This short news story focuses on an experimental diagnostic test that researchers hope can be used to detect stomach and esophageal cancers. The test would work by measuring chemicals found in a patient’s breath that suggest the presence of a cancer.
The story wisely notes that the findings are preliminary and that more work needs to be done before the test can be used on patients outside of clinical trials. We applaud the story for taking a cautious approach to the research.
On the other hand, the story leans heavily on a news release (which we also reviewed) and includes no independent perspective on the research. The story also could have been made stronger by including information on cost, the study design behind the findings, and how the research fits into the broader suite of research on using breath to diagnose cancer in general and stomach cancer in particular.
As we also noted in our review of the news release, we think the story spent too much time speculating on the benefits of this test, when the research just doesn’t support those speculations.
New tools that can help us more easily identify health problems can be exciting, and have the potential to significantly impact many lives. But news stories ideally explain to readers exactly how new diagnostic tools could make a difference, and how big that difference might be. For example, what could early diagnosis actually mean for patient survival and well-being?
This is a tough call, since this is early research. But heeding our mantra “if it’s not too early to talk about benefits, then it’s not too early to talk about costs,” an estimate would be useful here.
We were glad that the story did provide a number, which was that the test seemed to be “85 percent accurate.” But, that’s not enough information, because this is trickier than it sounds. Health professionals need to know both a diagnostic tool’s sensitivity — which is how good the test is at positively identifying people who actually have the problem — and the tool’s specificity – how good the test is at ruling out people who don’t have the problem. Presumably, the breath test has an 85 percent success rate in regard to its sensitivity (though that’s not entirely clear). But we don’t know the tool’s specificity — and it’s difficult or impossible to tell how useful a diagnostic tool is without understanding both. (According to the news release, the test had a sensitivity of 80 percent and a specificity of 81 percent.) And, as well, this test was performed on people who had advanced cancer–not early cancer–which may not be as detectable, and the numbers may be different.
Yet, the story quotes a researcher (in a line from the news release) as saying that the breath test could mean earlier diagnosis and treatment of stomach and esophageal cancers, with a higher survival rate. This is speculation–we don’t know if this test picks up early forms of these cancers, nor if earlier detection means a higher rate of survival, or just getting bad news earlier. The accuracy of the test when used to detect early cancer in healthy-seeming people is also likely to be far lower, as our review of the news release points out in some detail.
The test itself–if an exhaled breath is all it involves–is unlikely to carry immediate harms. But with any lab test used to detect disease, the potential harm is if the test fails the patient–by either flagging them as having cancer when they don’t, or flagging them as not having cancer when they do. The first scenario can cause anxiety and lead to extra, unnecessary testing, and the second means the cancer is going untreated. We always hope to see a nod to this in any story about testing, and it’s why we often say “there is no such thing as a simple medical test.”
As noted above, the story suggests that the results of this trial provide evidence that a breath test could detect early stage cancer. However, that is not what the trial tested.
In addition, the story tells readers that the study involved “more than 300 patients.” But we need to know more: How many of those patients had stomach or esophageal cancer? This makes a big difference in terms of understanding the foundation for the claim of 85 percent accuracy — whether one is talking about specificity or sensitivity. (According to the news release, 163 of the patients involved in the study had been diagnosed with stomach or esophageal cancer, while 172 showed no evidence of cancer when they had an endoscopy.)
No disease mongering here. We were glad to see the story discuss the prevalence of these cancers.
The story does not appear to include input from independent sources.
The story made it clear that this test is to help improve upon the current method of endoscopy: “A breath test could be used as a noninvasive, first-line test to reduce the number of unnecessary endoscopies.” It could have been clearer that finding cancers “earlier” — especially in healthy-seeming people — would be a huge leap from what is demonstrated here.
The story does a good job here. It not only notes that “the test needs to be validated in a bigger group before it could be used on patients,” but makes clear that the study findings were presented at a conference. Further, the story explains that “Study results presented at meetings are generally considered preliminary until they’ve been published in a peer-reviewed journal.” That’s valuable context for readers.
This isn’t the first research to address the possibility of using breath tests to diagnose cancer, nor is it the first research to look at using breath tests to diagnose stomach cancer. A quick search finds other journal articles that use various techniques to approach the challenge of diagnosing cancer using a patient’s breath (like this one, or this one). This may well be the first breath test to look at these specific chemicals in the breath, or the first breath test to show this degree of accuracy. However, we don’t know that, because the story doesn’t place the new research into context: How does it fit into the broader continuum of research on using breath tests to diagnose cancers?
The story appears to be based largely, if not solely, on a news release issued by the European Cancer Organization, which organized the conference where the research was presented. See our review of the news release.