January 2010: Yearly wrapup

Now that 2009 is over, it is interesting to look at the growth of VAERS reports over its 20-year history. Here is a graph of the number of reports that were added to VAERS each year. You can see this result for yourself by going to the MedAlerts search page and graphing "Year of Appearance". Note that 1990 has a small number of reports because it was the first year of operation and does not represent a full year of data collection. In the remaining graphs of this report, I will ignore 1990.

The number of reports added in 2009 was, as expected, another record-breaking year. However, what is particularly noticeable about this graph is the sudden jump in the past 3 years. Why is this?

Let's go back to MedAlerts and graph the individual vaccines over the last 20 years. Below is the graph of "Year of Appearance" and "Vaccine". Unfortunately, there are over 70 vaccines, so the result is a jumbled mess that is hard to interpret.

Although it is hard to tell what is going on, some vaccines do make a jump in the last three years. But rather than try to isolate those vaccines, let's look at it a bit differently: let's look at "Vaccine Type" instead of "Vaccine".

The reporting of "Vaccine Type" organizes vaccines according to the diseases that they address. Some vaccine types are straight-forward (the YF vaccine is the only one of type "Yellow Fever"). Many vaccines have multiple types (the MMR vaccine has the types "Measles", "Mumps", and "Rubella") and many vaccine types have multiple vaccines in them (there are 19 different vaccines with the type "Diphtheria"). If you want the full list of types associated with each vaccine, see the list of MedAlerts abbreviations. Graphing the vaccine type is useful because there are three times fewer vaccine types than vaccines, so this reduces clutter and may provide some insight into what is going on. Here is that graph of the "Year of Appearance" and "Vaccine Type".

In the above graph, each line falls into one of four different categories:

1. The lowliers at the bottom of the graph that never have many VAERS reports and haven't grown lately. These include Anthrax, Cholera, Encephalitis, Lyme, Plague, Rabies, Smallpox, Tuberculosis, Typhoid, and Yellow Fever. These types are not interesting and can be ignored.
2. The middle group, which do have significant number of VAERS reports, but have not increased in the past few years. These include Measles, Mumps, Rubella, and Polio. These types are also not interesting and can be ignored.
3. One particular type that has shot up dramatically in the last year and spikes to the top of the graph on the right (you guessed it: Influenza). This is interesting, but not relevant to the question of the last three years.
4. The remaining types, which show a noticeable rise in the past few years and are detailed below.

What does this tell us? First of all, it is well known that Tetanus, Diphtheria, and Pertusis are usually given together, so it is no surprise that these three curves follow each other. Some vaccine types are new (HPV and Shingles) so their rise in recent years is the result of their introduction. Much more information needs to be known before any conclusions can be drawn. Nevertheless, it is instructive to see these trends, and MedAlerts provides some tools needed to do this.

Tip of the Month for using MedAlerts

MedAlerts graphs are very powerful and can be changed on the fly. For example, each line of a graph can be removed simply by unchecking its name in the legend area below the graph. This makes it easy to remove clutter quickly and find the relevant curves, as I showed above. Another graphing tip is that you can see the data behind a particular point in the graph by placing the mouse over that point and waiting for a "tooltip" to appear. You can also change many other aspects of a graph without redrawing the web page because Medalerts graphs are "live" and very powerful. See the graphing help page for more information.

<< 12/2009: HPV vs. Flu adverse event rates 2/2010: MedAlerts vs. CDC Wonder >>