Spencer Bailey: Vermont’s all-payer model — at what cost?

Editor’s note: This commentary is by Spencer B. Bailey, MBA, a reimbursement and health economics consultant who has provided advisory services to hospitals, health systems and medical device companies for over a decade. He resides in Burlington with his family.

[E]very Vermonter agrees that improving the quality and controlling the rising cost of health care in Vermont is a good thing, but the all-payer model cannot and will not accomplish either. Vermonters are rightly concerned that the same organizations, and even the same people, who have failed in their current roles to control health care costs or improve quality will be entrusted with an even greater role going forward. State leaders have provided no credible evidence to show how these organizations are more likely to succeed under the capitated all-payer model. Many of the program’s flaws are already well documented, but most concerning is the absence of an escape plan if the all-payer model doesn’t work.

With respect to cost savings, the all-payer model places its faith in the ability of accountable care organizations (ACOs) to better coordinate care to reduce costs. The problem is that the largest ACO in Vermont, OneCare, has already tried and failed to do just that. OneCare failed to prevent what is termed in the industry as “excessive cost increases in its Medicare population” in any year since its inception, despite a handsome $9 million annual budget, equivalent to over $160 per Medicare patient. Indeed, only about 25 percent of ACOs nationally generate shared savings and nearly as many have gone in the opposite direction.

Other constituents in Vermont health care reform have fared even worse. Just as OneCare was failing to reduce costs, from 2010 to 2015 under the Green Mountain Care Board’s oversight, the budgets of Vermont’s hospitals increased by 26 percent. Related initiatives such as Vermont Health Connect and Vermont Information Technologies Leaders (VITL) will have spent over $500 million, or over $800 per resident, by the time the all-payer model takes effect in 2018. The state has failed to show how these massive administrative expenditures have reduced the cost of care or improved the quality of the health care its citizens receive.

If Vermont hospitals and ACOs, enabled by administrative largess and expensive IT infrastructure cannot reduce the cost of patient care today, how will they do better under an all-payer model?

With regard to quality, all Vermonters should be disturbed by the consensus definition of quality care as outlined in the all-payer model. The measures of success of the all-payer model do not evaluate whether health care services are provided in a safe or effective way and would be oblivious to catastrophic declines in patient outcomes. Only a handful of diseases are included in the draft all-payer model guidance – depression/suicide, substance abuse/addiction and a chronic care “basket” of conditions: uncontrolled diabetes, hypertension and unplanned hospital admissions (but only for patients with multiple chronic conditions). Why were only mental health, substance abuse and chronic conditions included? What about obesity, which is more costly than smoking? What about vaccination rates, which are dangerously low in Vermont? Hospital-acquired infections? The list goes on. Under the all-payer model’s definition of quality care, every Vermonter will be asked about depression, drug abuse or smoking at every doctor’s appointment, but there is no credible evidence that these questions will translate to better patient outcomes.

How can the state ensure that providers won’t divert inordinate resources toward improving specific “quality-measured” services at the expense of other areas of care? Once quality metrics are established, how can these be tied to payments or incentives for providers? For example, will a primary care physician be expected to prevent smokers from smoking or make diabetes patients modify their diets to receive a full paycheck?

Most discouraging is the all-payer model’s access measure, which asks whether patients “have a doctor” and whether they can “make timely appointments.” Succeeding on this measure in no way improves access to health care. What if the doctor you “have” is 35 miles away or more? As the exodus of Vermont’s physicians continues, will rural Vermonters have access to nearby care? Why won’t unscheduled office visits and ER wait times be included in the timeliness measures? These are just a few questions that highlight the questionable measures of access in the all-payer model.

Finally, and most importantly, there is no exit strategy. Vermonters, rightly, do not trust their leaders to make rational decisions about health care policy. Vermont Health Connect, after a quarter of a billion dollars, is literally a bridge to nowhere. In the face of this reality, Vermont leaders refuse to admit defeat and adopt the federal exchange, which is more functional and costs less than a tenth what Vermont Health Connect has. Instead, they have pursued political blame-games with vendors, hired consultants to evaluate the extent of the failure and are still unable to see a sunk cost for what it is. If the all-payer model fails as spectacularly as Vermont Health Connect has, will Vermont leaders be as stubborn to admit defeat?

All indications point to yes … To implement the all-payer model, the state, ACOs and health care providers will have spent hundreds of millions of dollars in additional administrative costs. They will have taken on hundreds of employees and will have created new bureaucracies which will be unpalatable to eliminate if an all-payer model fails. What would happen if providers are driven out of state, new taxes are needed to close budget shortfalls and there is no improvement in the quality of care provided or worse, care suffers? What mechanism will Vermonters have to say “no more” if the Green Mountain Care Board refuses to admit defeat? In short, will the all-payer model become something that is both untenable and immoveable?

The all-payer model is a risky, unproven model for health care funding that is being implemented in a small state with a frail rural health network. Each of the ingredients of all-payer has a mixed performance record – ACOs, CMS quality measures, the Green Mountain Care Board, VITL and others. Any rational decision-maker would see that combining them is not only risky, but even likely to fail. And with no contingency plan, Vermont will be stuck with the all-payer model for a very long time, regardless of the outcome.