An older woman with osteoporosis fell at home, developed back pain, then went to urgent care and was diagnosed with a “muscle strain.” During a follow-up visit in my office, she told me she’d had transient urinary incontinence after her fall, and on exam she had midline back tenderness, hip flexor weakness, and hyper-reflexive knee jerks. This woman had suffered a vertebral fracture, not a pulled muscle.

One of my patients with asthma, heart failure, and a recent ventral hernia repair went to his pulmonologist complaining of worsening dyspnea. His head-to-toe exam was described as completely normal. Later that day, he came to my office so winded he could barely climb onto the exam table. He was not only in severe heart failure but had obvious peritoneal signs, and when he was admitted to the hospital, was found to have a perforated bowel.

An elderly gentleman presented to my office with hypoxia, hypotension, and patchy dullness to percussion of his back. I sent him to a local emergency room for presumed sepsis, but a single-view chest X-ray showed “prominent vascular markings,” so he was admitted for diuresis. While in the hospital, his notes described a normal lung exam, but after ongoing clinical deterioration he was sent for a CT scan. He had multi-lobar pneumonia, an infection from which he ultimately died.

Recalling these stories is infuriating; our patients deserve better than this. Yet these events don’t represent a career or even a decade of mishaps. They all occurred during the past six months.

Physician colleagues have lamented the demise of the physical exam. Some have compared the exam to a ritual, one whose devaluation is harmful to both physicians and patients. Others have documented trainees’ inability to take accurate histories and detect abnormalities on exam. There is a known problem. Absent from this literature, however, is any sense of outrage or concern for patient safety, and I find that baffling.

Physicians’ collective reaction to the decline in examination skills has been cool, a sort of resignation to a new form of practice. A commentary on clinical skills education—after witnessing interns fail to palpate their patients and identify heart murmurs and upper motor neuron signs—correctly argued that physical exam skills need to be modeled by educators. The authors then, however, called for research querying whether improvements in clinical skills are actually essential for modern-day patient care. They definitely are. One of my patients died as a result of being improperly examined. A study on cardiac auscultation—after demonstrating residents’ inability to correctly identify cardiac pathologies on exam—warned about a downward spiral in clinical skills but argued that specialty certification may need to include clinical skills testing. It certainly must.

I am a flawed physician. I have misdiagnosed patients, failed to recognize important exam findings, and made prescribing errors that were caught by pharmacists. I am also mindful that exam findings can evolve over time. What may seem like constipation one day may look like peritonitis the next. However, I am worried that we are glibly turning out generations of clinicians who don’t know how to perform physical exams without considering our patients’ well-being and safety. I do not mean to impugn all physicians. Many of my colleagues can identify subtleties on exam that inform care in meaningful ways. I believe, however, that our medical education system is permitting trainees to advance without having mastered fundamental clinical skills. Not simply an inability to distinguish between pneumonias and effusions, but an inability to identify abdominal catastrophes, fractured spines, and sepsis.

I am primarily a clinician and a bedside teacher so lack specific expertise in medical education. I am beginning to believe, however, that we must radically reconsider how trainees are assessed and advanced and which criteria are used for professional certification.

First, the Liaison Committee on Medical Education (LCME) must mandate that medical schools guarantee their graduates’ clinical competence. This organization’s “Functions and Structure of a Medical School”—guidelines to which institutions must adhere—state only that curricula must include content and clinical experiences related to organ systems, phases of life, and various stages of care. Nowhere does it dictate that students be able to perform accurate physical exams or use them to diagnose disease. Those skills lie at the very root of our work. This must change.

Second, the National Board of Medical Examiners (NBME) must re-instate the clinical skills portion of the United States Medical Licensing Examination (USMLE) and make it both challenging and accessible. When this section of USMLE Step 2 was mandatory, clinical assessment was required for advancement and learners understood that simply memorizing question “stems” and biochemical pathways was insufficient. However, they rightly complained about having to travel great distances at substantial expense to take this exam. The NBME ought not only reinstate this part of USMLE, but “deputize” faculty at medical schools to administer it. Demonstrating acquisition of clinical skills as part of medical licensure ought to be both expected and unburdensome.

Third, we must continuously observe residents performing exams and faculty must have the requisite clinical skills to teach them. Lack of oversight by attending physicians has contributed to a decline in examination skills, yet academic faculty routinely attest to residents’ abilities without directly observing them. Further, willingness to teach ought not be sufficient qualification to teach. I have worked at four academic institutions and not one assessed faculty members’ clinical skills before permitting them to oversee learners. We need to ensure that our strongest clinicians have the greatest hand in teaching and make sure that they’re offering bedside observation and instruction.

Finally, recertification must include clinical skills assessment. I recently sat for the internal medicine once-in-a-decade exam. The day I took that test I could have delivered spontaneous lectures on glomerular disease or autoimmunity and the lung, yet even after I’d passed, the American Board of Internal Medicine would have had no way to know whether I could actually examine and diagnose a patient. There is a substantial gap between meeting recertification requirements and being a competent physician, and our governing bodies need somehow to fill it in.

Several colleagues have read this essay and raised important and provocative questions. Is a “top-down” approach to improving exam skills appropriate, or should individual institutions press the effort? Who determines which exam skills should be assessed and what clinical competence looks like at any given level of training? How would cases and evaluations be standardized across testing sites? Wouldn’t adding “high stakes” components to existing licensure and accreditation exams simply compound physician burnout? Are there enough “master clinicians” remaining to help better train our learners, and would institutions be willing to support their teaching efforts? And, finally, with the advent of point-of-care ultrasound, are nuanced exam skills still valuable?

As an “old school” generalist, I believe that high-quality medical care hinges on performing accurate and diagnostic physical exams. This conviction was hard baked into my training. I understand, however, that well-meaning colleagues may disagree with me. It is time for a sober assessment of whether we’re imparting adequate and appropriate physical examination skills to physicians of the future and if doing so is still a shared priority.

Mike Stillman is an internal medicine and rehabilitation medicine physician.