An older article, but interesting in light of the replicability crisis that has come out since. FTA:
"It is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgment of trusted physicians or authoritative medical guidelines. I take no pleasure in this conclusion, which I reached slowly and reluctantly over my two decades as an editor of The New England Journal of Medicine."
Physician here (intensive care, about 10 years of practice).
> It is simply no longer possible to believe much of the clinical research that is published
Absolutely agree. I've tried arguing this with colleagues—that we should be more careful about the clinical conclusions we draw from single studies—but I'm just met with resistance. We just added a new drug to the standard treatment of heart failure, based on a shady Novartis-funded study that decided, in a post-hoc analysis, to exclude something like 50% of the patients after-the-fact (and oh look! now we can show a mortality benefit!). It spits in the face of everything I've learned about the scientific method and clinical research. This is the PARADIGM-HF study and the drug is ENTRESTO in case anyone wants to look into it more. Caveat: it might be a beneficial drug for many patients.
> It is simply no longer possible...to rely on the judgment of trusted physicians or authoritative medical guidelines
I'm going to disagree with the author on this one. It is possible to trust and rely on your physician, and in many cases you have no choice. Guidelines are just that: guidelines. Not a replacement for careful thought. And I actually like the direction that many guidelines are adopting: one that brings more continual minor updates rather than infrequent major changes. This is a complicated discussion, and depends on what illness you have that leads to your interaction with our health care system, but my advice is to seek out doctors that seem to communicate well with you.
From briefly reading the paper and the post-hoc analysis, it seems that they were very clear about their post-hoc analysis and its limitations. In this situation, whoever are reading into it more than they should (could be novartis and their marketing, could also be the physician colleagues of yours) are the ones that should be criticized, not the authors themselves.
Edit: here are some quotes from the author of the post-hoc analysis
"Referring to the fact that these observations came from a post hoc analysis, Pfeffer said: 'I normally draw a line here, and I wouldn't cross this line if we had things to do for these patients.'
But he added: 'This is a growing part of the heart-failure syndrome. And if we have something that can help these 40% to 50% of people with symptomatic heart failure and an impaired prognosis—if we can improve their prognosis and take care of the safety measures—then I will go below this line by stating that our observations in the Americas—that spironolactone was associated with reduced CV deaths and hospitalizations for CHF—should be taken into account.'"
Your doctor Googling for 30 seconds (not quite on the correct topic) is no substitute for careful research, either. I trust my doctor, more or less, on many matters; but never more on prescriptions, or surgery (a whole other scandal.)
I'm an anesthesiologist. I'm always interested to hear the story of colleagues on hackernews. Would like to ask you a few questions about your projects. Reach me on albin.stigo@gmail.com
"It is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgment of trusted physicians or authoritative medical guidelines. I take no pleasure in this conclusion, which I reached slowly and reluctantly over my two decades as an editor of The New England Journal of Medicine."