Keys to reducing maternal mortality: circumstances of maternal deaths are investigated in Indonesia.

For every 100,000 live births in Indonesia, as many as 400 women die. In some regions of Indonesia--including the province of West Java--maternal mortality rates are believed to be even higher.

Can maternal mortality there be reduced by changing individual behavior? If so, how? Can health service and local government agencies implement any policy, training, budgeting, or procedural changes to prevent maternal deaths?

To answer such questions, investigators from the Center for Health Research at the University of Indonesia used qualitative research methods to better understand the experiences of 63 women from geographically diverse regions of West Java who experienced obstetric emergencies--53 of which were fatal--in 1994 and 1995. (1) Using an innovative method of qualitative data collection called the Rashomon technique, researchers conducted in-depth interviews with an average of six witnesses to the emergencies, including family members, neighbors, village officials, traditional birth attendants, and health care personnel. The witnesses shared what they had observed and their interpretations of the causes of the obstetric outcome, and their detailed accounts were then compared to develop a summary of circumstances surrounding the event. Finally, these accounts were coupled with evidence such as clinic records, police reports, death certificates, and other documentary materials. Using all of this information, physicians and researchers assessed why the death occurred and how such a death could be prevented in the future.

Three themes emerging from the research helped to illuminate epidemiological data showing that the major causes of maternal mortality in West Java are bleeding, infection, and eclampsia:

1. Bleeding, Witnesses' perceptions of blood loss varied and affected how they acted during the emergency. Insufficient appreciation of blood loss could delay the management of postpartum hemorrhaging. But accurate or exaggerated perceptions of blood loss could cause fear or confusion, which could lead to either action or paralysis. Lack of blood supplies and the inability of some health care professionals to administer a transfusion also contributed to death due to hemorrhage.

2. Infection. Unhygienic conditions at delivery contributed to postpartum infection. Furthermore, Javanese culture promotes various postpartum practices that are thought to benefit the mother but are extremely dangerous. These include inserting herbs into the vagina ante--or postpartum; the traditional healer inserting her hand into the vagina during birth and into the uterus after birth to extract the placenta; and the mother sitting for hours after birth with her back to a pole and her legs stretched out in front, with weights on either side of the feet to prevent movement. Infection was also commonly associated with abortion, which accounted for five deaths in the study. Methods of abortion, usually performed by traditional healers, commonly consisted of various herbal drinks to induce contractions, rough massage of the uterus, or the insertion of objects into the vagina to pierce the placenta.

3. Eclampsia. Untreated, the retention of fluids and hypertension associated with preeclampsia progresses until it affects the brain, causing headaches, blurred vision, vertigo, and then convulsions and coma. In many cases, relatives and friends of afflicted women noted particular symptoms but failed to recognize them as dangerous.

Other themes that arose included the surprising finding that many women died after receiving care over a long period of time. Difficult births were often seen as progressing too slowly while, paradoxically, the symptoms were too fast to manage. Often the care was inadequate, either because the birth attendant hesitated to act or to stay with the patient, or because vital supplies were unavailable. "Also, many women die because simple first aid is not given, because lifesaving skills are not practiced on the road between services, and because traditional healers see time as a potentially healing rather than a potentially threatening factor in case management," noted the researchers. Tragically, in several cases, the study found that women were essentially "too poor to live": Family members recognized the seriousness of the obstetric emergency, but did not act because they feared the cost of lifesaving care would be too great to bear.

Modern medical professionals failed to provide adequate care during many obstetric emergencies. Sometimes this failure was due to inadequate training. Sometimes it was due to a lack of materials, medicines, and blood supplies. Referral systems also were problematic. Clinics often quickly referred a patient to a hospital but failed to stabilize the patient's condition, to give appropriate first aid, and to ensure that appropriate, responsible people accompanied the patient. Clinics, hospitals, blood banks, and government agencies were all semi-independent institutions that often failed to communicate well and in a timely manner. At times, practitioner errors or deliberate negligence exacerbated such weaknesses in the system.

Among recommendations made as a result of the qualitative study were:

* Training of community-based providers about essential emergency obstetric care. The government should provide training about essential emergency obstetric care and lifesaving skills to village midwives and to community health center midwives and paramedics who are likely to attend births. Emergency obstetric care should involve treating symptoms of "shock," giving injections (including antibiotics and sedatives), and performing manual extraction of the placenta and simple curettage. Training in lifesaving skills should include pre- and post-evaluation of skills.

* Training of hospital-based providers about lifesaving skills and provision of essential comprehensive obstetric care. Hospital midwives and nurses should be trained to administer blood and blood product transfusions and provide other comprehensive emergency obstetric services without delay. General practitioners should be trained to manage postpartum hemorrhage and perform appropriate surgery. They should also be trained to treat preeclampsia or eclampsia and to manage prolonged labor, malpresentations, and neonatal emergencies.

* Training of midwives and hospital personnel to emphasize quality of care. These health care providers should be trained to improve communication with clients and to better educate clients about signs of danger during pregnancy and delivery.

* Supervision. Supervisors must support, guide, train, and assist medical staff to identify and solve problems. Meanwhile, clear job descriptions for staff should be accompanied by checklists to help supervisors monitor performance. Supervision should rely more on close and direct observation of staff activities. Supervisory responsibilities should include not only ensuring quality care but also making available appropriate training and information, education, and communication materials, as well as medical supplies and equipment.

* Raising the level of community preparedness for obstetric emergencies. Community-based information and education efforts should emphasize the potential risk of obstetric complications for all pregnant women and the need for individuals and family members to recognize danger signs. "It is obvious that the primary cause of delay in obtaining adequate care is failure to recognize danger signs, or the tendency to seek treatment from traditional healers whose methods are of little or no efficacy, and in many cases may exacerbate a bad situation," investigators noted. "The key to overcoming this situation is to educate the community to understand the risks and danger signs of pregnancy and delivery, and to mobilize them to obtain appropriate care quickly."

Reference

(1.) Iskandar M, Utomo B, Hull T, et al. Unraveling the Mysteries of Maternal Death in West Java: Reexamining the Witnesses. Depok, Indonesia: Center for Health Research, University of Indonesia,