Alexandria

As an internal medicine intern, I remember working at the VA hospital. It was utterly depressing. The care varied widely, and most of us would not have wanted to have any care at that hospital. Making rounds and finding a patient dead and cold in bed. Unable to find your ICU patient because they got into a wheelchair, unobserved by the nursing staff, to go outside and have a smoke. So, when stories started to surface that the VA provides the best health care in the US started to surface, I was skeptical. Besides its high rankings on quality, it also provided care at a cost lower than Medicare as noted in the link above from a Time article by Doug Waller.

Tom Bock, commander of the American Legion, has another idea: allow elderly vets not in the system who are drawing Medicare payments to spend those benefits at a VA facility instead of going to a private doctor, as is now required by Medicare. “It’s a win-win-win situation,” he argues. Medicare, which pays more than $6,500 per patient annually for care by private doctors, could save with the VA’s less expensive care, which costs about $5,000 per patient. The vets would receive better service at the VA’s facilities, which could treat millions more patients with Medicare’s cash infusion.

However, does Medicare Advantage (MA), the privately run arm of Medicare, which competes with Medicare for patients with standard Medicare have better quality? In some urban areas, MA must also compete with other MA programs. We know that in these areas, MA can have lowered costs. While MA usually costs more than standard Medicare, in these urban areas it can cost less. How would it compare on quality? Aaron Carroll links to this study by Trivedi and Grebia which suggests that the VA is superior. In a study that looked at data from 2000-2007 they found…..

The VA outperformed MA plans on 10 of 11 quality measures in the initial study year, and on all 12 measures in the final year. In 2006 and 2007, adjusted differences between the VA and MA ranged from 4.3 percentage points (95% CI, 3.2-5.4) for cholesterol testing in coronary heart disease to 30.8 percentage points (95% CI, 28.1-33.5) for colorectal cancer screening. For 9 of 12 measures, socioeconomic disparities (defined as the difference in performance rates between persons in the highest and lowest quartiles of area-level income and education) were lower in the VA than in MA. Across all measures, the mean interquartile range of performance was 6.7 percentage points for VAMCs and 14.5 percentage points for MA plans.
CONCLUSIONS:
Among persons aged 65 years or older, the VA health-care system significantly outperformed private-sector MA plans and delivered care that was less variable by site, geographic region, and socioeconomic status.

The usual caveats apply. It is one study. It does not cover all aspects of health care. It is looking at process more than outcomes. The VA population may be unique in many ways. It is a retrospective study. Still, this is a very large study. It looks at some of the big drivers of cost, diabetes, heart disease and cancer, in our medical system. This suggests that competition between health plans does not necessarily lead to either better quality or lower costs.

Yet, in other circumstances, it appears that competition can improve quality. When hospitals within the same health plan are forced to compete against other hospitals, quality has been found to improve. Bloom, Cooper, Propper, et al have published a series of papers on competition within the NHS in the UK **. The NHS, in 2006, instituted a policy that aimed to promote hospital competition. What these authors have found is that quality improved without a significant change in costs. To be fair, this has been disputed by other researchers in the UK, but the results have held steady for multiple studies. If these studies hold, it suggests that we need to utilize effective competition, and that competition may be best focused on providers rather than health plans.

** I have provided a link to a rebuttal by the listed authors. At the bottom they have links to many of their studies. It should be noted that while this was a joint rebuttal, the research was conducted separately by multiple teams. This was just an easy way to provide those links rather than doing them individually. Much of the dispute around their studies surounds their methodology. Unless you have an abiding interest in statistics, and are very well caffeinated, read at your own risk.