Because the most glaring geographic variations in health care use have been observed in specialty and end-of-life care, policymakers have had trouble coming to terms with the work of John Wennberg and his Dartmouth colleagues. The questions the Dartmouth researchers raise about spending and quality are too disruptive, too threatening. Specialty and end-of-life care are too sensitive for the blunt instruments of policy to address.

But with publication today of the latest edition of the Dartmouth Atlas, focused on chronic care, Wennberg and company carry their analysis to the most fundamental, everyday questions about the organization and delivery of care. The new Atlas joins the battle for health system improvement on a level that everyone can understand. It engages with questions that have immediate meaning for Congress, consumers, and many medical professionals and throws its weight behind policy proposals that already have widespread support. Wennberg is no longer a lonely voice crying out in the wilderness.

Building on earlier work, the new Atlas explores variations in care during the last two years of life among patients with one or more chronic conditions. Some of the work was previewed in Health Affairs’ November/December 2007 twenty-fifth anniversary issue. But while the approach is familiar, the 2008 Atlas throws down a new challenge to the conventional wisdom. The eye-opener here is that “higher utilization and spending in ambulatory settings, skilled nursing facilities and home health care was associated with higher utilization and spending for inpatient care.”

Not only did increased use of less intensive services fail to offset hospital use, but primary care, per se, did not seem to improve care coordination or overuse. “Simply increasing the number of primary care physicians alone will not improve coordination. Spending on ambulatory visits, many of them to primary care physicians, is positively correlated with inpatient days and inpatient physician visits,” the authors found. “There are no tradeoffs.”

The Atlas derives a menu of policy prescriptions from these findings, several of which align with and reinforce an emerging nucleus of ideas about attacking delivery system fragmentation at its roots. One of the options is “a scenario in which CMS (the Centers for Medicare and Medicaid Services) would offer shared savings partnerships to providers that agree to coordinate care among the various sectors” – a notion that echoes the Medicare Payment Advisory Commission’s (MedPAC’s) interest in bundled payment (a discussion recently reviewed on this blog).

More intriguing still, the Dartmouth authors weigh in strongly in favor of exploring “medical home” arrangements that have been explored and promoted by the American College of Physicians, the American Board of Internal Medicine, and MedPAC (see transcript cited above), among others.

Among proponents of the medical home, there is some controversy about whether the concept should be described and designed as a simple add-on to primary care that would allow maximum participation among generalist physicians who assumed some responsibility for coordination; or as a more sophisticated capability that would entail “systems-based practice” and information management and might be able “to reduce overall use for by the 10 percent to 20 percent of patients who use the most services,” as ABIM puts it. But this high-end version would not be easy to implement for many small physician practices.

The Dartmouth finding that “simply increasing the number of primary care physicians alone will not improve coordination” is an important contribution to a growing policy debate about medical homes. If a new national policy conversation about health reform does in fact occur in the wake of next November’s elections, there may be an opportunity to consider delivery system organization as a basic element, along with cost and coverage. A magical, payment-driven transformation of the delivery system into integrated entities cannot be assumed to be inevitable, as it was in the early 1990s. If there’s going to be real change, it will have to be organized from the ground up. The new Atlas ought to be an invaluable navigation aid to the next wave of pioneers. Perhaps the next step is to understand how primary care works in low-spending areas.