In 1991-1992, 230 Japanese women died while pregnant or within 42 days after their pregnancy ended; the resulting maternal mortality rate of 9.5 deaths per 100,000 live births is somewhat higher than rates in other developed countries at roughly the same time. Moreover, according to the results of an inquiry initiated by the Japanese government, nearly two in five maternal deaths that occurred in medical facilities could have been prevented. Findings from the inquiry highlight how inadequacies in Japan's system of delivering obstetric care contribute to the risk of pregnancy-related death.1
A 15-member study group was convened to conduct the inquiry, using data from death certificates to identify maternal deaths. The group investigated the details of each case by sending a questionnaire to the medical facility that provided care to the woman or, in instances where the woman had not received care, to the coroner's office. In addition to examining the characteristics of women who died and the causes of death, the group assessed characteristics of medical facilities involved, such as staffing patterns and the availability of laboratory services. Finally, the researchers invited a panel of 42 medical specialists to review the records of women who died in medical facilities and assess how preventable each death was (using the categories impossible to prevent, difficult to prevent, not difficult to prevent or indeterminable).
In all, 230 maternal deaths occurred in Japan--9.5 for every 100,000 live births--during the period 1991-1992. While most of the women (197) had received care and died in medical facilities, 22 died outside a facility; information on 11 women was unavailable. Among women who died in hospitals or clinics, 80% had received regular prenatal care; 58% had given birth previously. Thirty-seven percent of these women had a cesarean delivery, the same proportion delivered vaginally and 26% died before giving birth. The most common cause of death among women who had received medical care was prenatal or postpartum hemorrhage (38%); a variety of other direct and indirect causes each accounted for fewer than 15% of deaths. Similarly, postpartum hemorrhage was the most frequent cause of death among women who had not received medical attention.
Of the 197 deaths that occurred in medical facilities, the panel of medical experts deemed 72 (37%) preventable: Panel members unanimously characterized 19 deaths as not difficult to prevent, and at least 70% of these specialists characterized an additional 53 deaths in this way. Thirty-two deaths (16% of those occurring in medical facilities) were judged possibly preventable, and the remainder unpreventable. As was the case overall, most preventable deaths were attributable to hemorrhage.
Japanese hospitals and clinics that provide obstetric care have, on average, only one obstetrician on duty at a time, and a majority have no anesthesiologist. As a result, one doctor often serves as both obstetrician and anesthesiologist, and this was true for about two-thirds of women whose deaths were preventable (46 with hemorrhage and three with complications from anesthesia). In addition, the medical experts determined that 63% of preventable deaths were associated with deficiencies in hospital care, 10-13% with deficiencies in ambulatory or inpatient care and 50% with failure to meet basic practice standards.
The rate of unpreventable deaths was highest (12.9 per 100,000 live births) in facilities with at least four obstetricians, but the highest rate of preventable deaths was found in hospitals and clinics with only one obstetrician (4.1 per 100,000). Preventable maternal deaths due to hemorrhage also occurred with greatest frequency at facilities with only one obstetrician (3.8 per 100,000), while no such deaths occurred at facilities with four or more obstetricians.
Additional analyses provided further insights into the relationship between maternal deaths and hospital or clinic characteristics. For example, the proportion of preventable deaths declined as the number of either obstetricians or anesthesiologists at a facility grew. Furthermore, the rate of preventable deaths was considerably higher at facilities that transfer patients elsewhere before they die (56 per 100,000) than at facilities to which patients are referred (four per 100,000). Finally, facilities where maternal deaths occurred were unlikely to have basic laboratory services available around the clock.
As a result of their inquiry, the researchers propose several changes in Japan's medical delivery system. They recommend the establishment of regional obstetric facilities with increased physician coverage 24 hours a day. They further suggest that all facilities providing delivery care be staffed with at least one obstetrician and one provider of related nonobstetric care, and that these facilities be equipped to perform "essential laboratory services." Another proposal, stemming from a lack of uniformity in reporting of maternal deaths, is that all death certificates include standard classifications of obstetric and pregnancy-related deaths. Finally, the study group urges the government and the Japanese Society of Obstetrics and Gynecology to collaborate on the development of "clear community practice standards that delineate specific staffing and laboratory services necessary in each type of medical facility." While acknowledging the inevitability of some maternal deaths, the group members conclude that a systemic approach to change should reduce the occurrence of these events.--D. Hollander
1. Nagaya K et al., Causes of maternal mortality in Japan, Journal of the American Medical Association, 2000, 283(20): 2661-2667.