MCHB/EPI Miami Conference — December 7 - 9, 2005

All Around the Globe — Transcript

KATHERINE BERGER: I was originally concerned about time when I was noted that I was the fifth person to present, but I guess now I don't have to worry about that, which I'm thankful for.

My talk is going to be presenting about abortion related morbidity and mortality in Bangladesh during 1996-97. Bangladesh is a relatively poor and predominantly Islamic nation. The nation is approximately the size of Iowa , but has a population of nearly 144 million, which is just under half the population of the entire United States . The nation's largely agricultural and has 64 districts. There're an estimate 4 million live births per year.

Maternal mortality in Bangladesh is high and estimates range from 6.2 deaths per thousand live births in the early 1980s to a more recent estimate of 4.3 deaths per thousand live births in 1996. The proportion due to abortions has also historically been high with estimates as large as 25 percent. However, considerable improvements have been made with respect to maternal health. Over the last 20 years the fertility rate has decreased by nearly one half. Over the last 25 years, a number of changes have taken place with regards to maternal health services. Primarily, family planning services have been expanded which has resulted in the large decrease in fertility rates. Part of these changes have included the increased availability of abortion procedures.

Beginning in the mid 1970s the government slowly started introducing menstrual regulation services. And menstrual regulation refers to the use of a syringe to extract the contents of the uterus up to ten weeks gestation in order to restore menstruation. During the last 20 years these services have been expanded and extended throughout Bangladesh and the government has trained over 12,000 physicians and other health care providers, mainly family welfare visitors which are the primary health care providers in the rural areas and their females will provide reproductive health care.

The impact of menstrual regulation services on reducing maternal mortality and abortion related deaths has been significant, although determining the extent of the impact has been difficult due to the scarcity of data on abortions. Abortions remain generally underreported, not only in Bangladesh but in South Central Asia as a whole and deaths from abortion related complications are difficult to ascertain. And despite the safety of increased availability of menstrual regulation procedures, women in Bangladesh still lack access to these services based on differences in location, financial status, and despite access many women still seek traditional providers because of convenience. And these women are at greater risk of complications that could result in death.

We can get some estimates of the impact of abortion on maternal mortality from studies conducted in the 1970s and 1980s. The only study to look at comprehensive rates at a national level was completed in 1978 to 79 by a team of researchers that interviewed health workers in 795 health centers around Bangladesh . Nearly 26 percent of all maternal deaths ascertained in the study were estimated to be a result of abortion-related complications.

At this time almost half of the reported abortions were performed by untrained birth attendants through insertion of a foreign object into the uterus, most commonly a root or a stick. And it was also estimated that if services had been available for women to obtain a medically approved abortion from a trained provider, nearly 84 percent of the deaths would've been prevented.

Additionally, some estimates have come from studies in the Matlab area of Bangladesh , which is the rural sub district 30 kilometers from the capital city. In 1977 the area was divided into two areas, it was a research and comparison areas to determine whether carefully executed, very basic family planning programs could reduce fertility in maternal mortality. After about 10 years they determined that 15 percent of maternal deaths were attributable to complications resulting from induced abortions. Maternal mortality did decrease over that time period, but the proportion of deaths due to complications of abortions did not.

Due to the scarcity of data on deaths among reproductive aged women on maternal mortality and abortion complications the Bangladesh Institute for Research Promotion of Essential and Reproductive Health Technologies in conjunction with the CDC conducted a study that was modeled after the 1978-79 national survey.

The study was conducted using similar methods of interviewing health care providers, but was much more comprehensive in the amount of data collected and the number of facilities visited. Nearly 4800 facilities throughout Bangladesh were visited and cases were collected from all but 4 districts out of 64. Case records were obtained from interviews with health care providers and hospital records. For each known maternal death or abortion complication case a questionnaire was administered. The study covered approximately 79 percent of government facilities and 54 percent of community health workers. Overall, it was estimated that case reports were obtained on 43 percent of the deaths in that past year. A total of nearly 29,000 deaths to women over 10 years of age were reported in the survey and nearly 30 percent of them were maternal deaths. Of the maternal deaths, almost 17 percent were attributable to complications resulting from induced abortions.

Overall, the study showed a direct decrease in maternal mortality rate for women over 10 years of age, but the ratio of deaths per 1000 live births appeared to remain unchanged from estimates obtained in early studies. This suggests that the decline may largely be attributable to decline in fertility rates, not a decreased risk and the risk of death for women who become pregnant remains high. The study confirms and highlights many of the problems faced in Bangladesh with respect to maternal mortality and provides a valuable tool for assessing abortion complications and understanding the magnitude of risk associated with pregnancy termination. So the purpose of this study was to describe women who had a complication or death due to abortion and to determine the risk associated with medically approved and non-approved abortion procedures.

The subset of data from the Bangladesh survey on abortion morbidity and mortality was utilized for this study. There were approximately 31,000 case reports detailing abortion complications and deaths. Information was collected on the location, source of information, and outcome of the abortion complication. Various demographic characteristics were also ascertained, such as education, age, and with respect to the abortion procedure method, provider, the number of attempts, and weeks gestation were collected. Overall, women with abortion complications were primarily age 20 to 29. Approximately 5 percent of all women died from complications and the highest death rate was experienced by women who were 10 to 19. Nearly 8 percent of these women died.

When we look at additional characteristics of the women with abortion complications, nearly 45 percent of the women were considered to be poor and these women experienced a 58 percent increase in risk compared to women with higher economic status. Fifty-seven percent of the women were illiterate and had an 89 percent increased risk compared to women who completed primary education or greater. While married women did not account for a huge proportion of the women, they did experience over a three-fold increase in risk compared to married women. Nearly 75 percent of the women who had abortion complications resided in villages or rural villages. And these women had a nearly four-fold increase risk compared to women residing in larger populated areas.

Women did not always have a successful first attempt at abortion. Almost 14 percent of the women attempted the abortion two or more times. The percent resulting in death decreased with the number of attempts. The women obtained repeated abortions for a number of reasons, likely due to incomplete abortions for the first attempt or complications resulting from previous attempts. The beneficial association between attempts and the outcome of death may have been due to the additional treatment received by the women or may have been due to healthier outcomes of the first attempt that allowed the women to be physically capable of obtaining the second procedure.

When we look at the providers of abortions we can see that health care professionals provided about 53 percent of the procedures. Family welfare visitors were the majority of trained providers and had the lowest rate of death. Untrained providers accounted for almost 50 percent of the procedures. Traditional midwives had the highest rate of death.

When we look at the methods for abortion, we can see that almost 59 percent were medically approved. Fifty-five percent of these were menstrual regulation. And almost, about 40 percent of these were non-medically approved, with a large proportion attempting abortion with either a plant root or a stick. Menstrual regulation showed the lowest rate of death with only 1.7 percent and women who squirted traditional herbs or liquids into the vagina, they had the highest rate of death.

When we look at the methods and who provided those methods, we can see that almost 75 percent of the procedures that were performed by menstrual regulation were performed by family welfare visitors. Only 1 percent of these resulted in death.

When we look at the use of a plant root or a stick, almost 30 percent of these were performed by the woman herself and almost 16 percent of these resulted in death. Looking at women who attempted their abortion multiple times, it was interesting to note that 70 percent of these did use menstrual regulation as their last procedure and only 1 percent of these died. If they didn't use menstrual regulation, nearly 8 percent of these women died.

When we look at the overall risks associated with procedures, other than menstrual regulation we can see that if the woman did not use menstrual regulation she had almost a nine-fold increased risk of death. These results support and agree with earlier studies with respect to the dramatic decrease in mortality that would be seen if menstrual regulation was the method of choice for all women. However, over 50 percent of the women in the study did not choose menstrual regulation as their final method and method alone also did not determine the risk faced by women obtaining an abortion. This risk was modified by weeks gestation and by marital status significantly.

We can see that for a woman who was married and who obtained an abortion at greater than 10 weeks gestation, the risk actually decreased for methods other than MR, and what this really indicates is that the benefit of using menstrual regulation decreases some with the increased amount of gestation. Just the risk associated with obtaining an abortion greater than 10 weeks gestation is also evident for menstrual regulation, but menstrual regulation's still significantly protective.

What we do see that's very interesting though is that for unmarried women the risk just escalates, irrespective of weeks gestation. These women have almost a twenty-seven fold increased risk of death if they use a method other than menstrual regulation. And this may have been likely because of the large social stigma surrounding unmarried pregnant women in Bangladesh and many of these women may be going to extreme measures to terminate their pregnancy. They're likely also to have less access to safe facilities and may want to keep their pregnancy private and avoid health care workers. These women do have a very high risk of death and effectively targeting these women will pose a challenge for future program efforts in Bangladesh .

So in conclusion, we can see that 5 percent of women overall with abortion complications died. Low economic status, illiteracy, rural residence, and being unmarried significantly increased the risk of death. While this is not surprising it should be noted that these risk factors represent a large proportion of the population and indicate that a large number of women still face significant risks from abortions. Also over half of abortions were provided by trained providers and completed using MR, which leaves 50 percent that also weren't. And women not using MR had an eight-fold increased risk of dying. And what we can see is that by drawing in all these factors a larger picture emerges. And simply making menstrual regulation available may not be enough to combat the high proportions of maternal deaths resulting from abortion complications. Trained health care workers were essential for assuring the safety of menstrual regulation and differences in provider, weeks gestation, and marital status result in different levels of risks faced by women obtaining an abortion through menstrual regulation.

There were some limitations with this study. The first would be recall bias. Interviewers in the study had--those who were interviewed had more trouble recalling deaths 6 to 12 months ago, compared to 0 to 6 months ago. So the further out the death was the less likely those interviewed were likely to remember it. So there's likely many deaths missing. Also, there's a large proportion of missing data. Nearly 50 percent of the abortion complication cases are issuing type of method and various other demographic information and this could easily impact this magnitude of the risk faced by the women and/or estimates of mortality. The third weakness was the characteristics of the study cohort. The women in the cohort were included based on the result of serious complication following abortion, thus these women already had a high risk of death from complications and the true risk for all women obtaining an abortion may be greater or lower than that produced in the study. And for a public health impact, we can see that increasing our understanding of abortion associated morbidity and mortality in Bangladesh it's necessary to increase the availability, understanding, and use of safe abortion procedures.

And I'd just like to acknowledge the Bangladesh Institute for Research for Promotion of Essential and Reproductive Health Technologies and the CDC, who were--conducted the study and instrumental in providing with the data. Thank you very much.