Subscribe to DSC Newsletter

Much of the evidence that evidence-based medicine assumes is baloney

I, and many others, for some time have said that the concept that medicine is practicing on the basis of 'Evidence Based Medicine' is a 'Joke' !!  Period.

Now, a recent paper from SCIENCE NEWS clearly backed this up, calling the evidence for EBM 'Baloney'....

Another recent paper addresses it this way: Evidence based medicine: a movement in crisis?"

FULL CITATION:  BMJ 2014; 348 doi: http://dx.doi.org/10.1136/bmj.g3725 (Published 13 June 2014) Cite this as: BMJ 2014;348:g3725

This paper goes on to suggest that we "make a RETURN to developing REAL EVIDENCE BASED MEDICINE" ...... and in their conclusion discuss, to paraphrase:

"Although some  argue for the rejection of evidence based medicine as a failed model,  (our group) argues for a return to the movement’s founding principles—to individualise evidence and share decisions through meaningful conversations in the context of a humanistic and professional clinician-patient relationships........for this many stakeholders will need to be involved, working together:  patients, clinicians, educators, producers and publishers of evidence, policy makers, research funders, and researchers from a range of
academic disciplines......................."

SO MAYBE THERE IS HOPE ...... although this last source does do a "mis-justice" to the possibilities, I think,  in their discussion of "Algorithms" and "Decisioning" values for effective medical diagnosis and treatment (I do not think the authors fully understand the power and possibilities of modern 'Statistical Learning Algorithms' in that the individual, yes the individual, can be predicted with their uniqueness ...?)

Our book, about to be released, discusses, among many other things,  this issue, EBM in Crisis  as 'Tarnished Gold' (as it really is NOT the basis of "Gold Standards" in medicine)', and how effective, real  'EBM can be part of the future'  -  IF the Healthcare Field will "grab onto" what is now available in technology and knowledge ....... 

http://www.amazon.com/Practical-Predictive-Analytics-Decisioning-Me...

I suggest that everyone take a good look at the above 3 citations  -   the future for 'Effective Medical Care' is at stake ...

Tags: EBM, algorithms, baloney, based, crisis, decisioning, diagnosis, evidence, medicine, patients, More…stakeholders, treatment

Views: 802

Reply to This

Replies to This Discussion

I was the one who forwarded these articles to Gary, but I am not so negative about Best Practice research and Evidence Based Medicine.  Doing good research in this area is _hard_ so I have a lot of sympathy for researchers working in this area.  And, I think from a Public Health (not individual patient) basis, this work has had a lot of value.  And a lot of valuable work has been done by non-governmental organization such as the National Quality Forum to develop measures of quality and cost-effectiveness.  (I have served on several NQF evaluation committees.)

It's worth keeping in mind that the methodological issues described here are not exclusive to this area - they pervade a lot of the social and biological sciences.  Social Psychology (my Ph.D.) is now reexamining a lot of the fundamental Social Psychology research because much of the work can't be replicated.  Failures to replicate don't get published - and you have to deal with the publication system we have now to try to improve this. 

Measuring statistical power and effect size has been discussed by researchers for decades - it just doesn't seem to have affected actual practice.

It is certainly true that the volume of documentation for "best practice" is huge - and that it changes frequently.  How can practitioners keep up?  One interesting approach is the IBM Watson system.  WellPoint (the Health Insurer) is working with Watson to evaluate research and assist practitioners in optimizing patent care.  This technology is sophisticated enough to deal with some, if not all, individual differences.

So, I think EBM methodology and practice can be improved (the BMJ article listed a number of steps to take) and we certainly shouldn't abandon this approach to accept instead only "clinician insight and intuition" - both evidence and clinician expertise are valuable and should be used.

David

SHAKING TREES DO NOT CAUSE THE WIND .......an interesting "Section Heading" in the book TARNISHED GOLD-THE SICKNESS OF EVIDENCE BASED MEDICINE (by Steve Hickey Ph.D. and Hilary Roberts, Ph.D.).

As they point out in their book, "EBM has no cohesive theoretical structure" ....... at least two views predominate:

 

1.  "Philosophical notion of 'social construction'" -  "... Medicine, disease, and diagnosis are determined by social interpretation rather than underlying reality"... e.g. there may be no real scientific facts but instead socially constructed ideas followed by practical testing..... AN APPROACH USEFUL FOR 'AUTHORITARIANS who wish to CONTROL’.....

 

2.  "The world is REAL and SCIENCE is a DESCRIBING REALITY..."  -  

 

Practitioners of EBM may have a misunderstanding of the concept of causation.   Most things that happen in our universe have a "bottom line cause", although that 'cause' may not be sufficient to produce the event, e.g. in our purposes here the illness / disease.  Here are two examples:

 

A.  A person may be infected with the TB bacterium, Mycobacterium tuberculosis, but may, or may not exhibit symptoms of tuberculosis.  Some people may have such a strong immune system that this system prevents the bacterium from multiplying, thus the disease remains inactive; but others with the bacterium develop a full blown case of TB.

 

B.  DOES Christmas cause the SALE OF TREES?   People purchase Christmas Trees many days, many weeks (and some many months !!!) prior to the arrival of Christmas.  Thus some people looking at this phenomena, but not fully aware of the history, may conclude that BUYING TREES brings on Christmas.    But as the TARNISHED GOLD authors point out on page 64 of their book:  ".....regardless of how people learned to buy Christmas trees, it is clear that social and economic questions can have different characteristics to the linear cause-then-effect of classical physics".

 

And we have not even considered yet, in this discussion, the effect of GENETICS or the effect of what we might call STATISTICAL MAGIC.......

So that leaves us with the question:  DO SHAKING TREES CAUSE THE WIN

D ?

I have to  agree more with David.  Not that Gary doesn't have a good point - evidence based medicine (EBM) is important but does not yet take us to where we could be.  Good research is hard to do, as David pointed out.  EBM is a step in the right direction and when the designs are good, useful information is generated for populations.

We need to move ahead now and start using what we know from EBM and then use predictive analytics to produce useful algorithms for predicting to individuals, rather than to population means, as it has been in the past.  Excellent physicians are doing this informally already - using their intuition plus EBM when working with individual patients.  That intuition will not be totally replaced by predictive analytics, but good intuition will be augmented.  Poor intuition may be replaced by excellent prediction tools.

Finding good data is the hardest part of conducting any good research.  Data are protected as they should be, but until access is granted, research remains stymied.  It is therefore important for those with access to data to possess the most up to date research and predictive analytic skills.

It seems like there are two primary flaws in the current EBM decision making model:

  1. The source of the model is biased.  Managed Care and Pharma data are the most detailed sources for EBM & both have natural biases: Managed Care's goal is reduced cost & Pharma's is increased drug sales.  Neither of these are exactly aligned with the patient's goal, so statistical models will struggle to compensate without an unbiased means to "keep score".
  2. There isn't a robust feedback loop that tracks the recommended vs. actual clinical decision and most importanly WHY the doctor chose a different treatment.  As with any field dominated by experts, capturing this information is difficult, but critical to identifying where the "model" can be improved.

Factor in other variables, like malpracticve litigation or technological advances, and the doctor may also pay a significant role in the patient outcome. 

The recommended actions like:

  • "Independent funders must increasingly shape the production, synthesis, and dissemination of high quality clinical and public health evidence" or
  • "Patients must demand better evidence, better presented, better explained, and applied in a more personalised way"

are noble, but difficult to achieve without good tools for understanding the clinical decision by those who are impacted by the clinical decisions.  The best evidence of patient outcomes are from direct observation by patients and they have the poorest decision support tools of all involved.

The most resonable method to put patients back in the driver's seat is to move away from third-party payer systems, so everyone involved gets paid using the same definition of sucess.  Without this change, there will always be some level of error in the system that is due to the misalignment of the definition of success.  By making this change, a financial incentive is created to develop these tools and the advent of "fit-bits" & cloud computing has made the development achievable.

Evidence-based medicine is, in my opinion, important to follow and pay attention to, however, the benefit to any individual patient is limited.  To say that physicians now are practicing evidence-based medicine is only the case for a limited number of doctors.  There are so many distractions for physicians who are seeing a full schedule of patients every day, that to say that a majority are keeping up with the evidence is idealistic, but not likely the case.  There are so many variable factors that are in play that influence the decisions physicians make in a specific case, that the likelihood of following the current "evidence-based" protocol is limited.  

RSS

On Data Science Central

© 2019   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service