Visualizing the Opioid Epidemic

It’s been a while since my last blog post, for a couple reasons: the Ultimate season began and took up much of my free time, and it has been almost two years since the 2016 election- so there’s not a lot of new political data to analyze. Hopefully, now that Ultimate is over for the year and the midterms are right around the corner, I will be able to do some more analysis in the coming months. I’m especially interested in continuing to look for electoral anomalies, seeing as how the Russians appear to again be meddling in the midterms.

But until then, I wanted to share a visualization that I made of the rapid rise of the opioid epidemic:

Figure 1: Annual drug poisoning rate per 100,000 people by county. The epidemic started in Appalachia and New Mexico, but has since spread to nearly every corner of the country.
Figure 1: Annual drug poisoning rate per 100,000 people by county. The epidemic started in Appalachia and New Mexico, but has since spread to nearly every corner of the country.


The data for the animation above comes from the CDC, and covers the years 1999 through 2015. One caveat of note: this data is the total drug poisoning rate, including overdoses from non-opioids. But it’s safe to say that the bulk of the deaths in the animation come from opioids. Just consider Rio Arriba County, New Mexico: at first I thought that the high rate there was a typo in the data, but it is in fact mostly related to heroin and other opioids.

I should also note that the CDC doesn’t give exact numbers for the drug poisoning rate, but gives a range (e.g. 10.1-12.0 per 100,000). I used the lower end of the range to be conservative, meaning the actual rate is higher than it is shown in the animation. Because the maximum range given by the CDC data is just “greater than 30”, in some counties (like Rio Arriba) the true rate is actually much higher.