Measles & SmallpoxThis is part III of Measles week. In Part II (“Emerging disease”) I talked about the origin of measles; in Part I (“Introduction”), I posed the question of why measles case-fatality rates dropped so dramatically over the first half of the 20th century (example chart of death rates here). Today I’m going to quickly cover some of the possible explanations that are probably not right; or that, at the least, don’t explain the whole thing.

Those inadequate explanations are:

  • Poor surveillance. Did the medical establishment gradually, say, lose interest and stop recording measles deaths over the decades?
  • Sanitation. There were huge advances in sanitation, especially in water treatment, in that period. Was that why measles deaths declined?
  • Viral changes. Did the virus mutate and become less lethal in that period?
  • Antiserum. In that time-frame, a new treatment for measles became available. Was antiserum treatment responsible for saving tens of thousands of patients’ lives?

(I’m going to repeat my disclaimer from the first post here: I’m not a historian, a medical doctor, or a measles researcher, and I’m more than happy to be corrected.  Anyone who has actual information on this, please let me know.  If you have an opinion, no offense, but I’m not interested unless you have data to back it up.)

All of these are actually fairly plausible answers, and probably some of them did play some role. But:

• Surveillance –identification of deaths due to measles, and notification of the authorities–probably didn’t change significantly. Certainly surveillance isn’t, and never has been, perfect, especially when comparing disease frequencies over time:

… because of changes in understandings of the disease and contemporaries’ presentation of the data, the size of that effect and its role in mortality decline more generally elude us. Our analysis of the past depends on numbers that reflect not only changing treatments but also changing understandings of disease at the end of the nineteenth century. 1

And as I noted in Part I, in Britain measles did temporarily stop being a notifiable disease around 1919 — but when official surveillance kicked in again in 1939, measles was still there (chart), with the case-fatality rates right on the same trajectory as before.

In any case, changes in surveillance aren’t likely to account for a 100-fold reduction in death rates; you just don’t miss that much. And while you can probably miss a lot of deaths,2 that’s something that’s more likely to be a factor early on, while the public health system is still getting established. In 1917, for example:

The deaths reported as due to measles give a very inadequate and incorrect idea of the real number due to this disease, for it is well known that most of the deaths are not due to the original disease, but to a complicating bronchopneumonia, and many phsyicians who have failed to report the original disease do not mention it on the death certificate. 3

How to spot measlesSo the later, lower death rates, which are the unexpected and surprising ones, are the most likely to be accurate.

• Sanitation. No one could argue that sanitation isn’t important. But it’s most important for water-borne diseases. Measles is a classic respiratory disease; it doesn’t need water to spread, and sanitation doesn’t stop it from spreading. Of course, we know that the frequency of measles cases didn’t drop, just measles deaths.

What about protection against secondary infections after measles infections? That’s more plausible, because by itself, measles doesn’t kill at high frequency; it’s the secondary infections that kick in afterward, in the immune-suppressed and weakened patient, that are fatal. Here, sanitation probably helped some; diarrheal diseases were occasional causes of death after measles infections. But they were far from the most common. Usually, measles patients died of pneumonia. And again, these respiratory infections are the kind that sanitation is less effective in preventing.

So sanitation undoubtedly had some effect; it’s probably heavily responsible for the overall reduction in childhood mortality (chart; note the scale compared to the measles charts) during that period.  But that background reduction in overall mortality is much less than the effect on measles deaths.  So sanitation alone probably isn’t nearly enough to account for the whole reduction in death rates.

Measles week
Part I: Introduction
Part II: Emerging disease
Part III: Not the answers
Part IV: Some of the answers
Part V: What about the vaccine?

Did the virus mutate? Certainly that could happen — there’s precedent for viruses evolving that degree of change in virulence, in that kind of time-frame.

But modern measles virus can still kill children at the high rates of the 1910 epidemics, when the infections occur in the Third World. The virus hasn’t (apparently) lost its virulence; it’s the environment of the modern developed nation that lets us survive.

(That’s the conventional answer, anyway. I’m not completely convinced, and I would be really interested in seeing the sequence of some fossilized 1910 measles virus. Perhaps, like the 1918 influenza, there’s some frozen in the permafrost somewhere?)

Antiserum treatment. (This story was new to me, and when I ran across the discussion of it I thought briefly I had found the explanation. But probably not.) In the early 1920s, doctors began to transfer serum from people who had recovered from measles, into measles patients.4 This transfers antibodies, which are protective, and it turns out to be a really good way of stopping an ongoing measles infection in its tracks. This  serum treatment was used pretty widely:

In the London County Council,over a period of more than 10 years up to 1943, 66 litres of convalescent measles serum and normal adult serum from over 3,000 donors were used to inoculate 36,000 contacts5

But even thousands per year in the London area is still not very much, in the context of hundreds of thousands of measles cases per year.  Serum treatment doesn’t seem to have been widely-enough used to make a big enough dent in country-wide measles deaths. A fascinating round-table discussion in 19456 specifically considered and dismissed this as having enough effect:

DR. W. GUNN said that measles epidemics usually occurred every second year and reached their height in London about the middle of March, when the number of admissions to L.C.C. hospitals reached its highest level … and there did not seem to be any ready explanation why measles, a more contagious disease, was so slow in coming to a head. He did not think the use of sera had had any significant effect on reducing mortality, because so far few persons received injections. 6

(My emphasis) And that fits with the observation that there was not a sudden, abrupt drop in rates in the late 1920s, as antiserum started to be used, but rather a continuous gradual decrease from 1915 to 1955.

(Incidentally, in hindsight, I wonder how the serum was treated.  There’s no evidence that viral diseases like hepatits C or hepatitis B, or HIV, were spread via these treatments, but I don’t know if that was luck or design.)

So there are some possible explanations that, I think, don’t account for a large part of the drop in measles case-fatality rates. Tomorrow I’ll cover some of the answers that are more likely to have had bigger impacts.

  1. Condran, G. (2008). The Elusive Role of Scientific Medicine in Mortality Decline: Diphtheria in Nineteenth- and Early Twentieth-Century Philadelphia Journal of the History of Medicine and Allied Sciences, 63 (4), 484-522 DOI: 10.1093/jhmas/jrn039[]
  2. Especially since deaths due to measles-induced immune suppression may lag the actual measles infection by some time[]
  3. Observations on Measles. Charles Herrman. Archives of Pediatrics 34:38-42 (1917[]
  4. ZINGHER ABRAHAM (1926) Convalescent whole blood, plasma and serum in the prophylaxis of measles. JAMA 1180–1187.[]
  5. COCKBURN WC, HARRINGTON JA, ZEITLIN RA, MORRIS D, CAMPS FE (1951) Homologous serum hepatitis and measles prophylaxis; a report to the Medical Research Council. Br Med J 2:6–12.[]
  6. Butler W (1945) The Fatality Rate of Measles: A Study of its Trend in Time. Journal of the Royal Statistical Society 108:259–285.[][]