Thursday, July 12, 2007

El Sicko

I haven't seen "Sicko" by Michael Moore. I hear that it is powerful enough that it might convince my fellow US citizens to establish a decent health care system. People who follow the issue understand how badly the US health care system functions.

I am interested in one specific point. Cuba vs the USA. Cuba is an astounding case, since it is a very poor but very healthy country. I would agree with many commentators that noting this was a rhetorical error on Moore's part. Most people in the US are barely willing to admit that, say, Canada, is better in some way, and are absolutely appalled by praise of Cuba (even if it is based on standard data sets).

One astounding fact is that the Cuban infant mortality rate is actually lower than the US infant mortality rate ! The latest World Health Organization rounded statistics show Cuba at 5 per thousand and the US at 7 per thousand.

I would like to mention another statistic from another source -- deaths of fetuses which weigh more than 500 grams or deaths of third trimester fetuses*, that is, late miscarriages. This is relevant as the classification of a death as a fetal death rather than a live birth followed by a death reduces the measured infant mortality rate. Different countries use different systems. I wondered if the low Cuban infant mortality rate could be explained by describing deaths which would be considered infant deaths in the USA as fetal deaths.

Data from the U.N. Demographic Yearbook (warning pdf) strongly suggest that this is the case. Cuba has a much higher rate of late fetal death per pregnancy than the USA
in 2002 (latest year available) the USA had 3.3 late fetal deaths per thousand pregnancies and Cuba had 14.0 late fetal deaths per thousand pregnancies.

This difference dwarfs the difference in infant mortality. A 500 gram fetus has a much better chance of becoming a one year old child in the USA than in Cuba.

Given the numbers, I have no doubt that the low Cuban infant mortality rate is partially a statistical artifact. I mean they are still doing very well given their per capita GNP or situation in 1959, but not as well as the standard number suggests.

* watch out, there is a footnote. The number for Cuba is for fetuses which weigh over 500 grams, for other countries, including the USA it must be based on gestational age so the two numbers might not be comparable, but I spent a long time looking for them so here they are.

1 comment:

Michael Sullivan said...

I wonder also about who is doing the measuring of these things in Cuba. It was very difficult to get good data of this kind out of the USSR, and much of what we thought we knew about life there pre-wall-fall turned out to be very wrong, because we were fed lies by the politicians there.

Cuba is probably less closed than Soviet Russia, but it's still remarkably secretive and totalitarian. I doubt we can trust these figures. I've seen plenty of anecdotal evidence to suggest that glowing accounts of health in Cuba could well be very dubious.

Unless these figures come from data collected by outside agencies with full access to primary data, and who *checked* that data carefully, I don't think we can know whether Cuba's mortality rates are really as good as they say they are, or even in the ballpark.