The First Global Symposium on Health Systems Research officially kicked off this morning is idyllic Montreux, Switzerland.  Throughout the day yesterday, there were a number of interesting satellite sessions  where meeting participants were able to have smaller discussions prior to the launch of the overall conference on more specific topics.  I briefly attended two sessions, one on the measurement of indicators for tracking policy and heath system indicators for maternal, newborn, and child health (MNCH) and another wherein the Rockefeller Foundation launched a new report, entitled Catalyzing Change, which is a report commissioned by the Foundation from the management consulting firm McKinsey (of which I am an alumnus) to look at the question of how much it costs –  specifically the “system costs” –  of scaling up health insurance to the national level in a select group of countries.

During the first session the meeting participants spent a great deal of time discussing the pros and cons of various MNCH indicators and the challenges associated with attempting to deduce the effectiveness of policies aimed at improving these indicators.  In the second session, the authors of the report argued that by their assessment the health system costs of expanding health insurance coverage in a select group of countries are “relatively small”.  What struck me after attending these two, relatively distinct discussions, is just how difficult it is conduct good and meaningful health systems research – the focus of this conference – when the outcomes that most people would agree that health systems should target – reductions in mortality and morbidity, improvements in financial risk protection, and improvements in patient satisfaction – are so imperfectly and so incompletely measured – if they are even measured at all.

As health system researchers we are severely limited in what we are able to monitor so when end up focusing on what we can measure: frequently process indicators such as the number of people with health insurance coverage or the proportion of women who receive four or more antenatal care visits.  But sometimes we forget that these are just indicators, which we think might influence the outcomes we care about but this may not always be the case.  They are means to an end but not the end in itself.

We should only care that people have health insurance coverage if having this coverage improves health, reduces the proportion of households who suffer financial impoverishment as a result of a sick family member, or improves patient satisfaction with the services they are receiving.  Just having a health insurance card is not enough.  We should only care if women receive four or more antenatal care visits if these visits actually are effective at improving the mother during her pregnancy, reduce maternity related morbidity and mortality, and improve neonatal and child health.  Standing in line for a few hours to speak to a low quality health care provider – or worse to find out that the provider is not present – is not particularly useful, and might represent a significant waste of public resources.

It the absence of such data, we will continue having to make the assumption that higher levels of coverage of these indicators are making a difference.  But as we all know, making assumptions can frequently lead to an incorrect or an incomplete view of the picture.  My big hope from this conference will be a frank discussion of how data collection efforts about health system outcomes can be improved and strengthened around the world.

Sometimes small efforts can lead to big improvements in data collection efforts, such as asking about patient satisfaction, collecting information on household incomes and expenditures, or just getting more detailed data on the patients presenting and the services provided (e.g. quality indicators).  The incremental costs associated with adding these additional dimensions to existing data collection efforts may not be enormous but the incremental benefits from such efforts may be large.

Nandini Oomman recently asked in a blog post on the Center for Global Development’s Global Health Policy blog about whether research can make health system strengthening sexier, but I am left wondering if health system research itself will ever be sexy enough for the needed investments in data to be made?

Share on Facebook

8 Responses to “How do you do research on health systems when we don't actually know what they do?”

  1. Karen Grepin says:

    How do you do research on health systems when we don't actually know what they do? #hsr2010

  2. How do you do research on health systems when we don't actually know what they do? Karen Grepin on: Global Health Blog

  3. RT @KarenGrepin: How do you do research on health systems when we don't actually know what they do? #hsr2010

  4. @KarenGrepin asks: How do you do research on health systems when we don't actually know what they do? #hsr2010

  5. RT @NandiniOomman: @KarenGrepin How to do research on health systems when we don't really know what they do? #hsr2010

  6. How do you do research on health systems when we don't actually know what they do?:

  7. […] Health System Research was held this week in Montreux, Switzerland, and has been covered by NYU’s Karen Grepin and CGD’s Nandini […]

  8. […] make HSS sexy; by Karen Grepin, on the long and bumpy road to universal coverage  (see also here ), by Save the Children UK, with a couple of interesting posts, and by Kate Hawkins and others on […]

Leave a Reply

Analytics Plugin created by Web Hosting