Sorry for the delayed response. Medical school tends to provide just enough time to make comments, but not so much to respond more indepth. This response hasn't been mulled over as much as I'd like. I still have a few private messages to respond to, which are aging quickly. I hope to get to those over Thanksgiving break. Anyhow...
No problem man. This has taken me a while to respond to as well.
Cardiac arrest survival as a function of ambulance deployment strategy in a large urban emergency medical services system.
Persse DE, Key CB, Bradley RN, Miller CC, Dhingra A.
Resuscitation. 2003 Oct;59(1):97-104.
PMID:14580739
This is interesting, but I'm not sure about the statistics. They see significance when they use a Fisher's exact test, but when they report confidence intervals, their survival to hospital discharge is 0.02-1.06, and their alive at 1 year is 0.5-148.6. They also only got 24 "resuscitation attempts in the "uniform deployment" model, so it seems like it might be a little underpowered.
If there is a real difference here, it's very interesting -- because traditionally we've seen the factors that determine outcome be things like CPR, defibrillation, etc. Not typically ALS therapies. It seems like the tiered response group here were quicker at starting IVs and intubating -- but the p value for defibrillation is 0.29. Although they couldn't capture time to AED defibrillation, which might actually explain the difference in survival, without needing to invoke differences in deployment.
I don't think that EMS is some out-lier whereby experience doesn't matter like it does for the rest of medicine. Time and time again, experience is shown to matter in medicine. This should be the base assumption for EMS until something shows different.
On the surface this sounds reasonable, and maybe the burden of proof should be on those who say that an all-ALS model is better. I'm not sure. But if we're talking only about cardiac arrests, the idea that ALS intervention impacts outcomes isn't supported by the bulk of available evidence.
There's a point to be made here, and it's that if we have restricted ALS resources that are busier, we're also giving something up. Maybe we're no longer getting 12-leads on borderline presentations of MI, e.g. "weak and dizzy all over", "abdo pain", and we're no longer getting pain control to the same number of people, and perhaps we're getting less aggressive treatment of respiratory distress (ALS shown to have an improvement on outcome versus BLS-symptom relief in one OPALS subgroup). So if we have a tiered model, we may be giving something up as well.
]The effect of paramedic experience on survival from cardiac arrest.
Gold LS, Eisenberg MS.
Prehosp Emerg Care. 2009 Jul-Sep;13(3):341-4.
PMID: 19499471
The numbers are a little messy here as well. Lot's of confidence intervals that cross unity. They're reporting a 95% confidence interval for the OR of 1.00-1.04, and this is looking at a decent number (n=699).
I mean, it's hard to know. There might be something there. But it might be just noise. When they're starting to make statements like,
"We found that every additional year of experience of the medic in charge of implementing procedures such as intravenous line insertions, intubations, and provision of medications was associated with a 2% increase in the likelihood of survival of the patient (95% CI: 1.00–1.04). The number of years of experience of the paramedic who did not perform procedures but instead was in charge of treatment decisions was not significantly associated with survival (odds ratio [OR] 1.01, 95% CI: 0.99–1.03)."
It's hard to know if they have something. It would be great if they did. I'd love this -- but it's hard to be convinced. And it's really hard with all this King County stuff to know if we can generalise it to other systems -- because not a lot of environments have the same level of PAD or community CPR, and very few communities are posting similar survival rates.
(I'm assuming here that survival is survival to discharge, or 1 year or something relevant -- they don't explicitly state that, but I'm willing to give them the benefit of the doubt)
It could be that the region just has that good a system that they've removed a lot of the noise associated with differences in CPR / PAD / First response, and now we're seeing the real benefit of ALS. Or it might be nothing.
Out-of-hospital endotracheal intubation experience and patient outcomes.
Wang HE, Balasubramani GK, Cook LJ, Lave JR, Yealy DM.
Ann Emerg Med. 2010 Jun;55(6):527-537.e6. Epub 2010 Apr 14.
PMID: 20138400
(This study used survival to discharge as the outcome for multiple conditions associated w/ intubation.)
I like this better than the other two papers. This is a little more convincing. There's a certain argument that the patients treated by the providers with greater ETI experience might be closer to bigger trauma centers -- but it seems like when they looked at the rural responders as a subgroup the association was even stronger.
"All-ALS" has not been shown to be better than a tiered approach with fewer ALS resources. All-ALS is more expensive. How can one justify the cost of an all-ALS system? Shouldn't we use the cheaper option until more evidence is available to justify the increased expenses?
I'm not sure if all-ALS is much more expensive. Are the tiered response agencies running dual-medic? If so, a lot of all-ALS systems run medic / EMT. Staffing costs may also be minor, at some places there'a a 20% difference in starting EMT versus starting medic pay.
It's hard to argue here that we're using the most medically rational deployment structure. There's a lot of demand from taxpayers for a paramedic response to every call, and I don't think this area has been adequately researched to know which model is superior.
All-ALS, in a fashion similar to places outside the US does make sense. In the US, it tends to be all or nothing. There is not reason to send a paramedic with the US skill set to every patient. Now, if you want to send a paramedic with high yield skills - e.g. pain management, 12 lead, NTG, albuterol, ASA, etc. to every patient and then reserve a paramedic with a "higher" skill-set, then great. Maybe AEMT should become the standard ambulance staffing with Paramedic back-up?
Maybe. It's hard to know. I wonder if maybe we shouldn't keep the current training structure but have paramedic as the base-entry level. We could remove RSI, give better instruction on NIPPV, perhaps introduce some decent BiPAP, retain 12-lead capable providers on each ambulance, but cut back on the scope a little. Then develop a high tier medic with advanced airway management skills, and focus them on the critical calls.
I think we need to be able to deliver a provider to every call that can obtain and interpret a 12-lead, and give pain control. But maybe they don't need to be able to intubate.
Anyhow, I wish I had time to dig through more sources, but I've got to get studying for a bioethics final... a;ewhg;lakfhgl
Thanks for the references. I may have to re-think my position a little.
