Last Friday a colleague of mine was talking to me about a recent news piece they were listening to on today fm. The radio station was reporting on the results from recent research conducted by the charity “Save The Children”. (Report is at http://www.savethechildren.org/site/c.8rKLIXMGIpI4E/b.8585863/k.9F31/State_of_the_Worlds_Mothers.htm) The station was reporting that according to the charity “women in Ireland are almost three times as likely to die in pregnancy and childbirth as those in Norway”. An amazing claim. Is it true though?
At first the result could simply be due to the fact that the maternal mortality ratio (MMR, maternal deaths per 100 000 live births) in each country is so low. And as such, the 3 times factor could be nothing more than a statistical blip. So, I looked the numbers up. In recent years (since 1990 and according to the World Health Organization ) the MMR of Ireland varied from 2 per 100,000 to 9 per 100,000. In Norway the range was from 4 per 100,000 to 9 per 100,000. Rather similar. On average no significant difference. In fact in 2005 in Ireland the MMR was 2 per 100,000 and in Norway is was 9 per 100,000. That’s over 4 times more!
So it would seems this is not the statistic they were looking at. There was another statistic, know as, the “Lifetime risk of maternal death”. This is defined as “The probability that a 15-year-old female will die eventually from a maternal cause.” In otherwords, it is the probability a 15-year-old girl will die from giving birth rather than from other causes. A small number in the developed world.
The “Save The Children” reports claims that in Ireland this probability is 1 in 5,500 and in Norway is it 1 in 14,900. This is indeed 2.7 times smaller than Ireland’s probability. Why is this so different (if the MMR is so similar)?
Imagine two countries, identical in every way, except that in one country the fertility rate is twice that of the other. In this country the lifetime risk of maternal death is twice as high. Even though everything is exactly the same. This is because each women is having “two rolls of the dice” so to speak at dying from giving birth. Therefore the fertility rate needs to be looked at.
In Ireland the rate is 2.02 whereas in Norway is 1.86. This is a 8% difference. Therefore if the probability in Ireland was 1.08 times bigger then we can say there is no real meaningful difference. But it is not 1.08, it is 2.70.
Something is not making sense at this point. The MMR and the fertility rate is nearly the same but yet the difference in the probability is huge. What is going on here? The table in the report by “Save the Children” gives the source of the data. That is “UNICEF, The State of the World’s Children 2012, p.124”. (You can download it at http://www.unicef.org/sowc2012/pdfs/SOWC%202012-Main%20Report_EN_13Mar2012.pdf)
The UNICEF report gives the probability for Ireland as 1 in 17,800 and Norway as 1 in 7,600. This is backwards to what was reported! It is Norway that is 2.3 times more dangerous (according to this statistic and World Health Organization). So what went wrong?
When you look at the table you see that Hungry has a probability of 1 in 5,500 (about 6 places away from Ireland’s row on the table) and no country has a probability of 1 in 14,900 but 10 row’s later from Norway’s position is a probability of 13,300. I do wonder if someone or something (software) copied the numbers incorrectly. Maybe because the document is a pdf and is missing some data in the table. Or simple human error when copying and pasting.
Although, we were still expecting the probability to be the same, and it would seem now they are hugely different again but in favour of Ireland. Ultimately I suspect the 17800 number is some kind of error (a statistical blip or a simple human error). It is rather odd too that “Save the Children” used 2012 UNICEF data when 2013 and 2014 are available.
Let us see what these two more recent reports say. In 2013 the lifetime risk in Ireland was 1 in 8,600 and in Norway it was 1 in 8,100. In 2014 the lifetime risk in Ireland was 1 in 7,400 and in Norway it was 1 in 7,900. This now seem more like what we were expecting. The difference between Ireland and Norway is less than the difference in the fertility rate. The fertility rate differ by 8% and these differ by less than 8%. It would seem that rather than Ireland being nearly 3 times more dangerous it is in fact….. exactly the same!
We should not reject the hypothesis that the lifetime risk probabilities are the same for Ireland and Norway. In fact, they seems to be more or less than same, at least in the last two years, with the exception of a (possible) blip in 2012.
I’ll email “Save the Children” later today and see what happens…
UPDATE: Another colleague of mine sent me the 2015 UNICEF data. And it seems this is what was used in the report! I don’t know why the reference didn’t read UNICEF, The State of the World’s Children 2015, p.124. Although it did footnote the face that the reported used more recent data if it was available. And in this 2015 data Ireland had a higher MMR (2.25 times higher) together with the fertility rate explains the probability as far as I am concerned. The numbers are so small and vary so much year to year it doesn’t seem meaningful to compare a country like Ireland with Norway. I think I’ll leave the last words with my colleague who sent on the 2015 report to me…
“The purpose of this report is to compare the developed world with the developing world, it is not meaningful to compare developed countries.”