When I was a resident, faculty appeared in the morning for teaching rounds and presentations of admitted patients and then disappeared into their offices or labs. They were available if we needed them but we rarely did, partly out of pride and partly because the faculty were rarely of help with the acute general medical problems we confronted. These faculty usually had subspecialty interests and attended on the ward for only one or two months per year. If the attending was a cardiologist, we would move quickly past the patients with autoimmune disorders and infectious diseases and focus on the patients with congestive heart failure or valvular heart disease. If the attending was an oncologist, we would focus on the patients with cancer and order consults on the patients with complex problems involving the kidneys or heart. These attendings were congenial, remote, and typically engaged in research and would enliven our patient care discussions by holding forth about their particular areas of research. They rarely observed my conversations with patients or my physical exams but would sign their names at the end of my notes to acknowledge that they had fulfilled their responsibility for the overall management of the patient. Faculty were paid mostly on salary from state funds allocated for teaching, and they seemed to attach little importance to the billing of patients.
On the one occasion when I was informed that the clinic attending would observe my history and physical on a patient to provide a required evaluation, the entire experience seemed oddly forced and unsatisfying for both of us; this was a task ordained by some higher power that had little relevance to my role as a resident or my attending’s role as a clinician–investigator. Even the patient seemed mystified as to why two doctors were crowded together around his small exam table.
Much has changed since those days, and faculty currently provide far greater supervision of residents than I experienced. Each specialty is a bit different in the way it does this, with some giving in-house 24-hour-per-day supervision of residents’ activities and others providing less supervision, particularly at night. Changes in supervision have been mandated to meet accreditation requirements for residencies and to meet requirements for billing of faculty involvement in procedures and in the evaluation and management of patients. Clinical billing by faculty has become an increasingly important contribution to the funding for clinical departments since my era. In some cases, concerns about patient safety have also driven the changes.
Throughout this period of almost 35 years, there has been ongoing concern about the influence of residents’ medical education on the quality of care they provide to patients, because medical education takes place in the context of patients with serious illnesses, where an incorrect decision or poorly performed procedure could have grave consequences. Faculty have struggled with how to provide the right balance of oversight and independence so that a resident could learn through experience without making serious mistakes.
Literature reviews of clinical supervision have been published in 20001 and 2012.2 The review by Kilminster and Jolly1 included articles on supervision in nursing, social work, teaching, psychology, and counseling as well as physicians’ education; Farnan et al2 restricted their review to supervision in graduate medical education. One problem that becomes quickly apparent in reviewing the literature is the lack of a definition of supervision. Kilminster and Jolly suggest that supervision includes management, education, and support and that, in medicine, it should include patient and trainee safety, monitoring, feedback, education, and advice. In the articles reviewed by Farnan et al, researchers defined supervision in a variety of ways, sometimes including actual faculty visits with the patient and observation of the resident, and other times including only review of the resident’s presentation without a confirmatory visit to the patient or observation of the resident’s activities. These are very different levels of supervision and would likely have different impacts on the clinical outcomes of care and the education of the residents. Both reviews concluded that supervision had a positive effect on patient outcomes and the education of trainees. However, Kilminster and Jolly stated that current supervisory practice in medicine has very little empirical or theoretical basis, and Farnan et al noted that no validated instrument is available to quantify the connection between the quality of supervision and clinical outcomes.
There is ample evidence that residents have concerns about their supervision and how a lack of supervision has had negative impacts on the quality of care they provide.3 Yet faculty are increasingly being pulled away from the supervision of residents to cover the responsibilities of residents who have duty hours limitations or to provide documentation of patient care activities to support clinical billing requirements.
To illustrate what may be possible when we provide intensive supervision of an entire resident–patient encounter, I thought I would relate a recent personal clinical experience.
I was focused on observing an entire encounter of a resident with a patient in the emergency department, a luxury made possible by the temporary presence of an extra physician in the emergency department. I walked into the room with the resident, introduced myself, and took a chair to observe.
The patient was a middle-aged man who had fallen outside of a grocery store on the way home. He had bruises and abrasions, and the triage nurse was concerned about imminent alcohol withdrawal and asked the resident to prioritize this patient’s care so that she could administer medication.
The story unfolded that the patient had stopped drinking about a week before. He fell while carrying groceries because he was light-headed, and now he felt weak and shaky. I watched as the resident continued to pursue the diagnosis of alcohol withdrawal even as I noted inconsistencies in the story the patient was telling, such as the long time interval before symptoms, which was not consistent with typical historical and physical findings of withdrawal. I watched as the resident examined the patient and identified bruises over the chest and abdomen that, to me, looked to be of various ages and not what I would have expected from a recent fall. But I said nothing and observed. Finally, the resident completed the examination and explained to the patient that we would be doing some tests and would begin treating his symptoms.
I found a quiet place where the resident could present the patient’s case and summarize his findings. He had concluded that the patient had alcohol withdrawal, and wanted to give an intravenous benzodiazepine injection to prevent worsening of the withdrawal and a possible seizure, and to order routine laboratory tests to assess the liver function, plus some imaging studies to rule out a head injury. I listened quietly, and when he was done, I asked him whether there were any inconsistencies with the diagnosis of alcohol withdrawal. He admitted that the timing was unusual, but suspected that the patient had not been totally honest in relating his history.
I then prompted him to enlarge his differential diagnosis, but before he could do so, the nurse interrupted us, asking for the order for the benzodiazepine. She felt that she had waited long enough and needed to interrupt our discussion. The resident was about to comply with the request, but I suggested that we could wait a bit until we finished discussing his presentation. We then went on to explore other possibilities. What had caused all the bruises? What caused the fall? What about the patient’s color (a slight jaundice)? The exchange led to consideration of hepatic failure, sepsis, and internal bleeding with hypovolemia and syncope. Alcohol withdrawal fell lower on the list of possibilities.
We then discussed the patient interview, the exam, and the diagnostic approach, and I asked the resident to verbalize his thinking. He explained how he felt pressured by the nurse’s request and the initial evaluation that had occurred during the triage process. He felt that there was little danger in treating possible alcohol withdrawal compared with the risk of a seizure. We talked about the time pressures and the social pressures and how they could lead to premature closure of the diagnostic considerations. We discussed how to find a balance between the need to act and the need to consider the key possibilities and recognize the inconsistencies in our thinking as well as an awareness of the time constraints and the need to move on to other patients. As it turned out, our patient was in acute hepatic failure with gastric bleeding and renal insufficiency and did not have alcohol withdrawal. The fall had likely been caused by low blood pressure from blood loss. The slower, more careful approach had yielded a correct diagnosis that might have eluded us had we hurried to treat the presumptive alcohol withdrawal.
I realized I had rarely had the opportunity to observe a resident throughout the history, physical, diagnostic, and therapeutic workup except during care of our sickest patients, which we usually performed together in a resuscitation room. Most of the time, I would be receiving an abbreviated summary of the findings during a patient presentation and would have only a few minutes to confirm these findings by visiting the patient. Occasionally, I might detect an inconsistency in the thinking or findings, but that occurred relatively rarely because of time constraints. And although I might have been aware of the adverse impact of time pressures and social pressures, I might not have been able to help the resident develop communication skills to resolve the conflicts effectively. However, here was a case where something important had happened for the patient, for the resident, and for me as if by magic, because of my ability to observe and provide feedback. It felt almost like chemistry class where we would add white reagents into a beaker with a solution and mix them together, and then suddenly some colored solid would mysteriously appear.
I have had several more experiences where my ability to sit and observe yielded the opportunity to detect diagnoses different from those under consideration by the resident, or to identify factors in the environment of care that contributed to a potential error, and they reminded me of a finding by Mylopoulos et al4 in their study of expert diagnostic practice: “Physicians must remain sensitive to the possibility that a seemingly straightforward diagnostic case may bring unanticipated challenges, therefore becoming an opportunity for learning.” If expert faculty are not present to help identify these opportunities, the resident’s progress toward competency and expertise may not occur.
In the 35 years since I was a resident, we have not made enough progress in defining and providing effective faculty clinical supervision or feedback, based on the recent reviews I cited. We have barely begun to identify what supervision can offer. Why is this? Is it because clinical faculty must spend their time documenting their own activities in patient care rather than observing and teaching residents? Is it because our current promotion and reward systems provide little recognition for faculty excellence in bedside teaching, feedback, and evaluation? Is it because our medical educators have migrated to the simulation labs and standardized patients where one can control various important variables in the patient encounter? Is it simply that everyone has too much to do in too little time? I realize that many clinical cases will not yield productive learning opportunities for residents, and the study of education in the clinical setting is difficult and rarely funded, but if we do not focus our attentions on the opportunities of the faculty–resident–patient relationship and how to get the most out of faculty supervision, we will lose out on opportunities that occur every day to improve care in our teaching institutions. Significant faculty presence at the bedside allows for opportunities to observe the communications, thinking, uncertainty, and problem solving that give meaning to our relationships with our patients and our students.
I am hopeful that as residency programs identify milestones for residents that will mark the achievement of certain competencies, there will be a concomitant recognition of the role that experienced, expert faculty must play in observing and recognizing the achievement of the milestones. If faculty could observe residents, undisturbed by other patient care responsibilities, as I recently did, they might also be able to identify the factors in the environment of care that adversely influence residents’ performance and the clinical outcomes of their patients. The opportunity for expert clinicians to take a step back and observe trainees could greatly enrich the educational environment and quality of care and create more of those moments of magic when we do something important for our patients, our students, and ourselves.
David P. Sklar, MD
Editor’s note: The opinions expressed in this editorial do not necessarily reflect the opinions of the AAMC or of its members.