Blog | Thursday, July 16, 2015

Lessons learned by studying sore throats for 35 years

Thirty-five years ago I started collecting data in a non-acute emergency room. Over approximately 3 months the residents enrolled slightly more than 300 patients into the initial sore throat study. Spending the next 3 months learning how to analyze the data, I began a long journey that continues today.

Learning medicine rarely includes having epiphanies. Learning medicine requires patients and reading. But we who teach medicine can help our learners speed that process if we help them focus on some key features.

My colleagues and I often cite Judy Bowen's classic article: Bowen, Judith L. “Educational strategies to promote clinical diagnostic reasoning.” New England Journal of Medicine 355.21 (2006): 2217-2225. In that article she introduced many to the concepts of problem representation and illness scripts.

Lesson #1 We need to be precise in defining illness scripts. Problem representations depend upon illness scripts as these 2concepts interact. Back in 1980 when this journey started, I thought of a sore throat rather simply. Today I understand that there are a series of potential sore throat illness scripts.

Script #1 Acute sore throat 3 days or less in duration with or without various other symptoms (cough, coyrza, fever and difficulty swallowing) and various examination features (exudates, adenopathy, measured temperature, erythema, tonsillar swelling). The discomfort should be in the throat, not the external neck.

Script #2 Non-resolving sore throat that starts as an acute sore throat, but worsens rather than improves. It may include asymmetry of the neck, unilateral tonsillar swelling, deviated uvula, persistent fever, rigors and/or night sweats.

Script #3 Neck pain but no actual throat pain. This usually is recognized as different from the other 2 scripts. Script #1 and #2 have very different implications. However, few clinical educators have taught that difference, and probably few have considered sore throats enough to distinguish between scripts #1 and #2, yet the underlying differential diagnoses are quite different.

The problem is actually more complex, because over the years it became quite clear that pre-adolescent pharyngitis differed greatly from adolescent/young adult pharyngitis (Mitchell, Michael S., Annalise Sorrentino, and Robert M. Centor. “Adolescent Pharyngitis A Review of Bacterial Causes.” Clinical pediatrics 50.12 (2011): 1091-1095.) While the general scripts are the same, the potential etiologies differ. Adolescents/young adults have a much broader infectious differential diagnostic spectrum and that differential has major implications.

We should apply the general concepts here to teaching clinical medicine. For example, consider your illness script for community acquired pneumonia. How long should the patient have symptoms and at what point does the history no longer fit CAP? How do classify patients who do not improve with adequate antibiotic coverage? Do we think differently about atypical infections from classic bacterial pneumonia? And you can ask some further questions.

Take chest pain, how many illness scripts could you develop for chest pain? How does one proceed to match these illness scripts against a careful patient problem representation? The illness scripts should influence the data you collect.

We must define data carefully. A wonderful question asked in Japan helped me understand that many listeners had a different understanding of the term data. In medicine I consider 4 classes of data: the history, the appropriate physical examination, laboratory tests with accurate interpretation, appropriate imaging studies.

Each illness script should include at least the first 2 data classes, but often we should add the third and fourth. And the illness script influences how we describe the patient, and what data from classes 3 and 4 we need. But the illness scripts also influence our history and physical data acquisition.

Lesson #2 We must appropriately keep an open mind as we learn new concepts that might apply to the clinical problem. I learned this slowly over the last 13 years as the Fusobacterium necrophorum story started to crystallize. Again I did not have an epiphany, but rather I developed a profound curiosity about this bacteria and its potential role in explaining some adolescent and young adult pharyngitis.

This concept seems simple, but actually we resist changing our understanding of disease. Two great examples come to mind – Helicobacter pylori causing ulcer disease and beta-blockers treating systolic heart failure. We resisted these changes even as the evidence supporting the changes were becoming very clear.

Lesson #3 Demographics can matter. In pharyngitis, I have already written about the differences between pre-adolescents and adolescents & young adults. Some illness scripts differ by gender, or socioeconomic status, or country of origin. We should not ignore these factors.

Lesson #4 Every medical problem is complex. Many physicians and patients consider sore throats as a simple medical condition. But our sore throat illness script should include some red flags. Duration of symptoms is a red flag. Sweats or rigors is a red flag. An asymmetric bulge in the neck is a red flag. Every medical problem has red flags and we must do a better job of teaching those indicators of seriousness.

As educators we have a responsibility to understand clinical problems in a different way and to teach clinical medicine in the way expert clinicians think. I suspect that some readers can identify a few clinical educators who approach clinical teaching in such a manner, but that most educators do not have this understanding. Am I right or wrong?

db is the nickname for Robert M. Centor, MD, FACP. db stands both for Dr. Bob and da boss. He is an academic general internist at the University of Alabama School of Medicine, and is the Regional Associate Dean for the Huntsville Regional Medical Campus of UASOM. He still makes inpatient rounds over 100 days each year. This post originally appeared at his blog, db's Medical Rants.