India may have recorded an estimated 565,900 stillbirths in 2023 among pregnancies of 22 weeks’ gestation or longer, according to a new analysis published in The Lancet Regional Health – Southeast Asia.

The estimate is substantially higher than the 347,300 stillbirths estimated when only fetal deaths at 28 weeks or later are counted, the threshold commonly used for international comparisons.

The study, based on the Global Burden of Disease (GBD) 2023 analysis, provides what researchers describe as the first state-level estimates of stillbirths in India using the World Health Organization’s definition of a stillbirth at 22 weeks’ gestation or later.

The findings suggest that limiting surveillance to pregnancies of 28 weeks or more could miss a substantial part of India’s stillbirth burden.

The researchers estimated India’s stillbirth rate at 25.9 per 1,000 births at 22 weeks or later, compared with 16.1 per 1,000 births at 28 weeks or later. In other words, including earlier gestation stillbirths increased the estimated burden by about 1.6 times.

The paper notes that the WHO’s ICD-11 definition considers a stillbirth to be a fetal death at 22 or more completed weeks of gestation. However, the 28-week threshold continues to be used for international comparisons because data on earlier stillbirths are limited.

Uttar Pradesh accounts for more than a third

The burden varied sharply across India’s states.

At the 22-week threshold, the estimated stillbirth rate ranged from 9.3 per 1,000 births in Mizoram to 38.2 in Uttar Pradesh, a more than four-fold difference.

Uttar Pradesh also had the largest estimated number of stillbirths, with 202,500 deaths at 22 weeks or later, accounting for 35.6% of India’s total. Bihar accounted for another 13.8%, while Madhya Pradesh contributed 8.7%.

Five states — Uttar Pradesh, Bihar, Madhya Pradesh, Jharkhand and Chhattisgarh — had stillbirth rates above the national average under both gestational-age definitions.

The researchers said the geographical differences underline the need for interventions tailored to local health-system conditions rather than relying solely on national averages.

At the same time, they cautioned that the estimates for earlier gestation stillbirths should not be interpreted as precise counts because India has relatively little primary data capturing deaths between 22 and 28 weeks.

The study says the estimates at 22 weeks “have relied heavily on statistical modelling and extrapolation from later gestational ages and covariates”.

Official data may miss many stillbirths

The researchers also compared their estimates with India’s Sample Registration System (SRS) data for 2023.

They found that the national late-gestation stillbirth rate reported by SRS was 2.3 times lower than the GBD estimate. The discrepancy was particularly large in Andhra Pradesh, Bihar, Telangana, and Jammu & Kashmir and Ladakh.

By contrast, estimates of neonatal mortality — deaths during the first 28 days of life — were much more closely aligned between the two systems.

The researchers therefore argue that stillbirths remain considerably less visible in India’s health data systems than neonatal deaths.

They note that the problem extends beyond SRS. HMIS data were not used in the GBD modelling because of “known issues of coverage and quality”, while the researchers cite earlier evidence of substantial under-registration of stillbirths.

The study also points to a recent survey across nine Indian states in which only 1% of stillbirths were registered, despite stillbirth registration being mandatory.

Data gaps complicate the picture

The researchers analysed 122 data sources, including six National Family Health Surveys, 29 published studies and vital and sample registration data covering 2,684 location-year observations.

They used statistical modelling to standardise data collected using different definitions of stillbirth and estimate rates for both 22-week and 28-week thresholds.

But the authors stress that the 22-week estimates have wider uncertainty because of the scarcity of direct data.

They write that the inclusion of earlier stillbirths “has also highlighted that very limited data availability and quality continue to be constraints to make precise estimates for many locations, including for India”.

The study also found a modest inverse relationship between state SDG3 scores and stillbirth rates, meaning states with higher health-related SDG performance tended to have lower stillbirth rates. The researchers, however, caution against overinterpreting this association because most states had relatively high SDG3 scores.

Also read: Severe anaemia in pregnancy more than triples stillbirth risk, finds study of 2 lakh Indian women