What did Bangladesh do?

For decades, Bangladesh has centered reproductive, maternal, and newborn health in its public health agenda. By expanding health infrastructure and reducing barriers to care, Bangladesh has managed to improve key indicators of health system coverage. Interventions provided along the progression from family planning care to antenatal care, delivery care, and postnatal care have been instrumental in reducing neonatal and maternal mortality in Bangladesh.

Improvements in reproductive, maternal, newborn and child health indicators

Contraception access

Bangladesh has a long history of prioritizing family planning programs, beginning in the late 1970s. A now famous research project was conducted in the subdistrict (upazila) of Matlab, in which several villages were given expanded access to family planning services including home visits, method consultations, and follow-up counseling.¹ Years later, not only did the women in this program have fewer children, but they also demonstrated improved care-seeking behavior, had improved economic outlooks, and their children had lower mortality rates.¹ The success of this program inspired a nationwide expansion of family planning services centered around the door-to-door approach led by female health workers.¹

From 1993 to 2018, modern contraceptive prevalence among married women increased from 36.6% to 51.9% as shown in Figure 12. ² Over this same time period, the demand for family planning satisfied by modern methods grew from 55.0% to 70.3%.² These increases are both primarily driven by the use of the pill, injections, condoms, and female sterilization, which respectively make up 48.9%, 20.6%, 13.9%, and 9.2% of all modern contraceptive use among married women.² Use of traditional methods has remained relatively constant, between 8% to 10% from 1993 to 2018.²

As contraceptive use has increased, women in Bangladesh have generally had fewer children. The total fertility rate in Bangladesh declined from 6.9 live births per woman in 1970 to 4.5 by 1990 and 2.0 in 2019.³ Shown in Figure 13, declines in age-specific fertility rates were observed for all age groups, with particularly strong decreases for women ages 20 to 24. Women who gave birth by age 15 decreased from 14.0% in 1993 to 9.8% by 2018, reflecting the decrease in marriages before this age.²

Antenatal coverage

Women in Bangladesh are receiving antenatal care earlier and more often as shown in Figure 14. The percentage of women receiving at least one antenatal care visit from a skilled provider nearly tripled from 31% in 1996 to 92% in 2018, while the percentage of women receiving at least four antenatal care visits from skilled providers rose from 6.9% to 47% over the same time period.² Antenatal care often began earlier as well, with 37.1% of women in 2018 receiving their first antenatal care visit during their first three months of pregnancy, compared with only 13.7% in 1997.²

As many women receive at least one antenatal care visit, and more are receiving at least four antenatal care visits, maternal and newborn health experts have recommended expanding antenatal care coverage as a key area for quality improvement. One specific recommendation was regular follow-up visits to upazilas by the government and partners to monitor antenatal care coverage. Experts interviewed emphasized the need to identify and refer patients with potential complications and high-risk pregnancies during antenatal care.

Antenatal care quality

In addition to antenatal care coverage increasing, the quality of health care provided at these visits has improved or maintained high levels according to several key indicators. At antenatal care visits in 2018, 93.2% of women had their blood pressure taken, 88.1% of women were weighed, 72.1% of women provided a urine sample, and 65.7% of women provided a blood sample.² These all represent increases from 2004 in which these services respectively occurred at 32.1%, 73.1%, 45.2%, and 32.1% of antenatal care visits.² More than half of women who received any antenatal care visit in 2018 had at least one visit at a private facility, with the majority conducted by doctors as opposed to nurses or midwives.²

In-facility delivery

Although half of all births in Bangladesh still occurred at home in 2018, the rate of homebirths steadily declined in the previous decade, highlighted in Figure 15.² In 1999, 91.0% of births occurred at home in Bangladesh, compared with 50% by 2018.² The most rapid growth during this period was among births at private for-profit facilities, which only accounted for 3.0% of births in 2000 but constituted 32.0% of births in 2018.²

Skilled birth attendants

In line with increases in facility-based delivery, the percentage of births attended by a skilled provider (e.g., doctor, nurse, midwife) grew from 7.7% in 1996 to 52.7% in 2018.² Of these births, 72.9% were specifically attended by a doctor—reflecting the growing portion of births that occurred in private for-profit facilities.² This trend coincides with a steady decrease in attendance of traditional birth attendants, which were present at 57.2% of births in 1996 but only 9.8% of births in 2018.²

Cesarean section

The rate of cesarean sections (C-sections) increased in Bangladesh from 2.4% in 1999 to 32.8% in 2018.² Although a C-section is a valuable intervention aimed at saving maternal and newborn lives when medically necessary, overuse of this procedure presents undue medical risk. The World Health Organization stance is that a C-section rate above 10% is not associated with a reduction in MMR.⁴  As shown in Figure 16, Bangladesh has far exceeded this rate in recent years. Of C-sections in Bangladesh in 2018, 57% were elective C-sections and 43% were emergency C-sections.⁵ The C-section rate in private facilities was particularly high at 78.6% of births, compared with 35.2% in public facilities.² A recent study in Bangladesh identified that the final decision to undergo a C-section was made by the doctor 74.6% of the time, by the mother and her family 19.3% of the time, and specifically by the mother 6.2% of the time.⁶ Interviews with doctors about C-section decision-making suggest that they often chose to perform C-sections because of their personal preference for C-section delivery, pressures they faced from patients who wanted C-sections, and because the procedure can often be completed faster than a potentially lengthy labor and vaginal delivery process.⁷ Patients often reported that they elected the procedure because of peer pressure from others who had received C-sections, fear of discussing preferences with doctors, and an inflated sense of safety in regard to the procedure.⁷ This is reflected in another study that found 16.3% of C-sections in Bangladesh were performed specifically to avoid labor pains or out of convenience.⁶ Despite the rising levels of elective C-Sections, the expansion of C-section services in medically necessary situations has nonetheless contributed to Bangladesh’s status as an Exemplar.

Maternal postnatal and neonatal care

Coverage of postnatal care for mothers within four hours of birth increased from 25% in 2007 to 49% in 2018.² Similarly, the percentage of mothers who received no postnatal care decreased from 65.1% in 2007 to 46.8% in 2018.² The duration of stay in health facilities after birth also substantially increased, largely linked to the increase in the C-section rate, as 97.7% of mothers who gave birth via C-section stayed in the facility for at least three days, compared with only 9.8% of women who gave birth vaginally.² From 2014 to 2018, the percentage of women who received postnatal care from community health workers declined from 12.7% to 0.6%, while the percentage of mothers who received postnatal care from doctors, nurses, or midwives increased from 36.9% to 51.4%.²

More neonates in Bangladesh are receiving care, and this care is increasingly occurring earlier in their lives. In 2018, 38% of neonates received care within their first hour of life, compared with 19% in 2011.² The percentage of neonates not receiving care decreased over this period from 58.9% to 46.6%.² Key early life indicators have also improved, with the percentage of children initiating breastfeeding within one hour of birth increasing from 8.5% in 1994 to 59.8% in 2018.² Other key neonatal care indicators had high coverage in 2018, with the cord examined for 85.2% of births, temperature measured for 87.4% of births, and breastfeeding counseling occurring for 80.2% of births.²

Connecting reproductive, maternal, newborn and child health indicators to mortality reductions

Fertility decline is a major driver of NMR/MMR decline

There have been substantial decreases in fertility in Bangladesh, as the total fertility rate decreased from 4.5 live births per woman in 1990, to 3.2 in 2000, and further to 2.0 in 2019.³ Increased fertility rates translate to more high-risk pregnancies via shorter birth intervals, higher birth parity, and elevated birth rates among teen girls and older women.

Our analysis, using Jain’s decomposition method, isolates the impact of fertility declines on maternal and neonatal mortality. Fertility decline alone in Bangladesh was found to explain 44% of the MMR reduction and 47% of the NMR reduction.⁸ These mortality reductions due to only fertility decline correspond to 3,302 fewer maternal deaths in 2017 and 40,804 fewer neonatal deaths in 2019 than would have been expected if fertility levels had stayed constant since 2000. This decomposition approach attributes other improvements to ‘safe motherhood’ initiatives, a term used to collectively refer to improving intervention coverage and services such as antenatal care, in-facility delivery, skilled-birth attendance, and emergency care services. Together, fertility decline and safe motherhood initiatives led to 15,966 maternal lives saved in 2017 and 160,448 neonatal lives saved in 2019 compared to what would have been expected if fertility rate and care coverage levels had remained constant since 2000. This result, illustrated in Figure 18, highlights the impact of fertility decline as a key driver of MMR and NMR decline, in tandem with other health care indicators commonly associated with MMR and NMR decreases.

Lives saved by intervention analysis

An analysis using the Lives Saved Tool was conducted for Bangladesh, using 15 maternal and 21 neonatal interventions, to quantify the impact of increased intervention coverage from 2000 to 2020. The Lives Saved Tool uses estimates of intervention coverage, effectiveness of interventions, and estimates of NMR/MMR to model the contribution of each intervention on mortality reductions.⁹ The interventions included in this analysis were able to collectively explain 31.5% of the MMR reduction and 58.4% of the NMR reduction observed in Bangladesh over this period.

Interventions near the time of delivery were most impactful for saving maternal lives, with C-section, parenteral administration of uterotonics, and magnesium sulfate administration emerging as key interventions, respectively saving 13,219, 11,896, and 4,379 maternal lives from 2000 to 2020.

Neonatal lives were predominantly saved by interventions occurring near the time of delivery and after birth. Key interventions near the time of delivery such as C-section and neonatal resuscitation respectively saved 173,529 and 55,375 lives from 2000 to 2020, while crucial interventions such as case management of sepsis or pneumonia and clean cord care respectively saved 134,833 and 61,642 lives over the same period.

Hierarchical neonatal mortality decomposition

The rapid reduction in neonatal and maternal mortality is driven by interconnected distal, intermediate, and proximate factors. Distal factors reflect “upstream” social, political, and economic influences. Intermediate factors include programs and services that influence people’s ability to access high-quality health care as well as household and individual contextual factors that influence health-seeking behavior. Proximate factors include variables immediately relevant to individuals seeking health care, playing more direct roles in influencing health outcomes.

Using these distinct categorizations, we conducted a multivariable hierarchical decomposition analysis for the period 2000 to 2018. In doing so, we aimed to produce findings that could yield more specific insights than the fertility-based analysis presented earlier, which used the general category of “safe motherhood initiatives,” while still reflecting the influence of broader variables that cannot be captured in an intervention-based Lives Saved Tool analysis. This analysis was only conducted for NMR – and not MMR – due to the relative rarity of maternal mortality as compared to neonatal mortality, resulting in scarce individual-level MMR data.

Key proximate drivers of NMR reduction include the proximal factors of modern contraceptive use and pregnant women attending at least four antenatal care visits, which respectively accounted for 41% and 24% reductions in NMR from 2000 to 2018. The intermediate factor of C-section was also a key driver as it explained 37% of NMR reduction during this period. Further upstream, electricity access explained 29% of the NMR decline from 2000 to 2018. These factors, along with others such as maternal education and increased maternal height, helped explain the NMR reduction in Bangladesh despite the negative factors of household crowding, urban dwelling, and general unavoidable risk of neonatal death.