Highlights

Selected contributions to the assessment of key indicators in family planning, abortion, and fertility:
Indicator Selected methodological contributions
and usage in global monitoring
What did we find out?
Some examples
Family planning FPET (Alkema et al. (2013), Cahill et al. (2018)):
  • The Family Planning Estimation Tool (FPET) is a Bayesian hierarchical time series model to estimate and project contraceptive use and unmet need.
  • FPET is used for national level monitoring by the UN Population Division, for annual reporting for the FP2020 initiative, and, going forward, the FP2030 initiative.
From UN 2021 updates:
  • Globally among married women, the number of users of modern contraceptive methods increased from 420 (95% CI 405-435) to 712 (678-746) million from 1990 to 2020.
  • The number of married women with an unmet need for modern contraceptives changed from 200 (189-211) to 216 (197-236) million in the same period.
  • Unmet need is high in sub-Saraharan Africa at 25.7% (24.2-27.3) in 2020.
Abortion and
unintended
pregnancies
Bayesian demographic accounting model (Bearak et al. (2020a), Bearak et al. (2020b)):
  • Accounting model jointly estimates incidence of unintended pregnancy and abortion.
  • Model is used by the Guttmacher Institute and WHO, for example in Adding It Up.
From Bearak et al. (2020b):
  • In the period 2015–19, there were 121 (80% CI 113–136) million unintended pregnancies annually, with 61% (58–63) of unintended pregnancies ending in abortion, totalling 73 (67–82) million abortions annually.
Sex ratio at birth Bayesian hierarchical time series mixture model (Chao et al. (2021b)):
  • Model to estimate the sex ratio at birth and missing female births, and construct scenario-based projections.
  • Methodology is under consideration for use in the UN World Population Prospects.
From Chao et al. (2021a):
  • Up to 2020, 50 (95% CI 41-60) million female births are missing globally due to sex selective abortion.
  • If all countries at risk of SRB inflation experience a sex ratio transition, we project an additional 22 (12-40) million missing female births between 2020 and 2100, with a sizeable contribution of sub-Saharan Africa.
Total fertility rate BayesTFR (Alkema et al. (2011), Raftery et al. (2014)): From the World Population Prospects 2019 highlights:
  • The total fertility rate is projected to be 1.94 (1.68-2.26) globally in 2095-2100.
  • The largest reductions are projected in sub-Saharan Africa where the projection suggests that fertility will fall from around 4.6 live births per woman in 2019 to 3.1 in 2050 and further to 2.1 in 2100.
Selected contributions to the assessment of key indicators related to stillbirths, maternal mortality, and mortality among newborns, children, and youth:
Indicator Selected methodological contributions
and usage in global monitoring
What did we find out?
Some examples
Maternal mortality BMat (Alkema et al. (2017)) and BMis (Peterson et al. (2019)):
  • BMat is a Bayesian hierarchical temporal regression model to estimate the maternal mortality ratio.
  • BMis is a Bayesian hierarchical bivariate random walk model to estimate misclassification of CRVS reporting of maternal deaths.
  • BMat and BMis are used in UN MMEIG reporting.
From UN MMEIG (2019):
  • About 295,000 women died during and following pregnancy and childbirth in 2017.
  • The vast majority of maternal deaths (94%) occurred in low-resource settings.
Stillbirths BHTSRM ((Wang et al. (2020) , Hug et al. (2021)):
  • We developed a Bayesian hierarchical temporal sparse regression model (BHTSRM) to estimate stillbirths.
  • The BHTSRM is used by the UN IGME.
From UN IGME (2020b):
  • One stillbirth occurs every 16 seconds, which means that about 2 million babies are stillborn every year.
Mortality up to age 24 B3 (Alkema and New (2014)) and other Bayesian hierarchical time series models: From UN IGME (2020a):
  • Under-five mortality rates have declined by almost 60% between 1990 and 2019.
  • Still, more than 5 million children died before reaching age 5 in 2019, and nearly half of those deaths were neonatal deaths.