But Which Half?

By Kim Bellard, January 30, 2013

Advertising lore credits John Wanamaker, the department store magnate and marketing pioneer, with the famous quote: “Half the money I spend on advertising is wasted; the trouble is I don't know which half.”  It turns out he could have been talking about spending on health care.

The British Medical Journal, through their Clinical Evidence initiative, recently reported that they’d analyzed 3,000 medical treatments that had been studied in controlled, randomized studies.  It turns out that for half of those treatments, we have no idea how well they work.  Indeed, only about a third of the treatments were found to actually be beneficial or likely to be beneficial.  The rest are likely to be harmful.

Sadly, this does not come as a surprise.

We know we don’t know enough.  The vast number of medical treatments have never even been studied in a true clinical trial.  Worse yet, sometimes even when there is clear empirical evidence about which treatments are most effective, that information does not always sway physician behavior, or does so only very slowly (for example, see this study on the use of heart stents versus medication therapy).

There is no shortage of reports of unnecessary or even harmful care.  It’s even scarier when that care is associated with high costs.  In no particular order, one could cite recent controversies with spinal fusions, hip replacements, or chemotherapy drugs.   There can be lots of money at stake for manufacturers, drug companies, and health care providers.  That kind of money can distort the question of what is truly in the best interests of the patient.

Many employers, payors, and researchers have been pushing for “evidence-based medicine” for many years now.  EBM focuses on making sure that treatments have appropriate research to support their effectiveness, and in getting the word out about such treatments.  One of the many initiatives from ACA was the Patient-Centered Outcomes Research Institute, which is charged with conducting research to provide such evidence and funded by a $1 head tax on people covered by insurers.  And, of course, AHRQ probably is wondering why we need a new organization to focus on EBM, given their many efforts on effectiveness.

In time, this may all become much easier, as more patient data become electronic and more connected, and we can make more use of computing power to track what truly happens to patients under various courses of treatments.  I mentioned a couple examples of this in my last blog, citing Optum/Mayo’s new initiative and meta-research studies in lieu of clinical trials.  Another example comes from Archimedes Inc., a firm founded by David Eddy, who was one of the early pioneers of evidenced-based medicine.  Archimedes claims to use its advanced mathematics and computing prowess “to run clinically realistic virtual trials on any population and create compelling evidence to make decisions in health and economic outcomes research, policy creation, clinical trial design, and performance improvement.”  Apparently HHS thinks they can, as it hired Archimedes last year.

Most physicians I know are very bright, care very much about their patients, and work hard to stay current on the medical literature.  Unfortunately, the latter is virtually impossible to do, given the sheer volume of that literature.  Even when there are clear results about which treatment is truly the most effective, the research doesn’t usually come with a guide as to how physicians can implement the associated changes to their practice routines.  It’s as much of a question of change management as it is the evidence to make the change.

It would seem that the situation is tailor-made for clinical decision support tools, which seek to provide clinicians with information on treatment options, potential outcomes, and possible contra-indications at point-of-care.  Unfortunately, we may not quite be ready for them.

Last summer The Annals of Internal Medicine published a study on clinical decision support systems by Bright, et. al.  They did a meta-analysis of studies on CDSSs, and found ample evidence of their efficacy in improving process measures, but sparse results on their impact on clinical or economic outcomes.  Whether this is due to the limitations of the underlying studies, the CDSSs themselves, or how they were used by clinicians is unclear. 

Similarly, KLAS Research recently released results of their survey of health care providers on their satisfaction with clinical decision support tools.  The results cited a general level of frustration, especially due to lack of integration with EHRs and “alert fatigue” caused by ineffective targeting of alerts. 

Worst yet, according to new research from the University of Missouri, patients don’t seem to trust treatment recommendations from physicians who use CDSSs, believing them to be less capable than physicians who make decisions unaided.  Patients don’t even like it when physicians consult with other physicians before making a recommendation!  They think their doctor should know everything.  I blame television for this – on medical shows like Grey’s Anatomy or House physicians pull up the most obscure diagnoses and treatments strictly from memory, without ever having to consult any reference materials.  Nobody’s memory is that good. 

Clinical decision support systems aren’t going to replace doctors; they are simply tools to aid health care professionals, much as a stethoscope or a thermometer does.  One can imagine a future where CDSSs -- and EHRs -- fit seamlessly into patient visits, providing real-time, interactive information while with the patient.  The line between evaluation, documentation, and clinical decision support should blur, in order to more accurately diagnosis patients and determine the best course of treatment.  

In the meantime, it’s somewhat of a crapshoot.

A recent study by Deloitte indicates that 62% of Americans believe that, in fact, over 50% of U.S. health spending is wasted, which is up from the already high 51% in 2009.  The message about necessary spending may be getting out, but consumers may be getting the wrong idea – only 18% thought the problem was not using evidence-based treatments, versus 69% who blame fraud and abuse in the payment system.  In other words, the problem can be blamed on greedy crooks, not on well-meaning health care providers.  Defensive medicine and unnecessary paperwork were each also cited by about a third of respondents. 

I agree that fraud, defensive medicine, and inefficient administration contribute cause us to spend money we shouldn’t, and each should be addressed, but I suspect more of unnecessary spending comes from well-intentioned treatments that aren’t really best for the patients.  As professionals, health care providers should be more stringent about basing their treatment recommendations on evidence that truly supports them.  More importantly, as the people whose health is going to be impacted by those treatments, it’s incumbent on us to demand that evidence.

Maybe one day we’ll have Star Trek’s tricorder to non-invasively diagnosis or even Star Trek Voyager’s holographic doctor to treat.  Maybe someday nanobots will fix all our ills without our even being aware of their work.  All that is in the future.  For right now I’d settle for simply being able to know the odds that a recommended treatment will actually benefit me.