Measuring quality of maternal care from women’s perspectives: development and validation of a participatory-weighted composite index using nationally representative survey data from Burkina Faso
This study demonstrates that while women's perspectives on maternal care quality can be operationalized using existing survey data from Burkina Faso, current national surveys capture only a quarter of the dimensions women value, leading to significant discrepancies in identifying who receives poor-quality care compared to conventional researcher-defined indexes.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
For decades, the global effort to save mothers' lives has focused on getting more women into health facilities. The logic was straightforward: if a woman gives birth in a hospital rather than at home, she is safer. In many parts of the world, more women are indeed walking through clinic doors than ever before. Yet, despite this surge in visits, the number of mothers dying remains stubbornly high in places like sub-Saharan Africa. This gap between visiting a clinic and surviving childbirth points to a different problem. It is not just about access; it is about what happens once a woman is inside. The quality of the care she receives—whether the staff are skilled, whether they treat her with dignity, and whether the facility is clean—has become the critical missing piece. For too long, the tools used to measure this quality have been designed by researchers and policymakers, not by the women themselves. They count the medical tasks performed, like checking blood pressure or administering vaccines, but they often miss the human elements that women say define good care.
A team of researchers in Burkina Faso set out to fix this disconnect. They wanted to know if the standard national surveys used to track health could actually capture what women value, and if not, how to build a better measure using the data that already exists. Their work began not in a lab, but in conversation. They sat down with thirty-two women of childbearing age and asked them to map out what makes maternal care good. The women did not just list tasks; they described a complex web of factors, from the distance to the clinic and the skill of the midwife, to the respect shown during labor and the availability of clean water. The researchers then asked these women to weigh the importance of each factor, creating a hierarchy of what truly matters to them. This process, known as fuzzy cognitive mapping, allowed the women to define quality on their own terms, rather than accepting a list of medical checkboxes.
With this new understanding in hand, the team turned to the 2021 Demographic and Health Survey for Burkina Faso, a massive dataset collected from thousands of households. They treated this survey like a puzzle, trying to fit the women's definitions of quality into the questions the survey actually asked. The result was a stark revelation. The national survey, which is considered the gold standard for health data in low-income countries, captured only about one-quarter of the dimensions the women identified as essential. While the survey was excellent at recording medical procedures—such as whether a blood sample was taken or a newborn's heartbeat was listened to—it largely missed the human experience. Questions about how a woman was treated, whether she felt respected, or if the facility was clean were either absent or only partially represented. The survey also failed to capture many structural issues the women deemed vital, such as the adequacy of the building's infrastructure or the availability of necessary medical equipment.
Undeterred by these gaps, the researchers built a new score, a composite index, using only the parts of the survey that did align with the women's views. They took the twenty-nine survey questions that matched the women's categories and combined them into a single number. Crucially, they did not treat every question as equally important. Instead, they weighted the questions based on how much the women in the initial study said those factors influenced the quality of care. A category like "medical care before, during, and after childbirth" received a heavy weight because the women saw it as central to their safety, while other factors received less weight. This created a "women's perspective index" that reflected the priorities of the mothers themselves, rather than the priorities of the statisticians.
When the researchers compared this new index to the conventional way of measuring quality, they found a surprising result. At the national level, the average scores were nearly identical. Both methods suggested that, on average, women received about three-quarters of the care components that could be measured. However, the story changed when they looked at individual women. The two methods did not agree on who was receiving good care and who was not. The conventional index and the women's perspective index ranked women differently, with only a moderate level of agreement. This meant that a woman might be classified as receiving poor care by the standard method but good care by the women's method, or vice versa. The difference was most pronounced for the most vulnerable women. Those with no education and those living in the poorest households showed the largest gaps between the two scores. For these women, the choice of measurement tool changed the picture of the care they received, suggesting that standard surveys might be missing the specific struggles of the most marginalized.
The study also tested whether the results held up if the researchers changed how they combined the data. They tried weighting the questions equally, as is often done, and used a statistical method to find patterns in the data. The average scores remained stable, hovering between 0.71 and 0.80, which suggests the overall level of quality is robust regardless of the math used. However, the ranking of individual women still shifted, indicating that the way questions are weighted does matter for identifying who is being left behind. The participatory approach, which used the women's own weights, captured a slightly different reality than the standard approaches, one that was more sensitive to the specific dimensions of care that women value.
The findings suggest that while current national surveys are useful, they are incomplete. They are like a camera that has been set to focus only on the technical aspects of a scene, leaving the human faces blurry. The researchers concluded that to truly improve maternal care, we cannot just tweak how we combine the existing numbers. We must change what we ask in the first place. The surveys need to include questions about the reception women receive, the cleanliness of the facilities, and the respect shown by staff. Until the data collection tools are revised to include these human elements, the picture of maternal care quality will remain distorted, potentially hiding the very inequities that need to be addressed. The path forward is clear: to measure quality as it is experienced, we must listen to the women who live it, and then build our surveys to reflect their voices.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.