Moving Us Closer To Osler
A Miller Coulson Academy of Clinical Excellence Initiative

How to Improve Your Physical Diagnostic Skills With a Digital Database


To sharpen diagnostic skill, take a minute to cross check the physical exam against images and lab results in the electronic record - if not congruent, then why?

Many have noted a general decline in physicians’ diagnostic skills. At least some of the decline is attributed to easy access and overreliance on various lab and imaging technologies. As a result, many patients have lab and imaging tests performed even before they have had a thoughtful history and physical exam completed. Too often, this leads to devaluation of the contribution of the history and physical exam such that neither the history nor the physical exam is performed with sufficient thoroughness and precision. I am sure many of us can cite examples in which clues from the history or physical exam led to a diagnosis that would have been delayed or missed if incompletely selected lab studies or poorly directed images alone were relied on. Case in point: a man presented to the emergency department with flu-like illness and generalized joint pain. Were it not for a difficult to find petechial rash and one small pustule on a finger, a test for Neisseria gonococci might have been delayed or missed since he had no other symptoms associated with the infection. In fact, as occurred with this patient, each component of the diagnostic process – the history, physical exam, and lab testing – benefited the patient and contributed to his efficient, safe, and effective care.


In order to build my own physical diagnosis skills I have, ironically, come to rely on the lab and imaging technology that are sometimes overused. Much of it is accessible through the electronic medical record, and in many instances exists as a longitudinal record. Many patients have records stretching over years for comparison to previous results and to records of the history and physical exam. I use it to test or sometimes validate my various skills at inspection, palpation, percussion, auscultation or location of neurological findings, ability to detect pallor or icterus of the conjunctivae, or ability to detect the spleen or liver edge, and so on. I also find it a valuable adjunct to teaching the value of thorough and precise exam. Let me illustrate.


A middle aged, slender woman was admitted to our Internal Medicine residency inpatient service. She had experienced syncope and was found to have had a pulmonary embolism. Hypertension was noted as part of her past medical history. Heart exam was described as normal by the admitting team. Electrocardiogram (ECG) met voltage criteria for left ventricular hypertrophy, the heart silhouette appeared borderline enlarged on the PA and lateral chest x – ray (CXR). We went the bedside. In fact her heart exam was not entirely normal. There was a subtle abnormality, unrelated to her small pulmonary embolus, but nonetheless, important. Her point of maximal impulse (PMI) was displaced 2 cm to the left of the midclavicular line. Her heart rate was 90, blood pressure 160/87. The rest of her cardiac and vascular exam, including the fundi was normal.


Our first concern of course was her pulmonary embolism. However, the finding on cardiac exam was also important, and an opportunity for the team to appreciate the value of careful examination and attention to detail. At the bedside, the team was able to correlate the physical finding (a displaced PMI) with the finding on CXR and the ECG. Later in the conference room, we looked back over previous ECG and CXR reports and images. We discussed an alternative situation – easily imagined – where she was seen in clinic for initial primary care and management of her hypertension. A displaced PMI would have been evidence of end organ damage due to hypertension and could have justified more intensive investigation (perhaps an echocardiogram) and intensive effort to assure attainment of blood pressure goals. It was also apparent to us that reliance on ECG or CXR reports alone to assess end organ damage would have been misleading since interpretations of past ECGs were sometimes equivocal regarding the presence of increased voltage and CXR interpretations varied. Some CXR reports, even recent ones, were interpreted as having a normal heart size though the cardiac silhouette was more than half of the width of the thorax on posterior anterior projection – evidence of enlargement.


Perhaps in one respect, the access to the digital database of images and laboratory tests is akin to the autopsy of the past. By the time Osler became “Osler” and authored his path setting textbook, “The Principles and Practice of Medicine,” he had performed hundreds of autopsies. Unlike Osler, we aren’t performing autopsies frequently, if at all, but in many instances we have something of considerable teaching value – serial images and lab tests. They are readily available to examine and correlate with our physical exams, giving us an unmatched ability to improve, validate, become more confident in our skills, and sometimes even correct for the limitations of the technologies themselves.