Stillbirth, defined as a baby born with no signs of life including absence of a heartbeat, no respirations and no movement, is one of the more common adverse pregnancy outcomes.1 Nevertheless, when pregnancy outcomes are considered, especially in low and middle-income countries (LMIC), stillbirth data are often not collected or reported.2,3 To overcome this neglect, the Lancet, first in 2011 and again in 2016,1, 2, 3, 4, 5 with a series of papers, raised awareness of this outcome. In addition, since then, the World Health Organization (WHO), the United Nations (UN), the US National Institutes of Health (NIH) and many other organizations, have emphasized more accurate reporting of stillbirths.6,7 The United Nations and WHO have adopted an Early Newborn Action Plan (ENAP) that includes reporting of stillbirths as an important component of country health statistics.8 As part of the ENAP, national goals for reductions in stillbirth have also been established. Over all, the ENAP proposes that each country reach a goal of no more than 12 stillbirths per 1000 total births by 2030.

The stillbirth rate estimates for high-income countries are often calculated based on data on individual stillbirths collected directly and presented in the vital statistics reports.5 However, reliable stillbirth estimates for many other countries, and especially LMIC, until recently, have rarely been available.9,10 Hug et al, in a landmark 2019 publication in the Lancet, suggest that before the recent release in 2016 of estimates from the UN Inter-agency Group for Child Mortality Estimation, the only global estimates of stillbirths were published by WHO and the Global Burden of Disease Study (GBD).11 As noted, most previous stillbirth estimates for high-income countries were obtained with stillbirths collected as part of prospective registration of all births, and then these data were smoothed with regression techniques. However, Hug et al state that for all other countries, estimates were only obtained from regression models using indirect measures to estimate approximate stillbirth rates.11 Note that actual counts of individual country-level stillbirths were often not included for LMICs. To provide a sense of the quality of the actual stillbirth registration data available, Hug et al. state that among the 195 countries for which stillbirth estimates were generated, 22 countries had no stillbirth data, an additional 38 countries lacked high quality stillbirth data, and in many others the data is likely inaccurate.11 Most countries lacking high-quality stillbirth data are also high burden countries: 32% of those without any data were in sub-Saharan Africa.

Another issue related to understanding LMIC stillbirth rates is that the definitions of stillbirth used by individual countries differs.11, 12, 13, 14, 15 Most importantly, the lower gestational age and the lower birthweight used to define a stillbirth varies substantially. Many LMIC, but not all, use the WHO international comparison definition, which includes a lower gestational age limit of 28 weeks and a birthweight of 1000 grams, but these lower limits are not used consistently across all countries. The US, for example, uses 20 weeks of gestation and many European countries use gestational ages between 20 and 28 weeks and birthweights between 400g and 1000g as their lower limits.5 Since many fetuses are born dead between the gestational ages of 20 and 28 weeks, and at less than 1000g, if 20 weeks rather than 28 weeks and 400 or 500 g rather than 1000 g were used as the lower gestational age and birthweight limits for international estimates, the stillbirth estimates might increase by 20% or more. In any case, using a 28-week cutoff, the UN, WHO, the World Bank, the US Central Intelligence Agency (CIA) and many other organizations now have available on-line, country-specific estimated stillbirth rates by year, dating back to 2000, and in some iterations, to 1990.16

Our goal in this paper is to understand whether stillbirth rates have improved in selected individual LMICs since 2000 and to describe the extent of those improvements. We understand that in the absence of high-quality stillbirth data in many countries for some of the reasons noted above, at best, we will only be able to approximate the extent of the changes in stillbirth rates in individual LMICs over time. Nevertheless, we believe the effort is worthwhile even with these data issues, since we are interested in trends over time and for the most part, the methodologies used to create these estimates have not changed substantially.