Get a Clue

Do you recall the first time you treated a real patient? Not a classmate or a manikin, but a genuine consumer of emergency medical services?

I have vivid memories of my first alarm as a paramedic student. It was a “sick” call and shouldn’t have been an ALS job, but my preceptor thought he was doing me a favor by letting me play with the LifePak 10. Perhaps he anticipated the entertainment value. I didn’t disappoint him.

Consumed by the notion that This Is Not A Drill, repeat, This Is Not A Drill, I approached my morbidly obese patient. Dutifully, and with inordinate precision I placed the electrodes at the periphery of his doughy torso. Following a clockwise click of the monitor’s power switch, the tiny screen came to life, revealing slightly irregular emerald undulations that were much less distinctive than I’d expected. There were lots of little bumps, some of which were bigger than the other little bumps. I decided it was atrial fibrillation?an arrhythmia that occurs when the heart’s upper chambers don’t pump effectively. There was a whole section in our textbook about that. Feeling like the offspring of Christopher Columbus and Marie Curie, I announced my interpretation to my preceptor and our driver (also a medic).

My mentors were very kind?at first. One of them explained that the cause of those “little bumps” wasn’t defective atria, but a flabby patient who jiggled when he moved. Then they reminded me that even if the guy had presented in a-fib, his pulse was in the ‘80s, so he wouldn’t need that IV I was preparing. Oh, and the irregularity I saw was probably the patient’s fight-or-flight response to my awkwardness.

Things got worse during our next case, when the medics asked me if I’d found more atrial fibrillation. Uh…no, heh heh, pretty funny. They didn’t stop tormenting me until our fourth or fifth call. It was as if Johnny and Roy were being played by Don Rickles and Andrew Dice Clay. No mas! I’d learned valuable lessons about field-vs.-textbook tracings, hasty diagnoses and superficial assessments.

Some would say, invoking a tiresome medical maxim, that I’d heard hoofbeats and thought zebras instead of horses. Wrong. At that point in my training, I’d never treated a “horse,” much less a “zebra.” My biggest mistake wasn’t misinterpreting an EKG, but rather seeing what I wanted to see. If there’s an adage that addresses my behavior, it’s “to look within and adjust the mechanism of perception.” Even if Beat Generation poet Gary Snyder was stoned when he wrote that, his words remind us that prehospital assessment can be skewed by a bias to find a treatable malady.

It’s easy to understand how incorrect or incomplete assessments can lead to errors of commission. So much of our training centers on oral, practical and written games of Name That Disease. Occasionally, students are reminded not to over-treat?the National Registry’s static cardiology station comes to mind?but often we’re drilled to match signs and symptoms to pathology rarely diagnosed in prehospital settings. Combine that with most EMS curricula’s very limited pharmacology and physiology, and doing harm becomes a distinct possibility.

I think errors of omission in the field are even more common. If we assume it’s horses whenever we hear hoofbeats, but we abbreviate interviews and exams that confirm or refute such conclusions, then our complacency could become the most threatening presenting problem.

A 2008 New Mexico study examined a fundamental aspect of prehospital assessment?determining whether patients need ambulances?and concluded EMS providers make wrong decisions approximately 10% of the time.¹ I find that troubling because it’s much less risky to err on the side of caution when responding to emergencies. Are we overestimating our expertise, or just being complacent?

According to author, attorney and paramedic W. Ann Maggiore, it’s a bit of both.² Maggiore highlights several cognitive aspects of assessment errors and argues caregivers need a better understanding of inadvertent biases and assumptions. Maggiore explains we often arrive at diagnoses prematurely, then look for reasons to confirm our initial impressions, instead of maintaining an open mind about etiologies. Add a dash of overconfidence?EMS providers substituting street experience and limited skills for comprehensive medical knowledge?and we become dependent on script-driven, self-serving patient profiles. “We need to focus less on algorithmic protocols and more on developing critical thinking and reasoning skills,” Maggiore argues.

A subset of errors of omission raised by Maggiore, and also by Jerome Groopman in How Doctors Think,³ is when we force presenting problems to fit familiar patterns, i.e. we hear hoofbeats, see stripes, but continue to think horses. Let’s look at an example:

An alert, middle-aged female complains of nausea, dizziness and abdominal cramping the morning after consuming approximately 10 alcoholic beverages (the “hoofbeats”). She denies other history and substances. She’s in the medical field and is convinced she’s dehydrated (“horses”), certainly a possibility after alcohol consumption and vomiting. She requests a fluid bolus. You comply, without signs of hypovolemia (“stripes”), because you don’t see any harm and your patient validated your initial impressions. After 500cc of normal saline, she feels a little better and is discharged, only to return a few hours later with a hot appendix.

Dr. Groopman suggests a what-if dialogue between patients and physicians to fine-tune pattern-driven diagnoses. The three questions he highlights should be a reality check for prehospital providers, too:

• 1. Does anything not fit the diagnosis??broadens the scope of etiologies through emphasis of differentials.
• 2. What else could it be??raises possibilities that aren’t being reinforced solely by specialization or recent experience.
• 3. Might there be more than one problem??opposes “search satisfaction,” a premature sense of closure from first-pass pattern-matching.

The psychology of assessment is an important topic, one Maggiore and Groopman address comprehensively. My goal is to supplement their efforts with an unscientific sample of prehospital assessment errors I’ve encountered repeatedly in the field and as an administrator. I’m omitting ambiguous respiratory presentations, e.g. CHF vs. COPD vs. pneumonia, because that subject merits more thorough treatment than space permits here (see “Shortness of Breath” by Murphy, Colwell, Pineda and Bryan in the February 2009 issue of EMS Magazine).

I wish I could tell you I’m just a casual observer of the following mistakes, but I’ve made most of them. Maybe I can discourage you from doing the same.

Ineffective Interviews

During one of my first ambulance rotations as a medic student, I was directed to assess an alert, middle-aged male we found on a New York City street complaining of a leg injury. I had trouble getting specifics from him about the nature and location of his discomfort, so…I asked him if he was having any respiratory distress. Yes, I did, just like that. As embarrassed as I am about having used those words (and the gleeful response it got from my preceptors), my bigger mistake was posing a leading question.

According to MediaCollege.com, a leading question is one that “prompts the respondent to answer in a particular way.”⁴ Psychologist Elizabeth Loftus discovered that leading questions have an even bigger impact on subjects, altering their memories of events.⁵ What does that mean for you and me? Patients might interpret our questions about specific symptoms as prompts for “right” answers, then acquire those symptoms, or memories of those symptoms, as self-fulfilling prophecies. That’s exactly what happened to my patient. I watched as he became a diff breather?one I had created through suggestion.

What I should have done is ask an open-ended question, rather than a leading one. “Does anything bother you besides your leg?” would have helped me determine the scope of his problem. If his answer were, “No,” I could have proceeded with a more focused assessment. Some follow-up questions might have been mildly suggestive, e.g. “How does this feel?” but not as biased as “Is this painful?” or “That doesn’t hurt, does it?”

Asymptomatic Arrhythmias

I think the hardest part of learning EKGs is to not overreact to them. When we label a sequence of squiggles as an arrhythmia, curriculum-driven conditioning clouds our brains with directives to do something. That imperative can be unproductive, or even dangerous.

Rate and perfusion?not rhythm?dictate treatment. My opening a-fib anecdote is a good example of a patient who didn’t need rhythm-specific care, because his pulse was neither too fast nor too slow to prevent adequate oxygenation. One of the most common errors I’ve seen students make during practical exams is selection of electrical or pharmaceutical interventions to convert arrhythmias in the 60-90 range.

We’re taught to classify heart rates below 60 as bradycardia. That sounds ominous, but most people tolerate even 50 bpm without difficulty. Prehospital therapeutics like transcutaneous pacing, atropine and epinephrine that accelerate contractions shouldn’t be considered unless a patient is symptomatic. The consequences to our customers of over-treating their bradycardia are at best painful, and at worst fatal. In Taigman’s Advanced Cardiology, the author describes the making of a “cardiac cripple” through zealous administration of chronotropic meds.⁶

Let’s also consider rates and rhythms relative to individuals’ histories. I don’t think I’d function too well if my heart rate dropped below 50, but I know healthy people?athletes, mostly?who maintain pulses in the 40s. Sometimes prescription medications are responsible. A good medical history and knowledge of pharmacology should highlight those exceptions.

Two special situations involving stable patients are:

• New-onset arrhythmias. They need doctors, not medics, to rule out pathology.
• Ventricular tachycardia. Prehospital intervention prevents or postpones decompensation.

Misleading EKGs

Sometimes we expect too much from our monitors. They have alarms for abnormal heart rates, blood pressures and O₂ sats, but not for misinterpreted tracings.

Take artifact, for example. Inadvertent patient movement?like in my a-fib scenario?or electrical interference can make it much harder to read underlying rhythms. Filters on heart monitors are set to suppress such “noise.” However, digital averaging techniques employed by those filters can distort other parts of the EKG, creating what Froelicher and Myers, co-authors of Exercise and the Heart, label as “dangerously misleading” representations of waveforms.⁷ For example, a person having an acute MI might present with normalized ST segments, while another with ambiguous signs and symptoms could show exaggerated ST elevations. In some cases, ST depressions are represented as elevations, and vice versa.

The good news is that we have some control over the degree of filtering on each EKG. We can begin in monitor mode?usually the default setting?and view a “cleaner” signal that helps us identify rhythms, then switch to diagnostic mode, a more accurate representation of waveforms (and the default for 12-lead EKGs), when we suspect cardiac conditions that might cause deflection of ST segments. In general, three-lead diagnostic-mode tracings are visible on printed strips only, not on screens.

There are even worse ways we can misinterpret EKGs: We need to know when we’re viewing rhythms through pads or paddles we’ve placed on patients’ chests. Some monitors automatically select Pads when they’re applied, but might require manual switching to Paddles. If our screen displays Lead II before the cables are attached, or we connect the cables but continue to view the EKG through discarded pads or paddles, we might see a tracing that looks like ventricular fibrillation instead of the actual rhythm, or find an artificial asystolic waveform that camouflages true v-fib.

Check your monitors’ manuals for details about modes and leads.

Problematic Pulse Oximetry

Pulse oximetry measures the saturation of hemoglobin, a protein in red blood cells that delivers oxygen to tissues. Oximeters became popular field accessories in the early ‘90s. Portable, noninvasive and easy to operate, oximetry offered a seductive combination of high tech and low risk. It wasn’t until I started medic school a year or two later that I learned about the limitations of this equipment.

A common shortcut was substituting heart rates displayed by oximeters for palpated pulses, which is unwise, considering the absence of tactile feedback from a digital readout. When we check pulses manually, we sense not only rate, but also strength and regularity. Even when the rate is within a normal range, irregularity could be a sign of instability. We’d certainly want to know the source of frequent ectopic (out-of-place) beats, for example.

Another issue with pulse oximetry is its inability to detect aberrant forms of hemoglobin that don’t carry oxygen, yet still affect the saturation measured by these devices. For example, carboxyhemoglobin?hemoglobin transformed into a useless compound by carbon monoxide?causes a one-for-one percentage increase in oxygen saturation readings. That means your average cigarette smoker with a carboxyhemoglobin concentration of 6% will show a peripheral O₂ sat six points above the actual. High concentrations of methemoglobin?another non-oxygen-carrying variant of hemoglobin created by various diseases, drugs and toxins?cause pulse oximeters to read 85%, regardless of the true O₂ saturation.

Poor perfusion, nail polish, fungus or dirt under fingernails, motion and some arrhythmias can also distort readings.

Hypoglycemia: “The Great Impersonator”

EMS providers study a small subset of illnesses that might present in the field. Of those, we can reverse only a few. Low blood sugar is one of them. Thanks to compact, efficient, inexpensive glucometers, we can easily spot hypoglycemia and treat it with sugar or glucagon. Our failure to consistently do so contributes to the 7% of altered mental status (AMS) patients who present at EDs with hypoglycemia.

Prehospital caregivers aren’t the only ones who overlook that potentially disabling or even fatal condition. A 1996 article in The American Journal of Emergency Medicine reminds physicians to consider hypoglycemia as the cause of neurological deficits, “even when the findings seem to be explained initially by other etiologies.”⁸ Like what, for example? According to a 1996 New York study, hypoglycemia is misdiagnosed as CVA, TIA, seizure disorder, brain tumor, narcolepsy, multiple sclerosis, psychosis, sympathomimetic drug ingestion, hysteria, depression, traumatic head injury and symptomatic bradycardia. No doubt we could add other conditions hypoglycemia mimics, like intoxication, for example (alcohol consumption also can cause low blood sugar).

I think it’s particularly challenging to remember that trauma patients might be victims of hypoglycemia. We tend to focus on mechanisms of injury and often fail to consider that AMS may have lead to trauma, instead of trauma having caused AMS.

Pain in the Assessment

Picture this:

While lifting a backboarded patient, you feel sudden, sharp pain radiating from your lower spine to your left leg. You drop your end of the board an inch or two to the ground, then assume a position worthy of a circus contortionist. Your partner calls two ambulances?one for you and one for your startled patient, who suddenly looks a lot healthier than you do.

The medic assigned to you straps you onto your very own spineboard “just in case,” then initiates a collegial discussion about the prevalence of back injuries in EMS?just what you needed to hear. “It’s something we all have to deal with,” he counsels, before dropping the Dr. Phil façade and focusing on his paperwork. You want to tell him you hurt a lot more than he does right now, but you confine your complaining to an intermittent moan that blends with ambient ambulance noise.

You’re triaged at the ED, then bedded in a bay. Almost an hour later you’re medicated for pain which, if anyone had asked, you would have labeled as the worst of your life. Welcome to the patient’s side of pain assessment, a quality-of-life issue that needs attention.

Studies from 1995 and 2000 showed that EMS providers frequently underestimate patients’ pain because we don’t ask them about it, or we interpret their responses according to our own experiences and beliefs. Other research during that same period revealed that failure to administer analgesia prehospitally causes sufferers to wait an average of 90 minutes for pain meds in the ED! That sounds like a problem EMS should be able to address.

It is, according to McLean, Domeier, DeVore, et al.⁹ Their 2002 study showed that both verbal and numeric rating scales were effective tools for pain assessment?the first step in managing pain. Most of us are taught variants of the Borg scale?“On a scale of 1 to 10, where 10 is the worst you can imagine, etc.”?to assess all sorts of complaints. I prefer 0 to 10, where zero represents no discomfort at all. The breadth of the scale is less important than offering patients a way to grade and communicate their pain.

Higher Med-ucation

When I was teaching paramedics, I’d tell my students pharmacology is no less important than cardiology in our scope of practice. I’d urge them not to settle for a working knowledge of only the 60 or 70 drugs in their curriculum. Etiologies of patients’ illnesses often are much clearer when we recognize and mentally categorize their medications.

It had taken me a while to grasp that. Trying to memorize indications, contraindications, side effects and doses for over 100 drugs seemed like an insurmountable obstacle when I was in school. It was particularly frustrating to study meds that weren’t part of most EMS systems’ formularies. Why, for example, did I have to know that metoprolol is a beta blocker with antihypertensive and antiarrhythmic properties if I wasn’t going to be administering it? (I sure was wrong about that!)

Now I understand what an important prehospital assessment tool pharmacology can be. Even when patients don’t know or don’t want to say what they’re being treated for, their prescriptions usually give that away. I recognize most of the names, but I also carry a pocket guide to help me with new brands of old drugs. I usually question patients about medications that don’t match their history, particularly when those meds help explain presenting problems.

Sometimes our most useful assessment tool is a willing suspension of disbelief. Bias, assumptions and devotion to rote principles narrow the scope of curiosity and promote complacency. To perceive half a pattern is to see no pattern at all.

Embrace knowledge gained from atypical patient presentations. Accept the limits of technology. And never stop asking, “What else could it be?”

References

1.      Brown L, Hubble M, Cone D, et al. Paramedic determination of medical necessity: A meta-analysis. Prehosp Emerg Care 13(4):516-527, Oct 2009.

2.      Maggiore A. Escape faulty thinking. JEMS 33(1):117-125, Nov 2008.

3.      Groopman J. How Doctors Think. Houghton Mifflin, 2007.

4.      Leading questions. www.mediacollege.com, accessed March 29, 2011.

5.      Loftus E. Leading questions and the eyewitness report. Cognitive Psychology 7:560-572, 1975.

6.      Taigman M. Taigman’s Advanced Cardiology. Brady, 1995.

7.      Froelicher V, Myers V. Exercise and the Heart. W. B. Saunders, 2006.

8.      Luber S, Brady W, Brand A, et al. Acute hypoglycemia masquerading as head trauma: A report of four cases. Am J Emerg Med 14(6):543-547, Oct 1996.

9.      McLean S, Domeier R, DeVore H, et al. The feasibility of pain assessment in the prehospital setting. Prehosp Emerg Care 8(2):155:161, 2004.

Bibliography

Abbuhl F. Time to analgesia for patients with painful extremity injuries transported to the emergency department by ambulance. Prehosp Emerg Care 7(4):445-447, 2003.

Alonso-Serra H, Wesley K. Prehospital pain management. Prehosp Emerg Care 7(4):482-488, 2003.

Banh K, Tsukamaki J. Hypoglycemia. www.cdemcurriculum.org, accessed March 31, 2011.

Huszar R. Basic Dysrhythmias. Mosby, 1994.

Jasen J. The neural approach to pattern recognition. Ubiquity (ACM) 5(7): Apr 2004.

Kamat V. Pulse oximetry. Indian Journal of Anesthesiology 46(4):261-268, Aug 2002.

Lynch L. Hypoglycemia. www.drllynch.com/hypoglycemia.htm, accessed March 31, 2011.

Padder T, Udyawar A, Azhar N. Acute hypoglycemia presenting as acute psychosis. Am J Emerg Med 14:543-547, 1996.

Pattern recognition. Wikipedia, accessed March 17, 2007.

Raynovich B. Teach others to avoid leading questions. www.jems.com, accessed March 29, 2011.

Rubin M. What not to do in EMS. EMS Magazine 36(10):105-113, 2007.

Mike Rubin, BS, NREMT-P, is a paramedic in Nashville, TN, and a member of EMS World’s editorial advisory board. Contact him at mgr22@prodigy.net.