15 April 2011

Unhealthful News 105 - Stillbirth and statistics that mean nothing

Several news organizations ran stories about a highly touted Lancet series about stillbirths, defined as loss of a fetus/baby after 28 weeks of gestation.  The reports were an amazing muddle of useless and obviously confusing statistics, with blame to be shared between reporters who write without bothering to think and for the researcher-touts for pushing muddled information into the press.

The problem that is immediately evident is that talking about this category of outcomes as if it were a homogeneous group is totally misguided.  Some of the stillbirths that occur before viability are cases of genetic programming deciding that something is wrong the the fetus (perhaps due to an environmental impact but possibly just inevitable genetic or developmental problems).  Such spontaneous abortions usually occur earlier in pregnancy – a large portion of pregnancies end that way – but there is still a thin tail that occur later.  The report suggests that women with optimal nutrition, healthcare, etc. still experience stillbirth at a rate of just under one half of one percent of pregnancies.  So the aggregate numbers about the problem – the number of stillbirths per day or year or whatever that was the headline of most of the stories – are meaningless because some of these event are basically unavoidable.

The stories report that most of the events occur in less wealthy countries and that of those, half occur in childbirth.  Deaths in childbirth are radically different from other stillbirths.  Presumably almost all of those babies are perfectly healthy, but something goes wrong with the physical act of birthing that kills them.  This is obviously a problem and has some fairly obvious solutions; it has almost nothing in common with babies dying in utero.

The report from ABC News, in trying to make the case that poorer countries are highest risk, reported, "Collectively, Pakistan, India, Nigeria, China and Bangladesh account for half of stillbirths worldwide."  This is a classic case of a statistic that is missing the denominator.  Those countries account for nearly half of all births worldwide (well over 1/3), so it is not all that impressive that they account for half of the stillbirths.  This is why epidemiology sticks to statistics with a denominator – number of events per something.  Without it, such statistics are just ways to create misleading drama.

Most of the reports repeated the statistics that stillbirths claim more lives than HIV/AIDS and malaria.  But this is really an unfair comparison.  Either source of death is bad, obviously, but losses of fetuses are just not the same as a community-destroying epidemic that wipes out a large number of people in their prime.  The reports talk about how devastating and alienating the experience can be for the mothers, which makes sense, suggesting social interventions would be valuable.  But it still does not seem as bad as a young child losing his parent, and possible household income, to AIDS.  Also, I doubt it is actually "disenfranchising", as one report wrote – I assume the grieving mothers are still allowed to vote.

Obviously the many deaths that are in excess of those that might be inevitable – rates vary by more than an order of magnitude – suggest a dramatic lack of basic health (mother's nutrition, infection status, etc.) and prenatal care specifically.  But we knew that.  Mothers' health problems in poor countries are adults' health problems more generally, and they cause huge damage to welfare, productivity, social structure, development, etc.  And problems that cause some stillbirths by hurting fetuses almost certainly cause a lot more cognitive losses in the next generation, which to be blunt will ultimately damage poor countries a lot more than the deaths of some babies.  Maybe it is the bias of someone who survived birth and infancy, but it seems to me the most compelling case for most anti-stillbirth measures (apart from emergency obstetric capacity) is that they make existing people healthier in general and surviving babies smarter.

As a final point, I could not help but notice that the Canadian news report chose to emphasize smoking and obesity as reasons for stillbirths.  That seems to kind of miss the point and the content of the report.  But I think it might be Canadian law now that every public statement about health problem must blame the problem on smoking.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.