Possible heart attack symptoms require a physician’s attention, but determining who is actually having a heart attack and who is not is challenging. A global research team combined three commonly used tests into a single laboratory score that appears to do a better job of differentiating between higher and lower heart attack risk than do other diagnostic efforts. The news release does a nice job of describing the validation study but could have offered more detail about the make-up of the scoring system and about possible conflicts of interest among the researchers.
About one-in-four deaths in the United States each year stems from heart disease, so quickly diagnosing a heart attack is important. However, not all individuals who go to their doctors with what appear to be heart attack symptoms will experience a heart attack in the short term. Many people arrive in emergency rooms with chest pain. Being able to accurately and efficiently identify very low-risk patients is important because they can be spared additional testing and be can sent home sooner.
Differentiating those at-risk individuals from others is critical, and this relatively simple diagnostic strategy may be another good addition to the physician’s toolkit.
Although the release characterizes the chemistry tests that make up the scoring system as “common,” it offers no information about the cost of each or the likely cost of aggregating them into the recommended scoring system.
The text gets specific about more accurate predictions stemming from the scoring system. But other benefits are not well explained. For example, although the lead of the news release describes the proposed laboratory score as “safer,” it clarifies that statement in a rather murky way lower in the release by noting that the use of an existing diagnostic test, high-sensitivity cardiac troponin, has not been demonstrated to be safe. Why that is the case is never explained.
Also, the study design was not reported in the release. In fact, the study was retrospective, so investigators don’t actually know whether outcomes were better among those who underwent the new test vs. the standard troponin.
The text is specific about the number of heart attacks that would have been missed by the scoring system (1 out of 727) in this population of 4000+ patients who went to emergency rooms with possible symptoms.
This indicates a high sensitivity, which is good because it means that the test is not missing people likely to have a heart attack or die within 30 days. However, the other potential harm from a diagnostic test is from false positive results — which could lead to unnecessary hospitalizations and treatments. This statistic is not addressed in the release, though it is presented in the study.
The reader will encounter useful detail about the number of patients, the employment of data from patients whose heart attack outcomes were already determined, the global nature of the effort, and the relatively short-term nature of the design. It would have been useful, however, to have described the three tests that were combined to create the single predictive score.
Also, the accuracy of the predictive score depends on which cutoff the investigators use (scores ranged from 0 to 5). Rather than reporting statistics such as sensitivity and specificity to compare tests, it would be preferable to report the area under the receiver operating characteristic (ROC) curve — a tool researchers use to evaluate and compare the performance of diagnostic tests. However, this was not done in the study. They also calculated separate scores based on different high-sensitivity troponin assays. This was not addressed in the news release.
No disease mongering here. When a person lands in the emergency room with apparent heart attack symptoms, physicians struggle to identify those who are truly having a heart attack. The release provides some context on the types of tests used to diagnose heart attacks, and where they are lacking.
The funder—the Canadian Institutes for Health Research—is clearly identified. Conflicts of interest among the many coauthors, however, are not mentioned. The original research article offers a lengthy list of potential conflicts; for example, the two coauthors quoted in the news release have begun the process of patenting the laboratory scoring system discussed in the research. And the first author of the paper, one of those quoted, has received research funding and honoraria from a number of drug companies.
The text mentions two alternatives: a single blood test, called a cardiac troponin test, and “multiple blood tests over several hours.” It makes the case that the proposed laboratory score is superior to both.
The release omits discussion of other strategies such as electrocardiograms, exams and patient history typically used in acute treatment facilities. Stress tests are not as readily available. It does raise the question of the timing of the troponin test that was used to calculate the score — ideally, we’d like to know that results were quickly available if we’re seeing this as a way to shorten emergency room stays.
The release notes that the score is based on commonly available tests.
The novel element here is the strategy of combining three tests into a single scoring system. The news release makes that point clearly.
The text and quotes seem appropriate to the research.