Comparative Effectivenss, Healthcare Cost Reduction, and Virtual Decision Trees

I have posted a number of previous notes discussing how to assess comparative effectiveness and efficacy of various medical treatments (see: A Look at Deloitte's Healthcare Reform Pyramid: A Strategy for Reducing Costs, Is My Chemo Working? -- FLT PET Scans May Provide the Answer, Using Diagnostics to Monitor the Effectiveness of Treatment). The topic of comparative effectiveness is important for lab professionals because lab tests will necessarily be a key element in any such program. Dana Blankenhorn, who frequently blogs about the health industry, discussed this topic and its relationship to cost-containment in a recent note (see: Industry commits to comparative effectiveness). Below is an excerpt from the note with boldface emphasis mine:

$2 trillion. Insurers, drug companies and other market players [arrived at this potential savings in healthcare expenditures] by estimating what current health care inflation is doing to costs, then promising to shave that inflation by 1.5%....The significance is that this is twice what the President estimates it will cost to extend coverage to the uninsured. Advocates on both sides will cite it as meaningless, but it’s proof the industry wants to be seen as being in the reform camp. While some think the number was pulled from the air, it was likely pulled from the work of Dr. Ralph Snyderman, now helming Proventys, which is building “virtual decision trees” to guide best practices and deliver the kinds of savings he was able to achieve at Duke. All of which means industry and government have joined together in believing that comparative effectiveness can be made to work at the point of care, and committed to moving toward it.

Here some more information about Proventys and virtual decision trees (see: Proventys stars University chancellor in birth of a salesman):

The main way Proventys [assists physicians care for cancer patients] is by building virtual decision trees from clinical trial data, then selling the results to insurance carriers and self-insurance pools for use at cancer centers. “Collect patients’ data as they undergo chemotherapy, before and after, and  mine that data to see what were the most important predictive factors” for success, “that’s the way to do it” [said Dr. Snyderman]. Dr. Snyderman is confident that insurers will gladly pay for a service that gives them greater confidence in a positive result at the lowest possible cost. “What we need to look for, given the way the reimbursement system works, are people who have incentives to gain on the bottom line with this, who will share savings with the developer,” he explained.

I believe that physicians, in general, have an aversion to use the use of decision trees or care maps, or whatever you want to call them when the are used to regulate their treatment of patients. I don't think that calling them "virtual  decision trees" makes them in any way more appealing. This distaste on the part of physicians is derived, in part, from the face that they are viewed, appropriately, as recipes for cookbook medicine and partly from the fact that they take extra time to review and follow. I believe that are two ways to persuade physicians to utilize decision trees when treating cancer patients. The first is for insurance companies to provide financial incentives for close alignment of the treatment of, say, cancer patients to the recommended regimens. This might occur on the basis of the skill and knowledge of the physician or by close attention to the applicable virtual decision tree. The second is by providing easy access by physicians to the relevant on-line virtual decision trees. Ideally, this can be accomplished using smart cell phones equipped with a web browser. Of course, office PCs can also be used for this same purpose.