Last update: January 2023

Around 1.9 million stillbirths – babies born with no sign of life at 28 weeks of pregnancy or later – occurred worldwide in 2021. Many of these might have been prevented with proper care. According to the latest data, the global stillbirth rate last year was 13.9 stillbirths per 1,000 total births. This equates to 1 in 72 total births resulting in a stillborn baby, or one every 17 seconds. Still, this number may be an underestimate, as stillbirths are often underreported.


Stillbirth data

Notes on the Data

Definitions of indicators

Stillbirth: A stillbirth is a baby born with no signs of life after 28 weeks of completed gestation.

Stillbirth rate is defined as the number of babies born with no sign of life at 28 weeks or more of gestation, per 1,000 total births.

Data sources and methodology

In the interest of identifying global and national trends over time and making comparisons across countries, UN IGME and its CSEG (Core Stillbirth Estimation Group) use a model-based procedure to obtain nationally representative stillbirth rate (SBR) estimate for all UN countries. Input stillbirth data are obtained from four main sources: administrative systems (e.g. vital registration systems, birth or death registries), health management information systems, household surveys, and population studies. Given that stillbirth data is not readily recorded or available in all countries, it is necessary to include factors related to stillbirth as covariates (e.g. socio-economic factors, demographic and biomedical factors, perinatal outcome markers, and indicators of access to health care) in the stillbirth estimation model. Definitions used for stillbirth reporting can vary, and therefore the UN IGME uses a standard “28 weeks or more of completed gestation” for stillbirth definitions. Data recorded using common alternative definitions (e.g. definitions using birthweight, or 22 or 24 weeks of completed gestation) are adjusted to a 28-week definition prior to model fitting. Increased variation in estimated due to definitional adjustments is accounted for in the estimation model. To ensure high quality, representative data are used data quality is assessed prior to model fitting, and data not meeting pre-defined inclusion criteria are withheld from the model. To reconcile differences across data sources and better account for the systematic biases associated with the various types of data inputs, members of the Core Stillbirth Estimation Group (CSEG) of UN IGME have developed a new approach to estimate trends. A hierarchical Bayesian model with spline smooth is used to produce a stillbirth rate estimate for each country-year. A more detailed explanation is available in the explanatory notes (available in Arabic, English, French, Spanish and Russian).

More details on the data and methods used in deriving estimates are available in CME Info