Critical Appraisal of the Medical Literature

[JWebar]

Dear SDN colleagues:

Doing my residency and now in my fellowship, I'm starting to get more and more into evidence based medicine. But even with journal clubs and such, im starting to realize this subject is much more complex that people usually give if credit for. It's so easy sometimes to get fooled by a faulty rational and skewed conclusions, and even severely biased articles sometimes end up in the "big journals", so it's up to us to fight our way through the jungle.

I was wondering if anyone knows of a good course (online hopefully, free or payed) that teaches you how to really stripe down an article and to analyze it thoroughly.

As always, thanks in advance 🙂

---

Members don't see this ad.

 

[BLADEMDA]

Dear SDN colleagues:

Doing my residency and now in my fellowship, I'm starting to get more and more into evidence based medicine. But even with journal clubs and such, im starting to realize this subject is much more complex that people usually give if credit for. It's so easy sometimes to get fooled by a faulty rational and skewed conclusions, and even severely biased articles sometimes end up in the "big journals", so it's up to us to fight our way through the jungle.

I was wondering if anyone knows of a good course (online hopefully, free or payed) that teaches you how to really stripe down an article and to analyze it thoroughly.

As always, thanks in advance 🙂

Here is my advice: take every article with a grain of salt. The best journal is the NEJM and I am more willing to change my practice after a large, peer reviewed study in that journal. Still, I like to see at least 1 or 2 other studies confirming the conclusion of the one in question. Finally, does the study line up with at least some anecdotal data/experience?

The more you are at this game the more you see the data tilt one way then tilt back again 20 years later. Be skeptical but remain open to new ideas.

I am not recommending you complete your training and NEVER change a thing in your practice over the next 20 years. On the contrary, what I am saying is don't abandon everything you have learned and practiced for the latest new article on the topic. Remain skeptical but open to new ideas, concepts, etc.

 

[Aether2000]

It is important to remember a couple of things about EBM: 1. The lack of Level 1 evidence is not at all the same as the absence of EBM supporting a technique or medication or approach. If all that is available is Level 3 evidence that supports such and there is no Level I, Level II, or equivalent Level 3 evidence to refute those studies, then you have achieved the premise of EBM- that the highest level of evidence is evaluated and is supportive (even if insurance companies disagree due to their own perverse and profit driven method of evaluating EBM). 2. The presence of one Level 1 study cannot absolutely refute 20 other lower level (2 and 3) studies that drew the opposite conclusion. In such cases, wait for corroborating studies. 3. Statistical level of evidence supporting a study without any analysis of clinical efficacy (e.g. NNT, effect size and its precision (such as 95% confidence interval) for each primary and secondary outcome, MIC, MCID, and CMD, SEM, etc) make the study useless in EBM. There absolutely must be some level of clinical effectiveness measure employed in clinical studies. 4. Bias, especially in GRADE recommendations, is huge. The data may not at all support GRADE recommendations rendered by authors that desire a specific outcome, and in fact may be contrary to the data.

 

[BLADEMDA]

It is important to remember a couple of things about EBM: 1. The lack of Level 1 evidence is not at all the same as the absence of EBM supporting a technique or medication or approach. If all that is available is Level 3 evidence that supports such and there is no Level I, Level II, or equivalent Level 3 evidence to refute those studies, then you have achieved the premise of EBM- that the highest level of evidence is evaluated and is supportive (even if insurance companies disagree due to their own perverse and profit driven method of evaluating EBM). 2. The presence of one Level 1 study cannot absolutely refute 20 other lower level (2 and 3) studies that drew the opposite conclusion. In such cases, wait for corroborating studies. 3. Statistical level of evidence supporting a study without any analysis of clinical efficacy (e.g. NNT, effect size and its precision (such as 95% confidence interval) for each primary and secondary outcome, MIC, MCID, and CMD, SEM, etc) make the study useless in EBM. There absolutely must be some level of clinical effectiveness measure employed in clinical studies. 4. Bias, especially in GRADE recommendations, is huge. The data may not at all support GRADE recommendations rendered by authors that desire a specific outcome, and in fact may be contrary to the data.

Summarize: Most studies are bull****.

Begin your practice as your were trained to do so but remain open to new ideas, concepts, etc.

 

[Aether2000]

Yepper!!!