The role of obstetric factors in perinatal mortality trends

Trends in concurrent maternal and perinatal deaths at a teaching hospital in Ghana: the facts and prevention strategies

OBJECTIVES

To determine the trend of concurrent maternal and perinatal mortality at the Korle-Bu Teaching Hospital (KBTH), Ghana, and to propose measures for its prevention.

METHODS

A retrospective study, from January 1995 to December 2002, of all concurrent maternal and perinatal deaths in which the woman was 28 weeks' gestation or more (or, if gestational age was not known, the baby weighed 1000 g or more) and died either undelivered or in the perinatal period (within 1 week of delivery) at the KBTH.

RESULTS

Over the 8-year study period, there was a total of 93 622 deliveries at the KBTH with 108 concurrent maternal and perinatal mortalities, giving a ratio of 115.4 concurrent maternal and perinatal deaths per 100 000 deliveries. More than 80% of the mothers who died had little or no formal education. Of the 108 mothers, 22 died undelivered. The leading cause of death was a medical condition in pregnancy along with eclampsia/gestational hypertension. Of the 86 delivered mothers, the leading cause of concurrent death was a medical condition in pregnancy. Approximately two-thirds (72/108) of the perinatal deaths were stillbirths. Over the study period, there was a rising trend of the obstetric disaster of losing both mother and baby.

CONCLUSIONS

There is a rising trend of concurrent maternal and perinatal mortality at the KBTH. It is suggested that a regular antenatal clinic be established with both an internist and obstetrician to jointly see and manage women with medical problems. There is a need for improved and adequate resources to improve outcomes for both mother and baby. A waiver of user fees for maternity services may be one way to improve access for needy and at-risk mothers. Concurrent maternal and perinatal death is the latest negative reproductive health index of the deteriorating socioeconomic situation in developing countries and needs to be tackled decisively.

Rates of cesarean section and perinatal outcome: stillbirths

Based on their yearly clinical statistics for 1983-1989, the cesarean section and stillbirth rates at National Maternity Hospital (Dublin), where the patients are generally white, and University Hospital (Newark), where the patients are predominantly black, were compared. After adjusting for confounding factors, such as unregistered mothers admitted subsequent to fetal demise, differing rates of < 2500 g births and lethal congenital defects, the overall rates of intrauterine fetal demise were identical. When analyzed according to birth weights, in all weight groups, the stillbirth rates were significantly lower in Newark than in Dublin. However, for every 1000 births, more < or = 2500 g fetuses died in utero in Newark than in Dublin. This trend was reversed in the > 2500 g group, where the respective rate at National Maternity Hospital exceeded that of University Hospital more than twice. The latter result largely derived from an almost 7-fold higher rate of intrapartum deaths in the > 2500 g weight group at National Maternity Hospital as compared to University Hospital. When projected against prevailing American statistics, which reflect a stillbirth rate about twice as high among blacks than among whites, the results suggest that the management patterns favored in Newark, including a relatively liberal cesarean section rate (17.5% versus 5.8%), affected the rate of in utero losses favorably.

Rates of caesarean section and neonatal mortality

A comparison of the yearly statistics of National Maternity Hospital (Dublin) and University Hospital (Newark) was undertaken for the years 1983-1989. The findings appear to indicate that after the elimination of major confounding factors, the substantially higher rates of Caesarean section in Newark (17.5% versus 5.8%) did not bring about a measurable reduction in the rate of neonatal losses. The impact of paediatric care upon the respective neonatal outcomes could not be assessed on the ground of the reviewed data.

Measurement of fundal height

Fundal height should not be assessed by comparing the uterus with anatomic landmarks on the maternal abdomen because of the following reasons: 1) the inability to assess fundal growth accurately when the uterus is measured in finger breadths or as a fraction of the distance between two landmarks; 2) biologic variability in the placement of anatomic landmarks on the maternal abdomen; 3) disagreement among clinicians and researchers on the precise relationship between the fundus and anatomic landmarks at each stage of gestation; and 4) variability among women in the number of gestational weeks at which the uterine fundus reaches the umbilicus. A rational basis for continuing the practice of comparing fundal height with anatomic landmarks on the maternal abdomen does not exist. Inexpensive and more accurate techniques of measuring fundal growth are readily available.