Adrenal insufficiency - protocols in the future?

Or, if you work in states other than just the Nutmeg State, the present!

One development in EMS in the past few years has been the push in many states to get prehospital providers to carry and administer steroids to people with various forms of adrenal insufficiency. Unlike cardiac arrest, intubation, or cervical spine injury, there hasn't been a great deal of talk in the EMS blogs about this subject, so I was hoping to address that deficit!

First off, why should I care about adrenal insufficiency?

Well, one answer is that some one is going to make you care. Most likely, that someone will be the CARES Foundation. This group was started by the parents of a child with congenital adrenal hyperplasia (CAH) (relax if you aren't familiar with CAH - we don't see it very often in the ED!). Back in 2009, they started focusing on the role of EMS in treating children with CAH. If we take a look at this map of the United States that shows which states have EMS protocols for treating CAH crises:

We can see that the advocacy efforts of the CARES Foundation has been quite successful in changing protocols in the northeast part of the country, with one small exception - Connecticut!

The more astute readers will say - "Hey! Out of every 16,000 live births, only one will have classic CAH! That doesn't sound too common." Well, yeah, you're right. On the other hand, there are a lot of people out there in the world with adrenal insufficiency, and they can get into the same problems.

Sooo... What is CAH, adrenal insufficiency, and stuff like that?

Let me explain the two forms of adrenal insufficiency you'll deal with. This is a crazy-deep topic, but there will be basically two kinds of steroid problems you'll see.

1. Little kids with a known diagnosis of CAH, and,
2. Adults with a diagnosis of adrenal insufficiency who tell you that.

I.e., this is how you diagnose adrenal problems these in the field:

Let me give a very brief review of the physiology of adrenal problems. So, this gets complex fast, but there are some main points. Here's a picture of how steroids get made in the body.

First, the body gets an infection, or gets shot, or even just sees a scary movie - somehow gets stressed out. Time to release the stress hormone - cortisol! The body follows a Rube Goldberg series of steps to make cortisol:

• The hypothalamus makes CRH. 
• When some of that CRH drifts by the pituitary, it makes some ACTH. 
• In turn, the ACTH triggers the production of cortisol from the adrenal gland.
• Eventually, the elevated cortisol levels shut down (temporarily) the production of CRH and ACTH.
• When the cortisol levels settle down, the process is ready to go again!

Kids with congenital adrenal hyperplasia (CAH) have most of the system in place and working, except for the last part - their adrenal cortex can't make the cortisol (and some of the other stuff). These kids are dependent for the rest of their lives on steroids. It's somewhat like diabetic kids, who are insulin-dependent. These kids are cortisol-dependent.

Adults can a variety of different problems with their steroid-making process: Maybe the have Addison's disease, like a former president did. Maybe their adrenal glands eroded during an infection. Maybe they're taking mondo amounts of steroids to prevent transplant rejection, and their body has "forgotten" how to make it's own steroids now. These folks also are on chronic steroids.

Nope. Try again.

So, generally, as long as the kids and adults are taking their "usual" daily dose of steroids, everything goes fine. Usually.

Okay, it sounds like we have the problem solved then!

Er, almost. The problem is that the body regulates the ebb and flow of cortisol over a wide range of levels. For example, your morning level is about 10 times higher than your midnight level.

When you get sick, no one is quite clear what the right level should be. When kids get sick, parents are taught that their daily dose of pills should be significantly increased, and are given detailed instructions (PDF download) about when to, say, double or triple their usual dose.If things get real bad, they have a emergency steroid injection kit, and the parents have practiced how to draw up and give the shot.

Good, the parents are all over it. What's left for us?

Well, even fairly small kids can't hang around their parents all day, every day. They may get stuck with. well-meaning relatives, day-care, or even a lowly babysitter. While the parents may be up to speed on the the when and how of injected steroids, it's asking a bit much to expect 16 year-old sitter to do that too.

Not what she signed up for.

Instead, she'll call 911, and shove the pamphlets, medic-alert stickers, and meds at you. Your turn.
 Naturally, you know all about assessing acute adrenal insufficiency, right?

What do adrenal-insufficient people in a crisis look like?

First off, these people already have a history of chronic adrenal insufficiency. That begs the important question: What exam finding indicates a patient has chronic adrenal insufficiency?

Answer

Next part in assessment is figuring out what else is gong on. Something had to trigger an episode of acute insufficiency; there are diverse causes, but all of them stress the body in some way. The general causes, though, are: Stopped steroids abruptly, fever/infection, a large trauma, vomiting. The New Hampshire protocols, for instance describe which patients to treat with specific therapy:

Patients will look shaky, gray, sweaty. The may be hypotensive and hypoglycemic as well.

The paramedic should look at the precipitating events, and treat those, whether it's a trauma, an asthma attack, a burn, a bad episode of gastroenteritis, or pneumonia. Second, they should address the symptoms of the episode: Zofran for vomiting, dextrose if they're hypoglycemic, IV fluids for hypotension. An EKG could show signs of hyperkalemia, and you should be ready to ask med control for calcium chloride.This is all stuff you should be doing now.

The specifc therapy for acute adrenal insufficiency is hydrocortisone, or Solu-Cortef:

Adult dose = 1 box, IM, IV, or IO.

Now, CT does not have any protocols so far regarding this medication. You can check out the protocols that RI, MA, NH, and NY have, though. (Those are all PDF downloads, BTW)

So, be ready for the change when it comes!

Hypoglycemic patients, and refusal of transport.

It's a fairly typical ambulance run - you're called for "low blood sugar," and on arrival the wife points you to a sweaty, pale man who is shaking and speaking nonsense. The finger stick is "LO," but a box of D50 brings him around fairly quickly. You soon determine that he had administered his insulin just prior to lunch, but some distraction prevented him from eating on schedule.

No problem sir, just sign right here!

Probably not the best AMA form to be using...

For such a common practice, you would imagine that this would be a fairly safe encounter. Most EMS providers are careful to ensure that the patient will another person with them, that they eat a meal before the ambulance leaves, etc. Nonetheless, there are still a number of potential pitfalls that await the unwary paramedic.

Let me point out 5 important aspects of the "routine" hypoglycemic patient encounter:

1. In general, this is a safe practice that patients are very satisfied with.
2. Hypoglycemia due to oral hypoglycemics can be prolonged.
3. Accidental ingestions of oral hypoglycemics by children, even if they are just suspected or possible, must always be transported, and admitted to the hospital.
4. Hypoglycemia associated with long-acting insulin is also dangerous, and should also be transported.
5. After a hypoglycemic episode, patients are especially vulnerable to a repeat episode, often showing symptoms at a higher glucose level.

Signing off patients is generally safe.

A number of studies, using a variety of methods and protocols, have shown that treating patients at home, and permitting them to refuse transport is a safe and popular practice. Studies conducted in the U.S. have more often used protocols that used an on-line medical control physician to talk with the paramedic, and often the patient as well. They also have used explicit inclusion criteria, and formal AMA paperwork. One study out of Buffalo used this inclusion check-off sheet:

Patients who met the criteria, and who had discussed refusing transport with the medical control physician, were given this form:

This level of documentation is very thorough, and likely prevents many potential lawsuits. The down-side of this elaborate protocol was that, over 3 years, less than 1/3 of the medics in that EMS system enrolled any patients, and only 36 patients total were enrolled.

Contrast this with one Canadian study, where only 34% of hypoglycemic patients ended up being transported. There was still a requirement for the paramedic to contact medical control, but there were no extensive inclusion criteria to follow. Instead, clinical discretion was expected of the medics. Perhaps that's why, in less than 1 year, 145 patients were found to decline transport - a far higher number than in Buffalo. Nonetheless, the practice was very safe, with recurrence of hypoglycemia happening just as often in the transported group as the refusals.

Even further from the U.S., both in protocol and in geography, is Finland. Researchers examined how well the protocols were working for them. Paramedics only needed to transport if patients "did not regain consciousness, were not able to eat after treatment, were pregnant, had type II diabetes treated with oral medication only or oral medication in addition to insulin, in case of insulin overdosage and variably in case the patient would be left alone or was a child (physician consultation)." In general, paramedics did not contact base physicians. Only 10% of patients ended up being transported!

Hypoglycemia due to sulfonylurea drugs must be transported.

Note that in the Finnish study, despite the willingness to leave many of the hypoglycemic patients at home, they required transport for patients who were taking oral hypoglycemics. Why was that?

Two reasons: duration of action, and underlying causes.

First, you must understand that not all oral diabetic medications are alike. There are five classes of oral medications: sulfonylureas, meglitinides, bigiuanides, alpha-glucosidase inhibitors, and the thiazoladinediones. While they all act in some way to prevent hyPERglycemia, only the first two classes are able to cause hyPOglycemia.

The most common sulfonylureas are glyburide, glimepiride, and glipizide, and they all have pretty long duration of action (24+ hours for both glyburide and glimeride). This means that if the patient accidentally took 2 or 3 pills extra today, 1 box of D50 and a PB&J sandwich may only temporarily bump the glucose, before the levels ground out again. This may not be a patient you can safely leave at home - they may have trouble for the next few hours or days!

Certainly, skipping a meal, or running for an hour longer may also promote hypoglycemia. These causes will be clear from the history. However, there are other causes of hypoglycemia associated with the oral medications, and the history may not be obvious.

Such causes are:

• In patients who are undernourished or abuse alcohol
• In patients with impaired renal or cardiac function or gastrointestinal disease
• Concurrent therapy with salicylates, sulfonamides, gemfibrozil, or warfarin

These comorbidities often play a role in hypoglycemia. If you have a patient taking glyburide who has had an episode of hypoglycemia, realize that this may be just the "tip of the iceberg." We will end admitting a good chunk of these patients, so don't leave them at home. The little evidence we have about the safety of not transporting is not encouraging.

Speaking of the oral medications; this isn't strictly on-topic, but while we're on the topic...

The suspected ingestion of a sulfonylurea by a child gets transported.

Obviously, if a child is symptomatic from such an ingestion, they will be transported. The danger lays with the child whom the caretakers think may have taken a pill or two out of grandmother's purse. Despite normal glucose levels at the time of evaluation by EMS, patients have been known to develop hypoglycemia many hours after ingestion. This delayed effect seems to happen when children are given dextrose or food to prevent hypoglycemia, but it just puts it off for a few hours.

These are simple cases - they will all be admitted, even if there is just the suspicion of an ingestion.

"Kid and glyburide means ambulance ride."

Hypoglycemia associated with a long-acting insulin should be transported.


Lantus (insulin glargine) was approved by the FDA in 2000, and was the first insulin that was effective for 24 hours. This was a great therapeutic development, providing patients a more natural delivery of insulin. Levimer (insulin detemir) is the other commonly-used long acting insulin

Unfortunately, most of the good studies that looked at EMS "treat and release" protocols were conducted either before, or soon after, the release of Lantus. As a result, it is unclear if it would be safe to allow patients on these long acting insulins to refuse transport - we just don't have the data yet.

In the ED, these patients will usually be admitted, so don't think too hard about transporting!


Acute hypoglycemia predisposes patients to more recurrences.


Episodes of severe hypoglycemia seems to make it easier for a patient to have more episodes in the short term. Patients seem to lose both the ability to feel when their sugar is low, and a weaker capability to send out catecholamines to raise the sugar.

As a result, patients may have symptoms of hypoglycemia at a lower glucose level than before. If they developed an altered level of consciousness at a level of 55 mg/dL the first time, they may need to drop down to 35 mg/dL the second time.

For that reason, especially if patients decline transport to the hospital, they need to be educated that they should keep their glucose levels higher than usual for the next few days. In truth, it could take a few months of blood sugars in the 160-180 range to significantly improve their hypoglycemia awareness.

What do the Greater Bridgeport protocols say?

Glad you asked.

This protocol can't cover all the possibilities - we need an educated prehospital practitioner in the field to look at the scene, the patient, and get the history. But when you call in, you'll now have a much better idea of the things we're listening for, and the ways we can avoid the pitfalls.

ResQ-Pod 2: The story, and device, continues...

ResQ-Pod 2: Pod Harder.

 This study caught my eye caught my eye because the lead author is Stephen Smith, the author of one of the best ECG blogs out there. He also writes for other blogs, and at least one EMS ECG blog follows his work. Dr Smith is, in short, the man.

 The study also caught my eye because it uses a retooled version on the ResQPOD, now called the ResQGARD. I know a lot of EMS folk swear by the ResQPOD, but the recent evidence has not proven its value. So, it's interesting to see "part 2" of the ResQPOD saga.

Because Part 1 worked out so well.

The ResQGARD works on the same general physiologic principle as the ResQPOD. It allows for normal, unimpeded exhalation, and does not provide any PEEP. During inhalation, however, it slightly increases the force required to draw in a breath. The actively expanding thorax normally acts as a sort of "suction" to also pull blood up from the belly, but with this added resistance to the air inflow, this "suction" effect is magnified.

And there is some animal and human data to back up the claims for usefulness in treating hypotension.

Red means more blood, I guess.

This device is evidently "cleared" by the FDA for treating "low blood circulation," and various studies have shown an ability to raise the blood pressure in, for example, blood-donors, or in other models of hypovolemia or hemorrhage. You can check out some background device in this article from Journal of Special Operations Medicine, the coolest journal that you aren't reading yet.

"Oh, is that JEMS you're reading? That's cute."

This trial had two parts. In the first part, the device was tested in a randomized, controlled, and blinded fashion in the emergency department for patients with hypotension due to various causes. The primary endpoint was the maximum change in SBP over the first 10 minutes after placement of the device. They enrolled 47 hypotensive patients. These patients ended up with diagnoses of dehydration, sepsis, or hemorrhage most of the time. You can see the average change in SBP in the table below, broken down by the cause of the hypotension.

When they looked at the overall results, they found that patients who had the real device had an average rise in SBP of 12.9 mmHg, while patients who got the sham device only had a rise of 5.9 mmHg, a difference of 7 mmHg. They tell us that the difference is statistically significant.

In the second part, the device was used by EMS, with no control therapy, in an unblinded manner. It's not much of a study, and they were really only looking at "feasability" of using the device by prehospital providers. They also had 47 patients in this arm of the study, and they determined that, yes, it was feasible to use, and that patients tolerated the device well.

It's hard to interpret the data on the change in blood pressure, etc., in part 2, since, as we saw in the results of the first part, the average blood pressure tended to go up with or without the device. The bar graph below shows that the pressure came up to a statistically significant degree, but it can't tell us if this was better than doin' nuthin'.

So, what can we take from this paper?

In the end, not much. Let me list the reasons why:

1. The difference in SBP is statistically significant, but unclear if clinically significant.
2. Some of the causes for hypotension have established, beneficial treatments.
3. Some of the causes require no treatment, and improve on their own.
4. The majority of the literature supporting the use of this device is written, in part, by the inventor of the device.

The average difference in systolic blood pressures was 7 mmHg, which is pretty small. How much do you care about raising the pressure by 7 points? In addition, it looks like almost everyone's pressure went up somewhat, with or without the device. Remember that 7 represents the average difference - some patients didn't improve much with the device, and some people had a big jump in pressure with the fake device!

In another article about the ResQGARD (or ITD-7) written by Smith, the authors present this table:

"Treatable" ≠ "Proven effective for"

Let's look at the causes they list. First off, heat stroke is about heat load and mental status - the primary treatment is cooling, and hypovolemia is usually not a significant component of the problem.

The proper treatment of true dehydration is, well, hydration. As for sepsis, while hypotension is a manifestation of the problem, there is a large amount of evidence that large amounts of IV fluids, delivered rapidly, saves lives. Simply raising the pressure through other means is not appropriate.

Regarding hemorrhage, there is a thicket of controversy about optimum treatment. While much of current practice emphasizes normalizing the blood pressure, a lot of evidence suggests that "permissive hypotension" may be the best (non-)treatment. Where the ResQGARD falls in this area is not clear at all.

Lastly, orthostatic hypotension is usually transient, and requires no intensive therapy; e.g. juice & cookies after blood donation.

Let me speak of the appearance of bias in the studies supporting the ResQGARD. The inventor of the device is Keith Lurie, a cardiologist. I have no doubt that he has aspirations to advance medical science and save lives. Unfortunately, as the inventor of the device in question, and the owner of the company that sells them, he has a vested interest in selling the device. And, while it's not the most expensive medical device out there, it costs real money.

Look at the references listed in the paper. Of the 23 studies that Smith et al. provide as references, 20 had Dr. Lurie as a co-author. That's a real conflict of interest. By way of example, check out this intervew that appeared with Dr. Lurie in an EMS blog. He had an interesting take in the failure of the ResQPOD to show an effect in the ROC trial.

Interviewer: "I think many of us who have been following the ResQPOD were surprised by the recent announcement by the National Institute of Health that the ROC PRIMED trial was stopping enrollment. ...  Considering that the ROC PRIMED trial was a prospective, multi-centered, randomized clinical trial with large enrollment, are you concerned about these results?

Dr. Lurie: "To directly answer your question, I am not concerned with the results, nor am I surprised."

To sum up: The ResQGARD appears to have a statically significant effect in hypotensive patients, but the clinical effect, as well as the appropriateness of this therapy, are unclear. I don't think any EMS service should be stocking up on these yet.

Why paramedics need to read EKGs, and not just read interpretations.

Two patients, 2 EKGs, and 2 very different stories! Both these patients came into the Bridgeport ED, 1 of them by EMS. One of them went to the cath lab, while the other got some tests and a sandwich.

So, here's the EKG for pt #1, a 40 y.o. male who was complaining of some brief, transient, and sharp chest pains:

And for Pt #2, a 60 y.o. male who had drank a 12-pack by himself, vomited, and now complained of chest pain:

So, who got the expensive metal in their coronary artery, and who got the expensive sandwich? Which one would you have called in as a PAMI?


One of the prominent roles for EMS is in identifying the ST-elevation myocardial infarction. The basic idea is simple ("time is myocardium"), but trying to decrease the time from symptom onset to balloon inflation is pretty complex, and involves multiple decisions, actions, and environments.

Now, the role for EMS is seemingly straightforward - give the aspirin, get an EKG, and call up the ED early if you have a potential cath lab activation. Heck, the protocol seems fairly black and white.

But our patients were not put on Earth to follow the protocol, were they?

Now, pt #1 would seem like an easy cath lab activation, but a close look at the ECG suggest benign early repolarization, or perhaps pericarditis, as the most likely culprit. He had a mess o' ECGs and troponins over the next few hours, and nothing came up funky. He appreciated the sandwich!

Pt #2 had the not-so-helpful computer interpretation; "Cannot rule out Anterior infarct." Logically, you can put that on every ECG you record, since you rule out an MI with sequential troponins, over many hours! But despite the patient's intoxication, this ECG is actually classic for a bad problem, an occlusion in the proximal LAD. When we got this ECG right after he rolled into room 5, I immediately activated the cath lab. The interventionalist found a complete occlusion of the LAD just proximal to the first diagonal.

D1 = First diagonal. Lotta real estate downstream from there!

If you take another look at EKG #2, there are impressive ST-segment depressions in multiple leads, and multiple regions. Up until now, the conventional wisdom had been that this was an NSTEMI, and did not require emergent intervention.

Research over the past 10 years has changed our perspective, however. The key is in aVR, the "forgotten lead." The definite ST-segment elevation in that lead suggests an occlusion of the proximal LAD, perhaps even the left main coronary artery. If you look at the picture above, you see that the Left Main segment is responsible for >75% of the blood supply to the left ventricle. Widow-maker, indeed.

This isn't in the guidelines - yet. In one recent article, a group of cardiologists and emergency physicians suggested additions to the currently accepted cath-lab activation criteria.

Let's focus in on that last one...

So, this is all proposed stuff for now - what do you do tomorrow when you're bringing in a guy with this EKG?

Hint: Not just drunk.

Well, follow the rest of the SHCGB guidelines - ASA, IV, O2, monitor, and grab some more ECGs during transport, especially if the symptoms change. Talk to triage about bring this this patient to room 5; they may not end up going to the cath lab, but I would prefer to see that patient sooner than later!

Prehospital blood pressures

Thanks to Emergency Medicine Literature of Note for directing me to a new article. It's a great blog, and Dr. Radecki makes it easy to feel current, and in bite-sized posts!

One of his recent posts reviewed the new study "Agreement between emergency medical services and expert blood pressure measurements." It's an interesting study to read, and it must have been fun for the research assistants (RAs).

The methodology was pretty elegant. In case you didn't read the abstract above, I'll hit the high points. The researchers wanted to assess the accuracy of BPs obtained by paramedics. 

In Phase I of the study, the RAs took blood pressures on EMS-transported patients just after they arrived in the ED, and compared those measurements with the last BP obtained in the field. In Phase II, they actually rode on the rigs, and took BP measurements simultaneously with the paramedics. Additionally, the RAs interviewed the medics about their measurement technique (phase I) and observed their techniques in the field (phase II).

Now, they used as the criterion of blood pressure measurement technique a document published by the AHA in 1993. It lays put exactly how wide the cuff should be, which of the 5 sounds should be used for measurement, and so on - pretty standard stuff. Of course, on a daily basis in the ED, most of it is observed only in the breach! For example, take this passage:

"Ideally the measurements should be made after a period of rest in a quiet, relaxed setting, not immediately after exertion or ingestion of coffee or during conversation; the legs should be uncrossed, with the feet resting firmly on the floor, not dangling, and the back supported, because any form of isometric exercise during the measurement will transiently raise the blood pressure level. Blood pressure levels are affected by environmental, emotional, and physical stimuli, so every effort should be made to standardize the conditions of the measurement, keeping extraneous influences to a minimum. Anticipation of pain or anxiety about the procedure and its outcome can raise the blood pressure level and potentially lead to overestimation of the usual blood pressure levels."

You can fill in here your own mental picture of the conditions under which most BPs are obtained in the field. Obviously, any prehospital BP measurements will far fall short of these "gold-standard" conditions, but it does make for amusing reading. Most of the BPs obtained in rooms 4 and 5 in my ED would not meet this high standard, I'm afraid.

The Phase I and Phase II results are summarized/combined in this graph:

In this graph, the measurements of the RAs are those along the x-axis, while the y-axis depicts the amount that the paramedics' measurement differed from the RAs'. For example, you see the blue dot right above where it says "150" on the x-axis? That dot means that the RA got a SBP of around 145, while the medic got a pressure about 45 points lower, or 100mmHg. And that blue dot sitting right above the 100? Evidently the medic measured a SBP of around 65 mmHg. Seems pretty bad, right?

Well, blood pressures change, they go up and down. Heck, even if the BP in the ED is normal, a lot of literature has found that prehospital hypotension predicts some bad things. (That's four links right there. Go check 'em out right now!)

Okay, how bout the Phase II results, where they measured both arms simultaneously? Here, it seems to spell bad news for the basic skills of the ALS providers. They found that:

"59.1% of the systolic measurements and 63.9% of the diastolic measurements were ≤ 5 mm Hg different than the expert measurement. EMS systolic and diastolic measurements showed a difference > 10 mm Hg from that of the expert at a rate of 13.6% and 13.9%, respectively. Nearly all measurements were within 15 mm Hg of each other." (My emphasis)

Hey fun fact: What percent of patients in the ED have a difference in SBP between their two arms of over 20 mm Hg? That is, if you took simultaneous measurements in both arms, just like they did in part 2 of this study, how often would the SBP be different by 20 points?

Turns out, about 20% of the time!

Singer and Hollander looked at this issue back in 1996, and found substantial differences when simultaneous arm BPs were measured.  In fact, one could surmise that the differences in blood pressure in the current study, when measured in both arms simultaneously, were actually closer in agreement than would have been predicted.

I applaud the researchers for tackling the prehospital environment, and for addressing basic, but essential, skills. Three things in particular concern me, however, about how this study was conceived and delivered.

1. I'm not sure why the research assistants are deemed "expert," and their readings taken as the gold standard. A true standard would be an intra-arterial pressure (as noted by Singer and Hollander). Short of that, I believe we can only say that the measurements between two people differed. As the methods are worded, they may be taken as somewhat hurtful to the pride of experienced prehospital providers.

2. The premise for the whole study is that these possible errors in BP measurement are material, that they matter. Even the capsule summary by the editors agrees with this premise, stating "Prehospital protocols often call for administering drugs to treat elevated blood pressure, or are otherwise dependent on an adequate blood pressure measurement." First off, I haven't heard of any protocols that call for lowering blood pressure, and it sounds like a bad issue for EMS to wade into. 

3. I agree with the editors on one point - we need "adequate" BP measurements, meaning it is good enough for the purpose for which it is used. I don't need to know if a SBP is 150 or  180 before I give a NTG tab, 'cause either is fine. That's an acceptable amount of error for me, in that context. Now, if the patient is pale as heck, sweating, and the SBP is 95 mmHg, I just got a whole lot more interested in accurate measurement of the BP, and I may even take it myself. I expect that most paramedics feel the same way!

Prehospital sepsis - new research

-->

Sometimes, it can be pretty discouraging to keep up on the EMS literature. Seems like everytime you hear of a new study, it’s showing that something that EMS does is either harmful, or just doesn’t matter. 

MAST, pediatric intubation, ACLS drugs, IV fluids for penetrating trauma – it’s enough to make a proud medic hang up her laryngoscope. 

This is no longer the standard of care?!

In a nice change of pace, a couple recent studies point out the value of EMS in treating sepsis, an area that hasn’t had much emphasis in EMS education and protocols in the past. Even better – the results show that EMS is probably making a difference, even without trying!

These recent publications tie in with a patient that AMR brought us in Bridgeport recently. The paramedic called med control, looking for some advice. He was bringing in an older gentleman from an ECF, and he was reasoning out some antiarrhythmic therapy. We talked on the radio some about the rhythm, but the most important thing he got across to me was that this guy was sick looking. That’s enough for me, and I met them at the door and we got things rolling. Fever, hypoxia, altered mental status, nonsustained VT – yep, he was sick! Three liters of NS and 3 antibiotics later, the ICU team was wondering what all the fuss was about.

Of course, if EMS hadn’t identified this guy, he might have slipped through triage, as his vital signs looked OK at the moment he came through there. We would have noticed when his pressure or sat dropped, but we would have been behind at that point, delaying his antibiotics and fluids for some time. 

One recent paper, hot off the presses, may illustrate the benefit of EMS in cases like this. 

The folks at U Penn wanted to see how EMS was affecting the care of these kinds of patients. They already had a bunch of data on the severe septic patients they were treating in their ED, and so they examined the records to see which of them were brought in by EMS.

What they found was that EMS only brought in 41% of the bad sepsis players. Of course, they were sicker than the 59% that came in by other means. The EMS patients were older, had higher initial lactate levels, and higher APACHE II score (rating system for critical illness). The EMS patients were also far more likely to be African-American, which usually predicts worse outcomes, no matter what ailment you’re talking about.

The two critical interventions in the treatment of sepsis are IV fluids and antibiotics, and the EMS patients got these in the ED about 40 minutes sooner than the “walk-ins.” That’s pretty big, considering that, in the ED, we're scrambling to get the antibiotics in within the first hour! 

No, they couldn’t show that EMS was saving any lives, but this study wasn’t designed to show that. But that really isn't the point either. Nowhere in the paper do they describe any specific education that EMS had, any QI process, or any change in protocols. EMS was just doin' what they was doin'.

Well, okay, maybe this isn't a big deal. We all know that using EMS is a great way to skip the line at triage if the long wait at the ED bothers you. Heck, maybe toe pain gets seen faster if it comes in by EMS. Is there any evidence that having a conscious EMT in the back of the rig changes anything?

In 2010 year there was another paper that showed similar findings, and added some others. 

The first author is hiding his paramedic roots.

The ED at Carolinas Medical Center looked at a similar group of ED patients, bad sepsis, and wanted to see if the people who came in by EMS got fluids and bug juice sooner than non-EMS arrivals. 

Just like the first paper, the EMS patients were a sicker bunch, but they got treated sooner. Interestingly, they had the same approximate differences in time to treatment – 40 minutes – that the U Penn group found.

The really neat thing about Studnek's paper was that the researchers also checked to see what the prehospital providers had written on their run-sheets. If the medics wrote the word “sepsis” somewhere in their impression, it turns out that the times in the ED for those patients were even better. Those EMS-identified patients got their antibiotics about 50 minutes earlier, and had resuscitation started an hour earlier! Again, just as in the preceding study, the researchers made no mention of any specific EMS sepsis protocol or pre-arrival alerts.

I'm fascinated by the results of these investigations.

For example, why would a medic record an impression for “sepsis” if there weren’t a corresponding protocol or pre-arrival alert? I'm imagining an analogous study: picture a system where the EMTs received basically no education on heart problems, angina, MIs - nothing.

As in, pre-Johnny and Roy

What if we then looked at the patients who had STEMIs, and then went to check if arriving by ambulance affected the door-to-ballon times? Furthermore, what if the patients who had a prehospital diagnosis of "heart trouble" got the cath lab even faster? That's kind of what is going on in these studies.

This is the benefit, I believe, of having educated and aggresive prehospital providers. The results above don't seem to reflect any new QI project or protocols. However, I think that Basics and medics are often at their best when they are “off protocol.” The medic who brought in my septic patient was calling med control to discuss treatment for salvos and runs of VT, but he was really calling in to tell me he had a “sick-as-something” patient.

Of course, we don’t have a “SAS alert,” but he was doing what he could with the system we’ve set up. When I was a resident up in New Haven, I had a medic call in a STEMI alert. When we looked at the ECG, however, it certainly did not meet any of the standard criteria for an activation. Looking at the patient, however, he certainly met classic SAS definitions – sweating, pale, cold, screaming in pain. A closer look at the ECG showed a classic left-main occlusion pattern. The medic had never heard of this rare pattern, but he knew that the ECG had some ugly things on it, and that the guy was circling the drain.

Bridgeport, I'm calling in with a CTD alert.

The potential for EMS to be involved in the care of sepsis is huge. Jeez, look at how much effort we’ve put into cardiac arrest, and for what? We might be just as well off with taxis and AEDs! (Insert a winking emoticon here…) 

Sepsis is a whole new area for EMS to show its worth, to make a difference. There is going to be more research on this; findings ways to identify septic patients in the field, start interventions, calling alerts.

One request in the meantime, though: Just don’t call in any SAS alerts for now!

***Late breaking*** 
Dug up this study, buried deep in the JEMS website:

 Decreasing Blood Lactate Levels in EMS Patients
By T. Ryan Mayfield, MS, NREMT-P; & Mary Meyers, MHA, EMT-P

Introduction: Research has shown that clearance of blood lactate is associated with better outcomes in patients with severe sepsis and septic shock. One of the primary treatments of these patients is administration of IV fluids. This study looked at blood lactate levels before and after EMS treatment to determine if there was a significant change.
Hypothesis: There will be a change in blood lactate levels between EMS and hospital lactate levels.
Methods: Paramedics were provided with and given training on the Lactate Pro blood lactate meter by Arkray Inc. This meter is FDA-approved and CLIA waived, and has shown a good correlation to hospital lactate tests. Between May 1, 2009, and Sept. 15, 2010, 134 patients with suspected severe sepsis or septic shock underwent blood lactate readings by EMS. Patients with a lactate reading of ≥ 4.0 mg/dL were considered to be in shock regardless of their corresponding blood pressure. Treatment was not dictated by this study and was administered according to EMS protocols.
Results : Of the 134 patients, 120 had hospital lactate levels available for comparison. Overall, hospital lactate levels were lower after EMS treatment. EMS patients were divided into groups that received greater than 1000 mL of fluid between readings (Group A), and patients who received between 250 mL and 1000 mL (Group B). Group A had a median decrease of 2.25 mg/dL (p = 0.0003) while Group B had a decrease of 1.1 mg/dL (p < 0.0001). Analysis used the Wilcoxon-Rank Sum Test.
Conclusions: There was a significant decrease in lactate levels associated with EMS treatment. Further, the group that received greater amounts of IV fluids had an even larger drop in lactate levels. These results illustrate the importance of EMS treatment and how it might impact patient outcomes. Further research and training needs to be done to expand the role of lactate in EMS, as well as reinforcing the importance of fluid administration to these patients.

Use of the Straight ("Miller") blade

A lot of people are talking about the future of endotracheal intubation for EMS, about when it should be done, allowed, or even if it should be taught. This is NOT one of those discussions.

Also not a discussion of the utter nonsense shown in this picture.

If you are going to intubate, however, you should know how to do it. And it shouldn't be just one way to do it, but more like 2 or 3. 

On the other hand, I started with the Mac #3 as a new medic, but quickly became a "The-#4-Mac-has-never-let-me-down" kinda guy. Still am, even as an ER doc.

Candid confession - I have no talent for the Miller
I'll be honest though. One of the chief reasons I'm a "#4 Mac guy" is that I never learned to use the Miller effectively. 

Sure, people say that "It's better for peds," or "It's better for trauma intubations." But I've had trouble sweeping the tongue with the tiny flange when I used it like a Mac, placing it on the right side, and trying to move the tongue to the left. 

Placing it in the midline was even worse; I had no control of the oropharynx, and the tongue would just flop around. And it didn't even make sense to me - if moving the tongue with a curved blade didn't give me the view I needed, why would smooshing the tongue help?

A completely mythical view of the cords.

Turns out that my instincts were not far off the mark! A number of anesthesiologists have come up with a better techniques for use of the straight blade for difficult intubations, avoiding any tongue-control issues, as well as providing clearer views of the cords.

Paraglossal approach with the Miller
There are some variations, but they all start with proper positioning of the head and neck, either in "sniffing position," as with your medical patients, or in neutral, in-line stabilization for your trauma patients. 

Much like a Mac, you place the blade into the corner of the mouth, and advance it along the groove between the tongue and the tonsil ("paraglossal"). Then, however, things go a little differently.

Levitan, on his excellent AirwayCam website, describes the paraglossal approach:

"Proper position is achieved with straight blades by deliberately directing the blade to the right paraglossal space. No tongue should be present to the right of the blade.  Full insertion of the blade should occur through the right lateral mouth, over the molar dentition, and while the distal blade may then be directed medially, the proximal blade should never be brought back towards the midline, otherwise it will hit the central incisors.

After the epiglottis edge is identified, the handle must be tilted forward (e.g., the tip moves backward, toward the posterior hypopharynx). The blade is then inserted slightly farther (~1-2 cm), and the tip passed under the epiglottis.  Once the epiglottis is “trapped” under the blade tip, the blade is rocked slightly backward (handle brought slightly more upright) and then the lifting force increased."

Note that the blade is to the right of the nose, and that the ET tube runs under (not through) the blade lumen.

"Tube delivery should be done using the extreme right corner of the mouth, and come up from below the line of site.  An adult tube will not fit through the lumen of a Miller blade (and should not be attempted)."

Here's a drawing of the technique, from a key article by Henderson:

Blade stays on the right side of the nose.

Look at that finger hooked into the corner of the mouth - that's a real helpful technique in normal intubations, and it's key here. You are not moving the blade to the left, so you need some help on the right to insert and manipulate the ET tube. (Great article - contact me by Twitter or the Facebook page if you want a copy.)

Does it work?
A study from 2010 in China confirmed the benefits of this technique. The title says it all: Prevention of dental damage and improvement of difficult intubation using a paraglossal technique with a straight Miller blade. Of course, we don't care that much about teeth when TSHTF, but it's a nice touch.


Another study (download here), done in 2008 in Canada, also showed that you could get a better view with this technique than with  the standard curved-blade approach. There is also a great discussion about the history of laryngoscopes, and how we ended up with the current designs.


Better view - but more difficult to place the tube? 
One trade-off of this better view may be that it is harder to actually place the tube. 

In the 2003 paper "Straight blades improve visualization of the larynx while curved blades increase ease of intubation," Spanish anesthesiologists noted that use of a Macintosh blade, while providing an inferior view of glottis, nonetheless made it easier to place the tube. (Download) 

Of course, if you don't have gottic view in the first place, it's going to be hard to place the tune


The Bottom Line

So, while the future of EMS ET intubation is a matter of much discussion, the need to be proficient in various techniques is not. Anybody who checks a set of blades at the start of a shift should know a number of techniques for using them. Hope this helps!

2013 update