Low birth weight – UNICEF DATA

Data Notes

1. Definitions

Two key indicators on birth weight are described in this section, (i) the prevalence of low birth weight (LBW) and (ii) the prevalence of newborns without birth weight data in national data sources.

Birth weight is the first weight of the newborn obtained after birth. For live births, birth weight should preferably be measured during the first hour of life, before significant postnatal weight loss occurs. Low birth weight is defined as less than 2,500 grams (up to and including 2,499 grams).

The prevalence of newborns without birth weight data reflects newborns who were not weighed and those who were weighed but whose birth weight was not captured by major data sources. data. Estimates of newborns without birth weight data from administrative systems (e.g., health management information systems) include unweighed births and weighed births not recorded in the system. Estimates from household surveys include births for which weight was not available in an official document (e.g. health card) or could not be recalled by the respondent at the time of the interview.

(1) Typically one calendar year for administrative sources and two to five years for household surveys.

(2) A valid birth weight is biologically plausible, defined as ≥250 g and ≤5,500 g.

2. Low birth weight data sources and methods

Nationally representative estimates of the prevalence of low birth weight can be derived from a range of sources, broadly defined as administrative data or nationally representative household surveys. National administrative data comes from national systems, including Civil Registration and Vital Statistics (CRVS), National Health Management Information Systems (HMIS) and birth registers. Nationally representative household surveys include Demographic and Health Surveys (DHS), Multiple Indicator Cluster Surveys (MICS), and other national surveys for which microdata are available for reanalysis.

The development in 2023 of the national, regional and global database with annual LBW estimates from 2000 to 2020 built on previous rounds of LBW estimation (2004 and 2019) and included the following steps:

a) Collect birth weight data from administrative and survey sources, and filter these data with inclusion criteria to ensure the quality of the input data.
b) Adjust survey input data to account for underestimation of low birth weight due to missing values ​​and crowding.
c) Produce annual LBW estimates using statistical modeling
d) Organize national consultations to validate input data and modeled LBW estimates, and refine the data/estimates as necessary to finalize the estimates.

The LBW database was developed through a collaborative effort between UNICEF, WHO and the London School of Hygiene and Tropical Medicine (LSHTM). The three partner organizations formed a steering group and working groups for the estimate, and were guided by the Expert Advisory Group (ECG) and the Guidelines for Accurate and Transparent Reporting of Health Estimates ( GATHER) throughout the estimation process. The Expert Advisory Group is an advisory group composed of global experts in preterm birth and birth weight measurement, including obstetricians, neonatologists, statisticians, preterm birth researchers, and program staff. The estimation of LBW, which was carried out alongside the estimation of preterm births for the first time in 2023, received ethical approval from the LSHTM. A basic description of the steps for estimating the LBW for 2023 is available in Annex 2 of the next UNICEF-WHO report on low birth weight estimates; and the complete and detailed methods are available in the LBW estimation protocol and results publications (1,2).

3. Low birth weight newborns in national data sources

3.1 National data sources

Data on low birth weight newborns come from the same sources used to estimate low birth weight, with some important caveats.

The low birth weight estimate used all collected data, which included multiple sources and multiple years of data for many countries. Given the problems of comparability of estimates of the percentage of low birth weight newborns in the data source for different levels of administrative data quality, only administrative data sources in the highest quality category, such as described in Annex 2 of the future UNICEF-WHO report on low birth weight estimates and surveys were used for this database. Since global and regional estimates were expected to cover a recent period, only the most recent source for each country, newer than 2011, was used. Globally, 12 countries had no recent data (i.e. the latest data was from 2011 or earlier) and 43 countries had no administrative data source in the highest quality category or survey.

3.2 Estimated national prevalence of low birth weight newborns in the data source

For countries with recent administrative data in the highest quality category, as described in Annex 2 of the future UNICEF-WHO report on low birth weight estimatesthe prevalence of low birth weight newborns in the data source was calculated by summing the number of live births reported as weighing <2500 g and ≥2500 g in a calendar year in the source and then by dividing the sum by the number of annual live births in the data source. the 2022 revision of the United Nations Population Division's World Population Prospects (UNPD-WPP) for the same year.

Estimates from household surveys represent the percentage of live births occurring during the survey reference period (e.g., in the last 2 years, in the last 5 years, etc.) among women of childbearing age as defined by the survey (generally women aged 15). –49 years), with no birth weight in the dataset. It should be noted that for some surveys, all births of each sampled woman are included in the birth weight data collection and for other surveys, only the last live birth is included in the data collection. For household surveys, the prevalence of low birth weight newborns in the data source was calculated by dividing the number of live births with a valid birth weight (defined as ≥250 g and ≤5,500 g ) during the survey reference period by the total number of live births in the survey reference period. the reference period of the survey. Prevalence estimates from surveys were weighted using the sample weight of women and the calculation of confidence intervals took into account a complex sampling design (i.e. clusters and strata).

3.3 Regional and global estimates

Regional and global population-weighted estimates were weighted using the 2021 annual number of live births from the 2022 revision of the UNPD-WPP. Population-weighted averages for a given region were generated by (a) multiplying the prevalence estimate for each country with data available during the required period by the number of births in 2021 in that country; (b) add up all country-specific products; and (c) dividing the sum of the products by the total number of births in all countries with data in 2021. Population coverage, or the share of the population for which an estimate is available, was calculated by dividing the number of births in 2021. in countries with data on the total number of births in each respective region in 2021. The standard used for minimum population coverage (i.e. minimum population coverage required to display the population-weighted regional estimate) is 50 percent.

References for data notes

1. Krasevec J, Blencowe H, Coffey C et al.Study protocol for UNICEF and WHO estimates of global, regional and national prevalence of low birth weight for 2000 to 2020. Open doors, resolution 2022, 6:80 a.m. (https://doi.org/10.12688/gatesopenres.13666.1)
2. UNICEF and WHO (2022). Technical note for country consultation on estimates of low birth weight and premature births.