Giving birth in a health facility, where trained staff can act if something goes wrong, has become the norm in Ghana. But the women with the fewest resources are still the least likely to reach one.
An analysis of nine national household surveys found that the share of births taking place in a hospital, health centre, clinic or maternity home rose from 42.3% in 1993 to 85.4% in 2022. That is roughly double, with most of the gain coming after 2006.
Even so, about one in seven births still happens outside a facility, and the gaps follow familiar lines. In 2022, 97.1% of births in the richest fifth of households took place in a facility, compared with 71.1% in the poorest fifth. The rate was 93.7% in urban areas against 77.9% in rural areas, and 98.0% for women with higher education against 72.3% for women with no schooling.
Where a woman lives also matters. In the Upper East Region, 97.7% of births were in a facility. In the Northern Region the figure was 69.0%, so about three in ten births there happened elsewhere.
There has been real progress. In 1993, only 23.5% of births in the poorest households were in a facility, against 70.4% in the richest. Among rural women the rate has climbed from 28.0% in 1993.
The researchers note that the rise coincided with free maternal care from 2008 and the growth of national health insurance, but the data cannot show that these policies caused it. They call for more health workers, health posts, better transport and stronger financial protection for poorer, rural and less educated women, especially in northern Ghana.
Research Scope
This article is based on a 2026 peer-reviewed journal article by researchers from the University of Cape Coast, the University for Development Studies, the University of Ghana and All Nations University, held in UGSpace. It uses WHO estimates drawn from nine nationally representative Ghana Demographic and Health Surveys and Multiple Indicator Cluster Surveys from 1993 to 2022, with sample sizes of between 3,896 and 7,845 women per round (7,845 in 2022). Because the surveys are nationally representative, the trend can be read as national, but the study is a research signal, not a full national survey or official statistic of its own. It describes patterns rather than causes, and estimates for smaller regions are less precise.