Prehospital antibiotics

[Brandon O]

As this topic has come up before, for your interest: http://www.thelancet.com/journals/lanres/article/PIIS2213-2600(17)30469-1/fulltext

No benefit on mortality. Not very long transport times though.

 

[E tank]

So, with abstracts, more questions than answers...I don't have access to the Lancet paper...

To begin:

"The intervention group received antibiotics a median of 26 min (IQR 19–34) before arriving at the emergency department."

So the medics recognized and started abx in a relatively short time.

Then:

"In the usual care group, median TTA (time to antibiotics) after arriving at the emergency department was 70 min (IQR 36–128), compared with 93 min (IQR 39–140) before EMS personnel training (p=0·142)."

Are we to take this to mean that the ED physicians took nearly 40 minutes longer than the medics to determine that the patient needed antibiotics after the medics were trained to recognize sepsis and 3 times longer than paramedics before the paramedics were trained?

This begs the question. Are the medics giving abx to people that don't need them or are they better at diagnosis than the ED medical staff? Not having the paper to read, we don't know how many if any patients were treated inappropriately. It'd be good to know what patients were excluded from the study...Meta analysis forthcoming hopefully.

Don't send those cephalosporins back to the pharmacy just yet...

 

[Carlos Danger]

I could be remembering incorrectly because I haven't looked at it in a while, but my understanding was that while recommendations call for anitbiotics within an hour of recognition of severe sepsis / septic shock, there is really little evidence that earlier administration results in better outcomes. In other words, administration within one hour is not necessarily better than administration within two hours, maybe even three hours. Which kind of makes sense when you consider we're talking about time from recognition, not necessarily time from onset. Sometimes these things take days to develop, and they are recognized at all different points on the timeline as compared to when the SIRS actually began.

If that's all true, then it makes sense that the 20 or 60 minutes saved by giving the ABX prehospital wouldn't show a benefit.

 

[rescue1]

I skimmed the entire paper, so a few thoughts.

This study takes place in the Netherlands, so there a few differences. Notably, almost 75% of patients were referred by their GP and many of them were already on antibiotics. Also, EMS in the Netherlands is run by RNs who have taken extra training, so that has some potential confounding there when looking at a US system.
Still, before training, only 14% of septic patients were correctly documented as such by EMS nurses, after training it increased to 41%. The TTA for the usual care group also dropped about 20 minutes after training.

There were also mostly short transport times.

The study population was healthier than many other sepsis studies, which focused mainly on septic patients already admitted to an ICU. Therefore, only a small amount of these patients had actual septic shock (around 3%). This probably also explains some of the long time to antibiotics for the usual care group (no EMS antibiotics) and also the low mortality rate.

The time to antibiotics for the usual care group began at triage, not at physician assessment.

The TTA in the usual care group, despite ranging from 36 minutes to greater than 240 minutes, did not impact survival either.

Older studies that looked at TTA dealt with much longer times than the usual care group. Those studies had TTAs ranging from 2-5 hours, as opposed to the one hour ish in this study.

I couldn't really find anything on patients who were "inappropriately" given antibiotics by EMS, which would seem like a very obvious thing to consider when using SIRS criteria, which isn't the most accurate test in the world.

So basically, who knows. Perhaps even earlier antibiotics are superior for a small group of very severe sepsis cases, or perhaps the system in the Netherlands is efficient enough that the improvements in EMS aren't helpful.
Most likely, as Remi said, antibiotics are better when given quickly but only up to a certain point, which was already reached by the EDs in this study.

Personally, I've yet to be convinced that EMS antibiotics would be useful except in the most rural of places.

 

[EpiEMS]

Mortality matters, of course, but if there’s reduced readmission, shorter hospital stays, etc. it could very well be worthwhile.

 

[rescue1]

Mortality matters, of course, but if there’s reduced readmission, shorter hospital stays, etc. it could very well be worthwhile.

All variables (mortality, length of stay, % ICU admission) were the same except for 28 day readmission, which was slightly less in the treatment group.

 

[E tank]

I skimmed the entire paper, so a few thoughts.

where did you find the whole paper?

 

[Peak]

Just a few thoughts, currently regulatory standards from CMS are that we need to draw blood cultures prior to antibiotic administration. Are field providers going to draw these? Who is going to grow out the culture? Will there be consideration based on destination as not all hospitals use the same blood culture bottles and laboratory equipment?

How would prehospital antiobiotics be chosen? It sounds like in most of these studies they are giving rocephin, but depending on the suspected source and pending susceptibilities we sometimes use maxipime, merrem, invanz, zosyn, unasyn, ampicillin, clindamyacin, vancomycin, and a few others as our first line antibiotic instead of rocephin.

Are systems who are giving or trialing antibiotics giving the adult fluid bolus of 30 mL/kg or peds intial bolus of 20 mL/kg, if so how are they evaluating the need for additional fluid boluses? Have any primarily 911 systems used lactate measurements either to trend or initially assist in determining sepsis?

 

[rescue1]

where did you find the whole paper?

Hypothetically, one could google a Kazakhstani scientist who got fed up with the payment requirements for research papers and made a website storing millions of free papers.

I would of course never encourage this kind of behavior.

 

[DrParasite]

Just a few thoughts, currently regulatory standards from CMS are that we need to draw blood cultures prior to antibiotic administration. Are field providers going to draw these? Who is going to grow out the culture? Will there be consideration based on destination as not all hospitals use the same blood culture bottles and laboratory equipment?

forgive my ignorance in this area, but aren't he blood cultures drawn, so you can grow the culture and administer the appropriate medication based on the type of infection? Basically what Peak is saying below, or are we saying generic penicillin for every septic patient?

How would prehospital antiobiotics be chosen? It sounds like in most of these studies they are giving rocephin, but depending on the suspected source and pending susceptibilities we sometimes use maxipime, merrem, invanz, zosyn, unasyn, ampicillin, clindamyacin, vancomycin, and a few others as our first line antibiotic instead of rocephin.

 

[rescue1]

forgive my ignorance in this area, but aren't he blood cultures drawn, so you can grow the culture and administer the appropriate medication based on the type of infection? Basically what Peak is saying below, or are we saying generic penicillin for every septic patient?

You would give empiric broad spectrum abx (in this study ceftriaxone, but in the US I think that's rarely given alone due to resistance) depending on your suspicion of where the infection originated. It takes time--more than 48 hours--to culture bacteria, which is obviously too long to wait for people with severe infections. However, once you have cultured the bug, you can switch from broad spectrum abx to something more specific. If you were to give prehospital antibiotics, EMS would need to draw blood for culture before giving antibiotics for this reason.

Penicillin isn't broad spectrum, more commonly you would use pipercillin/tazobactam, a carbepenem, or a combination of a few others.

 

[Akulahawk]

What rescue1 says is correct for the US. We give broad spectrum abx depending upon the suspected problem bacteria involved. One reason you might see different antibiotics used initially is that different areas very well could have bacterial that is susceptible to a different antibiotic than in another area. When they do the blood cultures, they don't just grow to identify the bug, they also determine what the but is sensitive/susceptible to. Once they know that, a more specific, targeted antibiotic can be used. Doing a C&S takes about 72 hours just because it takes time to grow the bug, isolate it, grow more of that, and then determine what kills it. When I discharge people from the ED on antibiotics, if we did any C&S studies on that patient, I tell them to take the prescribed abx unless they hear differently from us as we'll call in about 3 days if they need to change to a different antibiotic. No news is good news because that means they should continue taking the medication as prescribed and to completion.

 

[reaper]

This has been working for two years already.

http://www.jems.com/articles/print/...-hospital-sepsis-care-into-protocols.html?c=1

 

[Brandon O]

Just a few thoughts, currently regulatory standards from CMS are that we need to draw blood cultures prior to antibiotic administration. Are field providers going to draw these? Who is going to grow out the culture? Will there be consideration based on destination as not all hospitals use the same blood culture bottles and laboratory equipment?

Cultures drawn before antibiotics IF it will not delay antibiotic therapy.

I agree that some of the details may be challenging, particularly drawing sterile cultures, which is even difficult in the hospital sometimes. Contaminated cultures cause a bunch of headache.

How would prehospital antiobiotics be chosen? It sounds like in most of these studies they are giving rocephin, but depending on the suspected source and pending susceptibilities we sometimes use maxipime, merrem, invanz, zosyn, unasyn, ampicillin, clindamyacin, vancomycin, and a few others as our first line antibiotic instead of rocephin.

Obviously it would be a numbers game, looking for broad empiric coverage while recognizing the challenges of the austere environment (storage, no labs, etc). You won't cover everything, but you can hopefully get most of it, and do it safely. I would probably do something like Zosyn.

Are systems who are giving or trialing antibiotics giving the adult fluid bolus of 30 mL/kg or peds intial bolus of 20 mL/kg, if so how are they evaluating the need for additional fluid boluses? Have any primarily 911 systems used lactate measurements either to trend or initially assist in determining sepsis?

I imagine most EMS systems are already comfortable bolusing fluid.

 

[Tigger]

My medical director asked for some research on this two years ago when we were talking about adding pre-hospital antibiotics for sepsis and open fractures. We didn't, but we also lack clinical leadership that is willing to do anything that wasn't the standard of care in 2005 unless it sounds sexy. Antibiotics are apparently not sexy. Neither are vasopressors, but dammit we will RSI. Prehospital Ancef for open fractures is around in Colorado, mostly flight but several rural ground services as well.

Anyway,
Greenville EMS in NC has done a fair bit of research on their own pre-hospital sepsis guidelines. Most of it is retrospective, but if nothing else they found that they could draw cultures about as well their hospitals could and that their providers were correctly identifying patients that would benefit from prehospital antibiotics. I've attached some of their materials (all are public record).

 

[EpiEMS]

Prehospital Ancef for open fractures is around in Colorado, mostly flight but several rural ground services as well.

I'm intrigued by this. I know that TCCC has recommendations on prehospital antibiotics, but I would be surprised if there is a huge advantage in administering antibiotics prehospitally in the civilian setting, unless you're dealing with very long transport times. How long are these rural ground services driving?

 

[Tigger]

I'm intrigued by this. I know that TCCC has recommendations on prehospital antibiotics, but I would be surprised if there is a huge advantage in administering antibiotics prehospitally in the civilian setting, unless you're dealing with very long transport times. How long are these rural ground services driving?

A 90 minute transport time is not uncommon for us, just from a house. Generally we can't take open fractures to the local Critical Access Hospital as the injuries associated such a condition is beyond their self stated surgical capabilities. We have air, but it's the mountains soooo... Most of these protocols are designed for backcountry calls though. While not a particularly high volume, it might take us two hours to reach an ATV accident and then another hour to move the patient to a helicopter, if that is even possible. Sometimes it's not necessary either, if you have an isolated ortho injury we'll take you buy ground even if it takes three hours. Services in the "real" mountains often provide medics to SAR teams and they could be spending 6+ hours with these patients.

 

[Peak]

Cultures drawn before antibiotics IF it will not delay antibiotic therapy.

Honestly there is no reason why we can't get cultures before antibiotics. With the correct training cultures can be drawn anywhere, there isn't some great mystery to them. Even a few drops of blood with grow out in a culture, though it is far from ideal, and we should be getting some blood from IV starts. A second set of cultures is nice but we should have at least one, and I have drawn second sets either with their ABG or when starting an A line.

imagine most EMS systems are already comfortable bolusing fluid.

While most systems will start fluids they are not starting a 30 mL/kg bolus on adults. Flipping through the two sets of protocols most of our patients come over one states that patients should receive 30/kg before pressors but does not directly indicate a 30 mL/kg bolus, the other states that adults should receive 20mL/kg in sepsis. Typically unless the patient is hypotensive they come in with one bag of NS strung up and hanging to gravity.

 

[Peak]

I'm intrigued by this. I know that TCCC has recommendations on prehospital antibiotics, but I would be surprised if there is a huge advantage in administering antibiotics prehospitally in the civilian setting, unless you're dealing with very long transport times. How long are these rural ground services driving?

The current recommendation is to start antibiotics for open fractures within one hour of identification, so even if transport was only 30 minutes when you look at the time it takes to prepare antibiotics prehospital administration can be beneficial.

 

[Brandon O]

Honestly there is no reason why we can't get cultures before antibiotics. With the correct training cultures can be drawn anywhere, there isn't some great mystery to them. Even a few drops of blood with grow out in a culture, though it is far from ideal, and we should be getting some blood from IV starts. A second set of cultures is nice but we should have at least one, and I have drawn second sets either with their ABG or when starting an A line.

There is no great mystery to cultures, but some patients are difficult to get any blood from, and getting STERILE cultures takes some doing -- I presume even more so in the field.

To me a sincere attempt at a sterile culture requires sterile gloves, a chlorhexadine prep, a hat and mask, and a fresh stick. You can draw from an IV or other line if it is the FIRST thing that passes through that catheter, but not once it is already in place -- and there is actually some data that cultures from IV sticks are more likely to be contaminated even if it is a fresh placement. Two sets is close to essential, both to reduce false negatives (missed it) and false positives (one was contaminated). And a few drops is not enough; yield from blood cultures is directly linked with the volume of blood obtained. 10ml per bottle should generally be the goal.

This problem is probably not as meaningful the earlier you get in the chain of care, but you should see us up in the ICU scratching our head over what to do about the culture growing out coag-negative staph (i.e. skin flora, usually a contaminant, sometimes not). It increases healthcare costs and unnecessary workups and treatment.

Giving antibiotics in the field may be easier than drawing worthwhile cultures.

While most systems will start fluids they are not starting a 30 mL/kg bolus on adults. Flipping through the two sets of protocols most of our patients come over one states that patients should receive 30/kg before pressors but does not directly indicate a 30 mL/kg bolus, the other states that adults should receive 20mL/kg in sepsis. Typically unless the patient is hypotensive they come in with one bag of NS strung up and hanging to gravity.

I imagine most hypotensive patients will get bolused in the field. Anyway, this is a separate issue from the question of whether to give antibiotics.