Serum troponin is widely used in the acute setting to help distinguish between life-threatening conditions; however, serum troponin also represents one of the most widely over-ordered tests in the United States. Mild elevations in serum troponin, when ordered in asymptomatic patients, can lead to unnecessary testing, preventable radiation exposure, misdiagnosis, and financial toxicity to patients who do not have anginal symptoms. In one study, serum troponin testing was performed in 17% of emergency department visits in the U.S., representing 28.6 million visits, and 47% of inpatients (Makam & Nguyen, 2015, JAMA IM). Notably, serum troponin testing was performed in 35% of inpatients without anginal symptoms (Makam & Nguyen, 2015, JAMA IM). Therefore, there is an urgent need to increase awareness about the role of conditional probability in the use and interpretation of this test.
Evidence-based measures of diagnostic accuracy such as positive/negative predictive values and positive/negative likelihood ratios are usually not available at the point of care in the clinical setting. PPV/NPVs and +LR/-LRs run counter to the default approach for arriving at diagnoses using heuristics and illness scripts, methods that are susceptible to cognitive biases (Manrai et al, 2014, JAMA IM). Therefore, we embarked on a mission to design an interactive tool that will make it easier for clinicians to integrate measures of diagnostic accuracy into their understanding of myocardial injury. Ironically, the prevalence of mild troponin elevations with high-sensitivity assays has heralded an era in which the physical exam and the history of present illness — which can only be obtained by directly interviewing and examining the patient — becomes of particular importance in teasing apart myocardial injury without myocardial ischemia versus myocardial injury with myocardial ischemia.
Thirdly, there are a number of populations in which the interpretation of elevated troponins differs from that of the general population, such as patients on immunotherapy, patients with congenital heart disease, or patients with decreased kidney function. The presence of these conditions makes it difficult to interpret troponin elevations for cardiologists and non-cardiologists alike. Therefore, the final pillar of this project is advancing education on special populations and the significance of dynamic or static troponin values in these populations.
All members of the team are volunteers who contributed to the project alongside full-time clinical rotations. We hope that this resource will be useful to clinicians everywhere. Please do not hesitate to send us feedback or let us know if you would like us to speak at your institution about this initiative.
Ruey Hu, MD, MPH
on behalf the Yale Troponin Team