Tag: assessment

What the Heck is a General Impression?

I’m tired of all the mumbo-jumbo.

Here’s my beef. Every medical provider, particularly those who work in the acute setting (such as prehospital medicine, critical care, or the emergency department), talks about a concept familiar to us all: the overall, gut sense of how ill a patient appears. In EMS training this is often described as the “general impression,” the “view from the door,” “big sick vs. little sick,” or other euphemisms. It’s your basic opinion of whether a patient is doing okay or not, and it’s formed within the first moments of contact.

Whatever you call it, it’s important. In fact, this one factor is often what really drives your management decisions. If a patient looks truly sick, it may not matter what the vital signs show or how the history sounds; they’re getting zipped over to the hospital with bells on. Conversely, if they look really well, it’s hard to get excited even if they complain of “12 out of 10 pain” and their pulse is 100.

Here’s the rub: everybody acts like this quality is completely impossible to describe. If you tell me the patient “looked sick” and I ask what you mean, you’ll probably wave your hands and reply that it’s ineffable; that you “had to be there”; that you know it when you see it, but that it can’t be quantified and can’t be analyzed.

If true, that would mean it can’t be taught, either. New providers would have to learn to recognize this mystical patient presentation by dint of long, hard-earned experience.

And perhaps this is true. Certainly there are other aspects of patient evaluation and management that actually are too complex to reduce to simplicities. Indeed, one of the central skills of medicine, and one that humans are uniquely equipped to perform (hence the last one that computers will take from us), is our ability to extract a diagnosis from a large number of variables by recognizing subtle patterns.

But I doubt that’s true here. Why? Because you form your general impression within the first moments you meet a patient. There just aren’t very many factors that can come into play, because you haven’t obtained much information yet. The view from the door isn’t going to include ECG findings or subtleties of the OPQRST.

So I have a theory, and here it is. The entire mythical gestalt of your general impression actually involves only three things: the patient’s behavior, their breathing, and the appearance of their skin.

Behavior

The first thing you notice when you meet a person is their behavior. This mostly means two things: their mental status and their level of distress.

A sick patient may be unconscious, or visibly lethargic; healthy people are awake and alert, because the brain is one of the last things the body allows to shut down. They’re also not obviously loopy, such as profoundly confused or combative, unless they have a chronic condition such as dementia.

And if sick people aren’t so sick they can’t complain at all, then their complaints reflect their acuity. They scream, they moan, they are visibly distressed by pain or fear. They say things like they’re dying or can’t breathe or can’t see or can’t move.

Some interpretation is needed here, because appropriate behavior can depend on the circumstances. Malingerers may say they can’t breathe when they clearly can. Panic attacks may present with greater distress than the physiology warrants. A child is most reassuring when grabbing at your stethoscope and stealing your gauze. And an infant may be normal when he cries vigorously and sick when he sits in silence. But it all comes down to how the patient is behaving.

Breathing

The patient’s breathing can be evaluated from across a parking lot. You can’t auscultate or measure their oxygen saturation, but you can get a general idea.

Are they breathing at all? Are they laboring, wheezing, gasping agonally, gurgling through pulmonary edema? Are they chatting easily with the firefighters, or is sucking down air the sole focus of their attention?

Skin

Skin appearance is an idiot-proof and instantly recognizable finding.

The most common sick skins involve pallor and diaphoresis. Shocky or otherwise sympathetically-charged patients are starkly white and sweating like they’re in a sauna. It’s one of the most characteristic appearances of acute illness.

Cyanosis is next up. “Shortness of breath” in a patient who’s pink, warm, and dry is one thing, but it’s quite another when they’re turning blue.

Less common findings include the red-hot skin of fever, the yellow skin of severe jaundice, the dry skin of dehydration, and the dependent lividity of the very dead.

That’s all, folks

When you talk about a patient who looks sick, or “doesn’t feel right,” or has some other nebulous problem like being “toxic,” you’re not tapping into some vast, indescribable vault of clinical judgment. All you’re doing is using shorthand that refers to the patient’s behavior, breathing, and skin. (Notice how these factors are emphasized in our initial assessment.)

A gut belief that a patient has a big problem after a full work-up (including an H&P and diagnostics) is a different phenomenon, and suggests that your intuitive side is recognizing a larger pattern that your conscious self hasn’t yet been able to label. But that’s a distinct process from the instantaneous triage you perform when you first walk into a room.

There may be exceptional cases where something different sets off your alarm bells. But I bet most of the time, it can still be linked back to one of these three categories. (An example might be the frequent flier, well-known to you, who is usually stoic but today seems worried and wants transport. That’s a discrepancy in their behavior, ain’t it?)

Don’t believe me? Just think of how you tell the stories of your sick patients. I’ll bet you say things like, “I walk in, and he’s bent over gasping; his skin is completely soaked and looks whiter than copier paper.” Those are the factors that we recognize as important, and that’s why they’re so vividly evocative. They’re the colors we use to paint the picture of badness.

I may raise some ire by dismissing the voodoo surrounding the clinical gestalt, but here’s my challenge: if you believe there’s more behind your general impression of “sick or not sick,” then reply in the comments and tell me what it is. Maybe I’m missing or forgetting something. Maybe I’m doing it wrong and you’re doing it right. But if you can’t point to what’s missing, then I’m betting there’s nothing more to it after all.

Advanced CPR Techniques for Basic Providers

Handstand CPR

 

So you’re an EMT operating at the BLS level, and you understand that when it comes to cardiac arrest, you’re the man. Sure, you’ll call for the medics if you get there first, but the stuff that’s really important — compressions and defibrillation — well, that’s right in your wheelhouse.

But it may seem a little simple. Simple is beautiful, but maybe you’re wondering what else you can do to really master the art of resuscitation, especially when you’re out there on your own. Take it up a notch, if you will. And a lot of the cool stuff that’s being tried in the big world, such as pit-crew choreography and various supportive devices, are only available if your service makes a large-scale decision to adopt them. What can you do as an individual provider to absolutely ensure your peri-dead patients have the best chance of survival?

Here are some ideas.

 

Don’t Stop Compressions, at All, Ever — Seriously, Just Don’t

Hopefully at this point you don’t need to be convinced that stopping compressions is a bad thing. It truly is. The mountain of evidence is unequivocal: any time spent not-compressing kills people; each interruption in compressions kills people; pausing after compressions before defibrillating kills people; pausing after defibrillating and before resuming compressions also probably kills people; and so forth.

The trouble is that, despite this knowledge, we still stop all the goddamned time. There’s a lot going on during a code, and a lot of things you might want to pause for. But let’s go through a few and see if we really have to stop:

 

Stop for Pad Application?

As soon as you found the patient, you began compressions, right? As long as they weren’t wearing a honking seal-skin anorak, you can do that just fine over a shirt, blouse, or other light garment. (Hint: anoraks and similar loose outerwear can often just be pulled off the arms overhead, like removing a T-shirt.) Bam, in you went.

Now your partner needs to apply AED pads, though. Should you stop what you’re doing? Heavens, no. Let him work around you if he needs. He can unzip, rip, cut around your hands, tug the fabric out from under them as pressure lifts between compressions, and clear as much of the chest as he needs. Then he can simply apply the pads. No interruptions, no problem.

In some cases, a CPR-feedback device will be present, either combined with the pads as a one-piece unit, or as a separate “puck.” Either way this usually needs to go between hands and chest, but you should be able to slip it under there with (at most) a brief hiccup in the rhythm

 

Stop for Rhythm Analysis?

Unfortunately, if you’re using an AED (rather than a manual monitor like the medics are toting), you will need to stop compressing and come off the chest in order for the device to analyze the rhythm. Otherwise, the electrical motion artifact produced will confuse the computer. So as soon as the device tells you to stop compressions for analysis, clear the body — but don’t go far (in fact, I would simply hover), and as soon as it’s finished, get back on there.

You may need to stop for manual rhythm analysis as well, but some monitors have a filter that can allow the medics to “read through” compression artifact.

 

Stop while Charging?

So the AED finished analyzing and advised a shock; now it’s charging. Can you compress during this period? Yes. Both common sense (it won’t shock unless someone pushes the button, so… don’t push the button) and at least one study (albeit for manual, not automated defibrillators) have shown this to be safe. There are some AEDs that will get confused if you compress during this time, so know your gear. [Edit: per our “para-engineer” friend Christopher Watford, the Philips FR2+, FRx, and FR3 AED models, plus the Zoll AEDPlus and AED Pro, may complain and possibly halt if you try to compress while charging or shocking. Lifepak AEDs should be mostly okay. Chris and David Baumrind — two of the conspirators behind EMS 12-Lead — wrote a feature for JEMS discussing the behavior of various AEDs if you attempt these maneuvers. Required reading!]

Once the device has charged and is ready to shock, clear everybody except the compressor, ensure that they’re clear, and coordinate between the compressor and button-pressor. Something like, “I’m going to count to three, and when I say three, I’m going to come off and you’re going to press shock, okay? One — two — [come obviously clear] and shock — aaand back on.” The actual defibrillatory shock takes a fraction of a second, and the device will verbally announce once it’s delivered, so you can get back on the chest almost immediately after pressing “shock.” There is no residual “charge,” it doesn’t “take a while” to deliver, it’s a quick blip, so you’ll only need to clear the chest for a moment — no more.

 

Stop while Shocking?

As a matter of fact, do we need to clear the chest to shock at all, or can we keep our hands down, compressing continuously while the electrons flow?

Instinctively, most of us say “No thanks!” However, a little logic suggests the risk may be low. Electricity follows the path of least resistance, and if pads are properly placed and well-adhered to the chest, this path should always be through the patient’s chest. The alternate path up into your hands is much longer, and will only exist at all if you have a connection to the ground, which (if present at all) will probably run through fabric and other insulators. Since almost all AEDs now are biphasic — these use less current than the old monophasic devices — and since pretty much everybody wears rubber gloves while they compress, risk is probably quite small.

The evidence supports this somewhat. Consider these studies: Lloyd, Neumann, Sullivan (supports multiple-gloving in my view), Yu, and Kerber.

This idea has been gradually gaining traction, and some folks have already started doing it routinely, mostly of their own volition. Salt Lake City Fire has even been experimenting with making it a standard option during all resuscitations. For the most part, the worst adverse effect reported seems to be a tingling sensation, particularly if there’s a tear in your gloves. It’s reasonable to ensure that you’re wearing intact gloves, especially over prolonged efforts (multiple shocks may break down the material), and probably wise to double- (or triple-) glove. If there’s a feedback device between your hands and the chest the risk is even lower (or you could lay something like a rubberized blanket over the chest to totally insulate yourself, as in the Yu study).

Now, everybody has a story about a guy who knows a guy whose ex-partner’s bartender was touching a patient during defibrillation, got blown across the room and set on fire, and now can’t pronounce vowels. For the most part, this seems to be purely legend. The trouble is that there isn’t sufficient evidence yet proving it’s safe to make this an official practice on a top-down level; but that doesn’t mean you can’t make the decision for yourself.

If you have an arrhythmia (especially with an ICD or pacemaker), or another legitimate reason to be concerned about your own heart, it’s probably reasonable to pass. For everybody else, to paraphrase Dr. Youngquist of SLC Fire, this practice is probably safe for providers — if not yet for administrators. So you might not see this in your protocols for a little while, but I’ll bet it doesn’t say not to do it, either. The decision is yours.

(There is a possibility that some AEDs, particularly those with feedback technology, may detect the ongoing compressions and refuse to deliver a shock. Again, see above for more info.)

 

 

Stop for Ventilations?

Until you get some kind of tube into the patient’s airway, you’re going to have a hard time bagging any air in unless you pause compressions first. One option would be to simply skip it and perform continuous compressions, which is very reasonable, especially early in the code, or really whenever in doubt. But if you do pause to ventilate, take as little time as possible — pause, breathe goes in, exhale, second breath, and then immediately back into compressions (no need to wait for the second exhalation).

 

Go Faster — and Probably Harder

The currently recommended rate for chest compressions is “at least 100 per minute.” In other words, that’s not a target, that’s a minimum. Can you go too fast? Probably, but it’s hard, and it’s much easier to go too slow.

There’s an accumulating body of evidence, however, that points toward a more exact rate — right around 120/minute. Up to that number, more people survive if you push faster; above that number, fewer survive. It’s not for-sure yet, but in this business, not much is totally sure.

Since it fits the official “over 100” recommendation anyway, I now use 120 as my target rate, and I think you should too. It does mean that your old go-to songs for musical pacing, such as Stayin’ Alive (or perhaps Another One Bites the Dust) won’t work anymore, since those are matched to 100/minute beats. But 120/minute is simply twice per second, and most people can approximate that pretty well, or you can find a faster song (try this app for suggestions).

With that done, are you pushing hard enough? The recommendations are at least two inches deep in adults, so you should at least be hitting that. (It’s deeper than you think.) But as much as some people are willing to go wild on the rate, few people ever seem to challenge the depth. Unless you are an 800-lb gorilla and the patient a 70-lb granny, you are unlikely to cause meaningful damage, and there is a direct link between depth of compressions and cardiac output. Try to really aim for the mattress, and whatever depth you’re hitting, even if you think it’s pretty good, go a little deeper.

 

The Knuckle Hinge

Does it matter how you hold your hands against the chest? Maybe.

What really matters is that you provide good compressions, but hand position can affect that. What you should do is find a CPR mannequin and experiment until you figure out what works best for you. But while you’re experimenting, here’s something to try.

Most people lay one palm over the back of their other hand, and either interlace their fingers (as the AHA videos usually depict) or don’t (I don’t, since I find it somewhat awkward, but since it forces your arms to externally rotate, it can help encourage providers to lock their elbows). Either way, as you meet the chest, you’ll be making contact with the heel of a palm and one set of knuckles.

“Glue” these knuckles to the chest; they don’t move, so once you’ve found your position, you’re locked-in. But each time you compress, do allow your palm to lift off the chest, “hinging” at the knuckles as they remain in contact. Don’t come up very far — just enough that you could slip a sheet of paper between palm and chest — but get a little daylight in there.

What’s the point? One of the more common errors when otherwise high-quality compressions are performed is a failure to allow the chest to fully recoil. You can go deep, but if you don’t come all the way up at the top, you’re still not producing the largest possible stroke. What’s more, unlike poor depth, this isn’t always obvious by looking at the chest (either to you or to others), so the safest method to ensure full recoil is to actually lift off the chest. If you remove your hands completely, though, you tend to lose your place, and your hands can “wander” until you’re pushing on the patient’s feet or your partner’s face. The knuckle hinge allows the best of both worlds.

 

Assign a Monitor

Isn’t this tiring? Now you’re pumping away crazy deep, twice a second, full recoil, and not stopping for almost anything.

Even if you’re an Olympic decathlete, this will start to wear you out fairly quickly. You’re full of adrenaline, and you’re a rockstar lifesaver, so you won’t say anything, and perhaps you won’t even notice; you’ll keep plugging away. But before long, you won’t be pushing quite as hard or deep, or quite as fast, or maybe you’ll start leaning on the chest instead of recoiling all the way. I promise you will; many studies have shown this; and what’s more, you’ll probably still think you’re doing good work.

No problem. As long as we have adequate manpower (and in most places, there are plenty of people on scene at a code), simply assign one person to monitor the quality of compressions. If it’s you, your sole job is to sit somewhere with your head close to the action, staring at the up-and-down, and ensuring it follows all the criteria we’ve discussed. If it needs to be faster, you tell them to speed up until they’re on pace. If it needs to be deeper, tell them. If they ever pause for any unnecessary reason, yell at them like an Italian grandmother until they start back up. And once it’s clear that they’re fatiguing, you make them swap out, and ensure that the swap happens with minimal delay. The AHA recommends switching every two minutes, but use a smart approach; some compressors will last less, some more, and if you reach a mandatory pause (for rhythm analysis, say), you might as well change even if the current person has some juice left.

Depending on resources, they may be swapping with you, or there may be enough people sitting around that you can have a rotating pool of dedicated compressors. You can maintain the same person as monitor (the easiest method, if you can spare them), or just have each on-deck compressor act as monitor.

Useful tools for the monitor include a watch with chronograph, but even better would be a metronome. That way you can set up an audible pace (120/minute, remember) that any monkey can follow. A few services do carry actual digital metronomes, but if not, most smartphones have metronome apps available. (Find and download it now, not in the patient’s living room.) You can also throw an MP3 from an appropriately-paced song onto your phone, if nobody minds running a code to a soundtrack (probably not ideal when there’s an audience). The monitor person can keep track of other times as well, such as the ventilatory rate once an advanced airway is placed, total duration of the code, times of medication administration, and so forth. A pad of paper or strip of tape down the leg are helpful.

An electronic feedback device is a helpful adjunct to this role, and if resources are limited can replace it, but it’s not quite the same. If it is available, tracking the automatic feedback (and ensuring the compressor obeys) is the monitor’s job.

Whether or not a monitor is assigned, everybody performing compressions (really everybody at the scene) should understand that it’s still their responsibility to ensure quality. This is particularly important when it comes to eliminating interruptions, because even if there’s somebody to yell at the compressor when he stops, if he’s stopping all the time that’s still a lot of pauses. An effort should be made when assigning a compressor (who isn’t you), such as a first responder or bystander, to make them understand that they “own” their compressions, and it’s their responsibility to do ’em right and stop for nothing. The monitor’s job? Just to keep them honest.

 

Ask Why

Cardiac arrest happens for a reason, and even though it’s the most time-sensitive, treat-the-ABCs syndrome that exists, there are still times when you’ll never fix the problem without understanding the cause.

In a perfect world, you’d show up, compress, apply AED, shock, get a pulse, the patient sits up and hugs you, you transport and all’s well. In a realistic world (depending on your area), usually ALS shows up at some point and things take a more technical direction. But if you’re working the arrest for more than a couple minutes, have adequate manpower, but are still BLS-only, then your extra providers shouldn’t be sitting around twiddling their thumbs; they should be gathering information, planning the next step, and preparing for transport.

Ideally, one person is running the code. Either that person or somebody competent he delegates to should communicate with family or bystanders, examine available records, dig through the meds, whatever — try to determine both the history of the present event, and a reasonably-complete past medical history and medication list. Partly, this is for later management; the medics or the ED may need it. But it’s for you, too, because it may suggest your course of care.

Without an ECG, you haven’t got much to tell you what’s happening, except that the patient’s got no pulse. (Auscultating the chest may indicate whether a regular heart rhythm is present which is simply not perfusing — PEA, or if you’re a magician you may be able to “hear” V-tach — but you have to stop compressions to appreciate much.) You’re unlikely to be able to magically predict whether you’re dealing with V-fib versus torsades versus asystole. But you may be able to guess that certain correctable causes are present.

For instance, was the patient complaining of classic MI symptoms (crushing chest pain, nausea and vomiting, dyspnea) for twenty minutes before he finally became unresponsive? And he’s had two heart attacks before, with several stents placed? It’s a fair bet that he’s had another, which caused this arrest, and you may not have much luck getting him back until that artery can be opened back up. You can and should still work him initially on scene, but your mental goal should be delivering him to a PCI-capable hospital, so while you do your thing, stay on that track. If you get a few “no shock advised” messages with no pulse, or perhaps shock once or twice but he remains severely unstable, try to get him packaged as you continue your awesome compressions, notify the hospital of the situation and your suspicions, and get him over there. Try for ALS, who can perform a 12-lead ECG, which will facilitate this process (and your protocol may not permit you to divert to a more-distant PCI hospital otherwise).

Do you have reason to suspect hypovolemia as the cause of arrest? Is there obvious external bleeding… or is there a rigid and distended abdomen, perhaps with a story of abdominal pain or blunt trauma? In that case, you can push or shock all you want; you’re not going to refill an empty pump. Maybe chest trauma with a potential tension pneumothorax or cardiac tamponade? Transport ASAP to a trauma center (and perhaps ALS, since they can decompress a pneumo and give some volume if appropriate).

Is this a hemodialysis patient who missed two sessions, has been lethargic and sick-appearing, poorly-tolerating exercise, and finally fell asleep and didn’t wake up? Suspect hyperkalemia, a true “ALS-curable” condition, so if medics are available, work it until they arrive. If they’re on the dark side of the moon, transport with the best compressions you can manage.

Is the patient a known diabetic, taking insulin, and a story consistent with hypoglycemia? Check that sugar if you can, and if it’s something perverse like 7 mg/dl, get them to either ALS or an ER — both can administer intravenous sugar.

Could it be a hypoxic arrest? All arrests are hypoxic after a few minutes — dead people don’t breathe — which is why it’s usually reasonable to breathe for them (although far from a top priority). But if you walk in to find a post-drowning victim, or a hysterical mother saying her child choked and now has no pulse, you may have a cardiac arrest whose underlying cause is nothing more than hypoxia: their heart didn’t get enough oxygen, so eventually it gave up too. They still need compressions, and may need to be shocked, but most of all they need oxygen, so opening the airway and bagging in high-concentration O2 is a top priority. (Compare this against the post-MI patient above, who doesn’t need any oxygen at all until you have enough hands to provide it without delaying compressions and AED use, and even then doesn’t need much.)

Possible pulmonary embolism? Poisoning? Commotio cordis? The list goes on. The point is, if you have the resources to take a moment, gather some information, step back, and think, you can often do a pretty good job of guessing what brought you here, even without the benefits of the ECG. In some areas, your policies and protocols will dictate pretty clearly what decisions you can make, and it may not matter much. But flip through that rulebook now, because often times people assume it says more than it does (for instance, “closest appropriate facility” is more common than “closest facility”). When in doubt, you can always call medical control and make your case.

(As a general point of safety: continuing CPR while packaging and transporting emergently is difficult at best, and both unsafe and low-quality at worst. This should factor into your decision-making, as should the specific obstacles presented by extrication, and the potential availability of a mechanical compression device, which can make the process substantially easier.)

Just don’t ever try to argue that only ALS is allowed to think.

BLS is all yours, and cardiac arrest remains a fundamentally BLS problem. Own it.

Staying in Place: Compensation and Endpoints

Red queen running

 

Man’s leaning against a wall. He doesn’t move for hours. Just stands there not moving. Finally, someone says, “You been here all day — don’t you have anything to do?”

“I’m doing it,” he answers.

“Doing what?”

“Holding up the wall.”

 

And who’s to say he’s not? Maybe he’s working as hard as he can to make sure that wall doesn’t fall down.

In this situation, the man is a compensating mechanism. He is struggling to prevent changes in the wall; keeping that wall upright is an endpoint he cares to maintain, to sustain, to keep intact.

How do we know that the wall isn’t holding up the man? Because we don’t care about the man. Whether he leans or falls doesn’t matter much to anybody. But it would be a terrible thing if the wall collapsed. So we’ll let the man lean or shift in order to prop up the wall when it starts to totter — we’ll use him, adjust him, to compensate for any wall-changes. That’s why he’s there.

If the wall gets weak enough or tilts too far, though, he won’t be able to keep it up. He’ll try, but he’s not infinitely strong, and then maybe the wall begins to tilt or collapses completely. Since we know that under normal circumstances, he’s doing his best to prevent this, if we walk in and see that the wall is tilting, that is not a good sign. It may mean that despite his best efforts, the man has exhausted his strength and is no longer able to resist further wall-changes; or it may mean that, for some reason, the man isn’t doing his job properly. Either way, any further tilting will be unopposed, and will probably happen rapidly and uncontrollably.

 

Compensators and endpoints

This same dynamic plays out within the human body. As we know, living organisms seek to maintain a certain homeostatic equilibrium. We put our vital metabolic processes in motion and we don’t want them to halt or change, despite any insults or fluctuations imposed upon us by our surrounding environment. So our bodies struggle to keep all of our complex systems at an even keel, using a diverse and powerful array of knobs, dials, and other regulatory tools. Not too hot or too cool, not too acid or too basic, not too fast or too slow. Just right.

The kicker is this, however. Some of our physical parameters are more important than others. In other words, while some parameters have room to adjust, others aren’t negotiable, can’t change much, without derailing our basic ability to function and survive. Things like blood pressure (or at least tissue perfusion, for which blood pressure is a pretty good surrogate measure) are essential to life; your pressure can fluctuate a little, but if it drops too low, you are unquestionably going to suffer organ damage and then die. And yet there are many insults that could potentially lower our blood pressure if we let them: if we bleed a little, or pee a little, or don’t drink enough water, or sweat, or even just stand up instead of sitting down. How do we preserve this vital parameter despite such influences?

By compensating, of course. Our body gladly modulates certain processes in order to preserve other, more important parameters. So in order to maintain blood pressure, perhaps we accelerate our heartrate. In an ideal world, it might be nice if the heart were thumping along at — let’s say — a mellow 80 beats per minute. It’ll use little less energy and less oxygen than if it were beating faster. But it’s really important to keep our blood pressure up, and speeding up the heart can increase the pressure, so we gladly make that trade and induce tachycardia. (Many of these compensatory systems are linked to the sympathetic nervous system, our body’s standard “all hands on deck” response to stress and crisis.)

So imagine we find a patient who’s bleeding and notice that he’s tachycardic, with a normal blood pressure. This suggests a compensated shock; the body is using tachycardia to maintain that normal pressure we see; although his volume is lower than usual, the critical endpoint of adequate blood pressure is still intact.

But what if instead, we found him tachycardic and hypotensive? Well, that’s not good. We see that the body is trying to compensate, but we also see that the important endpoint — blood pressure — is falling nonetheless. The body would never intentionally allow that; BP is too important. So we recognize this as decompensated shock. The hypovolemia has progressed so far, and volume is now so low, that he can’t make up the difference anymore — the compensatory slack has run out — and any further decreases in volume will probably lead to an immediate and unopposed drop in pressure. There’s nothing more the body can do on its own; it’s out of rope.

The skilled clinician — or “homeostatic technician” as Jeff Guy says — uses this predictable progression to understand what’s happening in almost any crisis. Because primary insults are initially covered up by compensatory mechanisms, they may not be immediately apparent, and the earliest and most detectable signs of physical insult are usually nothing more than the footprints of the answering compensation. Thus, when when we encounter those, we know to suspect the underlying problem even if it’s not obvious yet. It’s like seeing brakelights flash from cars on the road ahead; even if you can’t see an obstacle yet, you know people are slowing down for something.

Obvious signs of decompensation usually show up late. Once the primary, underlying problem is revealed by failure of the corrective mechanisms, it’s often progressed so far that it’s too late to address. If you wait to brake until you can see the wreck itself, you might not be able to stop in time.

 

Two signposts for decompensation

There are two great ways to recognize which signs and symptoms connote decompensation.

The first is to understand which physical parameters are endpoints — which functions the body tries to preserve at all costs. These processes are only compromised as a last resort, so if you see them deteriorate, things are in the end-game; the body doesn’t intentionally sacrifice these for the benefit of anything else.

The second clue is more subtle. In this case, you observe a compensatory mechanism (not an endpoint), but find that it’s no longer successfully compensating — it’s failing, and starting to unwind and scale back, rather than doing its job. The changes in the compensatory system are inappropriate, resulting in less of what we need, not more. This happens when our systems are so damaged that they can’t even fix problems and pursue homeostasis anymore; our infrastructure, maintenance, and repair systems are breaking down. Consider this: we saw how tachycardia could be compensatory, but could bradycardia ever be beneficial in shock? Probably not. So if we found a shocked patient with bradycardia (and likely hypotension, the failing endpoint), we should be very alarmed indeed. There’s nothing helpful, compensatory, or beneficial about bradycardia in the setting of shock, so we recognize that the body would never go there on purpose. It’ll only happen when the machinery itself is falling apart.

Consider, for instance, Cushing’s Triad, the collection of signs often encountered after severe traumatic brain injury, when intracranial pressure has increased enough to squeeze the brain out from the skull like toothpaste. The triad includes hypertension, bradycardia, and irregular or slow respirations. What’s interesting is that, while all are a result of increased ICP, one of these is compensatory, while the others are merely the result of damage. Hypertension is the body’s compensatory attempt to force blood into the brain despite the elevated pressure in the skull. But bradycardia and bradypnea simply result from pressure upon the regulatory centers of the brain tasked with maintaining breathing and heart-rate. That’s why hypertension may be seen earlier, while the other two signs won’t usually manifest until the brain is actively herniating. One signals compensation, the other two decompensation.

Of course, there can be other reasons why compensatory mechanisms might fail, or at least exhibit lackluster performance. Some medications or other aspects of a medical history (potentially unrelated to the current complaint) might throw a wrench in the system. For instance, beta blockers (such as metoprolol and other -olol drugs) limit heart-rate as part of their basic mechanism, so patients with beta blockade often have trouble mustering compensatory tachycardia during shock states. That doesn’t mean they’re any less shocked; in fact, it means they’re more susceptible to hypotension, and that you must be especially on the lookout, because you won’t see one of the red flags (a rapid heart-rate) you might usually expect. Elderly patients with many comorbidities are generally not able to muster up effective compensation for anything, so they can deteriorate quickly, and without much fanfare. Ironically, healthy pediatric patients are the opposite: since they’re so “springy” and smoothly functioning, they compensate very well, with few changes in observable endpoints, until suddenly running out of slack and crashing hard because they’re already so far from shore.

Here are a few important compensatory signs, breakdowns of compensatory systems, and vital physical endpoints:

 

Appropriate signs of compensation

  • Tachycardia — increases cardiac output
  • Vasoconstriction (cool, pale skin) — raises blood pressure
  • Diaphoresis (sweatiness) — decreases temperature when necessary, but is often just a side effect of sympathetic stimulation
  • Tachypnea — increases oxygenation, CO2 blowoff, and cardiac preload
  • Fever — part of the immune system’s response to infection
  • Shivering — warms a hypothermic body

Inappropriate changes in compensatory mechanisms

  • Bradycardia — reduces cardiac output, rarely useful in illness; as a chronic finding may be the result of high levels of cardiovascular fitness (in healthy young patients) or medications (in sick old patients); but acutely, it is an ominous finding
  • Bradypnea — reduces oxygenation, CO2 blowoff, and cardiac preload
  • Hypothermia (or normothermia when a fever is expected) — suggests a failure of temperature regulation

Inviolable endpoints

  • Blood pressure — can elevate in stress states, but should not drop below resting levels
  • Mental status — except in the presence of a drug or similar agent directly affecting cognition, maintaining appropriate alertness and mentation are always a top priority for the body
  • Blood glucose — kept at normal levels in almost all situations, except when the regulatory systems fail, as in diabetes mellitus
  • pH — most of the cellular machinery fall apart if significant acidosis or alkalosis occurs
  • Low O2 saturation or cyanosis — although oxygen saturation can dip briefly without harm, and in some patients (particularly those with COPD, or long-time smokers) it may run low at baseline, a significant acute drop — or the clinical equivalent, which is frank cyanosis — is always inappropriate.

Managing STEMI Mimics in the Prehospital Environment: Video Lecture

A while ago we shared a PowerPoint presentation, Managing STEMI Mimics in the Prehospital Environment. This diverges somewhat from our prime directive around here by focusing on an ALS topic (ECG interpretation), but for the medics, it’s a topic that I think is important.

It’s also dense and difficult, in this case amounting to a 190+ slide presentation. In an attempt to unpack things a little, and to further explore our recent forays into multimedia content, we’ve got ahead and created a narrated slideshow walking through this subject.

This is still tough material, but as an overview it should be fairly approachable. The trick, of course, is to follow it up by viewing a large volume of pertinent ECGs to get some practice in applying the concepts. See our Links page for some great sources for practice strips, or visit the old standby, EMS 12-Lead — probably the best source on the internet for ECG education.

It’s broken into three parts, with total time of about 1:45. Treat it like a continuing education lecture, take your time if needed, and feel free to print the slides themselves for review. (Unfortunately, the lecture does assume at least a baseline ALS-level knowledge base, so if you’re just getting started with electrocardiography you may want to start elsewhere.) For any questions, throw ’em out here!

Part 1 (43:26):

Part 2 (34:06):

Part 3 (26:57):

Clinical Judgment: How to Do Less

 

It was around 11:00 AM when we were called to a local skilled nursing facility for a hip fracture. The patient was a 61-year-old male with mild mental retardation and several other issues, who’d fallen last night while walking to the bathroom. He was helped back to bed with moderate hip pain, and the staff physician stopped by to check him out. A portable X-ray was performed, which the physician interpreted as showing a proximal femur fracture as well as an associated pelvic fracture. This was communicated to us via a scrawled note and a cursory report.

The patient was found resting comfortably in bed, semi-Fowler’s and alert. He had no complaints at rest, although his pelvis and left femoral region were mildly tender and quite painful upon movement. No deformity was notable and there was no obvious instability. His vitals were stable and he was generally well-appearing, in no apparent distress. He denied bumping his head and had no pain or tenderness in the head or neck.

We gently insinuated a scoop stretcher underneath him, filled the nearby voids with towels and other linen, and bundled him into a snug, easily-movable package. Then we gave him the slow ride to his requested emergency department, a teaching hospital in town just a few minutes away.

We rolled into the ED and were lifting him into bed on the scoop when a young man entered the room, bescrubbed and serious-looking. I gave a brief report. As the words “pelvic fracture” left my lips, his mental alarms started visibly beeping and flashing, and he hurriedly asked, “What kind of pelvic fracture?”

“We don’t know. All we’ve got is the radiology note, which doesn’t say much.”

“Okay, but pelvic fractures can be a big deal. It could be … ” he sucked in air, “… open-book. There could be a lot of bleeding.”

I stared at him. “Well, sure. But he’s been stable since last night, and has a basically normal physical with no complaints at rest. He’s not exactly circling the drain.”

He didn’t seem to hear me as he briskly approached the patient and began poking him and asking questions. While we pulled our stretcher out of the room, he asked, “Does your neck hurt at all?”

Now that the patient had been stuck on a scoop stretcher for over twenty minutes, he thought for a moment and then shrugged. “Sure.” The doctor immediately ordered the placement of a cervical collar.

As we escaped, he was on the phone to the SNF, and the last thing I heard was him berating them with his urgent need to know exactly what type of pelvic calamity the patient had suffered.

 

What was the failure here? It was a failure of clinical judgment.

Clinical judgment is a phrase which means different things to different people, and often its meaning is so nebulous (much like “patient advocacy“) that it sounds good while saying nothing. But most would agree that it means something like this: the ability to combine textbook knowledge and personal experience, applying them intelligently to the current patient’s situation to yield an accurate sense of the possible diagnoses and the costs vs. benefits of possible treatments. In other words, it means knowing what the patient’s probably got and what to do about it, which is the heart of medicine anyway. So what’s all the fuss about?

In reality, when clinical judgment is mentioned, what’s often meant is something specific: the wisdom to know when something’s not wrong. Much of medicine is about planning for the worst, ruling out the badness, and looking for the unlikely-but-possible occult killer that nobody wants to miss. As a result, we often act as if nearly everybody is seriously ill, even when they probably aren’t.

On a practical level, most complaints — from chest pain to the itchy toe — could conceivably represent a disaster. Anything’s possible. So if we want to truly adopt perfectly mindless caution, we should be intubating every patient and admitting them directly to the ICU so that we’re ready when their skin melts off and their eyes turn backwards.

But we can’t do that, and we shouldn’t. So how do we know when to do a little less? Clinical judgment.

Clinical judgment is the acumen to assess a patient and say, “I think we’re okay here. Let’s hold off on that.” It’s what you develop when you have both the knowledge and experience to understand that a person is low-risk, and that certain tests or treatments are more likely to harm than to hurt them. That doesn’t mean that nothing will be done, or that more definitive rule-out tests will not occur, but it means you’re not freaking out in the meanwhile. It’s a triage thing.

Put another way, imagine the patient who you’re placing in spinal immobilization, or providing with supplemental oxygen, or to whom you’re securing a splint. They ask, “Look, I don’t much like this; do I really need it?” Well, I don’t know, rockstar — does he? If you’re simply acting on algorithms, reflexively doing x because you found y, then you really don’t know. How important is that oxygen? To answer that, you’d need to truly understand the benefits versus the potential harms, which means having a strong grasp of the mechanism of action, familiarity with the relevant literature (including the pertinent odds ratios, NNT and so forth), prior experience with similar patients, et cetera… only with that kind of knowledge do you really understand what’s happening. In essence, the patient is asking for the informed element of informed consent, something he’s entitled to, and you can’t provide it if you don’t have it yourself.

But when you do develop that depth and breadth of knowledge, you gain a special ability. It’s the ability to do less. When you truly understand what you’re dealing with, and more importantly, what you’re not dealing with, you can titrate medicine to what’s actually needed and stop there. Along with the knowledge comes the confidence, because you don’t merely know, you know that you know; in other words, you don’t need to take precautionary steps merely because you’re worried there might be considerations you don’t understand.

When it comes to withholding anything, even the kitchen sink, you might ask, “isn’t there risk here?” And strictly speaking, there is risk. But you can set that bar wherever you want. The important thing to grasp is that “doing everything for everyone” is not the “safe” approach; overtriage and overtreatment are not benign. All those things you’re doing have a cost. They may cause real harm. Even at best, they cost time and money, and subject the patient to unnecessary discomfort and inconvenience. We’d like to minimize all that whenever possible.

So, we return to the gentleman with the pelvic fracture. Strictly speaking, fracture of the pelvis has the potential to be life-threatening; certain types of unstable fracture can cause massive bleeding, along with damage to nervous, urinary, and other structures. So a textbook response to “pelvic fracture?” might be to treat it as a high-risk trauma.

But a patient with an unstable, severely hemorrhaging open-book pelvic fracture probably wouldn’t look like that. It would be evident; it would cause a number of apparent effects, such as pain and distress, shock signs, altered vitals, deformity or palpable instability. Except in bizarre cases or in patients who are clinically difficult to evaluate, big problems create big changes. While it’s true that we don’t know exactly what the X-ray showed, so one could theoretically argue for any conceivable pathology, there’s no question that the patient appeared stable, had remained unchanged for many hours, and had apparently been judged low-acuity after evaluation and imaging by his own doctor. In other words, let’s take it easy.

The question of spinal immobilization is another example. Strictly speaking, could we rule out the possibility of a cervical spine fracture? Well, no. Not without CT and MRI and even then who knows. But the fall was many hours ago, the patient was freely mobile and turning his head throughout that period, had no peripheral neurological deficits, denied striking his head or loss of consciousness, and quite frankly, had no pain until he spent twenty minutes with his head against a metal board.

It’s not often that you find a doctor more concerned about C-spine than an EMT. How did it happen here?

Despite the fact that we delivered the patient to a major tertiary center, it was nevertheless a teaching hospital, and the new interns had just hit the wards. While this particular clinician was undoubtedly smart and well-educated, at this stage he had about two weeks of experience behind him, and that is not conducive to providing judicious (rather than applied-by-spatula) care. He had neither the experience to know when to take it easy, nor the confidence in that experience to stand by such a decision.

We don’t want to take this concept to its extreme, which would involve doing very little for most of our patients. In the end, this is still emergency medicine, and emergency care will always involve screening for the deadly needle in the benign haystack. There’s also danger in simply becoming lazy and burned-out, and using Procrustean application of cynical “street smarts” to justify never bothering with anything. The real goal is to do the right things for the right reasons, no more, no less. And to get to that point, you have to put in some time.

Live from Prospect St: The Big Crunch (part 2)

Continued from Part 1

Since the two children appear generally intact, you ask your partner to evaluate them more fully while you head for the sedan to find the driver. Anticipating three transports, two stable and one potentially critical, you ask your dispatch to continue the P12, and also to ensure that police are en route (they are).

Arriving at the sedan, you find a middle-aged woman in the driver’s seat, alert. She is pink and warm, perhaps more diaphoretic than you’d expect for the ambient temperature, and does not initially notice as you kneel beside her. A firefighter is holding C-spine immobilization from the back seat.

When you greet her and pat her on the shoulder, she gives no response, but with more vigorous stimulation she looks over and acknowledges you distractedly. With multiple attempts and some yelling, you’re able to get answers to a few questions, but she is slow, tangential, and often ignores you outright. She gives her name as Samantha, but cannot or will not provide her last name; she is unable to describe the events that led to the collision; and she gives no medical history or current medications. She does state several times that she’s fine and would like to leave. When asked about her passengers, she mumbles “my kids” and mentions her brother several times. She endorses pain when asked explicitly, but does not specify where. She agrees that she drank “a little” alcohol; when asked about any drug use, she denies it vehemently.

Physically, she appears generally unremarkable. She is breathing somewhat shallowly but effectively, and her radial pulse is around 100 and slightly weak. Her seatbelt is not in place, but it’s unclear whether it was removed at some point. No gross trauma is apparent upon her head, face, or neck, and she does not complain or grimace upon palpation. She is uncooperative with a neurological exam, but demonstrates spontaneous movement of all four extremities. Her pupils are equal and seem appropriately small on this moderately bright day. Chest rise is generally equal and her abdomen is supple; no bruising consistent with seatbelt injury is visible. Her left knee is abraded and somewhat swollen. A sprinkling of dark blotches and streaks are noted on her left ventral arm in the antecubital region. Both frontal airbags are deployed; the windshield is cracked, but lacks a “starred” point of impact; and the plastic dashboard in the driver’s knee area is damaged and cracked. No blood or other damage is visible in the interior compartment. There are no child seats.

Your partner comes over. “The kids seem fine, just upset. One’s complaining of some abdominal pain, but it looks okay. They’re little troopers. Fire says they were wearing regular lap belts with the shoulder strap tucked behind them.”

When you wonder aloud whether there are more patients, he says, “There was nobody else in the car when fire arrived. The truck driver gave a statement to the police about how she was swerving across the road and plowed into him, but then he eloped.” He looks over your shoulder. “Oh, and the P12 is pulling up now.”

 

What is your treatment plan for these three patients? What are their respective priorities, any points of concern, and how could you shed additional light on their status?

Who will transport which patient, and to which destinations?

What special considerations should be made during documentation?

 

The conclusion is here

Live from Prospect St: The Big Crunch (part 1)

It’s 4:00 PM on a gloomy Friday in Chandlerville, and you’re the technician for the A2, a dual-EMT, transporting BLS unit dedicated to the city. Chandlerville is a small town, but densely populated, and its numerous industrial districts are frequent sources of work. 911 dispatch is directly through the fire department, which also sends a BLS fire apparatus to assist on all medical calls; your company’s ALS is also available by request. You are equipped with finger-stick glucometry, glucose, aspirin, and epinephrine.

After a “man down” call that ended in a patient refusal, you’re now returning to quarters. Just as you’re beginning to back into the garage, a tone sounds.

“Engine 3 and Ambulance 2, respond to 2108 Coastal Rd, the Empire Shipping Company, for an MVA. That’s two-one-oh-eight Coastal Road, in front of Empire Shipping, for an MVA. Engine 3?”

“Engine 3 is responding.”

“Ambulance 2?”

As your partner flips on the lights and pulls out to the street, he speaks into the radio: “Ambulance 2 has 2108 Coastal Rd.”

“Time out 16:01.”

Coastal Road is a long connector that wraps around the edge of town, and you glance at the map book to confirm that the 2000 block will be near the very end, about as far away as you can get in Chandlerville. Engine 3 is stationed in that district, however, so they arrive within minutes.

“Engine 3 to Firecom.”

“Firecom answering.”

“We’re off at 2108 Coastal Road. Two-car MVA, car versus truck. Multiple injured parties and entrapment. Start an ALS unit and a ladder for extrication.”

“Engine 3, you have a car versus truck, multiple injuries with entrapment. Break. Ladder 3, respond to 2108 Coastal Rd for the MVA; Engine 3 is on scene and A2 is responding. Time out 16:04.”

A few seconds later, your company radio dispatches Paramedic 12 to the same address, after Chandlerville Firecom contacts them via landline. The P12 starts responding, but they’re coming from two towns away, with an ETA of 10+ minutes. The field supervisor also starts rolling from an unknown location to assist. 30 seconds later, Engine 3 updates that they have an injured adult and several children.

Now very awake, you reflect that the nearest hospital will be Chandlerville Memorial, a 3–5 minute emergent transport (10 minutes otherwise). The nearest large tertiary center, Bullitt Medical Center — a Level I adult trauma center and a designated pediatric ED — is 15 minutes emergently (25 otherwise). The nearest Level I pediatric trauma center, however, is the Children’s Hospital, which is also 15 minutes but in the opposite direction; they do not receive adult patients.

Ladder 3 arrives on scene momentarily, and you pull up a few minutes later. As you park and call yourself out, you observe a Ford sedan with its front left corner smashed in, two feet of its fender and frame crumpled. This is evidently the result of driving almost headlong into the side of an 18-wheeler. It appears that the driver swerved right to avoid the truck, undercutting its rear wheels and “submarining” itself; the damage reaches the passenger compartment, but there does not appear to be significant intrusion. The truck itself seems minimally damaged.

As you jump out, a firefighter waves you down. “We’ve got three!” he announces. “Mom’s in the driver’s seat; she seems really loopy, probably drunk. Her door is just dented, we popped it open. But her kids are over there.”

Twenty feet away, you see two young girls, around 4 years old, each in the arms of a firefighter. They are crying loudly and clearly upset, with no visible injuries. The mother is hidden from sight in the sedan. The driver of the truck is nowhere to be seen.

 

What are your initial steps for addressing this scene?

Who appears to be the first priority for care?

What resources will you need? Which, if any, should you cancel?

 

Continued in part 2 and the conclusion

The Curious Incident of the Dog in the Diagnosis

 

“Is there any point to which you would wish to draw my attention?”

“To the curious incident of the dog in the night-time.”

“The dog did nothing in the night-time.”

“That was the curious incident,” remarked Sherlock Holmes.

Sir Arthur Conan Doyle, Silver Blaze

We can learn a lot from Sherlock Holmes.

If there’s anybody who better personifies the ultimate diagnostician, I don’t know who. Sir Arthur Conan Doyle, creator of the Holmes canon, was himself a physician and purportedly based his famous detective on Dr. Joseph Bell — who, it was said, could glean a dozen esoteric facts of a patient’s background, history, and complaints from a single glance. (Holmes himself, of course, was not a medical man; that role was played by Watson, the earnest physician who carried his stethoscope wrapped inside his hat.)

Holmes didn’t diagnose illness. Instead, he diagnosed crimes. But the methods were the same, so much so that among the countless fictional characters based upon the Holmesian archetype, some have been crime-solvers (cf. Monk), yet others have been medical doctors (House is the best). Perhaps we shouldn’t model ourselves after the man, who was a single-minded addict and misanthrope, but when it comes to diagnosis — something we can’t escape in medicine — he knows whereof he speaks.

 

The diagnostic method

Holmes tells us in The Sign of Four that detection involves nothing more than three skills: observation, deduction, and knowledge. Let us consider what he means when faced with, for example, a complaint of chest pain.

Observation: we perceive a middle-aged male, alert and seated upright, rubbing at his sternum with a pained expression. His skin is slightly pale, his respiratory rate is slightly elevated, and he is hypertensive. He complains of “tight” 4/10 chest pain whenever he breathes. Upon auscultation we detect diffuse, bilateral, biphasic wheezing. We note a history of coronary artery disease, diabetes, and COPD.

Knowledge: Chest pain in adults indicates a high risk for acute coronary syndromes. Pallor, tachypnea, and hypertension are consistent with this diagnosis. Sharp, mild, pleuritic pain is not, nor is wheezing, all of which are more consistent with a primarily respiratory etiology. But we also know that MI often presents atypically, particularly in diabetics.

Deduction: Both cardiac (ACS) and respiratory (COPD exacerbation) diagnoses top our differential. An ECG and biomarkers are needed to further evolve the odds.

 

So what just happened? We observed using our medical assessment — the history, physical, and diagnostic tests — thus yielding a collection of facts and data. We took the set of background knowledge we already possessed, regarding pathophysiology, epidemiology, and hazard ratios, and used it to “fill in the blanks” and provide context to our assessment findings. Finally, we connected the dots together and used deduction to decide what we’re dealing with.

Holmes knew this method well. He might observe your tattered boots, and using knowledge he possessed of typical wear patterns in the various trades, deduce that you make your living as a longshoreman. Simplicity itself.

Why is this a useful model for diagnosis? Because it highlights the fact that these three skills are entirely distinct, though all quite essential. Observation requires skill with the physical exam, the ability to take a nuanced history, the acumen to interpret diagnostics — it’s simply the trait of being aware. (Holmes, succinctly: “Data, data, data! I can’t make bricks without clay.”) Knowledge is knowledge: it’s memorized facts, what you learn in school or from books, and it gives us the basis to understand the raw material we discover in our assessment. Finally, deduction is the mental capacity to analyze, discover patterns, weigh odds, use your imagination, and extract from the vast pool of observation and knowledge the particular pieces that are actually relevant. (Holmes: “… to recognize out of a number of facts which are incidental and which are vital.”)

 

The hidden danger

Here’s the rub: we’re almost too good at deduction. Humans are excellent at finding patterns in anything. If I leave you my tea-cup for long enough, you’ll undoubtedly find an image revealed in the leaves.

That’s good — but it’s an error. Because there’s not really any image in the tea leaves. But if you’re good at observing details, and have a strong imagination, you’ll still “deduce” many wonderful things from it. Call it apophenia: people want to connect the dots, even when there aren’t any. So we create connections that may be true, but are not always true. We develop stereotypes. Simplifications. False associations.

In medicine, we’re especially prone to this. Because we do know that the human body is interconnected, and that patterns are the rule rather than the exception. Indeed, a large part of developing experience and clinical judgment is increasing your catalog of mental connections. Crackles mean CHF. Irregular pulses mean A-fib. People with Foley catheters have UTIs. Homeless people are drunk. Toe pain is a nonsense complaint. We can’t avoid making the connections, because just like when Holmes examines your boots, those connections are essential to doing our job. But at the same time, we need to learn when to reign them in, or we enter an inescapable diagnostic tunnel after the first moments of patient contact.

It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.

A Scandal in Bohemia

What’s the secret? Knowledge.

Sure, we gather the pieces from our assessment, and we automatically start to connect them together. We can’t help that; patterns jump out at us, we’re natural pattern-finding machines. But using our knowledge, we can look past those simplistic, eye-catching patterns, because knowledge tells us something more subtle: what’s missing.

She’s all false positives. See, that’s the trouble with naturals. They don’t see what’s missing.

Lie to Me, “Moral Waiver”

Okay, Friday night, a “man down” call for a homeless guy on the sidewalk. You’re already thinking: drunk. And the initial observations confirm it: he rouses sluggishly, slurs his speech, and pushes you away as he rolls back over. But then you open the mental box that you filled with this sort of thing in your training, and you reflect: where’s the bottle? And this is a strange spot — it’s cold, wet, public and unsheltered. And come to think of it, is that a medical alert bracelet? We should probably check this guy out a little more. Maybe take his blood sugar, look for any trauma, shake him awake and ask some questions.

The initial pattern recognition is there, but you don’t have to be a slave to it, because you know what else to look for. Even if five clues say one thing, if we don’t see five others that ought to be there, that tells us something different. Pertinent negatives, they call ’em in the business.

Maybe there’s nothing else; maybe the drunk is just drunk. But we’re too smart to make that kind of assumption. Because we know that getting it right doesn’t just mean registering the hits — it means checking off the misses, too.

Holmes would expect no less.

 

 

Glucometry: Clinical Interpretation

Continued from Glucometry: Introduction and Glucometry: How to Do it

Implementing glucometry into your overall assessment means understanding three things: when to use it, what the results mean, and when it fails.

 

Indications

First of all, by and large the only people with derangements of their blood sugar should be diabetics. The rest of us are generally able to maintain euglycemia through our homeostatic mechanisms, except perhaps in critical illness causing organ failure and similar abnormal states. Now, if someone injected you — a non-diabetic — with a syringe of insulin, you’d become terribly hypoglycemic, since it would overwhelm your body’s ability to compensate for the loss of glucose. But nobody’s likely to do that if you’re not a diabetic, unless it’s meant for somebody else and a drug error occurs, or I suppose if they’re trying to assassinate you.

With that said, people walk around who are diabetic and don’t know it. I’ve lost track of the patients I’ve transported who presented with signs suggestive of a diabetic emergency, denied a history of diabetes, and came back with a BGL of 600. Well, my friend, I have some bad news for you. “Everybody is diabetic, even if they’re not” is my attitude. Almost a fifth of older Americans are diagnosed, and the older and sicker they are, the more common it is.

Which brings us back to: who needs a BGL?

The most correct answer is anybody with clinical indications of either hypo- or hyperglycemia. As we saw, diabetes itself is really associated with hyperglycemia, which is why the classic signs of hyperglycemia are usually used to diagnose diabetes: polyuria (excessive urination, as extra glucose is excreted by the kidneys and brings water along with it osmotically), polydipsia (excessive thirst and water consumption, to replace the fluids urinated out), and polyphagia (constant hunger, since despite all the sugar floating around it’s not reaching the cells very easily). If your patient is complaining of those, you might be the first one to discover their condition. The diagnosis doesn’t require elaborate tests and imaging; a fasting glucose over 126 BGL tested on multiple occasions, or just once in combination with clinical symptoms, or a post-prandial (after eating) glucose exceeding 200, is the definition of type II DM. (With that said, I wouldn’t go around diagnosing your patients; that’s not your job, and you’re not quite that good.)

Once the glucose gets higher than the “renal threshold” — usually around 180 in average folks — the body starts to excrete it into the urine. This can actually be detectable by chemical dip-stick, or even by odor and texture at very high levels.

When hyperglycemia becomes severe and prolonged enough, we start to worry about diabetic ketoacidosis. Although burning fat and protein is not necessarily dangerous (some popular diets actually put you into a mild ketogenic state intentionally), extensive accumulation of ketones caused by a total lack of insulin (as in type I diabetics — DKA is rarely seen in type II) creates a metabolic acidosis in the body. This is when the long-term harm of hyperglycemia becomes a short-term hazard. DKA causes altered mental status, usually elevated states of confusion and disorientation, and combative behavior isn’t uncommon. Combined with the acetone odor that sometimes presents on the patient’s breath — which can smell like alcohol — DKA patients can seem suspiciously like drunks, and treating them like drunks is a great way to go down a bad path. (A word of wisdom: not only is everybody diabetic, but drunks are definitely diabetic.) DKA also frequently presents with symptoms of dehydration, due to the osmotic water loss in the urine; nausea and vomiting; and deep, rapid Kussmaul breathing to blow off the acidic CO2.

A few situations can cause short-term hyperglycemia, including stressors of any kind (there’s even “white coat hyperglycemia,” where patients tend to produce elevated sugars at the doctor’s office), but these typically won’t produce anything like the massive levels leading to DKA.

With all of that said, you need to really build up some glucose before hyperglycemia becomes symptomatic, and even more than that before it becomes acutely dangerous and unstable. That’s why as a rule, we’re more concerned with hypoglycemia, usually due to medication administration, physical exertion, or metabolic demand exceeding what was expected. Hypoglycemia again presents as altered mental status, in this case more often an inhibited rather than an elevated state: confusion, lethargy, disorientation, inability to focus or follow commands, weakness, headache, seizures, and eventually coma and death. The fun part is that the impairments can present as focal as well as generalized deficits: unilateral weakness of the limbs or face, speech slurring, poor gait, vision abnormalities, and more. In fact, hypoglycemia is a neurological chameleon, and can look like almost anything; it’s particularly notorious for imitating strokes, and for causing (not imitating) seizures. Interestingly, kids are particularly prone to hypoglycemia due to their gigantic heads, full of glucose-hungry brain.

Despite all this, the primary manifestations of early hypoglycemia are actually not symptoms of hypoglycemia. Rather, they’re caused by catecholamines — by the body releasing stress hormones, primarily epinephrine, in a response to the emergency. (This is not an irrational move: epinephrine helps us release and retain glucose.) As a result, we often seen the same signs we’d expect in anybody with a profound sympathetic stimulus: pale and diaphoretic skin, anxiety and shakiness, tachycardia and hypertension, even dilated pupils. Wise diabetics recognize the early signs of this sympathetic response and drink some Pepsi. As levels keep dropping, these symptoms combine with the neurological effects of glucose starvation to produce a confused, sweaty, increasingly stuporous individual. If left untreated, finally the sugar drops until we’re looking at the picture of impaired and diminished consciousness caused by true hypoglycemia. So just like always, the signs of compensation are our early warning system; once the body decompensates, it’s already late in the game.

To make a long story short, anybody with altered mental status, or any kind of general systemic complaint (weakness, fatigue, anxiety, nausea, etc.) should probably get their glucose tested, whether or not they have a known history of diabetes. This is true even if you suspect another cause, such as stroke. Not only can diabetic emergencies look like anything, they can also be comorbid; it is extremely common for patients to have another problem, yet also to bring a high or low sugar along for the ride, due to the illness throwing a wrench in their normal intrinsic and extrinsic glycemic homeostatic systems.

A number of years ago, there was some limited but compelling research that suggested poorly-controlled blood glucose (meaning not severe derangements but merely small deviations from the ideal range) was associated with increased mortality among an inpatient population with a wide variety of conditions. In other words, if you were hospitalized with something like sepsis, you were more likely to end up dying if your sugar tended to float around 160 instead of 110. As a result, it become trendy to practice extremely tight and aggressive glucose management for virtually everybody; diabetic patients were being tested every few hours and ping-ponged around using medication to keep their numbers textbook-perfect. More recently a number of studies have suggested that this may be less important than was thought, and in fact that excessive paranoia leads to a lot of iatrogenic harm from accidental insulin overdoses. This battle is still being fought in the hospitals, but for our purposes a reasonable take-away would be: when managing acute illness, from sepsis to head injury to cardiac arrest, once everything else is done it’s not a bad idea to check the patient’s sugar.

 

What’s the Number Mean?

So you’ve taken a blood glucose, either by capillary finger-stick or from a venous sample. Now what?

We mentioned that the “normal” range is something like 70–140. Diabetics seeking to control their condition and not have their toes falling off in a few years usually strive for tighter control of their BGL than is needed for acute care; a sugar of 175 is a little on the high side for a routine check, but a pretty meaningless elevation for our purposes.

All things are also relative, in that a given BGL must be compared to the patient’s baseline to predict its effects. In other words, poorly-controlled diabetics who are routinely sitting at 200 may become symptomatic of hypoglycemia at relatively high levels, whereas very well-controlled diabetics who usually run lower may be able to drop very low indeed without noticing it. However, a few rules-of-thumb are useful:

Non-diabetics usually become noticeably symptomatic below a sugar of, on average, about 53. (Diabetics, particularly those who are usually poorly-controlled, are more variable — their average symptomatic threshold is more like 78.)

After a recent meal, diabetics may demonstrate hyperglycemia to various degrees depending on whether they ate a Cobb salad or an entire chocolate cake. Non-diabetics should not exceed 200 or so. A few people can exhibit hypoglycemia after meals, due to alcohol consumption, “dumping syndrome,” or some other phenomena, but far more often they’ll exhibit similar symptoms without any true hypoglycemia; some people get shaky and sick due to postprandial epinephrine release.

After an unusual period of fasting (“haven’t eaten since yesterday”), non-diabetics should still have a largely unremarkable sugar. For diabetics, it will depend mainly on how much and what type of medication they’re using.

There’s usually a gap of 10–20 mg/dL between hypoglycemia that’s noticeable to the patient (i.e. sympathetic effects) and hypoglycemia that causes cognitive impairment (i.e. neurological changes). This is their safety margin, when they’re taught to eat or drink some fast carbs; if it keeps dropping they may no longer be able to take care of themselves.

But here’s the problem: the sympathetic “warning signs” can be mediated or impaired for various reasons. For one thing, if your body has to flip that switch often, you become numbed to it, and your hypoglycemic thresholds becomes lower and lower. And many patients with various metabolic and endocrine failures simply can’t muster much of a stress response — the same reason why the elderly may not produce tachycardia and other shock signs when they become hypovolemic. Finally, drugs like beta blockers that directly block sympathetic activity can seriously obscure hypoglycemia. Grab your nearest bottle of beta blockers and read the list of adverse effects: one will be hypoglycemic unawareness, a five-dollar term that means beta blockade can make it difficult to know when your sugar drops low.

Another important consideration in evaluating glucose levels is the expected trend. For instance, a BGL of 70 in a diabetic patient might not excite anybody. However, if you’re testing her because her nurse said that she just accidentally received four times her normal insulin dose, then a BGL of 70 should be alarming, because it’s probably going to keep dropping, and she doesn’t have very far to go.

To make a long story short, the clinical effects of both hypo- and hyperglycemia can vary substantially. What to do? It’s simple: assess the patient physically, obtain a history of their oral intake, medications, and metabolic demands (such as exercise), test their sugar if there’s any possibility of glucose derangement, and compare all those data against each other. A low number in the setting of obvious clinical symptoms is bad. A low number in an asymptomatic patient, or a normal number in a patient with highly suggestive signs and symptoms, should force you to bring out your thinking cap and weigh the odds.

What about treatment? Severe hypoglycemia needs ALS or the hospital — they’ll receive IV dextrose. Severe hyperglycemia needs the hospital only, where they’ll receive carefully-dosed insulin; this is generally considered too dangerous to administer in the field (although patients may have their own), so paramedics are reduced to giving fluid boluses, which may help dilute high glucose concentrations (not a very elegant solution) and is probably needed by a patient in DKA anyway, but isn’t really a fix.

What about oral glucose, in the cute little tubes we carry? Typically these are gels containing 15g of glucose, taken orally (either swallowed or held in the mouth — against the cheek or under the tongue — until it’s absorbed). Do they work? Sure. But it’s not much sugar and it’s not very fast. I found one source that suggests 15g of oral glucose should raise the BGL by 50 mg/dL within 15 minutes of administration — but I’ve never found it to be nearly that effective. In my experience, a bump of about 10 mg/dL per tube is about the best you can hope for in the short-term. If you need more than that, go with the medics and the IV syrup.

 

Testing Errors

When is a tested capillary or venous glucose unreliable? Usually it’s your fault.

Well over 90% of BGLs that test outside the maximum error range (remember, around 15%) are due to user error. Some of the main ones:

  • Your meter requires lot coding, and you failed to do so or used strips from the wrong lot.
  • You failed to clean the skin before lancing, contaminating the sample (not to mention creating an infection risk), or you had some D50 on your glove and it got mixed in there.
  • Rather than gently wicking the sample into the strip, you “smeared” the two together with mechanical pressure, interfering with the expected reaction process.
  • You drew blood from an arm with an IV infusion of D50, TPN, or other meds distal to it. Particularly when peripheral perfusion is poor, always try to sample at a different limb from any running drips.
  • You tried to reuse a non-reusable strip (gross).

Okay, okay, so nobody’s perfect. Factors that may not be as obvious include:

  • Temperature. The test reaction is designed to function within a specific temperature range, which is broad (often around 40–104 degrees) but not limitless, so don’t use them in freezing weather, and try not to leave your equipment ungaraged without climate control when it’s very hot or cold out.
  • Altitude. Just in case you’re an Everest expedition doctor.
  • Humidity. The strips have trouble when it gets very humid.
  • Air. The reagents in the strips will actually degrade if exposed to air for sufficient periods of time, so make sure that you keep them in their tightly-sealed case, and follow their printed expiration dates.
  • Time. If you draw whole blood and leave it around (much more likely to happen in the laboratory than in the ambulance), the erythrocytes will metabolize glucose at about 5-7% per hour.

The good news is that in many of these situations, internal error-checking within the glucometer will recognize the problem, and flash an error rather than a reading. Errors messages are usually numbered and can be informative, but each manufacturer uses different codes, so read the manual if you want to know what “ER2” means. (Hint: not enough blood in the sample is by far the most common.) Many of the other problems can be caught if you regularly check the meter using a known-value test solution, which you should be doing anyway according to most drug and safety agreements. (By the way, both the test strips and those vials of solution are usually meant to expire a few months after opening — the printed date is for an unopened bottle — so if they’ve around forever it’s probably time to retire them.)

What about physiological states that can interfere with the reading? We’ve discussed a few, but briefly:

  • Hematocrit. Anemia from any cause, including cancer or blood loss, causes falsely high readings. High crit, common in neonates, causes falsely low readings.
  • PaO2. Oxygen interferes with the electrochemical redox reaction; thus high concentrations of dissolved oxygen cause falsely low readings, and low PaO2 (i.e. hypoxia) cause falsely high readings, potentially masking a true hypoglycemia.
  • pH. Primarily in meters using the glucose oxidase enzyme, alkalosis will cause falsely elevated readings, while acidosis causes falsely low readings. The acidosis of DKA can therefore cause falsely low readings, masking the profound underlying hyperglycemia, so if the clinical picture screams DKA, don’t necessarily let the glucometer tell you different.
  • Macronutrients. High levels of circulating proteins or fats can cause falsely low readings due to dilution.
  • Hypoperfusion and inadequate circulation. See our previous remarks on this, and remember that venous sources will be more accurate than capillary.

Finally, are there medications that can interfere with glucometer accuracy? There sure are. These in particularly are highly device-dependent, with the glucose oxidase-type meters most often affected. Generally, the effects are not profound, but occasionally they may be clinically relevant.

  • Ascorbic acid. Better known as Vitamin C, some people take megadoses of this stuff, thinking it’ll cure their cold or flu. Depending on the meter it can cause falsely high or low readings, usually a minimal change, but at “megadose” levels the effect can be significant.
  • Acetaminophen. Also known as Tylenol. The effect is similar to ascorbic acid, but even more modest; it should only be considered in major overdoses, and even then the difference is unlikely to break 35.
  • Dopamine. Massive doses, such as might be used for intensive inotropic support, can modestly influence glucose dehydrogenase-based meters.
  • Mannitol. High doses can elevated the measured BGL by around 35.
  • Icodextrin. This is a dialysate solution used for peritoneal dialysis (not hemodialysis — this is where they pump fluid into the abdomen, let it sit, then drain it out), mainly in patients with diabetes. It metabolizes to maltose, which can cause falsely elevated readings in certain meters. There’s at least one tragic and unfortunate case report of a patient death resulting from massive insulin overdose due to this effect, not noticed until the true BGL was obtained by laboratory analysis. If your patient undergoes peritoneal dialysis, try to find out what dialysate is used, and if that’s not possible, it may be safest to assume their sugar is lower than you’re measuring.

 

Conclusions

After all this you’re probably thinking glucometry is so convoluted and rife with pitfalls that you’re better off just eyeballing how sweet your patients are. But don’t let me turn you off! This remains one of the best assessment aids we have, because diabetic emergencies remain some of the most common, most treatable, and most easily confused disorders that we encounter. We can’t perform exploratory surgery, and we may never see prehospital CT scans, but this is a diagnostic test that’s so cheap and simple, with such real potential to affect your decisions, that it should be available everywhere. If you maintain your equipment, learn how to do it right, and keep a few basic confounders in mind, it’ll serve you well as one of your most reliable tools.

Glucometry: How to Do it

Read part one at Glucometry: Introduction

So we want to know how much glucose is in our blood. How can we determine this?

Most modern systems involve a handheld electronic meter, which accepts disposable test strips. The general method:

  1. Insert a strip into the meter; this usually turns it on automatically, and the screen will indicate when it’s ready for a sample.
  2. Clean the patient’s fingertip with an alcohol swab.
  3. Using an automatic lancet (a spring-loaded needle), prick their finger-tip, drawing out a droplet of blood. You may need to push or massage the skin toward the puncture site in order to “milk” blood out, particularly if there’s poor circulation.
  4. [Optional] Many services recommend wiping away the first drop of blood and drawing out a second for your sample.
  5. Once you have a sizable, “hanging” drop of blood, apply it directly to the sample site on the test strip. It will wick inside and be absorbed.
  6. The meter will usually display some kind of count-down. Once it’s finished analyzing, it will show the blood glucose concentration (BGL) in mg/dL or mmol/L.
  7. Apply a band-aid to the site, and dispose of the test strip, lancet, and other bloody bits as appropriate.

What magic happens when you apply blood to the strip? There are a few methods.

(Skip this paragraph if chemistry wasn’t your favorite class.) As a general rule, the glucose in the sample is broken down by an enzyme (often glucose oxidase, or a version of glucose dehydrogenase). This reaction is proportional to the glucose concentration, and can be visualized by the accumulation of an indicator; the more glucose that reacts, the more color develops, and this is measured by a photometric transmission sensor. Alternately, in most current sensors, a more modern and somewhat more robust electrochemical method is used; here glucose is selectively oxidized, and electrons are pulled across a mediator to an electrode, which measures the current generated — either average, peak, or total depending on the type of analysis.

 

Results

Across the US, blood glucose is measured in the units mg/dL (milligrams per deciliter). In much of the rest of the world, the unit is mmol/L (millimoles per liter). This means that if your paramedic buddy from the UK is telling you about a diabetic he treated, the numbers may seem peculiarly low. Since we’re mostly Yanks here, we’ll be working in mg/dL, but if you ever need to convert to mmol/L, you can simply divide it by 18 (or multiply by 18 to get from mmol/L back to mg/dL).

Much like vital signs, textbook ranges for “normal” blood glucose levels vary. A loose range for practical purposes would be around 70–140, although ideally we should be under 100 most of the time, and routinely testing over 125 is not a great indicator for your health. Numbers will be elevated after eating, but non-diabetics still shouldn’t break 200 or so.

Although we’ll talk more about clinical interpretation later, in general it’s safe to say that the lower the number, the more each point matters. The difference between 70 to 50 can be profound, while the difference between 200 and 180 may be totally undetectable.

 

Accuracy and Precision

Glucometers have evolved through quite a few generations by now, and they continue to improve in robustness and reliability. Most diabetics use them regularly to track their sugar and thereby guide their diet and medications.

How accurate are they? Depends on who you ask. The American Diabetes Association says that at a minimum, they should give readings within 15% of the true value, and ideally manufacturers should shoot for an error of under 5%, at all concentrations. But percentages can be a confusing way to measure it, because as we observed, a 15% difference at a sugar of 500 (a possible range of 425–575) may mean little, while a 15% difference at a sugar of 60 (a range from 59, which is low, to 69, which is about normal) can be rather important. So the FDA says this instead: 95% of the time, for values below 100, meters should be within 20 points of the true value, while for values above 100, they only need to be within 20 percent.

Whatever the case, every meter varies, but generally they can be relied upon to fall within about 15% of reality, as long as no user errors or confounding factors (we’ll talk about those) are present.

 

Blood Source

Traditionally, capillary blood for glucometry is taken from the fingertips. This is painful, so most modern glucometers have been evaluated to determine their accuracy when blood is drawn from alternate sites. Any location with lean, vascular muscle close to the surface (i.e. not too much fat overlying, which you may not be able to penetrate with a lancet) can be usable — the forearm is the most common site. The research has shown that this practice is generally fairly accurate for routine purposes, but the danger is that BGL from the forearm lags behind that from the fingertips. It takes longer for these readings to approach reality — about 30 minutes, in fact, before you’ll read the same from the forearm as you’d read at the fingertip, and until then the numbers may be radically wrong (for instance, a reading of 145 when it’s really 50). So glucometer manufacturers recommend that diabetics always use the fingertip when there’s any question of hypoglycemia, when they’ve recently eaten, or any time when it’s important to have the most current and accurate figure. Obviously, this is always important for EMS, so we should generally stick to fingers.

On the other hand, in many areas it’s common for paramedics to start IVs and then use a drop of blood from the catheter’s flash chamber for glucometry. Briefly, like so:

 

A used catheter (needle inside)

 

The rubber stopper behind the flash chamber

 

Press on the rubber until a usable drop of blood comes out the end

 

This method works, saves you the trouble of lancing a finger, and spares the patient some extra pain. But it’s usually considered technically incorrect, because the blood in the catheter is venous, whereas glucometers are calibrated for capillary blood. See, since venous blood has already given up glucose to the tissues whereas capillary blood is still in the process of doing so, venous BGL is lower than from capillary sources — usually about 5–10 mg/dL. (If by chance you have a source of arterial blood, then that should be higher still.) However, after eating, particularly carb-rich foods, capillary sugar may be as much as 25% higher than venous, because of the extra glucose sequestered in the muscular tissue. (Stockpiling this fuel is why marathon runners like to “carbo load” before events.)

With that said, I’m going to make a controversial recommendation: in most cases, whenever it’s available, venous blood should be used instead of capillary blood. If someone has started an IV, then you should be using that instead of a fingerstick. Why? Despite the small and usually predictable difference, in sick people, it’s actually a more accurate result.

In sick people, circulation is often impaired; this is particularly true in situations like shock, sepsis, and the mother of all shock states, cardiac arrest. When perfusion is poor, the first thing we lose is the peripheral circulation, and it doesn’t get more peripheral than the capillaries of the fingertips. What does this mean? It means that in many acute patients, when it’s important to have accurate diagnostics, capillary blood sugars can be utterly, totally inaccurate. Since blood is no longer moving actively through the periphery, it tends to “pool” there stagnantly, letting the tissues chew through its glucose supply without resupplying it. This results in a falsely depressed capillary BGL even when the venous BGL is normal. Conversely, it’s also possible that in poor circulation, the distal capillaries are the “last to hear” about a drop in sugar, resulting in a falsely elevated BGL. But high or low — usually low — it’s not reliable. Anybody with impaired circulation should get a venous glucose if there’s a chance of it affecting care. (And if there’s no chance of it affecting care, then why do it?) By the way, this includes impaired local circulation, such as patients with PVD. Not that a diabetic would ever have PVD…

(Edited 6/12/12: A few commenters have pointed out that the practice of drawing blood samples from used IV catheters can present a safety risk; although modern safety catheters usually retract or obscure the needle, this is not a fail-proof mechanism, and pushing on the plunger can potentially lead to an accidental stick. We should all be sensible about this sort of thing, so be cautious and give a moment of serious thought to the conditions, equipment, and your technique before trying such a move — and of course be aware of any policies your service has on the subject.)

 

Coding and Calibration

The important business during glucometry is taking place in the test strip, where the actual chemical reaction occurs. Since this is a rather minute organic event, individual test strips tend to vary a little in their performance.

Traditionally, this is handled by lot coding. Each batch of strips (they come in packs of so-many) would usually include an electronic coding strip, which looks like a regular test strip, with some extra electronics attached. You insert it into the meter, and it automatically calibrates it for the current lot. If your device works this way, it is essential that you code your meter for the lot you’re using, and do not mix your strips with those from other lots; your results can be off by over 30% due to using the wrong code. However, many current glucometers no longer require coding, either by automatically self-calibrating using information in the strip itself, or by controlling manufacturing tolerances so that all strips are the same. Read the manual or check your policy!

Now, is a rose a rose, or are there different BGLs out there? Really, there are two that matter. When we prick the finger and sample capillary blood, we’re measuring the glucose concentration in whole blood — the raw, unmodified stuff running through your veins. We could also take that blood, centrifuge out all the big cells (particularly red blood cells), and measure the glucose in the plasma that remains. This latter method is how it’s done in the laboratory, and this is the gold standard for this type of test. (In the handheld glucometer, the test strip usually uses a filter to either absorb or lyse the red cells, but their presence still affects the measured concentration.)

Why does this matter? Only because whole blood BGL differs slightly from plasma BGL. Since the number is a concentration, and the presence of hemoglobin slightly dilutes the blood, plasma values are typically 5-15% higher than than whole blood values. In most of us it’ll be about 11%, but the exact difference depends on how much space your red blood cells are filling up, aka your hematocrit, so that estimate only works for people with a normal “crit” (around 45). The higher your crit, the larger the difference (and the levels of other circulating lipids and proteins can be relevant as well). The good news? In order to make home BGL readings comparable to laboratory readings, most glucometers report results as a “plasma equivalent,” either by assuming a normal crit and performing a quick mathematical adjustment, or by actually measuring the hematocrit. Some meters can be set to display either whole-blood or plasma equivalents, and ideally we should know which we’re looking at, but plasma is usually the default.

 

Ketones?

We know that when hyperglycemia becomes severe, the body often develops high levels of ketones in the blood and urine. (These are involved in a secondary metabolism that cells can use as an alternative to directly consuming glucose.) Lots of ketones in a diabetic is a corroborating sign of a highly elevated sugar, and suggests deterioration to diabetic ketoacidosis, a dangerous state involving a deranged pH.

There are handheld meters that can measure ketone levels, but simple glucometers can’t. However, many models have a feature where, if BGL is found to be over a certain level (often around 300), an indicator will light up with a warning like: ketones?

This is not indicating that ketone bodies are present, which the meter can’t know, but is merely a reminder that at these glucose levels, we should consider the possibility of their presence. Which, as clinical wizards, we already knew, so it doesn’t tell us much. (In fact, it’s more intended for patients, who may have the specialized strips with which to measure their ketone levels.)

 

Takeaway points:

  1. Glucometry can vary by around 15% even when it’s working correctly.
  2. Use venous blood (e.g. from an IV) rather than capillary blood (from a fingerstick) whenever possible.
  3. If using capillary blood, use a finger rather than alternate sites like the forearm.
  4. If your meter needs coding, make sure you do it.
  5. Remember that many conditions (such as shock, PVD, and a recent meal) can alter capillary BGL, and some (such as anemia or hyperlipidemia) can even alter a venous reading.
  6. Ordinary glucometers don’t measure ketones.

 

Tune in next time for a discussion of more clinical phenomena that can influence blood glucose readings, as well as interpreting and applying the results in real patients.

 

Editor’s note: Remember that although we often don’t cite specific references for our figures and data, if you ever want to know what studies or evidence we’re using to support our claims… just ask! We’re happy to oblige. This applies to all of our posts, but may be particularly germane for this one, where some specific and possibly controversial points have been made.