Here, Child Health Safety tackles the fascinating topic of measles incidence and mortality. CHS refers to “grossly false claims by the US Centers for Disease Control ['CDC'] – vastly exaggerating the threat measles as a disease poses” and accuses them in the title of lying.
CHS quotes (accurately, surprisingly enough) the CDC as stating that:
Worldwide, there are estimated to be 20 million cases and 164,000 deaths each year.
He then cites a figure from the World Health Organization that there were 226,722 reported cases of measles worldwide.
What happened next?
Did CHS mention the time difference between the two sources referred to? No. (The CDC page states that it was last reviewed in 2009 and the WHO figures cited are from 2012.)
Did CHS tell readers that there might be a difference between estimated actual cases and reported cases of measles that might at least partly explain this huge discrepancy? No. The possibility of under-reporting is not mentioned. In spite of the fact that there are gaps in the table linked to (not many in 2012, but including France, which reported 15,000 cases the previous year), and some columns where a zero might raise some eyebrows (for example, the Democratic People’s Republic of Korea had 3,550 cases in 2007 and zero in every other year, except where there were gaps).
Did CHS cite a WHO figure for worldwide deaths from measles? No. Interestingly enough, there is a figure for worldwide deaths from measles in 2011 here. That figure is 158,000, which isn’t that far removed from the 164,000 cited on the CDC page which CHS claims “vastly exaggerates” the dangers of measles and makes “grossly false” claims. How odd. CHS was very keen to compare the CDC’s estimate of worldwide cases with the number of cases reported to WHO, but neglects to compare the CDC estimate of deaths with figures from the WHO. Very strange.
Instead of comparing the CDC figures for deaths from measles with the WHO figures, as he did for cases, he compares the CDC figures for worldwide deaths with a hypothetical number of deaths based on the WHO figures for worldwide reported cases and the mortality rate in England. I don’t think I need to point out that England is not the world, or that there might be reasons why England might see a lower mortality rate than developing countries (but I will, apart from anything else: Measles case-fatality rates have declined in association with economic development and associated decreased crowding, older age at infection, improved nutrition, and treatment for secondary pneumonia).
Brilliant. Take the estimated cases and deaths figures from the CDC. Take the reported cases figure from the WHO. Ignore the deaths figure from the WHO. Pretend that the mortality rate in England will reflect the mortality rate elsewhere. Then simply call the CDC liars, accuse them of making grossly false claims, and assert that they have vastly exaggerated the threat of measles.
There were one or two other slightly odd things in the CHS post.
One was the mortality rate figure that ranged from 1 in 25,000 to 1 in 5,000 (which doesn’t match the DH figures I’ve seen or the Miller paper that CHS cites – I presume it should be 1 in 2,500 and CHS has copied and pasted a figure from the DH website without spotting the error).
The other was the figure of “162,000 more deaths”. CHS seems to think there should have been 2,000 deaths rather than 164,000 but I can’t see how to get to 2,000 from the figures used in the post. Using the WHO reported cases and the English mortality rate I get 45-90 deaths and using the CDC estimated cases and the English mortality rate I get 4,000-8,000. 2,000 deaths from 20,000,000 cases would be 1 in 10,000, which is not a figure referred to by CHS. I’m guessing what CHS has done is to divide the number of worldwide deaths estimated (according to the CDC) by the number of worldwide cases the CDC reported and multiplied it by the number of reported cases from the WHO table. CHS seems to have taken various figures that cannot be meaningfully combined and stitched them together to weave a misleading narrative based on inappropriate calculations that are not made clear to the reader.
And that’s numberwang!