Key lessons from the CODES project: A conversation with Prof Stefan Peterson

Key lessons from the CODES project: A conversation with Prof Stefan Peterson

At the Center we take pride in our work and collaborations with experts therefore, we bring you outcomes of our interfaces. In this issue we had a one-on-one conversation with Stefan Swartling Peterson a Professor of Global Transformations for Health at Karolinska Institutet and a visiting Professor at Makerere University School of Public Health. He shares his personal life and lessons from the CODES project one of the numerous collaborations we have had with him. A big appreciation to Stefan for taking off time to share a piece of his work with us. Enjoy on and happy new month our dear readers and followers.

Kindly, break down for our readers who is Stefan, and does he have a routine? 

Stefan is a product of Makerere, and her collaboration with Sweden – I came to Uganda to work for WHO placed in the MOH in 1998, and again back to MUSPH in 2004-7.  After 4 years+ as UNICEFs global chief of health I am now Professor of Global Transformations for Health at Karolinska Institutet, and a visiting Professor at MUSPH.

My routine is to start my days with the news – papers, radio, pods, social media you name it I consume it. Unfortunately, of late I hear of the terrible toll COVID is taking on Uganda, and hearing of the losses you all incur while the world has the solution makes me very mad.

In your world what defines happiness?

Happiness is about making my family safe and my kids developing optimally, but then also to make a difference for others, to make the world a slightly better place- now and for future generations.

In simple terms, how would you describe data-driven decision making?

In the words of my mentor Hans Rosling, “you need a clean windscreen when you’re driving”.  Data is about knowing the land, to be able to navigate. Yet so many of the important decisions we make in health care, education, and so many other sectors are taken with a dirty windscreen, when we turn the steering wheel on a feeling, rather than seeing where we are going.  Having local, up-to-date data available, analyzed at your level of decision making, lets you be a better manager, and take data-informed decisions. But I realize there is more than data affecting e.g., a District Director of Health Services when she makes her managerial decisions. But let her at least have the data, analyzed in a way that can be understood, in front of her and her team.

In CODES we actually analysed local data, and demonstrated where the “bottlenecks” in implementation were – on the “supply side” – e.g. clinics not open, out of drugs or providing insufficient quality care- or on the “demand side” – people not seeking care or not adhering to advise. We then tested whether this made a difference to coverage of key child health services like immunization coverage, randomizing 16 districts to intervention or control. And it did- intervention districts had way higher coverages than comparison districts!  This may well be the first randomized controlled trial of a district health management intervention – at least I could not find another one.

Implementation science encourages involvement of end users, from your experience how can children be involved in the model recommended by your work?

In CODES we also involved communities, and had discussions with community groups around health services, care seeking and how to improve the health status. Also based on local data, presented in one page reports with understandable figures. These community groups then also met their local health workers, and they made their own, local plans to improve both care seeking and service delivery.  This was highly encouraging.

Now moving from the general community – where we adults tend to dominate – to children, we need to approach children by themselves, in age-appropriate groups.  And depending on topic girls and boys separately.  This is not something we tested in CODES, but something we do via UNICEF in Sweden. Interestingly children and young people in Sweden ask us though: “Why should I share my concerns and experiences with you? What will you do with the information I give you?”.  These are legitimate questions, not just for children, but every time we contact a community, I think. In Sweden we are also finally introducing the Ugandan innovation “U Report” – to ask children via their phones their issues and priorities.

 What are some of the model countries with data driven decision making and what do countries like Uganda learn from them?

Unfortunately, there are very few if any examples of data driven decision making. However, I believe we need to push for it – unfortunately, things like digital health information systems let us down when we wanted to start using them to monitor health service utilization during the pandemic. Those systems were still too slow and could not provide timely data. So, we have work to do.

 Would the approach presented in your work be effective during the pandemic and how?

I firmly believe in the potential of better surveillance systems to guide local decision making – if you think back to Uganda’s first lockdown, when there were just a few cases in Malaba and Entebbe airport – why were schools closed all over Uganda then? And communities locked down and care seeking prevented?  A more timely, local decision making will be necessary.

 Any other learning points in your work that we have not captured in this engagement that you would like to share?

Looking at improving health services more broadly, I do believe firmly in implementation science, and putting policymakers, implementers, users, and local researchers together to define issues and knowledge gaps. For too long we have defined research questions as researchers, and only then gone to policy makers and implementers and told them to “get research into policy and practice”. When in fact we need to involve all these stakeholders from the start, to define the question. We tried this successfully from UNICEF in Pakistan in areas with low immunization coverage, finding that some issues were more managerial, while others would benefit from a 2–3-month rapid research piece. “Embedding research in programming” should become the new normal I believe. But this will also require line ministries, implementers, and universities to think and behave differently, for instance valuing not just my nice paper in a “high impact” journal, but also how I affect policy and practice locally.  To that effect I am happy to report that a spinoff from CODES has now found its way into Uganda’s HMIS, analyzing, and presenting “bottlenecks” to district managers.