Summary: In this issue, we profile two different types of accountable care organizations (ACOs), one led by a health system and another led by independent primary care providers. This profile focuses on the latter, Rio Grande Valley Accountable Care Organization Health Providers, based in Donna, Tex., which was one of 114 ACOs to join the Medicare Shared Savings Program in 2012. It has succeeded thus far by reducing hospital and emergency department use through more proactive management of chronic conditions—bringing down per capita costs of care for its assigned Medicare beneficiaries by about 14 percent. Participating providers are attracted by the opportunity to earn more income while tapping into the IT resources and other support services offered by the ACO.
Structure and Beneficiaries Served
Rio Grande Valley Accountable Care Organization Health Providers (known officially as RGV ACO Health Providers, LLC), located in the southernmost tip of Texas along the U.S.–Mexico border, was one of the 114 accountable care organizations (ACOs) to join Medicare's Shared Savings Program in 2012, its inaugural year. It was one of only four groups to take on the riskier of two payment models—exposing it to significant financial penalties if providers failed to curb spending for their designated beneficiaries, but also greater financial rewards if they succeeded.1 The ACO was launched by six independent primary care practices, with a total of 11 physicians and 10 midlevel providers. It has since expanded to include 13 practices, adding seven more physicians and several more nurse practitioners and other midlevel providers.
Together, the 13 practices serve about 8,500 Medicare beneficiaries. About half of them are dual eligibles, covered by both Medicare and Medicaid. Because of the high concentration of Medicare beneficiaries in their patient panels—and for the sake of simplicity—the practices don't target their care management and coordination strategies at ACO-covered patients, but instead deploy them across all patients.
Jose F. Peña, M.D., the ACO's CEO and chief medical director, is the driving force behind the organization. In 2012, he pitched the idea of forming the ACO to colleagues, who believed that making operational changes could have a big impact on patients' health, while also increasing revenue. The group agreed to take on risk as a way to "work under pressure, have some deadlines, and be focused on thinking and acting in an ACO way 24/7," Peña says.
The practices, which are located in the central part of the Rio Grande Valley, face tough challenges.2 Hidalgo County is one of the poorest in Texas, with more than a third of its residents living below the federal poverty level.3 Diabetes is epidemic among the largely Hispanic population: nearly half of the region's elderly patients have some form of the disease.4 Heart failure and chronic obstructive pulmonary disease are also common. When trying to engage patients, providers face several barriers, including low literacy rates, unhealthy diets, and a tendency to use the emergency department to manage care.
One factor that may have helped the Rio Grande Valley ACO succeed thus far is the relative lack of competition among primary care providers in the region, making it less likely it will lose patients to other providers. In addition, because it operates in a high-cost region, it had many opportunities to reduce spending: Medicare spent $11,195 per capita in 2012 in Harlingen, the hospital referral region that includes the central Rio Grande Valley, compared with $8,874 across the nation.5 Edwin F. Estevez, Ph.D., a consultant to the group and its former chief operations officer, says that "there was fat to lose" in terms of reducing costs related to hospitalizations, readmissions, and use of home health care.
All of Rio Grande's providers speak Spanish, and most share cultural backgrounds with their patients. Early on, the providers sought to educate patients about what an ACO is, the benefits of the model for patients, and their responsibilities for their health and care. (CMS also sent letters of explanation to ACO-enrolled patients, but Estevez says that most didn't understand them.) Other early initiatives sought to elicit patients' feedback and address points of dissatisfaction with their care. For example, when survey findings uncovered widespread frustration over long waits to see doctors, one practice designated a staff member to greet people at the door, offer them coffee or water, and perform an initial assessment to make sure urgent cases were seen first. As wait times have only slowly begun to decrease, the practice has found that this approach has greatly reduced patients' frustration.
Providers also found that patients missed appointments or didn't pursue care because they couldn't take time away from work or family duties. To head this off, practices now offer same-day appointments when needed, as well as evening and weekend hours. In many cases, patients failed to follow treatment advice because they didn't understand it. To address this, practices created longer appointment times to accommodate meetings with care coordinators, diabetes educators, or others as needed to review the care plan with patients and offer support in following it.
Prior to forming the ACO, the primary care practices had made ad hoc efforts to reach out to patients who missed appointments, had poor disease control, or were at risk for complications. Now, all practices have on-site care coordinators, whose salaries are paid in part by the ACO. The care coordinators focus on high-risk patients, including those who have been recently hospitalized and those with uncontrolled diabetes, to develop individualized care plans, connect patients to services such as home health care, and make follow-up phone calls or home visits as needed.
Rio Grande's leaders are aware that the ACO’s success depends to a great extent on whether providers can convince patients—particularly those with uncontrolled diabetes—to make dietary and lifestyle changes needed to improve their health. During its second year, the ACO created an American Diabetes Association–accredited Diabetes Education Center, which is supported by each of the practices and housed in a central location in Weslaco, Texas. Certified diabetes educators work with patients to help them understand their disease and encourage them to make the lifestyle changes needed to control its symptoms. They also have trained medical assistants in diabetes management approaches and assigned a nutritionist to meet with patients with poorly controlled diabetes. The center promotes community awareness of the disease, for example through participation in diabetes walks and collaboration with the local diabetes association.
Challenges: Performance Reporting
ACOs are required to aggregate and report performance data for 33 measures related to patients' and caregivers' experiences, care coordination and safety, appropriate use of preventive care (e.g., vaccinations and screenings), and outcomes for particular at-risk populations, including those with diabetes, coronary artery disease, ischemic vascular disease, and hypertension.
This requirement presented challenges for Rio Grande because the practices used four different electronic heath record systems across the six practices. Some of the needed data were stranded in free-text formats, or stuck on proprietary systems. After interviewing 40 different vendors, administrators decided no off-the-shelf platform would allow them to aggregate and report data as required by CMS, so they built their own repository for data collection and reporting, a process that is still ongoing and carries a $1 million price tag.
During the Shared Savings Program's first year, ACOs were not assessed based on their performance on the 33 measures, but were instead required only to report data for all measures and all participating providers.6 Still, the Rio Grande Valley practices saw some improvements after their initial year as an ACO on measures tracking the quality and safety of care and health outcomes. For example, from 2012 to 2013, more Medicare beneficiaries treated in Rio Grande's practices were screened for risk of falls (59.7% to 88.9%) and depression (53.8% to 78.2%), received pneumococcal vaccinations (81.1% to 93.3%), and had BMI screening and follow-up interventions (75.8% to 82.4%). In addition, more patients with diabetes achieved control of their disease—from 36.5 percent in 2012 to 48.3 percent in 2013.7
In 2012, ACOs received shared savings if they were able to reduce the per capita costs of care below a benchmark level, based on their past claims (adjusted to account for the illness of beneficiaries and national growth rates).8 Rio Grande Valley spent about $1.2 million on operations related to setting up and fulfilling the requirements of an ACO. It managed to reduce the per capita costs of care for its assigned Medicare beneficiaries by about 14 percent, mainly by decreasing the number of hospital admissions, readmissions, and use of home health care. One-quarter of its shared savings award has been reinvested in ACO operations and growth, allowing practices to hire additional medical assistants or other staff key to care management.9 The remainder was distributed to participating practices, with 70 percent toward bonuses to primary care professionals (both physicians and midlevel providers) and 5 percent toward hiring additional staff. The additional funds meant significantly increased earnings—around 30 percent—for providers.
Moving forward, Rio Grande must demonstrate continued improvement on measures of care quality while further reducing spending. Its performance will be compared against a growing number of ACOs, even as it adds new practices and hastens to bring them up to speed.
As the ACO expands, its leaders are recruiting physicians who are enthusiastic about the care delivery model and willing to devote their own time—several hours per week—to make it work. They also are continuing to develop the ACO's central data repository so it can capture data from each of the practices and generate monthly performance reports. Financial incentives are being given to practices that gain National Committee for Quality Assurance certification as patient-centered medical homes. The ACO has also purchased a van to help patients get to their appointments.
Leaders also made a strategic decision to focus care management efforts on the highest-cost patients—the 10 percent who account for 50 percent of spending—by addressing the drivers behind avoidable use of acute care services. The group hired a nurse practitioner and physician assistant, whose services are shared among the practices, to make regular home visits for high-cost patients—making sure their conditions are stable and watching for deteriorations that could lead to hospitalizations. During these visits, the providers also teach patients self-care techniques and encourage behavioral changes needed to improve disease control. These providers also serve as these patients’ first point of contact should they need help, giving the sickest, most vulnerable patients a trusted person to call instead of heading to the emergency department (ED).
Rio Grande's efforts have been aided by new partnerships with the local hospital and post–acute care providers. While it formerly took days, or even weeks, for the practices to be notified when one of their patients visited the hospital, they now receive daily reports on their patients' admissions, discharges, and ED visits—enabling care coordinators to follow up in a timely manner. The ACO also is working with a local home health agency to keep patients healthy and out of the hospital. "It's a win–win for everybody because the hospital reduces its 30-day readmission rates, the home health agency captures its episodic payment, we reduce Medicare costs, and the patient's health improves," says Yolanda Salinas, Rio Grande's chief operating officer.
In addition to coordinating care across different types of providers, Rio Grande is using the ACO model as a way to calibrate and strengthen the approach to primary care in their underserved community. It is doing so by sharing resources among practices, building capacity for care management and data reporting, and working together to develop targeted improvement strategies. The group appears to have succeeded thus far because it uncovered opportunities to reduce spending by better managing chronic conditions and focusing on patient engagement and prevention. The ACO leaders also achieved buy-in among participating physicians by customizing operational changes to meet the needs of individual practices, minimizing disruptions, and preserving physicians' autonomy. What's more, participating providers were attracted by the opportunity to earn more income while tapping into the IT resources and other support services offered by the ACO.
It's not acceptable for a patient with poorly managed A1c, or high LDL, to leave your office without talking to them about the problems, adjusting their medication. It's not business as usual: you need to take extra time to figure out what patients need. —Jose Peña, M.D., Rio Grande's ACO's CEO and chief medical director
Deliberate behavior change on the part of providers was also critical. "It's not acceptable for a patient with poorly managed A1c, or high LDL, to leave your office without talking to them about the problems, adjusting their medication," says Peña. "It's not business as usual: you need to take extra time to figure out what patients need." Still, it remains to be seen to what extent providers can convince patients to make the behavioral changes needed to bring down rates of diabetes and other conditions.
To encourage other small, physician-led ACOs, Peña says CMS should consider allowing more flexibility in the Shared Savings Program. The application process could be less onerous for small groups, and CMS could offer capacity training and financial assistance for infrastructural requirements to maximize small ACOs' potential for success, he says. Rio Grande has received some assistance through the Advance Payment ACO Model, through which CMS provides start-up resources to help physician-led and rural ACOs participating in the Shared Savings Program to build infrastructure such as IT systems or hire new staff. In all, 35 groups have thus far received this assistance.10
In the end, physician-led efforts such as Rio Grande's may be key to the success of the ACO experiment. "We are little, flexible, and quick—and we have that immediate touch to the patient," says Estevez. "That is the strength of our approach."
1 Medicare's Shared Savings Program, launched in 2012 under a provision of the Affordable Care Act, seeks to reduce health care costs by providing incentives for providers to work together in entities known as accountable care organizations (ACOs). ACOs take on responsibility for the quality and costs of care for a designated group of Medicare beneficiaries.
2 The health care market in McAllen, Texas (part of the Rio Grande Valley), received national attention after Atul Gawande, in a 2009 New Yorker article, used it to explore the causes of high per capita health care costs. See http://www.newyorker.com/reporting/2009/06/01/090601fa_fact_gawande?currentPage=all.
3 U.S. Census Bureau's Small Area Income and Poverty Estimates, data for 2012. See http://www.census.gov/did/www/saipe/data/interactive.
4 Of the 43,241 beneficiaries over age 65 in Harlingen's hospital referral region in 2012, 18,739 (43%) had diabetes. See Centers for Medicare and Medicaid Services, Geographic Variation Public Use File, http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Medicare-Geographic-Variation/GV_PUF.html, HRR Report—Beneficiaries 65 and older.
5 See Centers for Medicare and Medicaid Services, Geographic Variation Public Use File, http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Medicare-Geographic-Variation/GV_PUF.html, HRR Report—Beneficiaries 65 and older.
6 During subsequent years, ACOs will begin to be assessed based on their performance on the measures, not just on reporting data.
7 Diabetes control is assessed based on five factors: hemoglobin A1c levels, LDL levels, blood pressure levels, smoking status, and use of aspirin.
8 See http://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/Downloads/Shared-Savings-Losses-Assignment-Spec-v2.pdf.
9 CMS has not given the RGV ACO Health Providers permission to make public the amount of shared savings it received; the agency will publish the final reconciliation amounts for it and other ACOs in September 2014.
10 For more details on the Advanced Payment ACO Model, see Centers for Medicare and Medicaid Innovation, http://innovation.cms.gov/initiatives/Advance-Payment-ACO-Model/.