Hot Off the Press: Troponin Testing and Coronary Syndrome in Geriatric Patients With Nonspecific Complaints: Are We Overtesting?

Nonspecific complaints, such as “fatigue,” “dizziness,” and “not feeling well” are relatively common among elderly patients presenting to the emergency department (ED).1, 2 The appropriate workup of such complaints is not always evident. When employed in populations with very low risk of disease, our tests can result in many more false positives than true positives, frustrating physicians and potentially harming our patients. However, a significant number of elderly patients presenting with nonspecific symptoms are ultimately diagnosed with serious or life-threatening conditions.3 Elderly patients diagnosed with acute coronary syndrome (ACS) are more likely than younger patients to present without chest pain, including up to 20% who present with weakness as their chief complaint.4, 5 Therefore, the troponin presumably has some role in the workup of elderly patients presenting with nonspecific symptoms. This study by Wang and colleagues uses a chart review to explore the frequency of ACS in elderly patients presenting with nonspecific complaints and the utility of troponin testing in this population. This is a chart review study that identified patients aged 65 years and older presenting to the ED with a nonspecific chief complaint in whom the treating physician decided to order a troponin. Of 594 elderly patients with nonspecific complaints, troponins were ordered in 416 (69%), of whom 52 (12%) were positive. However, only five patients were determined to have ACS. The ED troponin was 80% sensitive and 88% specific for ACS, with a negative predictive value (NPV) of 99.7%, but a positive predictive value (PPV) of only 7.7%. Strategies to improve the accuracy of medical chart reviews include training abstractors before starting the study, explicitly defining inclusion and exclusion criteria, precisely defining variables of interest, using a standardized abstraction form, holding periodic meetings to resolve disputes and review coding rules, monitoring the performance of the abstractors, blinding the abstractors to the study hypothesis, and testing inter-rater agreement between multiple abstractors.6 Impressively, this study utilized all of these recommended methodologic techniques. However, there are inherent limitations to chart review methodology that limit the reliability of the reported findings. Clinicians may have altered the chief complaint after the troponin value was reported. For example, in the context of a positive troponin, the physician might decide to emphasize the fact that the patient mentioned chest pain, despite an initial chief complaint of “fatigue” or “multiple complaints,” resulting in the exclusion of patients in whom troponin testing was truly valuable. Furthermore, the retrospective look at these data necessitated a dichotomous approach to troponin values (in which the test was considered either positive or negative at a specific cutoff). However, even before the introduction of high-sensitivity troponins, the troponin result has always had to be interpreted within the clinical context and often considering the trend of multiple values. In a chart review, we lose this crucial aspect of clinician judgment and therefore do not know how these troponin results would have been interpreted in real practice. For these reasons, although we believe the general conclusion that troponin testing in this population is low yield with many false positives, the precise numbers reported should be interpreted cautiously. They initially identified 1,146 potentially eligible patients. After excluding the patients who had a specific complaint listed and those with documented fever, they were left with a total of 594 patients. Of those, 412 (69%) had troponins ordered. The average age of the cohort was 78 years old, 58% were female, and 75% were admitted to hospital. The most common chief complaints were altered mental status (43%), weakness/fatigue (33%), and dizziness (21%). The troponin was positive in 52 patients (12.6%), but only five (1.2%) were adjudicated as having ACS at the index visit or within 30 days. Focusing specifically on the first troponin in the ED, this results in a sensitivity of 80%, a specificity of 88%, a NPV of 99.7%, and a PPV of 7.7%. Considering all troponins, the sensitivity was 100% (95% confidence interval [CI] = 48% to 100%), the specificity was 81% (95% CI = 77% to 85%), the NPV was 100%, and the PPV was 6.1%. As a general rule, tests perform poorly in populations with a very low prevalence of disease. ACS is relatively rare in patients without chest pain or dyspnea, and therefore we should expect the troponin to be a low value test in this population. However, ACS is more common in elderly patients, and the elderly are also more like to present with atypical symptoms, such as “weakness.”4, 5 The limitations of chart review data preclude any definitive rules, so clinicians will have to continue to use their clinical judgment to determine whether the nonspecific symptoms in the patient in front of them warrant testing for possible ACS. Dr. Art Sanders on the SGEM blog: I am concerned about the 19% mortality of elder patients with no documented ACS (15 deaths of 77 patients). The authors note “There is a wide body of literature demonstrating that troponin elevation predicts worse outcomes in a variety of noncardiac conditions.” Patients died from a variety of other causes such as sepsis, dehydration, etc. Would these other conditions have been diagnosed and warrant admission if the troponin had not been tested? The lesson might be that, especially with high sensitivity troponin, we may need to think differently about troponin testing – as an ED doctor, I may not diagnose the patient as NSTEMI, but admission to assess for the other serious conditions is warranted. If we discourage troponin testing, we may not be aware of some of these serious conditions, especially with atypical presentations in elder patients. Dr. Wand responds: We did not specifically look into how troponin would perform as a prognostic factor. However, I did go back into the original data set to look at the original presentations and reasons for admission… None of these 16 patients who had an initial troponin elevation and died within 30 days of the index visit had just an elevated troponin as cause of admission. All of them had other reasons that were found during their ER visit that cause them to be admitted. Again, I am unable to fully comment on whether the troponin added valuable prognostic information or risk stratification on top of the other tests that were abnormal for these patients as that was not our paper's intention. But I think your question is a interesting one as a potential follow up study- why do physicians order troponins- is it for prognostic purposes? risk stratification? to help admit? or concern for ACS? Christian H. Nickel (@replynickel): Great study, great podcast @TheSGEM – topic merits prospective study. HsTroponin algorithms (such as 1 hr rule out) are derived from patients with “symptoms suggestive of ACS” – should (IMHO) not be extrapolated to NSCs [nonspecific complaints]! Dr. Ken Milne – EBM and Rural (@TheSGEM) responds: Good history, followed by a directed physician examination and then judicious use of investigations. #SGEMHOP Daniel Jafari (@DanielJafari): Interesting paper. I wish the poll was a bit more nuanced (well appearing, I'll appearing). Paper's pop admitted 75% of the time, so likely sicker. Also, worth noting single trop 100% sensitive. Question remains: how many unnecessary procedures for 93% FP trops. Dr. Ken Milne – EBM and Rural (@TheSGEM) responds: Wait until high-sensitivity is widely available in North America. Combine that with an aging population, indiscriminate testing and a zero-miss culture, I’m concerned. Marc A. Probst (@probstyMD) responds: I agree, there is definitely a risk of doing harm by over-testing and overreacting to “positive” HS-trop. We may need to rethink the clinical significance of a positive trop with these high-sensitivity assays… more of a prognostic factor than diagnostic factor in many cases. Rick Body (@richardbody) responds: Definitely. The advance in the assays needs to be matched by an advance in our thinking. We need to understand more about what troponin is telling us. It's a marker of myocardial injury. Our job is to understand what caused that The yield of troponin testing is low in elderly patients presenting with nonspecific complaints, and there are many more false positives than true positives. However, limitations in this study prevent any strong recommendations for practice change. Physicians will need to continue to apply clinical judgment in deciding which patients require a workup for ACS.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.