Dec 11 2009
We are nearing the end of the second wave of the 2009 H1N1 pandemic, and are now a few months out from the release of the vaccine directed against it. Two topics have dominated the conversation: the safety of the 2009 H1N1 influenza vaccine, and the actual severity of the 2009 H1N1 infection. Considering the amount of attention SBM has paid the pandemic and its surrounding issues, and in light of a couple of studies just released, it seems time for an update.
2009 H1N1 Vaccine Safety
This week the CDC released a report that evaluated the safety record of the 2009 H1N1 vaccine. The first two months of the vaccine’s use were examined, from October 1st through November 24th using data from two of the larger surveillance systems monitoring the 2009 H1N1 vaccine’s safety: the Vaccine Adverse Event Reporting System (VAERS) and the Vaccine Safety Datalink (VSD). This report represents the largest, and to date best, evaluation of the 2009 H1N1 vaccine’s safety profile since its initial testing and release. The findings are reassuring.
We’ve talked about VAERS’ uses (and abuses) in the past. Nevertheless, used properly as a surveillance tool, a “canary in a coal mine,” it can be quite helpful. In that two-month span of time when 46.2 million doses of H1N1 vaccine were distributed, 3,783 adverse events associated with it were reported to VAERS. 204 of these events were classified as “serious,” including 13 deaths that occurred within 19 days of vaccine administration.
At first blush people may assume (unwisely) that the vaccine directly caused each of these reported events, and would thus yield an adverse event rate of 82 total adverse events and 4.4 serious adverse events per 1 million doses. This is indeed the assumption (and mistake) made by people claiming for instance that the flu vaccine has caused X number of deaths or Y cases of Guillain Barre Syndrome (GBS). Even taken (again, unwisely) at face value, these rates would be impressively low, particularly when compared to the risks of H1N1 infection, as we shall see later.
The story is even more reassuring once we look properly at the data. It bears repeating that VAERS does not (nor was it meant to) establish causation, it only holds the potential to suggest a correlation. We should also bear in mind that GBS, death, all adverse events in fact, occur at a baseline rate in the population in the absence of the vaccine (a hypothetical vaccine causing zero adverse events would still have a list of adverse events reported to VAERS reflecting the population’s baseline rates). Thus to even determine if there is a significant correlation between the vaccine and any given adverse event, we need to determine not only how many adverse events occur in relation to the 2009 H1N1 vaccine, but the number that occur above the expected baseline.
That having been said, let’s examine the most concerning number first, the 13 reported deaths. Each of the 13 are detailed on this chart. It’s very much worth taking a look. There is no discernable pattern to the ages of these unfortunate people, their underlying diseases, or their causes of death. 9 of these 13 people had significant underlying diseases, and one of them died in a car accident. Indeed, considering the population’s baseline mortality rate, it’s remarkable that only 13 people out of 46.2 million died within 3 weeks of receiving the vaccine by chance alone. This doesn’t definitively exonerate the 2009 H1N1 vaccine from these deaths (well, we can probably safely rule out the car accident), but it certainly makes its involvement highly unlikely.
H1N1 Vaccine and GBS?
What of the concern of Guillain Barre Syndrome (GBS) following vaccine administration? After all, at least one influenza vaccine in the last three decades has been shown to cause GBS in rare cases, and some poorly handled stories in the media have further elevated public concern.
The first two months of vaccine use saw 12 cases of suspected GBS reported to VAERS. Investigation into these reports has confirmed four of these to be cases of GBS, four were not GBS, and the final four are still under scrutiny.
Again, these cases require context. As the baseline rate of GBS is ~1/100,000 people per year, ~550 cases can be expected to occur in the US during the two months of this report. These 8 likely cases of GBS in 46.2 million doses of vaccine is certainly not higher (and is in fact far less) than what one would expect to see by chance. The VAERS database provides no reason to suspect the 2009 H1N1 vaccine has anything but chance correlation with cases of GBS.
H1N1 and Other Severe Adverse Events?
There is no correlation between the H1N1 vaccine and either GBS or death, but what of other concerning adverse events? An evaluation of the 204 serious events reported reveals a scattershot of diseases, none of which have a signal that rises above baseline rates.
The CDC report contains a similar analysis using data from the VSD, where 438,376 doses of the H1N1 vaccine had been administered and adverse events tracked. As with the VAERS data, no serious adverse events rose above their baseline rates.
In short, after the first two months of use and 46.2 million doses, the VAERS and VSD data fails to provide any evidence to correlate the 2009 H1N1 vaccine to any serious adverse event. Given the seasonal influenza vaccine’s similar record over the past several decades, that the 2009 H1N1 vaccine continues to display an exemplary safety profile is not unexpected, but it is reassuring.
How Severe is 2009 H1N1?
What of H1N1’s severity? What toll has it exacted? The CDC has made detailed information, updated weekly, available to the public on its Fluview website. Containing a wealth of information, there you can see 2009 H1N1’s unique and peculiar epidemiology, the unseasonable spikes in outpatient visits for influenza-like illnesses that have troubled our EDs for the last few months, and the trend of lab-confirmed influenza hospitalizations and mortality over time.
Hard numbers are also available. As of November 28th, at least 31,320 people in the US have been hospitalized and 1,336 have died from 2009 H1N1 since August 30th. The 2009 H1N1 has thus far claimed the lives of at least 250 children in between the traditional flu seasons, which is more than the two prior flu seasons combined.
This data is most helpful if viewed as the minimum confirmed impact of the disease, and as a catalogue of the most severe cases to date. What you will not find on the Fluview site is the actual incidence of influenza infection, the total number of people infected, including minor infections. This number is extremely valuable when trying to gauge the true severity of any infection, but fiendishly difficult to acquire.
A study published in PLoS Medicine this week contains one of the latest attempts to quantify 2009 H1N1’s severity to date. Drawing from the data of two US cities during the initial wave of infections between April and July, they estimated that of all 2009 H1N1 infections, between 0.16-1.44% will require hospitalization, 0.028-0.239% will require ICU care, and 0.007-0.048% will die.
This study has garnered a significant amount of attention, for its estimates of severity are considerably lower (thankfully) than those made by the President’s Council of Advisors on Science and Technology in early August. The accuracy and differences between these estimates, the inherent difficulty of determining the true incidence, severity, and future course of diseases like influenza warrants its own post, and I’ll not address this particular angle in greater depth here.
I’d like to instead reflect on what these two studies might tell us about the risks of contracting 2009 H1N1 compared to the risks of receiving the vaccine to protect against it.
On the one hand, we have a virus that has proven itself to be widespread and highly contagious, to have claimed the lives of at least 1,336 and hospitalized over 30,000. Conservative estimates from the PLoS study place one’s risk of hospitalization if infected at ~1/625, and risk of death ~1/14,285. Furthermore, though we have completed the second wave of the pandemic, a third wave is almost certain to come. A small minority of the population has thus far been infected, past influenza pandemics have featured a triple peak, and we have now entered the beginning of the traditional influenza season.
On the other hand, we have an inexpensive vaccine which is an excellent match to this strain, generates an appropriate antibody response in most people (particularly those in the highest risk groups for 2009 H1N1), and after over 46 million doses has yet to be significantly correlated with any severe adverse events.
There are still a lot of uncertainties regarding the rest of this influenza season. Will we have a third peak of H1N1, and if so, how severe will it be? Will it continue to preferentially afflict the young, or will the elderly suffer a greater impact than they have to date? How will the presence of 2009 H1N1 impact the normal flu season, will it be cumulative, or will 2009 H1N1 “crowd out” the seasonal strains? The list goes on, and it absolutely includes the possibility that with ongoing surveillance and studies we may identify a serious but rare side effect caused by the vaccine.
As time goes on we will continue to refine our knowledge of influenza, and these questions will be answered, but it is unlikely that the big picture will significantly change. Influenza is a virus with serious potential for harm that can be prevented by one of the safest interventions in modern medicine. Please, particularly if you or yours are in a high-risk group, get vaccinated; I already know far too many of the names on this list.
22 Responses to “An Influenza Recap: The End of the Second Wave”