This article got a ton of press - but it tries to take far too simple an approach to far too complicated an issue. I've done research like this, where you use zip code centroids and calculated distances to nearest hospitals, and it's just one way a blind man describes an elephant.
These authors look retrospectively at all the acute MIs in four California counties, then looked at hospital daily diversion logs for each day from each of those hospitals - and tried to merge them together to prove that if your nearest hospital was on diversion for a lot of the day you had your acute MI, you had worse outcomes.
Their final analysis says, basically, there's a 3-5% difference in 30-day, 90-day, and 1-year mortality if your nearest hospital is on diversion >12 hours in a day vs. if your nearest hospital is on diversion <6 hours per day. The between 6-12 hour diversion cohort performed identically to the <6 hour per day cohort. So, I don't know exactly what to make of this. Their 95% CI almost crosses zero. Something magical happens at 12 hours that changes your acute MI mortality risk. So, yes, what the authors are trying to prove is probably true - but this article's data mining and massage can only hypothesize the association, and doesn't prove anything.
"Association Between Ambulance Diversion and Survival Among Patients With Acute Myocardial Infarction."