New open access paper with Sanjay Basu (@StanfordMed), Jarvis Chen (@HarvardChanSPH) and @monjalexander (@UofT) in @JAMANetworkOpen:— Mathew Kiang (should be writing) (@mathewkiang) February 22, 2019
Q1. (a) Where is the opioid epidemic the worst and (b) which type?
Q2. What does that mean?
(Long-ish) Thread: 1/10https://t.co/H9HdjstB9J
For Q1, we can say a state's been strongly impacted by the epidemic if the mortality rate is high or if deaths are accelerating quickly. We defined it as both: the opioid-related mortality rate is (a) >10 per 100,000 *and* (b) >doubling every 2 years (>41% annual increase). 3/10 pic.twitter.com/eqr1rrZrg2— Mathew Kiang (should be writing) (@mathewkiang) February 22, 2019
For Q2: What does that mean? We estimated the implied life expectancy lost (LEL) at age 15. Nationally, LEL from opioids is .36 years—higher than car accidents (.30) or guns (.34). For synthetic opioids alone, it was .17 years. However, a lot of variation across states. 6/10 pic.twitter.com/opddfJOf1f— Mathew Kiang (should be writing) (@mathewkiang) February 22, 2019
Take-home 1: The opioid epidemic is rapidly evolving and geographically expanding with substantial population-level impact—on par with car accident deaths and gun deaths in the majority of states. 8/10— Mathew Kiang (should be writing) (@mathewkiang) February 22, 2019
Take-home 2: In some areas, synthetic opioid deaths outpaced heroin-related deaths, which suggests new policies should take into account the possibility for synthetic opioid epidemics that occur outside of the heroin supply, and low-heroin-mortality states are not immune. 9/10— Mathew Kiang (should be writing) (@mathewkiang) February 22, 2019