Ambulance Shortages… A Game of Numbers

I feel like we’re beating a dead horse every time we see a news article like this.

Yesterday I was updating my twitter stream and saw a tweet from @mfdalert regarding Memphis having no available ambulances.

“No Units Available”

Words made famous in the EMS industry most recently by Charlie LeDuff in Detroit, describing the 20+ minute response time of Detroit paramedics to emergency calls. A phrase when given over the air on the EMS channel makes me cringe a little, just because I know what ends up happening is us getting sent across the city for a (typically) non-life threatening (NLT) call.

I remember a year or so, maybe longer, where Cleveland EMS established a minimum ambulance level, that they would dispatch calls as received until they ran out of ambulances, and then would hold non-emergent calls until the minimum number of in service ambulances returned. It’s a great idea, banking on the dispatchers to accurately (and properly) assess the call, and give it at the appropriate time.

I still remember the cries of foul on the message boards when Cleveland announced this as an alternative to their shortage woes. “Well, what happens if the dispatchers code the call incorrectly?” “This will kill people.”

Strangely enough, these were the same people who were jumping for joy in their minds because they finally have a way to solve those crunches on busy days. In order to talk about how often a mistake will be made, we have to be able to accurately assess risk, which, I think, emergency services are poor at doing. In order to talk about risk, we should assess the job we do as a whole using statistics and percentages.

Let’s say an ambulance makes 18 calls a day, on average. Of those calls, 70% require an emergency response as coded, so 12.6 calls require lights and sirens. When the provider arrives, they find that it is immediately life threatening less than 10% of the time, so approximately 1.8 calls are immediately life threatening requiring a speedy transport. Cardiac arrest presents in 0.018% of those calls, of which we can save about 12% of that 0.018%, the remaining 88% either do not survive to discharge or are left down in the field.

So what do the number say, exactly? It means that we both over-triage by phone AND 90% of all calls we receive is an NLT.

My suggestion to this: copy the Cleveland model and improve it. If we receive a NLT call during the shortage, and a minimum number of ambulances is not available, let’s say we want to hold 20% of the total force in reserve for LT (life threatening) calls, in Memphis that would equal 6 or 7 ambulances, so when the level drips from 33 down to 7, we start calling private ambulance companies to handle the overage. This doesn’t mean that we don’t utilize some form of system status management, like a semi-static response model, where we relocate ambulances from less busy stations to more busy stations during high volume periods to lessen response time. With the arrival at MDT and GPS tracking technology, ambulances are no longer static vehicles that cover an assigned area and only that area, they now cover the whole region, and as a regional resource the deployment should reflect as such.

This article was written by rstine