The news of Dartmouth’s withdrawal from the accountable care organization (ACO) program, as reported in a recent article in The New York Times, caught many by surprise since the ACO model can be traced back to concepts initially proposed by Dartmouth researchers. Citing unsustainable financial results, Dartmouth discontinued its participation in the federal ACO program.
In explaining their decision to withdraw, officials at Dartmouth hypothesized that their care processes were already too efficient to afford significant opportunity for additional savings. Left unanswered is the question as to why the rest of the ACO program has failed to realize the savings originally hoped for. The answers may lie in the oil fields of the Middle East and with a simple bus token.
The theory behind ACOs is intuitively appealing: groups of physicians and hospitals who share financial incentives work closely together to improve efficiency, eliminate avoidable waste and share in the payer’s savings. Overlooked in the ACO theory is the fact that health care spending is a zero sum game. Reduced spending by payers occurs only if provider revenues decline. In effect, providers in an ACO are expected to reduce their revenues in the hope of receiving a portion of the foregone income back in the form of a performance bonus. It should come as no surprise that enthusiasm for such a bargain is not universal among the providers in any ACO.
At the root of economic theory is the concept of self-interest. Consumers, as well as suppliers, are presumed to act in their own self-interest. By examining two classic economic behaviors, we may discover clues as to why ACOs have failed to deliver expected savings. Neither economic behavior is necessarily occurring per se in the health care market, but distant cousins of these classic behaviors may be contributing to the intractability of health care spending. The two economic behaviors are chiseling under a cartel and the concept of the free rider.
In the 1980s, the Organization of Petroleum Exporting Countries (OPEC) attempted to limit oil production by its members in order to maintain prices at a profitable level. As long as everyone complied with the production limits, the worldwide supply of oil would remain low enough to maintain a relatively high price per barrel. The arrangement fell apart when oil-exporting debtor nations such as Nigeria and Venezuela routinely exceeded production limits to increase their own revenue. Rather than acting in the interest of the group, they were optimizing their own revenue. Saudi Arabia retaliated in kind, flooding the market with available oil, causing prices to drop to one-half of their prior levels.
The health care analogy to chiseling under a cartel is not centered on price, since prices do not rise and fall precipitously with sudden changes in supply. But consider an individual provider in an ACO whose own revenue arises in part from utilization that the group might deem unnecessary. Not unlike Nigeria or Venezuela, it is easy to imagine individual providers maintaining their utilization at higher levels, sustaining their own revenues, and acting in their own self-interest as opposed to optimizing for the group (ACO). And that brings us to the economic free rider.
An economic free rider is someone who enjoys the benefits of a good or service without paying for it. Let’s return to the individual provider who refuses to alter their own “production” in order to sustain their own revenue. If the rest of the ACO participants reduce avoidable utilization and generate savings for the payer, the individual who opted out becomes an economic free rider. If the ACO fails to achieve targeted savings, the free rider is no worse for wear. And in the worst case, if the ACO assesses a 2 percent penalty on its members should costs actually increase, the free rider has netted far more by maintaining their historic utilization than they would have had they played by the rules. In short, the ACO model fails to create penalties for non-compliant members that are sufficient to offset the economic incentives for self-optimization.
In our current study, the findings of which are slated for release in early 2017, we discovered that chronically ill patients who receive the large majority of their physician care from a single multispecialty group, cost 20 to 30 percent less than clinically similar patients who receive less than 50 percent of their physician care from any one practice.
This discovery points to a potential explanation for the failure by ACOs to reduce health care spending: loose aggregations of independent providers are no substitute for genuine multispecialty group practices. Physicians joined group practices before there were any ACOs, not because of them. They practice together, not alone. Their financial interests are conjoined, not independent. They routinely see one another’s patients. Consistency becomes an expectation, not a hope.
The evidence is emerging that it is the existence of genuine group practices, not 2 percent shared savings contracts that matter. Dartmouth, with its integrated faculty practice, has the key attributes of a multispecialty group practice. It’s conceivable that the episode costs for patients in the Dartmouth ACO may have been lower than those of its compare group. What is not clear is that other ACOs, which lack genuine multispecialty physician groups, will realize the savings potential once hoped for.
Self-interest is a powerful motivator. So much so that economic theory would collapse without it. No one should have expected Venezuela to voluntarily reduce its own revenue for the benefit of others. And if we give someone a free bus token, we should expect them to use it.
