Births in a facility, and how far that facility is

Almost every Indian birth is now recorded. Rather fewer happen somewhere equipped to handle it going wrong.

Institutional delivery rates have risen sharply over two decades, and that rise is the single biggest reason maternal deaths have fallen. But the national average hides districts where the nearest functioning facility is a two-hour journey on a road that floods.

The useful question is not what share of births are institutional, but what share of women live within reach of a facility that is staffed and open when labour starts. That means putting delivery data next to facility location and staffing, which are published separately and rarely read together.

What we measure

These are the indicators this topic is built from. Each names its source and the year that source refers to, because an official figure is not the same thing as a current one.

Indicator	Unit	Source and year
Institutional deliveries	% of births	HMIS / NFHS-5Data year: HMIS 2026 monthly; NFHS-5 2019-21
Infant mortality rate	per 1,000 live births	SRSData year: 2021 (latest bulletin)
Health workers	per 10,000 population	Rural Health StatisticsData year: 2023-24
Facilities with a functioning labour room	count	Rural Health StatisticsData year: 2023-24

What we publish

A district map joining delivery rates to facility density, with the gap between them as the headline measure.

The brief above is written and the sources are identified. What remains is the data build. Until it publishes, no figure for this topic appears anywhere on DataSpeaks — we would rather show nothing than show a placeholder.

Where these numbers come from

How we handle gaps

Where a source has not reported for a recent year, the gap is shown as a gap. We do not interpolate to make a series look continuous or a map look complete. Where two agencies publish different figures for the same thing, both appear and we say which is which.

Reuse

Charts from this topic may be reused with credit to DataSpeaks and to the original source named above. The underlying data remains under its own licence — most Government of India datasets permit reuse with attribution, some survey microdata does not.