Among ACO Pioneers, Data Reveal Diabetes Performance Paradox

For accountable care organizations (ACOs), the holy grail of performance is the triple aim of delivering better population health with greater patient satisfaction while achieving financial savings. But can all 3 be achieved at once?

That’s the question experts were anxious to find out in early October. For the first time, CMS released detailed financial and quality performance data from the first 2 years of the program for the Pioneer ACOs. The original Pioneers were 32 healthcare groups selected to test value-based reimbursement principles because of their experience in this area. That the number of Pioneers dropped to 23, and then to 19, was an early sign that achieving the triple aim might be more attainable in theory than in practice.¹

A closer look at the CMS data reveals that diabetes care, in particular, presents a paradox: in general, ACOs that scored higher in helping patients achieve diabetes health targets had modest financial performances.² Conversely, some ACOs with stronger financial performances rated in the lower half of the diabetes scores.

Helping patients with diabetes mellitus (both type 1 and type 2) achieve goals in blood pressure, glycated hemoglobin (A1C), and cholesterol, and maintain positive lifestyle habits is key to preventing more costly care over the long term. In fact, 6 of the 33 quality measures now used by CMS to determine Medicare reimbursement are related to diabetes, 5 of which make up a “diabetes mellitus composite.” Those 5 measures are:

• the percentage of patients 18 to 75 years of age with diabetes mellitus who had A1C <8.0%

• the percentage of patients 18 to 75 years of age with diabetes mellitus who had low-density lipoprotein control of <100 mg/dL

• the percentage of patients 18 to 75 years of age with diabetes mellitus who had a blood pressure <140/90 mmHg

• the percentage of patients with a diagnosis of diabetes who indicated they did not use tobacco

• the percentage of patients 18 to 75 years of age with diabetes mellitus and ischemic vascular disease with documented daily aspirin use.³

While these 5 elements are measured as percentages, the composite is listed in CMS data as a score; the average among the Pioneers for the second year (2013) was a 33.26.2 In 2013, among the top 6 ACOs in the diabetes composite score rankings, only Mount Auburn Cambridge Independent Practice Association saw a financial savings above 3%; this ACO saved 3.3%, or $3.6 million. The next best financial performance among ACOs with high diabetes scores was a savings of 2.6%, or $3.12 million, for Park Nicollet Health Services. (See Table.)²

On the flip side, the Pioneer ACO with the best financial performance, both on a percentage basis and in the dollar among of savings, was Montefiore. The ACO saw a year 2 savings of 7%, or $24.59 million. However, its diabetes composite was on the lower end of the Pioneers at 22.93, with only 4 groups scoring lower.²

Other ACOs with better-than-average financial results repeated the pattern: Monarch HealthCare saved 5.4%, or $14.61 million, but had a diabetes composite of 23.98; Steward Healthcare Network’s savings was 3.3%, but a more substantial $19.22 million overall, while its diabetes score was 28.46; and Beth Israel Deaconess Care saved 3.9% or $17.38 million, yet its diabetes composite was still below average at 30.90.²

Policy researchers were eager to dig into the data when it was unveiled in October, and will likely continue to look for factors, such as the makeup of the population or the scalability of ACO programs, to explain the disconnect between the diabetes scores and financial results.

An analysis from experts at the Brookings Institution, that appeared shortly after CMS released the data, noted that while all but 1 of the ACOs still in the program improved quality from 2012 to 2013, a troubling bifurcation between 2 types of Pioneers was emerging.⁴

“After year 2, there still does not appear to be a direct relationship between higher quality scores and level of savings or losses in the Pioneer Program,” said the authors, led by S. Lawrence Kocot, JD.⁴

For its part, Montefiore ACO is working hard to improve the measures that make up the diabetes composite, according to spokeswoman Tracy Gurrisi. A number of patients with comorbidities within the ACO’s population are definitely a factor, she said, noting that Montefiore’s composite improved from 2012 to 2013.

“To help address barriers like access to care and education about diabetes, we started hosting a series of Diabetes Day programs at Montefiore sites that target patients with poor diabetes control,” Gurrisi said.

“At Diabetes Day programs, patients are able to receive a range of necessary exams in a single location as well as receive education about self-management. Days include

EBDM

diabetic care appointments with primary care, specialists like nutritionists and ophthalmologists, and diabetic educators so patients see all providers in one location at one time. This tactic will enable immediate treatment or intervention when providers identify a need, instead of referring patients for future appointments.” References

1. Pennic J. Pioneer ACO dropouts: why are providers leaving in droves? HIT Consultant website. http://hitconsultant.net/2014/10/07/pioneer-acos-dropout-providers-leaving/. Published October 7, 2014. Accessed November 21, 2014.

2. Pioneer ACO year 1 and year 2 results. CMS.gov website. http://innovation.cms.gov/initiatives/Pioneer-ACO-Model/CMS results. Published and accessed October 8, 2014. http://innovation.cms.gov/Files/x/PioneerACO-Fncl-PY1PY2.pdf

3. 2013 ACO Quality Measures. Northwest Ohio ACO website. www.nwoaco.com. Accessed November 26, 2014.

4. Kocot SL, White R, Katikaneni P, McClellan MB. A more complete picture of Pioneer ACO results. Upfront. The Brookings Institution website. http://www.brookings.edu/blogs/up-front/posts/2014/10/09-pioneer-aco-results-mcclellan#recent_rr/. Published October 14, 2014. Accessed November 20, 2014.