As hematologists, we are privileged to work at the forefront of the molecular age of medicine. Diagnostic, prognostic, and predictive genomic information has been incorporated into our clinical practice for years and, most recently, has led to development of targeted therapies and incredible improvements in patient outcomes. The flood of data gathered through proteomics, glycomics, metabolomics, and genomics, coupled with the analytic capability to crunch all these data, holds the promise to help us deliver true precision medicine.
My fellows and I often joke that, in the near future, when patients come to the hospital, they will immediately undergo deep sequencing and whole-body PET/CT/MRI scans – regardless of the chief complaint.
Amid this technologic revolution, I fear that the ability to obtain and integrate a well-taken medical history is becoming lost and devalued, though. Here are three short vignettes, culled from my own referrals, as examples.
A 51-year-old African-American woman self-referred for recommendations regarding care of her sickle cell disease (SCD). She was diagnosed at age 48 after presenting with intermittent wrist, elbow, and knee pain accompanied by redness. The pain was so severe that she frequently visited the emergency department. She had a healthy childhood and was previously diagnosed with essential hypertension, for which she was on a diuretic. Her complete blood count revealed mild normocytic anemia. An outside hematologist interpreted her hemoglobin analysis as showing SCD and prescribed hydroxyurea.
Having no family members with SCD, the patient did not think her history was consistent with the diagnosis; after obtaining her history, I agreed. A repeat hemoglobin analysis showed hemoglobin A levels of 56 percent and hemoglobin S levels of 38 percent. Her rheumatoid factor levels were moderately high. I advised her to discontinue her hydroxyurea and see a rheumatologist.
“As we integrate the remarkable new tools of molecular medicine into the care of our patients, I hope we remember the oldest and perhaps most useful “-omics” of them all: history-omics.”
A 34-year-old woman was referred for advice regarding therapy for her essential thrombocythemia. She had a history of intermittent headaches and facial paresthesias that had been diagnosed as migraines. She also had a persistent platelet count of around 700,000/µL for several years. Molecular analyses for JAK2, MPL, and CALR mutations were normal, and her bone marrow biopsy showed increased megakaryocytes without fibrosis. She had poor tolerance to medications and eventually received interferon and low-dose hydroxyurea to lower her platelet counts to between 400,000/µL and 500,000/µL. Her headaches did not improve.
A history taken by the resident in the clinic revealed that the patient had been involved in a motor vehicle accident when she was 8 years old and had required splenectomy. This had not been previously noted in her chart.
Our examination of the peripheral blood smear revealed increased, occasionally large platelets and Howell-Jolly bodies. We felt that the patient had post-splenectomy thrombocytosis and discontinued her interferon. After discontinuation, she felt markedly better.
A 38-year-old man was referred for advice regarding continued anticoagulation. Two years earlier, he had been hospitalized for acute cholecystitis and developed sudden shortness of breath after cholecystectomy. After clinicians found a segmental pulmonary embolism (PE), the patient was treated with heparin, then transitioned to warfarin, which he had been taking ever since.
Thrombophilia tests at the time of the PE showed heterozygous factor V Leiden (FVL) and he was told he required lifelong anticoagulation. It became clear to us that his PE was provoked, and that the presence of heterozygous FVL was not an indication for lifelong therapy. He decided to discontinue anticoagulation after a discussion of the risks and benefits of continued therapy.
In each of these cases, the history alone was highly suggestive of the correct diagnosis or appropriate course of treatment. Misinterpretation of a laboratory test or a test result’s implication led to inappropriate therapy.
In the nearly 30 years since I began my fellowship, the age of molecular hematology has fully blossomed. Diagnostic conundrums have become easier to solve, genetic profiling has clarified prognosis, and a dizzying array of molecularly targeted therapies has improved survival and quality of life. I think back to the patients I lost earlier in my career and imagine, after only a few more years, how different their outcomes could have been. I certainly don’t long for the “good old days,” when we relied so heavily on intuition and experience.
However, the ability to obtain a medical history, integrate those findings with imaging and/or laboratory data, and synthesize the information into a coherent story continues to be an essential skill for a clinician.
As illustrated by the above vignettes, previous diagnoses should be approached with healthy skepticism – not to question the competence of others, but because it is a consultant’s job to independently verify. I have often questioned my own previous notes and found errors. Sifting through a voluminous medical record (even an electronic one) can be a monumental task, so I have tried to focus only on pathology, radiology, laboratory reports, and treatment flow sheets, if available.
Finally, to summarize using a phrase I am sure my fellows have heard often, “if all else fails, talk to the patient.” As I progressed in my career, I was able to ask more nuanced questions focused on patient experience, from which I could discern a more likely diagnosis.
Of course, any disorder can present with an atypical history, but typical clinical courses are called that for a reason. And, although it is challenging to memorize a textbook’s worth of disease signs and symptoms, I can almost always remember the signs and symptoms related to me by actual patients.
As we integrate the remarkable new tools of molecular medicine into the care of our patients, I hope we remember the oldest and perhaps most useful “-omics” of them all: history-omics.