Meeting the target but missing the point: The politics of childbirth data

Pressure to meet targets to improve health outcomes for women and their babies can have unintended negative consequences.

Rose, one of two nurse-midwives on duty in a hospital labour ward in Southern Tanzania, slumps down into a chair and lets out a sigh. The desk she sits at is covered with loose sheets of paper and register books. This is the documentation she is expected to complete about the birthing women under her care. Some of the documents are meant to guide decision-making about the care women receive – plotting a woman’s contractions on a chart called the partograph, for example, helps track progress of labour. But Rose fills out most of the documentation after a woman has already given birth, “once the mother and her baby have crossed safely”, as she describes it.

Understandably, this paperwork is not her favourite part of the job. It is time-consuming and she considers it a tedious task. When the workload is high, it can be near-impossible to complete all the reporting requirements. Working in a labour ward is predictably unpredictable – some shifts pass without a single new patient arriving, while other days Rose finds herself managing five women in active labour at once. When someone needs urgent care, documentation is the least of her priorities, she says: “When an emergency comes, you leave those things because you can’t deal with paperwork when there is a patient who needs help.”

However, her managers will not accept incomplete documentation. Data collected by care providers are reported to the district health office, before being sent to regional and national levels. The health of women and newborns is high on the political agenda in Tanzania, and data about their survival are closely monitored. Top-level Tanzanian government officials and health managers recently expressed a renewed commitment to ensure “zero maternal deaths.” As a result, Rose points out, data about mothers and their babies have become “political stuff”.

This is by no means unique to Tanzania. In the last few decades, the field of global health has rallied around achieving targets like the Sustainable Development Goals (SDGs). Particularly in donor-dependent countries, this drove an increase in what has been called “target culture”. Linking to commercial notions of cost-effectiveness, policymakers and donors want to know how many lives are being saved by specific policies and interventions, and at what cost. Tracking progress towards global goals has increasingly become a priority, and this requires data.

Targets come from good intentions and can have real benefits. Agreeing on a common goal can create momentum and can help hold leaders accountable for advances towards safer childbirth. Yet, some targets are unrealistic in contexts where women’s health is shaped by poverty and social disadvantage, and where health facilities face shortages of staff, supplies and essential medicines. Improving maternal and newborn health is complex – no single ‘silver bullet’ will fix it all. And, as is clear from the experiences of Rose and her colleagues, target culture may also create an environment where health care workers are afraid to be blamed if they report on negative outcomes.

So afraid, in fact, that they might not always report exactly what happened. A recent review of studies conducted in low- and lower-middle income countries found that manipulation of maternal and neonatal health data among health care workers was not uncommon. In order to meet specific targets or protect themselves from blame, health care workers might alter or fabricate some of the data they are expected to report. Although this was a sensitive topic among Rose and her colleagues, they described various examples of what they referred to as ‘cooking’ data. During shifts where it is not possible to check the heartrates of unborn babies as often as required for all women on the ward, one nurse-midwife explained, danger signs might be missed. But to avoid being blamed for any problems that might arise, the nurse-midwife can make it look like she listened to the foetal heartrate more frequently: “You come to listen to the mother, oh no! She was in distress [the baby was not well] since long ago. So you come here, you write data that doesn’t exist, now you are cooking up partographs.” To prevent being scrutinised as a result of incomplete data, or to protect themselves from blame after a poor outcome, health care workers can thus fabricate or alter the data.

“Target culture” can have unintended consequences on data quality. When expectations are not realistic and health care workers fear being blamed or losing their job, they can feel ‘cooking’ data is the only option available to them, resulting in data that do not reflect what truly transpired. We must consider how the workload of routine data collection can be made feasible and relevant to the health care workers doing it, so that the usefulness of the data to help improve quality of patient care is clear to them. Otherwise, we might end up meeting the target, but missing the point.

—

Quotes and observations were drawn from qualitative research in Southern Tanzania conducted by PhD candidate Jil Molenaar in 2023. To protect confidentiality, names have been changed and places are not specified.