During residency you refine the habits that will underlie the quality of your care, job satisfaction, and career progression. The following suggestions based on 36 years of emergency medicine practice in both academic centers and busy community hospitals.
They are excerpted from 9 years of end-of-shift comments provided as feedback to residents under my supervision. Not surprisingly, themes emerged over that time, and my comments and opinions on these recurring topics have been refined and restated for different residents over the years. Caveat: These are solely my opinions. Your mileage may vary.
DAY-TO-DAY RISK MANAGEMENT AND PATIENT SAFETY
“Remember that patient . . .”
Whenever you send an email or leave a messages containing any variation on "remember that patient", it is so kind (if appropriate) to put something like "nothing bad happened, but…" or "no problem, but…" first. Like most of us, I tend to immediately break into a cold sweat when confronted with this phrase, unless quickly reassured that no one suffered under my care.
Err on the side of a conservative approach to evaluation and disposition.
Most arguments against this benefit others more than the patient: (eg, self, nurses, consultants, bean counters, and turnover recipients.) "Worst case scenario" is our area of expertise, and pursuing that concern should be balanced only against the risks, costs, and complications for the patient of the tests and measures we employ to rule out “badness.” It is important to think twice before ordering tests for which false positives (think D-dimer) have potentially harmful consequences in time, dollars, or risk from precipitating more invasive diagnostic measures. As a trainee, you are developing your own approach, so in a teaching environment, I will often support more testing than I might do personally. This is how you refine your practice. Unfortunately, as we progress in our careers, the thing that often makes our approach to patients more conservative is a bad outcome.
Repeat vital signs and physical exams at discharge. I have often failed to check vital signs before I discharge the patient, worrying about it later, when finishing the chart. If we send home a hundred young people with a blood pressure of 98, some will come back. Responsible (and defensible) care does not require that you do something for everyone with such numbers, but we are obligated to take note and explain: “Ambulatory without symptoms at discharge, nontender, taking PO, no questions, return precautions discussed.”
Similarly, every patient should be examined prior to leaving, and that exam documented. A succinct note provides a picture of the patient’s course. It also effectively answers a common complaint from the dissatisfied, that they “never saw the doctor again, no one told me …” Plenty of patients in your career will leave with gastroenteritis and return with appendicitis.
Gut feelings. I can be comfortable with trainee decisions based on “gut” and “personal experience” when (tempered with a risk/benefit analysis) they lead to MORE conservative testing, therapy, or evaluation. Conversely, I am rarely comfortable relying on your “gut” or mine in doing less than the objective history and physical exam dictate. Actually, that’s a pretty good definition of both a residency’s goal, and appropriate and defensible medicine: “teaching you to do what the objective history and physical findings dictate a reasonable emergency physician should do.”
Reigning in your “gut” goes double for the inevitable unlikeable patients or those who return frequently to the ED. Following one’s visceral reaction is a frequent source of preventable error with such patients. I often find guidance in looking at what the nurses and I wrote on the chart. If the record says “substernal chest pressure and shortness of breath while climbing stairs” or “worst headache” -- even in a “frequent flier” or someone taking calls or texting during my exam -- I try to do what any reasonable physician would do for a less challenging patient with the same H&P. I ask myself, if I am wrong, and with the benefit of hindsight, what would someone else looking at the chart conclude that I should have done? Shrug off negative or dismissive thoughts early. Doing so avoids raising our own stress levels.
The same risks occur with minimizing (the mindset we sometimes adopt early in a patient’s course that there is little or nothing wrong). Minimizing is common in settings (like the military) where the majority of patients are young and healthy or where lack of primary care access drives “minor” chief complaints to the ED. There seem to be two ways in which those who tend to do this early in their training change their behavior. One is to watch for it, be objective, and accept whatever evidence- or experience-based suggestions you get. The other way is when your attitude hurts someone.
Evidence-Based Decision-Making: Clinical Decision Rules, Algorithms, Guidelines and Protocols
I strongly encourage the habit of reviewing the clinical decision rules, algorithms, and protocols in use in your ED every time you wander into the clinical territory they cover. I do. Such documents are developed to protect our patients, often in areas of our practice with high risk and low frequency. They provide the evidence-based odds of a major adverse event for an individual patient presentation and promote the workup required to find those very rare, very serious, entities. Besides optimizing care, in the medical-legal context failure to follow a departmental protocol/algorithm (with alleged harm) is a slam-dunk for the other side. Failing to follow guidelines (or justify why not) can result in a prima facie negligence determination, regardless of the other facts. Even in the absence of bad outcomes, regulators (from department chairs to employers, hospital committees, and federal agencies) find that auditing compliance with such guidelines is an easy metric that can blow back on providers in many ways.
“Handoff” patients. Turnover patients are a well-known patient safety issue.1 The "book" would say: go to the bedside with the departing doc, introduce yourself, do a brief H&P, and convey the plan. Re-evaluations are important for all patients. When any turnover patient’s course becomes at all complicated, one should do a complete H&P and test review, forming your own impression and bottom-line plan. “Complicated” in a turnover patient can be anything: the consulting service decides to change the plan or adds on new orders, the clinical course changes, you are being asked questions you can’t answer (by anyone) about “your patient,” unexpected results return, delays . . . When the consultant decides to discharge the patient that you were told needs admission, you should be able to critically evaluate and discuss that decision based on your personal knowledge of the patient (AND then explain it to the patient). I find that one cannot always do the best job for such patients (eg, push for admission or accept an alternative) without a personal understanding of the case. The liability and responsibility are yours. They are often sicker, certainly more complicated at this point, and deserve your full attention. Document carefully; inadequate handoffs were a factor in 24% of closed ED malpractice claims.2
Arguments against “moving the meat” (ie, going as fast as possible in order to see a maximal number of patients/shift). Your current job is to learn all you can from every one of your patients. Breezing through may not achieve that goal. Share the wealth. Every good case you acquire solo is one less for the juniors to see. A significant portion of your job is to teach and help your juniors get to where you are. Supervising their management of interesting cases optimizes both of your educations. That is how I learn on every shift. Few experienced docs can see above-average numbers of patients per hour safely. Sometimes you have to do so, but you get there AFTER your clinical skills have matured to knowing what shortcuts you can safely take for an individual patient and presentation. I’m not sure many residents are at that point. “Moving the meat” can be defined as a conscious overall decrease in the level of diagnostic and therapeutic interventions that you choose to apply. Like any management choice, there is a risk/benefit calculation. The risks are significant when choosing to short-change every patient you see.
MAXIMIZING YOUR EDUCATION
Look stuff up.
Before your opinions and practice patterns become “locked in” to what someone told you in med school or residency, it helps to adopt the habit of validating/challenging/questioning your developing beliefs and practice patterns. Ask yourself on occasion: “What is the basis for my practice patterns? Could there be another way? “
Similarly, you need a plan for when you are given conflicting information from staff and consultants. Recognizing that there is always more than one acceptable way to approach most clinical situations, your primary concern is to advocate for your patient. In any clinical situation when advice differs you have the right (and obligation) to ask for the evidence basis behind the suggestions. Ask the question and then do your own research. It won’t be long before the final call will be yours.
YOUR PRACTICE PATTERN: CARING FOR THE INDIVIDUAL PATIENT
Avoid cognitive biases.
Anchoring is the tendency to rely too heavily on information obtained early in the diagnostic process, with subsequent failures to adjust as new information become available. I struggle with this on every shift. Anchoring often starts with the chief complaint entered at triage (eg, The patient who says “I have a migraine” who actually means a really bad headache they’ve never had before). Accepting the patient’s explanation for their symptoms is one form of anchoring. Start with a non-directive history (“describe your symptoms”), look at the objective data, and consider the full differential before acting.
Confirmation bias in history-taking: Emergency physicians structure history-taking to elicit answers that point toward (or away) from the serious diagnoses. For example, with chest pain, I know I look harder for answers relating to the effects of exertion, rather than the effects of certain foods (I still remember the IM resident asking the diaphoretic, dyspneic chest pain patient what flavor Jell-O they had been eating when the pain began.) With headache, it’s “When was the last time you had a headache this bad?” “Did it come on really suddenly?” - not “Did your Grandma have migraines?” This reflects an appropriate bias toward not missing the serious etiologies and toward building a case for appropriate disposition.
Alternatively, we can be unconsciously biased to formulate our questions in a way that minimizes the characteristics most associated with significant disease. This may be when we are busy, distracted, or have formed an opinion of the patient from their past or present presentations. I prefer to err on the side of caution.
These and other cognitive biases represent some of the biggest threats to optimizing our diagnostic performance.3
Talk to prehospital providers and witnesses early.
Identify useful witnesses early and get their story. Don’t let them leave without learning what they know. When a patient comes in altered or unexpectedly seizes, arrests, or changes for the worst, the first step (along with the ABCs and quick interventions) is to catch the person who can tell you what has been going on. Bring them into the room and get as much information as you can. History is almost always of primary importance, and the intubated or obtunded patient is a lousy historian. Everyone has cellphones; there may be a witness you can call.
Physical examination. Although we correctly emphasize the importance of history, one thing I’ve seen fall off during residency is performing an appropriately thorough and focused physical exam. When I find myself proposing alternative diagnoses after a resident’s presentation, physical examination findings are at least as common a trigger as the inevitable variations in the history obtained.
A “negative” neurological exam does not mean no need for CT. On innumerable occasions, I have been presented with a “normal” neuro exam as a reason for omitting brain imaging. Nearly all patients who qualify for a scan using any of the accepted clinical decision rules predicting statistically increased risk for an intracranial process will lack focal findings. As a corollary, no one should be considered to have a negative neuro exam until they’ve gone through a series of maneuvers while standing and walking. I never cease to be amazed at the number of people who seemed fine to me supine, then fail an upright evaluation and turn out to have something serious. Stand them up and walk them.
Do them first. Think of the pelvic exam as just another part of your initial exam in the female patient with abdominal pain. Get it over with at the same time. Start with writing the orders for pregnancy test, cultures, wet prep, etc., and grab the ultrasound. Take a chaperone as you go into the room. This simple habit will avoid delays.
Do them often. More often than not, lower abdominal pain in a woman creates the need for a pelvic exam. While the value of the “routine” pelvic has been questioned in pregnant patients with an IUP on ultrasound4 or in young women with strong suspicion of sexually transmitted infections,5 it remains an essential part of the evaluation of women with lower abdominal pain. No matter how we (or the patient) feel about it, a pelvic or, as a minimum, a bimanual exam, can be critical, even in those missing their pelvic organs. I’ve seen too many missed diagnoses from failures to perform this simple procedure, and imaging is an unacceptable way to avoid it. I get many more surprises and alternative diagnoses from pelvic exams than from the exam of any other organ system. And don’t forget the rectal exam as another route to critical information in pre-pubertal girls and in men.
Work on narrowing your differential diagnosis by history and physical exam as opposed to testing. When presenting your patient with abdominal pain, the long textbook list of causes should be narrowed to “causes of abdominal pain we need to worry about in this patient at this time.”
Think twice before ordering “basic” labs, which are costly, delay throughput, can lead to morbidity from chasing false positives, and frequently fail to improve decision-making. An appropriate level of testing starts with a rational justification for each test (eg, anemia could be precipitating chest pain; dysrhythmias occur with abnormal electrolytes). Test ordering is a measure of your understanding of the case at the time and a reason that “resource utilization” is scored on the oral board examination. Over a career, the costs of a reflexive habit of “routine” testing are considerable. I do lower my threshold for “routine studies” when consultation and/or admission is likely.
Don’t forget stuff. A piece of your education we often fail to teach is this “nuts and bolts” management of the PROCESS of providing care. We all need a system to “close the loop” on tests and treatments either ordered or forgotten so that balls don’t get dropped. Everyone finds their own method. It’s easy to start with a simple list of what you have ordered or what you are planning to order and check it prior to discharge. Slowing down is often not an option, so having a system you apply EVERY TIME will save you (and your patients) delay and risk.
Slit lamp examination. Don’t bring me an eye problem without having done a slit lamp exam. The Wood’s lamp is not a tool for EM physicians who aspire to board certification. “The slit of light, when directed at an angle, accentuates the anatomic structures of the eye, allowing close inspection. The slit lamp provides greater magnification (10 to 25 times) and illumination than most handheld devices (eg, Wood's lamp), which is necessary to diagnose a number of traumatic and non-traumatic disorders.”6-7
Asking for help can be a “critical action” on the oral boards and is a step in a number of algorithms (eg, difficult airway, precipitous delivery, etc.). Medicolegally it can be considered indefensible if help was available, you didn’t call, and something went south. It would also feel really bad.
When you’re stuck for a diagnosis and/or disposition:
1. Repeat the H&P and review your differential – go back to the beginning.
2. Satisfy yourself that you have eliminated a need for intervention, further testing, or admission.
3. If the patient is too sick or uncomfortable to go home, or if significant life-threats remain on your differential, refer for admission or observation.
4. If stable for discharge: Explain to your patient what you have done, and the diagnoses you have considered and have or have not eliminated. Do the best you can to be sure they understand. Engage the family. Ask: “Do you feel well enough to go home?” If not, what will it take to get them that way? (See #3) Arrange appropriate follow-up or timed short-term return to the ED. Provide and document careful discharge instructions and return precautions. What I will often say is: “I don’t have a definite answer. I’ve given you my best assessment as to a likely diagnosis, the possible explanations that we have excluded, and the course I expect your illness to take. If something other than what we have discussed turns up, come back.”
Risk, pretest probability, and admission vs. discharge. Our job is “worst case scenario.” We are looking for the rare but life-threatening event with a common presenting complaint. While managing crises is something many of us live for, “crisis averted” is way better for the patient. When studied, emergency physicians commonly consider admission for patients whose risk for a serious outcome is in the 1-2% range.8 This approach often leads to what the inpatient services perceive as “false positive” admissions. If even 15% of such patients meet our inpatient colleagues’ definition of “appropriate,” many of them will consider us to be “wrong” 85% of the time. Such a mindset can breed reluctance to care for such patients.
One of the ways I look at the need for admission, advanced testing, or observation in patients with high-risk complaints and less-than-slam-dunk ED evaluations is Bayes theorem. Finding uncommon serious conditions first requires a high index of suspicion and sensitive tests (ie, history: “worst headache,” substernal chest pain with other concerning symptoms, comorbidities and risk factors, age, even gestalt). Refuting a positive SENSITIVE test for serious conditions (eg, subarachnoid hemorrhage, acute coronary syndrome), requires SPECIFIC tests, like CT/LP for SAH and high-sensitivity troponin for MI.
Calling consultants. I have observed a general reluctance among residents to speak to staff consultants. Like your ED staff, they are here to teach us and help care for our (often also their) patients. Specialists are often very familiar with their complicated patients and are usually willing to offer recommendations, facilitate follow-up, and provide insight. If they say “admit,” it results in a much shorter conversation with house staff. On those occasions when you make the call, ask yourself: “Did I learn something? Did I help my patient?” When you are in independent practice, that phone call will often be a stronger guide to appropriate treatment and disposition than anything we have taught you. Leading with “Thanks for calling back, I just have a question” can help relieve any defensiveness related to possible admission. Personally, I call frequently and learn something nearly every time. I don’t feel guilty. At 2 am, we are our patient’s best and only advocate. Our job is to get them the optimal disposition. “Call your doctor in the morning” is easy, but not always the best option. It is our equivalent of “if it’s not better, go to the ER.”
Arranging follow-up. Many of our patients lack access to quality ambulatory care. On occasion, we should take responsibility for (and do the work of) closing the follow-up loop. By extending the ED safety net, such a practice can improve outcomes and reduce return visits for a vulnerable subset of our discharged patients. Similarly, deferring the outpatient workup or consultation you believe your patient needs to their primary provider can be “kicking the can down the road.” If you can’t come to an answer after an ED evaluation, is the primary going to do it in a time-limited office visit? How many are in the ED because someone else hasn’t solved the problem? IMHO: If you know what the patient needs, try to help them get it.
Discharge instructions. A major purpose of discharge instructions is to demonstrate that we have considered potential complications, done what we can to detect or prevent them, and advised patients of what to look for and when to return. EHR-generated generic instruction sheets do not relieve us of exercising and documenting this responsibility.
SHIFT MANAGEMENT: AS EMPLOYEE, CO-WORKER, MANAGER
Stay ahead of the interns/students. While it is important for junior trainees spend whatever time they need with the patient and then present them, you are in charge of that patient’s care. A student H&P can take a while. Your patients, frequently in pain and potentially unstable, are being “seen” but not treated. Your sickest patient may be the one you haven’t heard about yet. There is no downside to getting in ahead of the intern/student rather than waiting for a presentation.
Department management. One set of skills that may be difficult to acquire in a busy academic ED is to learn to be the “captain of the ship,” functioning as staff by owning and managing all aspects of a busy department. Doing so requires a combination of situational awareness and assuming responsibility. A good example is noting the 60+-year-old with chest pain and shortness of breath in the queue and checking on them. Learn and follow the environmental cues that should drive you to the bedside: staff members hurrying (with or without a crash cart, IV cart, or ECG), louder voices, more than one nurse at the bedside, any question brought to you about a patient you haven’t seen, and other clues that may be unique to your practice setting. If another trainee or staff beats you to the bedside of the new “sick” patient, you need to continue to work on this skill.
Some day (or more likely, night) you WILL be the only doc in an ED getting dangerously busy. This is what Carey Chisholm called the “gold standard of emergency medicine practice”: alone in a community hospital, in the middle of the night. Skills include not waiting for “overwhelmed” but anticipating factors that will influence flow and creatively using all available resources to deal with them early. You may need to reach out to manage waiting room patients as well as those in beds still waiting to be seen. Getting orders started, providing pain relief, streamlining consults and admissions, and keeping track of patients you may not have seen or fully evaluated is a necessary skill. A reasonable exercise is to look at a crowded board of waiting patients and ask yourself: “What could I do now if I were solely responsible for this ED?”
Come in early. Habits you form in training will influence your practice long-term. In any setting it’s always appreciated when someone comes in a little early for their shift. Little things often influence how you are viewed, and can promote similar helpful behaviors in your coworkers.
Ignore co-worker behavior. One way I’ve successfully practiced in various settings for 36+ years without burnout was to studiously avoid paying any attention to how hard or how fast my colleagues on the same shift are working. I just do my job, optimize my efficiency, and limit the ability of any third-party to critique my practice. Our “metrics” are our own. They are commonly recorded and compared. Fee-for-service or RVU-based reimbursement goes a long way toward making my recommended approach more palatable. Consistent patterns of non-collegial behavior generally become evident to multiple members of a group practice; they get discussed and dealt with. Save yourself the aggravation, the job is hard enough. Obviously, silence is not appropriate with observed concerns for quality of care or safety.
Dr. Frumkin currently serves as volunteer faculty, Emergency Medicine Residency, Naval Medical Center, Portsmouth, Virginia.
Required disclaimer: The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, or the United States Government.
Copyright constraints: At the time of the creation of this work the author was an employee of the U.S. Government, and the work was prepared as part of official duties. Title 17 U.S.C. 105 provides that “Copyright protection under this title is not available for any work of the United States Government.” Title 17 U.S.C. 101 defines a United States Government work as a work prepared by a military service member or employee of the United States Government as part of that person’s official duties.