We need to get this right. As I have said last month, I continue to be dismayed by the evidence that health-care costs are not being controlled. For example, in Minnesota, one of nation’s top health-care managed states, Blue Cross and Blue Shield announced that it would not sell individual insurance policies next year due to concerns over cost. Skyrocketing health-care costs will affect the affordability of private insurance and the existence of public health-care programs such as Medicare, Medicaid, and subsidized insurance sold under the Exchanges.

My clinic, McFarland Clinic of Ames, Iowa, is involved in two ACOs, a private-insurance ACO with Blue Cross and Blue Shield of Iowa and a Medicare ACO associated with the University of Iowa. The numbers: We have about 24,000 patients in our private ACO and about 17,000 patients out of 75,000 total patients who are in the University of Iowa Medicare ACO. This year, Medicare expects to spend about $8,500 (this cost does not include certain high-cost patients or the cost of medications for the entire population) per patient per year on this population for a total of cost of $637.5 million. An average trip to the emergency room (ER) for a Medicare patient of McFarland Clinic costs $912, and a hospital admission costs $11,567. Roughly 30 percent of Medicare ER visits were deemed for non-essential emergency-room purposes when evaluated by a respected formula. Many chronically ill elderly patients have multiple health issues and are deemed so medically complex that a trip to the ER automatically results in a hospital admission. Finally, during a one-week study at the Ames emergency department, 70 percent of Medicare patients seen in the ER had contacted their primary-care physician’s office in the previous 24 hours. I am on committees for both ACOs.

The goal of this blog post and for my two ACOs is to determine how we reduce needless ER visits and hospital admissions. My answer is that we provide incentives to primary-care physicians both during office hours and outside of office hours in ways to care for Medicare patients that avoid them going to the ER, which often results in a hospital admission.

What actions performed by primary-care physicians (usually family physicians and internists) would accomplish this goal? Some tested actions include open-access scheduling (sick patients are seen the same day), a 24-hour nurse hotline with backup availability to talk to a physician before recommending an ER visit,  after-hour clinics during the evening, on weekends, and on holidays, and an urgent-care option with access to X-ray, lab, IV hydration, and sometimes CT scans. The goals would also include seeing all patients once a year and include a risk-stratifying process, seeing high-risk patients more often, and finally, good cross-coverage of patients so that, when the primary-care physician is away, the patient will be seen in a clinic with adequate records easily accessible.

Given the data available from Medicare, we can give accurate rates of ER visits, hospital admissions, and hospital re-admissions per thousand for individual clinics and down to the provider level. The proposal that McFarland Clinic is considering would provide incentives to primary-care physicians by tying some of the shared-savings distribution to the rates of these costly activities. Primary-care physicians who were successful in reducing these rates would be  rewarded at a higher level when shared savings are distributed.

Another incentive idea is tied to the submission of diagnoses to Blue Cross and Blue Shield and Medicare. Diagnostic codes help create the actuarial cost estimate for any given patient’s health care in a particular year. My brother-in-law is an actuary for Mutual of Omaha. I have always said that I did not learn to be an actuary in medical school, but, in a bizarre way, my accurate diagnoses submitted to payers does add to the actuary’s estimate in a way only I can affect. Once again, we can generate data regarding diagnosis codes used or dropped (used last year but not this year) at the provider level. This data also could be used to encourage primary-care physicians to accurately code. Accurate codes often correctly increase the estimated level by which shared savings is calculated. A higher cost expectation then makes shared savings more possible. As with lower ER rates, primary-care physicians could be rewarded if they more appropriately coded their patients’ visits.

I am sorry for this slow, boring slog through the interworking of ACOs, but I firmly believe that it is this tedious dissecting of these processes that will present incentive opportunities that will make ACOs work.

The next step for McFarland Clinic is a work group of primary-care physicians to devise a transparent, somewhat simple, understandable formula by which to distribute shared savings to primary-care physicians. The goal would be to create a formula that would drive certain actions that would increase the size of the shared-savings pie. This would be a win-win-win situation: Health-care costs would diminish; primary-care physicians would be rewarded for their efforts; and, finally — hopefully — health insurance premiums and public-payer costs would be more reasonable. This is what we have to get right.

I will keep you informed of the workgroup’s conclusions and ultimate results of our two ACOs.

In a later blog post, I will discuss the question of quality care vs. rationing of care as found in an ACO.

As an aside, with this post we move into our fourth year with this blog. I am pleased that our readership spans the U.S. and 33 foreign countries.