Vinay, I am a hematologist, not a cardiologist, but this could be seen across the room by any competent physician. My guess is they did one of those "fake" physicals which seem so ubiquitous these days where you wave a stethoscope at the patient through their shirt, glance at lab values, and make another appointment. Glad (if it is as reported, which I doubt) that they gave it to the AI which found something obvious. It is not clear why they still have jobs.
This is more a reflection on the current pool of doctors admitted for how many weeks they spent living in the trunk of their car rather than ability to be decent physicians which requires a certain level of mental acuity clearly missing here. But that would be a whole other article. Nothing more sobering than being on a medical school admissions committee these days.
Embarrassing: "... a mildly increased marker of cardiac injury." Sounds more like a sentence from Newsweek than an ostensibly peer-reviewed scientific journal. Moreover, any MD with acute care training should know a 45-year old with new SOB cannot reasonably be diagnosed with new-onset asthma until CHF is ruled out. Add a grossly abnormal EKG and 'marker', and it's a lock—admit for cardiac work-up. The Times has been irresponsible with medical coverage for years, and needs an educated MD to screen their work. It's just one reason I ended my subscription. Vinay: take the job, straighten them out.
This is really a disconnect between the physician at a bedside in the clinic and what constitutes as “novel” by the editorial staff at Nature Medicine.
This is shameful.
I suspect they (the physician-scientists who comprise the Nat-Med editorial staff) have not personally seen or worked up a patient in decades.
I don't really believe some of these stories. Did you see the one where supposedly there was a guy who melted his dog's tumor using AI-generated mRNA cancer vaccine targeted to his dog's specific tumor? This story was widely circulated a few months ago and I have heard zilch since. Like if that were real, wouldn't we have numerous other success stories like it by now?? I think a lot of the hype around AI is just straight fake unfortunately. Gotta justify those big IPOs/valuations, right?
I checked the author guidelines & it turns out Nature Medicine doesn't take case reports & this is a "Correspondence" article, which has almost no requirements other than length of article etc. and this description: "The Correspondence section provides a forum for discussion or to present a point of view on issues that are of interest to the readership of Nature Medicine. Correspondences should not contain new research data, nor should serve as a venue for technical comments on peer-reviewed research papers..." The rest reads to me as though it's possible for an article to be screened for interest & edited by the "editorial team," whoever they are (Media Relations??), without any further input from authors. Maybe the authors wanted more clicks than they thought a case report could provide, so they published it as a quasi-"news article," except it happens to be in Nat Med, & NYT did the rest. But the thing that was really noticeable was that more than twice as much article real estate is spent talking about their AI models for ECG & stethoscope, their 2 clinical trials evaluating those (this patient became part of the AI-enabled ECG study), this was the first & that was the first, than about the patient, who is kind of a prop for their policy pitch. No emphasis on the disease per se or what we could learn about diagnostics from the uniqueness of the case or the "atypical presentation." Rather, what are the barriers keeping AI from being ubiquitous in EDs?
Vinay, I am a hematologist, not a cardiologist, but this could be seen across the room by any competent physician. My guess is they did one of those "fake" physicals which seem so ubiquitous these days where you wave a stethoscope at the patient through their shirt, glance at lab values, and make another appointment. Glad (if it is as reported, which I doubt) that they gave it to the AI which found something obvious. It is not clear why they still have jobs.
This is more a reflection on the current pool of doctors admitted for how many weeks they spent living in the trunk of their car rather than ability to be decent physicians which requires a certain level of mental acuity clearly missing here. But that would be a whole other article. Nothing more sobering than being on a medical school admissions committee these days.
Embarrassing: "... a mildly increased marker of cardiac injury." Sounds more like a sentence from Newsweek than an ostensibly peer-reviewed scientific journal. Moreover, any MD with acute care training should know a 45-year old with new SOB cannot reasonably be diagnosed with new-onset asthma until CHF is ruled out. Add a grossly abnormal EKG and 'marker', and it's a lock—admit for cardiac work-up. The Times has been irresponsible with medical coverage for years, and needs an educated MD to screen their work. It's just one reason I ended my subscription. Vinay: take the job, straighten them out.
This is really a disconnect between the physician at a bedside in the clinic and what constitutes as “novel” by the editorial staff at Nature Medicine.
This is shameful.
I suspect they (the physician-scientists who comprise the Nat-Med editorial staff) have not personally seen or worked up a patient in decades.
I don't really believe some of these stories. Did you see the one where supposedly there was a guy who melted his dog's tumor using AI-generated mRNA cancer vaccine targeted to his dog's specific tumor? This story was widely circulated a few months ago and I have heard zilch since. Like if that were real, wouldn't we have numerous other success stories like it by now?? I think a lot of the hype around AI is just straight fake unfortunately. Gotta justify those big IPOs/valuations, right?
I checked the author guidelines & it turns out Nature Medicine doesn't take case reports & this is a "Correspondence" article, which has almost no requirements other than length of article etc. and this description: "The Correspondence section provides a forum for discussion or to present a point of view on issues that are of interest to the readership of Nature Medicine. Correspondences should not contain new research data, nor should serve as a venue for technical comments on peer-reviewed research papers..." The rest reads to me as though it's possible for an article to be screened for interest & edited by the "editorial team," whoever they are (Media Relations??), without any further input from authors. Maybe the authors wanted more clicks than they thought a case report could provide, so they published it as a quasi-"news article," except it happens to be in Nat Med, & NYT did the rest. But the thing that was really noticeable was that more than twice as much article real estate is spent talking about their AI models for ECG & stethoscope, their 2 clinical trials evaluating those (this patient became part of the AI-enabled ECG study), this was the first & that was the first, than about the patient, who is kind of a prop for their policy pitch. No emphasis on the disease per se or what we could learn about diagnostics from the uniqueness of the case or the "atypical presentation." Rather, what are the barriers keeping AI from being ubiquitous in EDs?