Editor’s note: In addition to Stephen Shortell (photo and linked bio above), this post is coauthored by Sarah Weinberger, a graduate student at UC Berkeley; Matt Chayt, an associate at Nossaman LLP; and Ann Marie Marciarille, an Associate Professor at the University of Missouri at Kansas City School of Law.

As defined by the Affordable Care Act and subsequent rulemaking, Accountable Care Organizations (ACOs) are accountable for the cost and quality of care for a defined group of patients. In return, ACOs are able to share in savings that may result from providing cost-effective care, and they sometimes bear risk for excessive spending as well.  While originally intended for Medicare beneficiaries, public-sector ACOs have drawn considerable attention from many states as a vehicle for potentially providing more accountable, cost-effective care to Medicaid and uninsured populations.  At least ten states have already launched or are scheduled to launch Medicaid ACO initiatives.

The final ACO rules published by the Centers for Medicare and Medicaid Services specify that federally qualified health centers (FQHCs) and rural health centers are eligible for participation.   This change in the original rules and regulations makes it potentially easier for these safety net providers to combine Medicaid and Medicare accountable care initiatives targeted to the dually eligible population, in addition to serving the uninsured and Medicaid populations.

But in addition to these opportunities, safety net providers also face particular challenges in providing accountable care. With the aid of a survey administered to safety net providers in two California counties, this post examines those challenges and offers some policy responses.

Special Challenges

Safety net organizations face many of the same challenges facing other organizations attempting to provide accountable care.  These include the ability to assume risk, to provide coordinated care across the continuum, and to collect, analyze and report data on performance.  But safety net organizations also face special challenges due to the nature of the populations that they serve, the relative lack of available infrastructure and financial resources, the lack of primary care providers and access to specialists, and the separation of many safety net providers not only from each other but from their local delivery systems.

Population served.  Safety Net providers including FQHCs, community clinics, public hospitals, and in some areas nonprofit community hospitals primarily serve the uninsured, underinsured low income Medicaid recipients, and the dually eligible populations.  These groups are generally less healthy, more likely to go without needed care, less likely to have a regular source of care, and more likely to visit hospital emergency departments than privately insured populations.  They are also more culturally diverse with low levels of health literacy and with more than half being people of color, resulting in language and other barriers in accessing care.  As a result, it will be difficult for safety net organizations serving these populations to achieve the same cost, clinical quality, and patient experience outcomes as those serving primarily well insured or more homogenous populations.   These challenges are likely to be exacerbated given the increased demand brought on by 30 million newly insured individuals in the face of a limited supply of providers.

Financial Resources.  With rare exception, payments to safety net providers have been historically less than for those serving better insured populations.   In 2010 alone, 39 states cut or froze provider payment rates.  While the ACA will initially raise government payment to primary care providers, their reimbursement will still lag private sector reimbursement to other providers.  There is also concern that hospitals serving safety net populations will lose more money from the discontinuation of disproportionate share payments than they will gain from increased patient revenue generated from the increased insurance coverage.  This concern may be particularly acute in states that do not opt into the ACA’s Medicaid expansion.  The net result is a concern that resources will not be available to implement electronic health records (EHRs), invest in primary care provider capacity, or develop partnerships with other providers needed to better coordinate care across the continuum.

Shortage of primary care and access to specialists.  The shortage of primary care providers is particularly acute in the inner city and rural areas, which comprise the majority of the safety net populations.  While there is need for a greater supply of primary care providers, safety net providers are often at a disadvantage in competing for them due to low payment rates, which translate into lower salaries for such providers.  Safety issues in the inner city and the lack of education and cultural amenities in many rural areas also contribute to the shortage.  Even where adequate primary care is available there is frequently a shortage of specialty care providers or a limited number of specialists who will accept the lower payment rates to treat safety net patients.

Fragmented care.   Many safety net organizations compete with each other for patients.  Even when this is not the case, they may not be well integrated into other private or public community-based systems of care.  Thus, they are poorly positioned to provide coordinated integrated care across the continuum of patient needs as incented by the ACA.  The challenge will be to create safety net systems of care with EHR connectivity that can follow patients from primary care to specialty care to hospital and post-hospital settings as needed.  This will require the development of practice guidelines, care transition systems, and associated quality improvement tools associated with the development of patient centered medical homes.  A special need in the safety net community is to integrate behavioral health programs with primary care.  In order to address the obstacles noted above, many safety net organizations will need to develop new partnerships with other providers in the community.   This will require developing leadership skills and new governance models than what currently exist.

The Readiness Instrument

To address these issues more thoroughly, we developed a readiness assessment survey instrument that can be used by all ACOs forming across the country, but may be particularly beneficial for those serving safety net populations. We distributed the 81-question survey to safety-net providers in two California counties, one in Northern California and the other in Southern California. The survey includes questions covering several categories, including establishing governance and leadership structures; forming partnerships, acquiring financial resources and establishing risk-bearing contracts; addressing legal and regulatory issues; managing clinical care; implementing electronic health records; reporting on performance; and assessing the impact of forming an ACO on the organization’s mission and populations served.  (More details on the construction, distribution, and results of the survey can be found in the appendix below and the links contained therein.)

Respondents achieved the highest score in the organizational mission and populations served category.  However, they still indicated the need for additional information on the new populations they would be serving under the ACA, including their socio-demographic characteristics, health care utilization, and health status.   In addition, they indicated that a shortage of primary care physicians, nurse practitioners, physician assistants, and other primary care providers to treat the expanded population would be a challenge.  They also cited the lack of home health and behavioral health resources.

In the financial resource and contracting area, respondents indicated deficiencies in being able to track utilization and costs under risk bearing contracts.  They also confirmed the need for more resources to make the needed investments to meet accountable care requirements.  Further, they did not feel that they were very well prepared to distribute shared savings that could result from participation in risk bearing contracts.

As expected, respondents expressed the need to build the organization’s capabilities with regards to electronic health record functionality.  This included the use of disease registries, embedding practice guidelines into the electronic health record, incorporating information from non-participating providers, and developing electronic patient communication and engagement tools.

In the area of managing clinical care, a major need identified was the integration of behavioral health programs into primary care.  This need is particularly important for the safety net population, many of whom experience mental problems associated with physical illness and disability.

Respondents also indicated that providing more accountable care would involve establishing new partnerships with hospitals, specialty physicians and existing integrated delivery systems.  They identified the need to engage these entities much earlier in planning discussions along with considering new categories of health care workers.

Respondents gave the lowest scores to legal and regulatory issues.   These included how they would protect their 501C3 tax exempt status, how they would distribute savings to avoid inducing physicians to reduce or limit medically necessary care, and how they would address a number of federal compliance issues.

Overall, respondents were more confident about their ability to meet the quality metric targets required for providing accountable care than they were about their ability to meet expenditure and cost targets.  Interestingly, higher scores on performance reporting capabilities were strongly and significantly associated with greater confidence in meeting the quality of care measures.

Policy Implications And Recommendations

The results of the readiness survey of safety net organizations in the two California counties confirm and add further specificity to the challenges faced by such organizations nationally in providing accountable care.  Since the majority of the increased insurance coverage and demand for care will come from safety net populations, it is imperative that public policy address the challenges these organizations will face.  As indicated, among the most important of these are payment/resource issues, work force capacity, electronic health record functionality, and legal and regulatory issues.  As a group, they will determine the ability of safety net organizations to provide the greater care coordination and care management demanded by accountable care requirements.

Payment incentives/resources. The survey respondents indicated that there was some ability to afford up to $2 million in startup costs to provide more accountable care, but there was little or no ability to invest as much as $10 million for such purposes, if that is what the costs end up being.  To address this issue, states could make additional advance payments to safety net organizations to assist with the startup costs involved in building needed infrastructure for accountable care, including investments in data collection, analysis, and reporting.  This is similar to the advance payment model in the Medicare shared savings program.

Coordination bonuses could also be provided to safety net providers who integrate behavioral health care in the overall care management platform, a challenge also identified by respondents.  This integration could take many forms, ranging from direct employment of clinical psychologists, social workers, psychiatrists and other mental health professionals to partnership arrangements with behavioral health organizations in the community.   One example of the latter is provided by the Minnesota Community Mental Healthcare Management (CMHCM) initiative which has demonstrated cost effective care for the less severely mentally ill.  The coordination bonus could be paid from savings projected to come in the future from fewer hospitalizations and emergency department visits.

Finally, state Medicaid agencies could consider bonus payments to safety net providers, including both public and private disproportionate share hospitals, who concentrate their referrals to high quality/low cost specialists.  This would create incentives for specialists to work with safety net providers and for safety net providers to push greater volume to those specialists who are the most cost-effective.  This would help address the lack of access to specialty care, identified nationally and by respondents as a particular barrier for many safety net organizations.

Expanding workforce capacity.  Given the current shortage of primary care providers, particularly in areas serving safety net populations, states should re-examine current scope of practice laws and regulations governing what various health professionals can do.  Consideration should be given to adopting a competency-based approach to what services nurse practitioners, physician assistants, pharmacists, and other professionals can provide, rather than relying exclusively on educational degrees or outdated certification/licensing laws.  For example, in many states certified medical assistants must be supervised by physicians.  Consideration should be given as to whether nurse practitioners can also play such supervisory roles.

In addition, legislation may be needed to implement training for new categories of health workers such as promotoras and community health workers.  These providers are already in existence and can be used to help fill the shortage of primary care providers, especially for the safety net population.

The above notwithstanding, the demand for care is likely to exceed supply for many safety net populations.  The rapid growth of retail clinics in drug stores and related outlets across the country represents an additional resource that can be used to provide basic primary care services such as immunizations, flu shots, treatment for minor illnesses, and — with appropriate training and monitoring — helping patients in the self-management of such chronic illnesses as asthma and diabetes.  The payment incentives noted above should encourage some safety net organizations to include retail clinics in their network and develop relevant staffing arrangements.  The clinics also can serve as an additional source of outreach for newly insured patients.

Technical assistance for implementing electronic health records. As identified above, the adoption and implementation of electronic health records is particularly challenging for safety net providers.  States should provide assistance to safety net provider organizations to allow them to take full advantage of the existing HITECH Act and other existing financial incentives (federal and otherwise) to adopt and implement electronic health records.  This could be done through a small percentage charge of 1 percent or less on health insurance premiums.  These funds would be put into a pool for electronic health record implementation support for safety net providers.  Some of these funds could be set aside for development of an electronic health record user support network for safety net organizations.  The funds could also be used to pay competitive salaries for trained information technology specialists to work in safety net organizations.

States could also provide discounted fees to safety net organizations who participate in health information exchanges.  This is particularly important given that most safety net organizations see a high percentage of patients with multiple chronic conditions and this population requires a high level of care coordination across providers and settings, including the ability to link information across the continuum of care.

Addressing legal and regulatory barriers. As noted above, safety net ACOs will likely need additional grants or loans — a strategy which could have the side effect of complicating shared savings payment arrangements where IRS proportional distribution requirements (to maintain tax-exempt status) are concerned.  Some relief or flexibility in applying these requirements may be needed.  Another alternative is for safety net organizations to become benefit corporations which while not offering tax exemption, allow specified social welfare objectives to be pursued without liability to directors for not maximizing profit to shareholders.


Organizations providing accountable care to safety net populations will require specifically targeted policies that promote the development of capabilities in care coordination and management, behavioral health integration with primary care, patient-centered medical home capability, and the capacities to coordinate care with specialists, develop electronic health record functionality, and navigate legal and regulatory issues.  The recommendations based on the above analysis in California need to be tailored to the characteristics of each state’s safety net population and composition of safety net providers.  The assessment instrument is one tool that can be used to track progress across the country in efforts to provide more accountable care for the nation’s vulnerable populations.

Appendix: Readiness Instrument

This appendix provides additional information on the readiness assessment survey instrument distributed to safety net providers, as described in the text of the blog post.  The full instrument is available here, and more information regarding the development of the instrument and provider responses to it is available here.

We convened a national safety net advisory committee to provide guidance in the development of a survey instrument to assess the ability of safety net providers to address the challenges of providing accountable care.  We also reviewed existing ACO readiness assessment instruments to incorporate current industry thoughts on what constitutes general ACO readiness into our survey.  After several iterations, 81 questions organized under nine categories were developed for the survey.  A nine point behaviorally anchored scale from 1 (low) to 9 (high) was developed for the responses.

Since California has the largest number and most diverse safety net population in the country, we selected two California counties to survey, one located in Northern California and one in Southern California; both serve a high number of Medi-Cal, uninsured and dually eligible populations.

The northern county serves 211,000 Medi-Cal, uninsured, and dually eligible patients per year, involving over 1.4 million patient encounters, as part of a delivery system that includes 15 hospitals, 29 community clinics, and over 1,700 physicians.  It is overseen by a twelve-member board.  The southern county provides care to approximately 212,000 Medi-Cal, uninsured, and dually eligible patients per year, involving approximately 600,000 patient encounters.  These services are overseen by a multi-stakeholder coalition of organizations including FQHCs, community clinics, and non-profit hospital providers.

Preliminary interviews were first conducted with safety net organization leaders in each county to explain the purpose of the study, learn about the challenges the county was facing, and identify a list of relevant respondents for the survey.   Respondents included the county-wide safety net organizational leaders, clinical and managerial leaders of FQHCs, community health clinics, public and relevant non-profit hospitals serving safety net populations, and health departments.  Using a standardized letter of introduction explaining the purpose of the study, the lead contacts in each county sent an email to the respondent list encouraging participation.  They also discussed the study with potential respondents at group meetings to ensure high levels of participation.  The survey was administered to individual respondents online through Qualtrics and took approximately 30 minutes to complete.

A total of 51 responses were received; 26 from the Northern California county, representing a 68 percent completion rate, and 25 from the Southern California county, representing a 36 percent completion rate.   The completion rate was lower in the Southern California county due to the original inclusion of a much broader range of survey respondents.   The survey demonstrated high internal consistency reliability.  The mean score for the overall assessment of readiness was 4.8 on a 9 point scale, indicating that respondents felt that they have some of the capabilities to provide more accountable care but considerable room for improvement.