"Medical Homes Do Not Save Money" says Al Lewis at the National Medicaid Congress

That's a quote from Al Lewis of the Disease Management Purchasing Consortium about a new look at the data from the famous North Carolina 'medical home' initiative, otherwise known as the Comunity Care of North Carolina (CNCC)/ACCESS program. He spoke today at the National Medicaid Congress and the intrepid Disease Management Care Blog was in the audience taking notes for its readership.

As readers may recall, this program's actuarially determined savings of hundreds of millions of dollars are commonly held up as the reason why health reform wouldn't be complete without the patient centered medical home. CNCC was examined by the Disease Management Care Blog back in February; it didn't pass judgment on the savings question because it went to the "Program Impact" part of the CNCC website and looked at all the .pdfs. Not only did the DMCB actually read the actuarial reports, it didn't understand how the calculations were done and, unlike many CNCC's advocates, it wasn't afraid to say so.

Before you think this is some more of the internecine battling between the disease managementroids on one side and the medical homabots on the other, readers should also realize that there has been no more visible critic of the disease management vendor industry than DPMC's Al Lewis. In fact, the Disease Management Care Blog suspects many of its colleagues are somewhat relieved that Al has turned his attention to the medical home. They could use a break from his unrelenting mocking of the industry's opaque, always-positive-ROI, cherry picking and illogical innumeracy. Simply put, Al Lewis is no lightweight and commands the respect of a number of highly regarded Health Plans that rely on him for help and insight.

So what did Al Lewis say? Like the DMCB, he thinks evaluators of population-based care programs should never accept anyone's conclusions at face value and should always perform a plausibility test - especially when savings totalling hundreds of millions of dollars are being claimed. He found the CNCC is quite wanting plausibility-wise. He noted that the data supposedly shows inpatient costs fell by 46%, ER visits by 25%, outpatient visits by 24% and pharmacy costs by almost 11%. In Al Lewis' experience in looking in successful programs, it is practically unheard of for decreases in claims expense for admissions to not be accompanied by a corresponding increase in outpatient physician visits. What's more, in any population, almost half of all admissions are not controllable thanks to trauma and unexpected random illness; in order to mathematically achieve a 46% decrease, practically all admissions for chronic illness would have had to be avoided. Al Lewis smelled a rat.

Things got particularly stinky when Al Lewis told the audience that he had used the Freedom of Information Act to pull utilization claims data for the North Carolina TANF population, which makes up the bulk of their Medicaid population. According to his analysis, physician office visits did go up, ER visits were flat, admissions for heart attacks were flat and asthma admissions went down. While it's difficult to translate events (visits and admissions) into expense data (dollars), Al's rough estimates show very modest savings that are far less than the 'official' numbers and outweighed by the program's expense. In fact, when expenses are included, the program lost money. Sound familiar?

The DMCB looked hard for holes in the analytic logic. There were some minor glitches, but nothing serious. That's for another post.

Al told the DMCB that Senator Kennedy's office is interested in learing more about his conclusions.

What does the DMCB think?

Determining savings in insurance settings is fundamentally a brain cramping exercise in measuring what hasn't happened and then trying to assign a cause. The DMCB doubts there is any 'single methodology' that can answer that question for any commercial or government insurer. Rather, the better approach is to triangulate on the results using a combination of actuarial, epidemiologic, statistical and research methods that may very well include the DPMC methodology. We know have another prespective on CNCC, and for that we should be grateful.

Once again, the real challenge is to find approaches to the care of individuals with chronic disease that use the best combination of care interventions but are reasonably generalizable in multiple care settings. The DMCB has questioned whether CNCC's approach will work outside of Medicaid. So, the question isn't the utility of endless debate over the past of the CNCC but finding the right future combination of disease management, medical home, information technology, payment reform and benefit designs that give patients their money's worth.

In addition, supporters of the patient centered medical home better work hard to get this puppy into the upcoming Congressional health reform bills and passed before there is any more bad news - not only from Al Lewis but from all those other medical home pilots currently underway.

Last but not least, getting news like this is a good reason to keep up with the DMCB.