This morning, on the wake of the #michiganstateshooting, I had to tell my daughter that there was a social media threat against our school system yesterday. (The threat was deemed not credible, but there will still be an increased police presence today at school.)
Research can help us fix our nation’s firearm injury epidemic - and maybe not in the ways you think. My new piece for @washingtonpost outlines 5 key questions we need answered:
1. What are the actual numbers (of injuries, defensive gun use, stolen guns, averted shootings, etc)?
Believe it or not, we don’t know. (This is why CDC websites mostly discuss death rates - that’s what we actually have semi-reliable data on.)
2. Who is at risk?
We can’t reliably identify what makes someone higher risk for gun suicide, homicide, mass shootings…. (Kudos to folks who are trying… but we need better data and prediction tools, both in the moment and over time.)
For the flu vaccine, we (in US) choose based on what’s circulating in the southern hemisphere. But we have no data to support that approach for Covid. FDA is proposing a June decision on dominant variant… ok.
2. Why bivalent, vs updated monovalent?
This is the big debate in the scientific community & lots of folks have strong theories. The OG (Wuhan) strain is clearly passé. Will monovalent (single variant) be more effective? Stay tuned.
But I will bet 1000:1 that (as with almost every one of these) there were missed signs. Someone that should not have been allowed to own a gun. Hatred & desperation.
They matched people in a HUGE database - > 6 million recipients of the 3 most common vaccines (Pfizer, Moderna, J&J) and > 6 million people with similar age/race/gender/co-morbidities/neighborhood characteristics who had not been vaccinated.
They matched well (tables 1-3).
They then examined, using standard statistical techniques, the likelihood (“hazard ratio”) of: 1. any death other than Covid (the primary outcome of interest) 2. hospitalization for trauma (vaccines shouldn’t affect incidence of trauma!)