In the last two months, no doubt the most important news is the influenza A(H1N1) epidemic. A few friends were concerned about me as a frontline worker in the hospital. When they learned that I was perfectly fine, the discussion drifted to why the virus had become less deadly over the weeks. As of 12 May 2009, over 5000 patients were confirmed to be infected. Fifty-eight Mexicans died, compared to 5 in the rest of the world.
The cause for concern is understandable and has been discussed extensively elsewhere (http://ccszeto.blogspot.com/2009/05/show.html, http://drkmchow.blogspot.com/2009/05/contagion.html). I will concentrate my discussion on why the mortality rate of influenza A(H1N1) appears to drop outside Mexico.
When the outcome of a viral infection changes, there are only four possibilities: (1) The virus has changed; (2) The host has changed; (3) The treatment has changed; and (4) The observation is an artifact.
The most successful microorganisms do not kill people. Viruses spread by droplets would transmit more efficiently if the hosts survive and walk around. The only microorganisms that benefit from debilitating the hosts are those transmitted by vectors. If you cannot move around, mosquitoes and ticks can bite you easily and transmit the microorganisms around. Therefore, we often see that viruses causing respiratory infections become less virulent with time. Comparison of earlier and later strains of the virus would answer this question.
Mortality from influenza A(H1N1) appears to be more common among adults and diabetics in Mexico. It remains to be seen if high-risk individuals are over-represented in the Mexican cohort. It is unlikely that the mortality rate can be explained by ethnicity since the mortality rate in Mexico is also declining.
Early use of antiviral drug (e.g. oseltamivir) and advanced organ support have also been proposed to be the cause of lower mortality in developed countries. While this may be true, I have not yet seen reports showing such a dramatic efficacy of oseltamivir. The point on organ support can only hold if a significant percentage of infected US citizens required intensive care. Another potential explanation is that the pattern of coinfection may be different in different countries. Again, further investigations are required.
Finally, I think the most probable explanation is that the observation is an artifact. We now know that the majority of infected individuals have mild symptoms. It is very likely that many infected people were never detected and reported, especially in the early stage of the epidemic.
As a young journal reviewer, I often reverse the sequence. I check if a study has major methodological flaws before learning what it is trying to say. More experienced academics, however, tend to see if a study asks an important question before reading on. After all, anything worth doing is worth doing well. The converse is also true.