Will Swine Flu Exhibit Multiple Infection Waves?
In the 1918 Pandemic we had three distinct peeks of deaths. One in mid July, one in Late October / Early November. and one in Late February (upper panel Figure One). Given the similarity between the current H1N1 Pandemic and the 1918 Pandemic I wondered if there was evidence of multiple waves of H1N1 infections.
I used a specialized mathematical curve fitting package called PeakFit which takes raw data and looks to see if there are multiple underlying distributions present.
I used Quest Diagnostics confirmed H1N1 infections in the US as the data. I did that because I could not find number of deaths at a data resolution of weekly reports. So I am assuming that the death rate is consistent and proportional to the number of confirmed H1N1 infections.
The PeakFit software tells us there are two distinct distributional peaks present in quest diagnostics H1N1 infection data (Panel Two). We have a peak in US confirmed swine flu infections in Late June, and a much larger peak in early November. Note that this is eerily similar to the results in the 1918 - 1919 Pandemic.
Chart from: www.peterosborn.com/Reports/pandemic.htm
IF the similarity observed is not coincidental, and that is a BIG IF, than the results suggest another wave of H1N1 infections beginning in late January, peaking in late February / Early March which happens to be coincident with the peak influenza activity in the US (Figure Two).
At the minimum, results suggest that H1N1 has already exhibited two distinct waves and that if there is going to be another wave it should exhibit during the typical peak influenza activity in the US.
The primary take-away is don't let your guard down.