Practice beyond the microscope, a memoir: My years as a doctor’s doctor

Joseph H. Keffer, MD

February 2021—When I entered the practice of pathology the role was described as being a “doctor’s doctor.” That reflected the fact that physicians turn to the laboratory for so many tests that define diseases and determine the response to treatment, a reality true today. But my memories are of the more personal interactions with clinicians over the years as I pursued the practice of pathology beyond the microscope, and often beyond the laboratory.

Today, medicine is highly specialized with subspecialties in every field. These physicians are so well trained that it may seem as though there is no need for the pathologist beyond the walls of the laboratory. Indeed, many common medical problems are so well researched that standard practices in diagnosis and treatment are universally available. Although I have been retired and removed from the modern hospital environment, I wonder about the complexities of molecular diagnostics today, CAR T-cell therapies, diagnostic tests for COVID-19 and its antibodies, and other breakthrough areas of medicine. I suspect there remains much that the doctor’s doctor can contribute. You just have to look, stick your neck out, and become involved.

My experience along these lines began as a young resident at Walter Reed Army Medical Center during the Vietnam War in its early stages, the ’60s. Many young men were being admitted to “Reed” with severe trauma and infected wounds, often presenting with spiking fevers, which raised the question of whether that was due to an infected wound or malaria, a disease we had not had much experience with since World War II. On my hematology rotation, I encountered many clinicians pressing us to find the malarial parasites on blood smears. Our lab had lost the art of the “thick and thin” smears formerly, the traditional way to diagnose the presence of diagnostic forms of the parasite. Thick smears were designed to layer the red blood cells so that one could hemolyse them, making it possible to see low numbers of these parasites that would establish the diagnosis. At the time, our lab was measuring the white blood cell count by an advanced method that used saponin to lyse the red blood cells, permitting a particle counter to enumerate the remaining white blood cells. It didn’t take much experimentation to devise the conditions to provide a method to detect very few parasites by centrifugation, allowing the clinicians to rule out malaria and rule in other forms of infection. This test endeared us to the trauma surgeons and infectious disease physicians challenged with the differential diagnosis of many patients. The method led not only to my first publication in a medical journal but to active participation with the clinical malaria team approach at Reed.

In a later phase of my training, while rotating in immunology, I had the good fortune to assist a young, recently drafted pathologist who had just completed his training at Columbia University. He was setting up a new test for systemic lupus erythematosus. After learning how to spell it, and marveling that my mentor had traveled to Baltimore to a meatpacking plant to obtain a calf thymus gland, I assisted him in cutting the tissue into 1-cm cubes and freezing them for subsequent use in making imprints on slides to establish the first antinuclear antibody test at Reed. We used immunofluorescent antibodies to identify patterns of nuclear staining along with the titer of the antibodies found to be virtually diagnostic of a disease that was readily recognized in its full presentation but difficult to recognize in the majority of patients with early or partial features. Rheumatologists were generous in extending appreciation for this advance, including the progressive sophistication of our recognition of different patterns of nuclear fluorescence in lupus and related collagen diseases. It was a lesson that there is much more to communication with clinicians beyond the printed result on a laboratory report.

After my service in the Air Force, I was privileged to enter private practice in South Carolina with an outstanding group of young colleagues, led by Joe Black, MD, who had established a reputation for high professional standards. Among their innovative practices was the inclusion of an admission complete blood count and, the classic for its time, the 12-test biochemical profile with review of each abnormal blood smear and the pattern of the biochemical tests by a pathologist. Each afternoon, the pathologist of the day would sign out abnormal blood smears and the abnormal biochemical scans. Often, the findings were correlated between the morphology and biochemistry. When striking findings would be recognized, a personal call would be made to the admitting physician informing her or him of the potential diagnoses raised by these findings. There were numerous cases of iron deficiency, vitamin B12 deficiency, hemolytic disease, leukemias, sepsis, hypothyroidism, hepatitis, occult myocardial infarction, and many other diagnoses raised for early confirmation and other related tests. While many of these were suspected, many were not, and it became apparent to us that many clinical problems were not recognized without our interventions.

The experiences with recognizing silent or atypical myocardial injury at a time when nonspecific tests such as the total creatine kinase, transaminases, and lactic dehydrogenases were the predominant laboratory tests to complement the electrocardiogram led us to pursue the more reliable creatine kinase isoenzymes early in their availability. I had set up the first total creatine kinase test at Reed in my rotation on clinical chemistry while a resident. We had thought at the time that it was a marvelous advance over the other tests, and it was, but only after later establishing the isoenzymes did we recognize how misleading the total was in isolation. This was a quantum leap in diagnostic sensitivity and specificity. It would lead later to my early investment in developing the cardiac troponin test while in Dallas at the University of Texas as director of the medical laboratories at Parkland Hospital. Finally, the test was truly sensitive and specific, sufficiently so that one prominent cardiologist, mockingly and skeptically, referred to our troponin results in a patient as a “Joe Keffer 5 gram Infarct.” It really was. He was wrong. Once again, there was demonstrated a need for an intimate investment in collaboration with the clinicians.