Accountable care organizations could dramatically affect radiology practice

March 6, 2021
Steven Gerst, MD, MBA, MPH

Decision support, employment models, new teams could remake incentives that today drive radiology ordering.

Decision support, employment models, new teams could remake incentives that today drive radiology ordering.

As new approaches to health information technology and evidence-based medicine affect payment methodologies, it’s time for radiologists to rethink their clinical practices and business strategies. Radiology will be affected not only because it reaches across the full spectrum of healthcare delivery, but also because of the movement of integrated delivery systems toward employing radiologists and the use of corporate teleradiology interpretations to reduce costs.

Three specific legislative measures will significantly affect the ordering of diagnostic imaging tests and determinations of appropriateness. These are the HITECH Act’s incentives for electronic health record (EHR) implementation, the Medicare Imaging Demonstration Project (under the Medicare Patient and Provider Protection Act [MIPPA]), and provisions in the 2010 Patient Protection and Affordable Care Act for the development of accountable care organizations (ACOs), which will take effect in January 2012.

The HITECH Act and Medicare Imaging Demonstration Project will help streamline radiologic electronic order entry and foster more appropriate use of diagnostic testing through decision-support systems. The HITECH Act will impose penalties on physician practices for noncompliance starting in 2015. If successful, the Medicare Imaging Demonstration Project, which begins in January 2011, may drive the adoption of decision support as part of this electronic workflow. Decision support systems may also form the missing “connector” between payers and providers who must document savings. Projections call for $5 billion in savings to be generated via ACOs. These monies will be reimbursed by the Department of Health and Human Services (HHS) under the Medicare Shared Savings Program.


The Patient Protection and Affordable Care Act of 2010, generally known as the healthcare reform law, is likely to have the greatest long-term impact on the future practice of radiology. This act creates incentives for hospitals, health systems, large multispecialty groups, and independent prac­tice associations to more closely integrate physician practices in order to generate savings that can be shared with the Centers for Medicare and Medicaid Services (CMS) under an ACO model of healthcare delivery. Much of which is expected to be reimbursed to ACOs by HHS over the next 10 years, if the ACOs can meet the quality and cost savings initiatives being developed under the newly formed Centers for Medicare and Medicaid Innovations (CMI”).

Congress has allocated $10 billion for the CMI to develop and implement the final rules for ACOs. Richard Gilfillan, M.D., was recently named acting director of CMI. He is the former president and CEO of Geisinger Health Plan and executive vice president of insurance operations for Geisinger Health System, an integrated health system with 750 physicians, three hospitals, and 12,000 employees. Many major health systems are already teaming with commercial insurance companies to form ACO models on behalf of employer groups. At least 10 pilot ACO projects are being developed right now.

Systems such as the Carilion Clinic in Virginia, Norton Health System (with Humana) in Louisville, KY, and Tucson Medical Center (with United Healthcare) have initiated pilot ACO projects through the Brookings-Dartmouth joint venture run by former CMS director Mark McClellan, M.D., of the Brookings Institution in Washington, DC, and Elliot Fisher, M.D., director of the Dartmouth Institute for Health Policy and Clinical Practice. Additional ACO pilots are likely to be announced in the coming months.

More than 70 major hospital systems and insurers are participating in learning groups with the intention of forming ACOs. Premier, the 2300-member hospital group, has an ACO Collaborative and Readiness program established. More than 200 hospitals have joined already.


At a high level, the ACO represents a business partnership between the payer (e.g., Medicare) and the provider to work together to bend the cost curve over time by reducing the inflationary trend in healthcare costs and sharing in the savings produced (see graph below). Commercially, Humana, UnitedHealthcare, and other payers are already joining with hospital systems to pilot ACO programs. They promote team-based delivery and financing models that encourage physicians, hospitals, and other healthcare providers to collaboratively manage and coordinate care for eligible beneficiaries.

Some integrated delivery systems and large multispecialty groups already consider themselves ACOs. They are waiting to take specific action until there are rules for application and certification in 2011. The government anticipates that savings will be generated by reducing duplicative services, improving productivity, minimizing paperwork, and improving cost efficiency. In doing so, providers stand to receive a share of the savings generated, including radiologists. Decision-support systems will play a key role in tracking and reporting on shared savings generated for providers and payers.

A recent Urban Institute paper on ACOs1 listed three essential characteristics:

• The ability to provide and manage, with patients, the continuum of care across different institutional settings, including at least ambulatory and inpatient hospital care and possibly postacute care;
• The capability to prospectively plan budgets and resource needs; and
• Sufficient size to support comprehensive, valid, and reliable performance measurement.


ACOs are designed to provide financial incentives to produce more efficient and effective care, combating the current perverse incentives of overutilization, overbuilding of healthcare facilities, and overuse of technology, all of which contribute to our uncontrolled healthcare cost spiral. Radiologists should be keenly aware of today’s significant changes in healthcare, and also become involved with new developments in their organizations, including the selection and use of clinical decision-support systems and the establishment of ACOs.

The ACO will require a governing entity that can receive shared savings. This will incorporate primary care physicians and specialists who practice under the ACO. To focus on quality and outcomes, rather than on maximizing reimbursement, the ACO must align the incentives of physicians and hospitals with those of the payer, focusing on improved patient care quality metrics, increased administrative efficiencies, and lower costs.

With a focus on bundled payment models and outcomes-driven treatment, ACOs will need to rethink approaches to radiology services. Making radiologists or other specialists (such as cardiologists) a part of ACOs may require a new infrastructure and a new approach to work, including an employment model. A model where radiologists are employees is likely to become an increasingly attractive option for new radiologists coming out of residency under an outcomes-driven model. It offers predictable shifts, a more scheduled work life, and a stable income.


According to a recent study in the Journal of the American College of Radiology, approximately 26% to 30% of imaging tests ordered are either unnecessary or inappropriate. 2 Since the annual cost of diagnostic imaging is now nearly $100 billion3 and growing 17% per year, 4 unnecessary testing represents nearly $35 billion in unnecessary costs. 5 Through its Medicare Imaging Demonstration Project, CMS is trying to determine whether decision-support tools can improve the appropriateness of imaging tests being ordered.

For office-based practitioners, a radiological decision-support system can provide guidance for ordering procedures according to clinically sound criteria. The system can also enable more effective electronic recordkeeping at the point of care. Additionally, an advanced decision-support system will allow a medical director to instantaneously modify rules based on clinical research, utilization review committee suggestions, hospital/health plan policy changes, local practice guidelines, and/or business rule requirements.

As ACOs evolve, payers and providers will want to share the rules to ensure transparency. Ensuring the most appropriate advanced imaging tests are ordered and authorized at the point of ordering will maximize payment to providers and improve revenue cycle management while also improving the quality of care delivered to the patient and minimizing unnecessary exposure to ionizing radiation.

Advanced systems will include the capability for automated precertification of diagnostic imaging procedures. Many believe this new approach will eliminate the need for third-party contracted radiology benefit managers. In fact, the pending Medicare Imaging Demonstration Project specifically prohibits the use of prior authorization.


Under the law establishing ACOs, a gainsharing arrangement allows Medicare (as the payer) to determine what savings have been generated for Medicare ACO beneficiaries based on a preestablished baseline such as 2010 claims data.

Decision-support systems will integrate with the ACO’s EHR to provide guidance for the most appropriate tests at the point of ordering based on criteria and protocols published by the American College of Radiology (ACR), the American College of Cardiology and other medical societies. This is expected to significantly reduce the reported 26% to 30% unnecessary or inappropriate tests currently ordered. By tracking changes in historical claims data, radiologists will be able to determine savings achieved under an ACO for shared reimbursement with CMS and commercial insurers.

An example might be when a primary care physician orders an MRI for a generalized headache with no other complicating factors or medical history. Under the ACR criteria, this study would be considered inappropriate or unnecessary. The ordering physician would be electronically informed of this at the point of ordering. In the current environment, the study may be performed and interpreted, yet payment may be retrospectively denied by the insurer based on ACR or RBM criteria as “not medically necessary.” This increases the radiologist’s accounts receivables and noncollectable fees, thereby decreasing his or her collection ratio and creating a problem in revenue cycle management.



According to Dr. McClellan and the

Commonwealth Fund



while the U.S. currently spends $2.4 trillion annually on healthcare (nearly 17% of GDP), by 2017, this number is expected to double to nearly $4.8 trillion. Currently, the entire world spends a total of $4.3 trillion on healthcare. While the U.S. represents only 5% of the world’s population, we spend more than 50% of the world’s healthcare dollars. Without the implementation of a projected $575 billion in health reforms under the Affordable Care Act, the Congressional Budget Office and CMS Office of the Actuary predict healthcare costs to be 31% of GDP by 2035 and 46% GDP by 2080.


The transparency generated by decision-support systems at the point of ordering would significantly decrease retrospective denials, consequently increasing a radiologist’s collection ratio by guiding ordering physicians to order only the most appropriate tests authorized for payment at the point of ordering. As ACOs develop, insurance carriers will incorporate their preauthorization rule-based protocols into the ACO’s decision-support system and electronically approve orders at the point of ordering, thus eliminating cumbersome retroactive denials and appeals. This will save the insurer money and ultimately reduce the need for expensive, labor-intensive utilization review departments, ultimately lowering HMO administrative costs.

Since payers and providers share in the savings produced under the ACO model, the elimination of costs related to preauthorization, denials, and appeals increases savings that can be documented by the ACO. For example, if the projected inflationary trend in advanced diagnostic imaging costs is 17% (as expected for 2011), and, after 2011 claims are tallied the actual medical cost inflationary trend under a qualified ACO is found to be lower, HHS may consider the difference as ACO savings and pay all or a portion of that back to the ACO provider (usually a hospital system where Medicare is the payer).

Similarly, commercial carriers that contract with integrated delivery systems and large multispecialty groups under ACOs will set a baseline trend to establish “savings” generated by the ACO. The difference between this trend and any lower total costs will be considered ACO savings.

Finally, the Medicare Imaging Demonstration Project allows for six months of baseline data collection before implementation of a decision-support system to identify and track behavioral changes in ordering patterns and costs. This is expected to be much more physician-friendly and more effective under an ACO model than under retrospective denial processes currently imposed on ordering physicians by contracted third parties.


ACO legislation is designed to force payers and providers to jointly identify the proper nexus for shared accountability for a patient’s healthcare and to produce quantifiable savings. Many of the specifics are left to the discretion of the secretary of HHS, which will allow the design of the program to evolve over time. This includes criteria to qualify for savings reimbursement.

The method whereby ACO savings are distributed to individual providers is left to each ACO to resolve. For example, if physicians submit traditional claims to Medicare under the RBRVS system while the hospital submits its typical DRG-based claim, what method can best be implemented to determine the savings divided among the providers and hospital unless the physicians are employed?

There is also the significant problem of revenue maximization. While a large portion of integrated delivery systems are private nonprofits, a number of large institutions and for-profit hospital chains may operate under publicly traded holding companies. Their interests are driven by shareholders focused on profit maximization rather than a retrospective shared savings reimbursement determined by HHS. Most major insurers participating in ACOs are also likely to be publicly traded. If this model is successful, new baselines upon which savings will be calculated will get progressively lower each year. While this is the goal of the Congress and HHS, unless profitability can continuously be improved for both payers and providers, other incentives will need to be designed by CMI to sustain this model over the long term.

Finally, with the ACO model, the projected cost savings may attract greater volumes of commercial employed members and patients. In this scenario, the payer stands to benefit, but a potential reduction in per patient revenue may affect the willingness of providers to participate. To counter this, payers and providers are attracting employers to the ACO model by offering ever-decreasing annual premium increases resulting from the more efficient ACO model. This will increase ACO market share at the expense of insurers and hospital systems that do not participate in an ACO model.


Numerous changes under health reform represent a significant movement toward improving electronic workflow and tracking of cause-and-effect outcomes in a fragmented healthcare delivery system. Electronic medical records and decision-support tools will play a major role in this regard.

Just as in the Medicare Imaging Demonstration project, decision-support may be one of the few, if not the only, currently available software systems that can take an established baseline of claims data and effect a change in physician ordering behavior in a positive-more cost-efficient and higher quality-manner. Additionally, these systems may be among the few available to document the savings produced on both a per-transaction basis (by physician) for reimbursement from HHS under the Medicare Shared Savings Program and under a commercially insured ACO model. These savings can then be reported back to employers to document more efficient quality measures for marketing/sales and to justify lower premium increases.

Despite some flaws in the model, Congress is investing $10 billion to initiate, develop, and manage the ACO model, which is expected to save and return possibly as much as $5 billion to ACOs in its first eight years. This is considered the foundation of healthcare reform and should foster ongoing ACO development. While many features of the ACO Medicare Shared Savings program remain undefined, this law will result in ACO contracts becoming a permanent option under Medicare. Radiologists need to consider how this will affect their practices and work within the system to protect income while improving patient quality and outcomes.

1. Devers K, Berenson R. Can accountable care organizations improve the value of health care by solving the cost and quality quandaries? Washington, DC: Urban Institute, October 2009.
2. Lehnert BE, Bree RL. Analysis of appropriateness of outpatient CT and MRI referred from primary care clinics at an academic medical center: how critical is the need for improved decision support? JACR 2010;7(3):192-197.
3. Iglehart JK. Health insurers and medical-imaging policy-a work in progress. NEJM 2009;360(10):1030-1037.
4. Bautista AB, Burgos A, Nickel BJ, et al. Do clinicians use the American College of Radiology appropriateness criteria in the management of their patients? AJR 2009;192(6):1581-1585.
5. CBS News, Sept. 24, 2009.
6. Schoen C, et al. Bending the curve: Options for achieving savings and improving value in U.S. health spending. The Commonwealth Fund. Data: Lewin Group Estimates, 2007.
7. Affordable Care Act Update: Implementing Medicare Cost Savings. As retrieved from: Page 2; accessed Aug. 5, 2010.