All you have to do is walk into a primary health center or health post and eyeball the record-keeping and you know that health statistics in Africa need serious “enhancement.” On September 25, 2015, the Global Partnership for Sustainable Development Data was launched—the partnership is a multi-stakeholder group seeking to harness the power of data and technologies to drive implementation, measure progress, encourage accountability, and achieve the Global Goals. In a recent CGD blog post, Amanda Glassman asks the new partnership for more clarity on how partners’ commitments to the initiative will meet the need and will within low-income countries to produce and use better data. How will these commitments, she asks, work together at the country level where limited capacity requires coordination rather than siloed approaches?

We agree. But we also want to emphasize that data collection is not just about funding and developing data systems in countries—it is about funding and developing health systems more broadly.

Because, after all, who will collect this data? In most African countries, isn’t it (mostly) primary health workers? And how does health service delivery work at the primary level? Doesn’t it typically take place in under-staffed, over-worked rural health centers, in which two to three health workers (tops) oversee the daily clinical care of upwards of a hundred patients on very low and often inconsistently paid out salaries, while simultaneously being responsible for all the record keeping and data collection on behalf of: a) the Ministry of Health; b) parallel vertical programs; and c) assorted clinical and operational research projects, all without a dedicated data manager on site? Health workers not infrequently end up entering or phoning in data on their own time (and sometimes on their own mobile phones using their own airtime). And sometimes they don’t find the time at all.

As an example, Yassi et al (2014) provide insight into the human resources barriers to data management when describing a joint South African-Canadian Occupational Health RCT in South Africa.

[1] They note that the research team “continues to seek the most convenient way for busy occupational health practitioners to systematically collect the data needed for operational purposes, as well as for our RCT, without adding burden.” Our experience suggests that this dilemma is shared not only by many operational research projects, but also by national information systems more broadly. Yet too often the human resource-related problems encountered during data collection get swept under the rug during “data cleaning.”

Likewise, Lesley-Anne Long of mPowering Frontline Health Workers notes that frontline health workers are frequently burdened with “unfeasible workloads [and] under-valued by their communities, other health workers and Ministries of Health. Add to these challenges the fragmented activities of the global health development players – where collaboration is vital but consistently fails to move from rhetoric to reality, and where funding so often ring-fences aspiration for impact.”[2] Long argues that failure to support frontline health workers, in particular community health workers, is one contributing factor as to why so many projects have failed to go to scale.  But the same argument can be applied to countries’ struggles with information systems in general, especially those relying on data collected at the primary or community level.

Improved data collection and management, then, is a human issue as much as it is a data systems issue. It is a labor issue. It is a workplace quality issue. It is a service delivery issue. And if new systems of collecting statistics put an unfair time and resource burden on health facility staff at the lowest levels of the health system, then the only thing that has been enhanced is the implementing/research/financing partner’s program portfolio.

We argue that at the forefront of efforts to improve data systems in African countries, the relevant questions are:

  • How can we create data systems that are realistic to the environment and human resources restrictions that exist in the real world?
  • How can we make data collection systems health worker-friendly and patient-friendly?
  • How can we achieve maximum data collection with minimum human resources burden?
  • How can we ensure data collection does not result in less time spent on patient care?

Let’s fund efforts to address these questions along with new data systems. Because without people, new data systems can’t possibly succeed.

 

[1] Yassi A, O’Hara LM, Engelbrecht MC, et al. Considerations for preparing a randomized population health intervention trial: lessons from a South African–Canadian partnership to improve the health of health workers. Global Health Action. 2014; 7: 10.

[2] http://1millionhealthworkers.org/2015/05/18/data-for-decision-making-series-the-importance-of-chw-data-collection-by-lesley-anne-long/