at 32 weeks gestation. She was a known insulin dependent diabetic for 11 years with poor glycaemic control. She presented with an elevated blood pressure of 180/130 and symptoms of imminent eclampsia. IUD was confirmed. She had repeated episodes of hypoglycaemia and altered level of consciousness and seizures. Initial emergency care was suboptimal. Despite subsequent ICU care, the patient remained comatose from anoxic brain damage and died.
Case 5: A woman was 36 years old, P3 at 21 weeks gestation. She was an insulin dependent diabetic who was diificult to control She was discharged after 2 weeks of hospitalisation. She presented soon thereafter with symptoms of hypoglycaemia. This was attributed to an omitted snack. The insulin dose was not altered. She died at home 2 days thereafter, most probably from hypoglycaemic coma.
Case 6: The third patient had poor antenatal attendance. She was 33 years, P2 at 37 weeks. She presented in labour and had a normal vaginal delivery. Twelve hours after delivery she was noted to be "dull". Hypoglycaemic coma was diagnosed. She died despite subsequent ICU care.
Case 7: The remaining woman had problems of hyperglycaemia. She was 40 years old, P1 at 24 weeks gestation. She was a difficult to control diabetic managed with oral hypoglycaemics at a level 2 hospital. There was history of herbal medicine ingestion. She died with hyperglycaemia, dehydration and presumed ketoacidosis.
Comment: These cases identify shortcoming in the management of insulin-dependent diabetics in pregnancy. Three of the 4 maternal deaths due to glycaemic control were readily preventable. It is recommended that 'brittle' insulin dependent diabetics in pregnancy be managed as far as possible as inpatients particularly if socio-economic circumstances are not satisfactory.
On review of the maternal cardiac deaths, the following shortcomings are recognised:
A failure to recognise symptoms and/or signs of mild to moderate cardiac failure. This results in failure of timeous appropriate and adequate therapy with resultant adverse outcome.
A failure to consider the diagnosis of peripartum cardiomyopathy and to a lesser extent mitral stenosis when faced with a dyspnoeic woman or unexplained cardiac failure, bronchopneumonia or deep venous thrombosis in the peripartum period.
A failure to consider the diagnosis of peripartum cardiomyopathy when a patient with suspected or diagnosed pre-eclampsia develops additional symptoms such as breathlessness, excessive pedal oedema, and bronchopneumonia or has unexplained anaemia.
A failure to do ECG and/or chest radiograph to assist in the diagnosis of cardiac disease.
A failure to call for specialist help and advise.
A failure to appreciate the severity of the disease. Women with valvular heart disease who develop heart failure early in pregnancy are likely to decompensate in the third trimester and peripartum period.
Such patients must be considered for some definitive therapy or prolonged hospitalisation.
Failure to admit/investigate women with suspected and/or known cardiac disease early in pregnancy to ensure that an accurate diagnosis of the cardiac lesion is made with an appropriate plan for pregnancy and delivery.
Noting the above shortcomings, the following recommendations are made.