Shared Decision Making as Part of Value Based Care: New U.S. Policies Challenge our Readiness

Abstract

Shared decision making in the United States is increasingly being recognized as part of value-based care. During the last decade, several state and federal initiatives have linked shared decision making with reimbursement and increased protection from litigation. Additionally, private and public foundations are increasingly funding studies to identify best practices for moving shared decision making from the research world into clinical practice. These shifts offer opportunities and challenges for ensuring effective implementation.

The healthcare landscape in the U.S. and shared decision making

The United States is inching its way from a fee-for-service payment system to a value-based system in which health systems are increasingly being incentivized to achieve the three aims of better health, better care and lower costs, as opposed to greater volume.¹ One example of a value-based model is the establishment of accountable care organizations, implemented as part of the 2010 Patient Protection and Affordable Care Act (herein referred to as the Affordable Care Act or ACA). In this model, groups of providers come together to form a healthcare organization that cares for a defined population. Reimbursement is tied to quality metrics and cost savings. During the last 5 years, the number of ACOs increased from 110 in 2012 to over 800 in 2016, providing care to over 28 million Americans.²

In this context, shared decision making is being recognized as a strategy to advance the triple aims and support value-based care.³ Shared decision making has the potential to eliminate waste within the healthcare system by ensuring that patients get the care they want and nothing more. Shared decision making is also tied to patient safety, and in some areas of the U.S., malpractice carriers are offering discounted rates to physicians who practice shared decision making as an alternative to the traditional informed consent process. Indeed multiple organizations have identified shared decision making as a key strategy for achieving patient-centered care, enhancing patient safety, and achieving the triple aim of better health, better care and lower costs.^4,5

In the last decade several U.S. policies have aimed to integrate shared decision making into the value equation, forging shared decision making from the research setting into clinical practice. Yet, there remains a need to ensure effective implementation and measurement. Despite decades of research, there is limited experience with shared decision making; while clinicians are generally supportive of the principle that patients should be engaged in the decision making process, there remains skepticism that shared decision making with a patient decision aid is of value.⁶ Moreover, with the exception of a few recent mandates, the lack of incentives for performing shared decision making remains a major challenge. In this paper, we will describe the recent federal and state policies to advance shared decision making, along with the research landscape to support next generation science in shared decision making. Finally, we will discuss the challenges for implementation and potential next steps for advancing high-quality shared decision making.

National and State Policies Promoting Shared Decision Making

While the future of the American healthcare system is unknown in light of the 2016 presidential and congressional election, shared decision making as part of value-based care is unlikely to become a partisan issue. Moreover, much of the legislation related to shared decision making in the Affordable Care Act has already been enacted. The National Quality Forum, designated by the ACA as the entity to define a consensus process to certify patient decision aids, commissioned a white paper on how to ensure the quality and safety of decision aids, and convened an expert committee to develop national standards for decision aid certification. The white paper and certification standards, funded by the Gordon and Betty Moore Foundation, were published in December 2016, though the mechanism for certifying decision aids has yet to be implemented. Second, the Patient-Centered Outcomes Research Institute (PCORI), also established through the ACA, has already allocated millions of dollars over the last 7 years to support patient engagement and shared decision making, including funding the development and testing of patient decision aids. A careful analysis of PCORI’s decision aid funding portfolio through January 2015 showed 56 projects with a decision aid as a component, comprising 17% of all projects funded by the institute.⁷ Finally, the Patient-Centered Outcomes Research Trust Fund, which provides funding for PCORI, also provides training and infrastructure support for patient-centered outcomes research and comparative effectiveness research through the Agency for Healthcare Research and Quality. Several of these projects include an emphasis on shared decision making in some form.

At the federal level, several payment models incorporate incentives for shared decision making. The Centers for Medicare and Medicaid Services (CMS), the federal agency that administers Medicare and that works in partnerships with States to insure individuals with limited resources, established new payment programs that are based on value, not volume. Providers can form an accountable care organization that participates in the an advanced alternative payment program that evaluates several quality domains to determine cost sharing, including patient-reported shared decision making and the quality of provider communication. Providers not participating in an advanced alternative payment model will soon be enrolled in the Merit-based Incentive Program, in which 4 components of care will be used to determine reimbursement: quality; resource use; clinical practice improvement activities; and advancing care information. Among the clinical practice activities, there are plans to incorporate measures of patient preferences and shared decision making. Taken together, in the coming years, all providers caring for Medicare beneficiaries may have some portion of their reimbursement tied to shared decision making.

In a more direct way, CMS recently stipulated shared decision making as necessary for reimbursement coverage of 2 procedures: lung cancer screening with low-dose CT scan and left atrial appendage occlusion (a procedure to reduce thromboembolic stroke in patients with atrial fibrillation) – in recognition that fully informed patients may decide differently about whether to have the procedures. The coverage determination for lung cancer screening with low-dose CT was the first to explicitly require a patient counseling and shared decision making visit, and the use of patient decision aids, prior to a patient receiving a written referral for screening.⁸ For left appendage occlusion, only a non-operating clinician can counsel patients through shared decision making. Yet, it remains unknown whether other future coverage decisions will also be tied to shared decision making. Additionally, as authorized in the Affordable Care Act, the Centers for Medicare & Medicaid Innovation is launching a pilot program to test 2 different approaches to shared decision making for 6 preference sensitive conditions (stable ischemic heart disease; hip osteoarthritis; knee osteoarthritis; herniated disk and spinal stenosis; clinically localized prostate cancer; and benign prostrate hyperplasia) -- the Shared Decision Making Model (clinician engagement of patients in shared decision making) and the Direct Decision Support Model (decision support organization engages patients in shared decision making outside of the medical setting).⁹

At the state level, Washington State has made progress in implementing legislation passed in 2007, which encourages a shared decision making process that incorporates a certified patient decision aid and an attestation of a shared decision making conversation between the provider and the patient as a preferred alternative to traditional informed consent processes and forms.¹⁰ In 2016, the State began the process of certifying patient decision aids, starting in the area of maternal-fetal care. Future certification will focus on decision aids for the management of: orthopedic conditions, chronic ischemic heart disease; low-risk prostate cancer; low back pain; and advance care planning. As an interesting consequence of this work, a regional insurance carrier is offering 5% discounts to clinicians who take a course on shared decision making communication and practice shared decision making. As the first state to certify decision aids, their experience has already informed the National Quality Forum’s defined certification standards and will likely continue to inform future national policy.

Evidence-based Guidelines Promoting Shared Decision Making

Clinical guidelines committees have increasingly incorporated recommendations for shared decision making as a best practice, especially when the evidence is either unclear or supports two or more reasonable strategies. However, these recommendations are mostly focused on sharing information or risk estimates, and less on the process of shared decision making. For example, in the much discussed guidelines for cholesterol management, shared decision making is recommended as the preferred strategy for people with low-moderate risk. Unfortunately, the recommendation was not linked to any decision aids, despite their availability and proven effectiveness – or education on how best to support patients in developing informed preferences, leaving clinicians to figure out how best to engage patients in this discussion.

Perhaps the first clinical guideline to endorse shared decision making was the American Cancer Society 2010 updated guideline on early detection of prostate cancer.¹¹ For asymptomatic men with at least a 10-year life expectancy, the Society recommended “an informed decision-making process” with a health care provider after the patient received information about the uncertainties, risks, and potential benefits associated with screening for prostate cancer. The Society noted that decision aids are helpful in preparing men to make a screening decision, but did not endorse a specific aid. Like other professional organizations, the guideline identified core elements of the information that should be provided to men as part of the decision-making process, but did not offer incentives or guidance to support clinician time or communication skills.

The United States Preventive Services Task Force (USPSTF) has issued preventive services recommendations that include an emphasis on shared decision making to varying degrees. For example, the updated recommendation about lung cancer screening includes a section on shared decision making and highlights key issues that clinicians should address with patients.¹² This Grade B recommendation is particularly important because the Affordable Care Act requires all preventive services receiving a Grade A or B from the USPSTF be provided to patients without copayment.

Research Landscape

Several federally funded research institutes have prioritized shared decision making. As previously discussed, PCORI has funded multiple studies of shared decision making for a range of conditions in cardiovascular care, maternal-fetal health and gastrointestinal diseases, along with many others. Appropriately, PCORI is now interested in dissemination and implementation research to understand how best to scale shared decision making. The National Institute of Health has funded multiple studies that evaluate the use of shared decision making for cancer screening and management, as well as to guide end of life care. Finally, the Agency for Healthcare Research and Quality is supporting shared decision making through several funding strategies, including investigator-led grants (e.g., development of measures of shared decision making), the John M. Eisenberg Center for Clinical Decisions and Communications Science to translate the science of comparative effectiveness and produce shared decision-making interventions, the SHARE Approach curriculum for training health professionals in shared decision making, and policy related to patient-centered care, including shared decision making.

Additionally, the investments of several private foundations have enabled shared decision making to leap frog onto center stage, informing national payment programs. As previously described, the Moore Foundation funded the National Quality Forum’s effort to produce standards for the certification of decision aids. They have also been instrumental in supporting Washington State’s efforts to certify decision aids and educate/train clinicians practicing in accountable care models. The Center for Medicare & Medicaid Innovation implemented a clustered, randomized controlled study called Million Hearts to determine whether financial rewards for reducing 10-year predicted risk of cardiovascular disease across a physician’s patient population is an effective model to reduce the burden of heart attack and stroke. In the intervention arm, provider groups measure patients’ 10-year risk for a cardiovascular event using a risk calculator. Though there are no specific interventions that involve decision aids, one of the pre-stated predictors hypothesized to influence performance is the use of shared decision making. As such, the intervention arm will receive support training on how to engage patients in this discussion about risk and the use of shared decision making.

Challenges for implementation

Despite the flurry of new policies and the robust funding of research, there are several challenges to implementation. In order to meet new policy initiatives, several provisions need to be considered, including attention to the certification of patient decision aids as well as the supports for effective shared decision making. There is a need for rapid certification of patient decision aids, at least for the conditions that are currently specified in the legislation above, including the 2 conditions for which CMS reimbursement is incumbent upon shared decision making. This certification is necessary for clinicians, patients and health systems to trust the quality of the data presented to patients and its source, ensuring that the information presented to patients is evidence-based, balanced, and free of bias. Moreover, certification standards can ensure that developers employ a user-centered design, assuring that the information presented is of import to patients, is comprehendible, and is presented in a manner that helps patients develop informed preferences.

Assuming decision aids for the above conditions are certified, there is a need to consider how clinicians and health systems will shoulder the cost, as many decision aids are likely to be proprietary. In the U.S., only a handful of developers of decision aids will be able to meet the requirements for certification, which are heavily based on criteria established by the International Patient Decision Aid Society (IPDAS) specifying that decision aids be developed and tested with patients and physicians, a necessary though expensive process. Additionally, without funding mandates, developers will likely be responsible for the ongoing maintenance of decision aids every 2 years or earlier should new evidence become available. If decision aids are not made public, the expectation will be that clinicians and health systems bear the cost, which in the short term can pose a financial burden.

Independent of the process for forging ahead with decision aid certification, there is a broader need to develop largescale knowledge of and experience with shared decision making. In a study by Herrin et al of strategies used by hospitals to engage patients, 70% reported using decision aids.⁸ The authors conducted a secondary analysis of 50 such hospitals, asking them to send examples of decision aids. Perhaps not surprisingly, 14 of 15 respondents were referring to patient education pamphlets or instructions on how to prepare for surgery; only 1 was actually using decision aids. Assuming decision aids are promoted and made readily accessible, there is still a risk that shared decision making will be relegated to the status of a patient hand-out or pamphlet, which is informational, but does not necessarily support engagement and alignment of healthcare decisions with patients’ preferences, values and goals.

To start, there is a need to educate and train clinicians about shared decision making, recognizing that many clinicians are skeptical of its value. In our experience, once clinicians gain experience with shared decision making, they are clamoring for tools that can help them with their conversations with patients. However, in the U.S., few clinicians are familiar with shared decision making in clinical practice. Despite the broader culture of patient autonomy, there are persistent concerns among clinicians related to sharing evidence with patients and shifting the responsibility of decision making to the patient. Some believe that they already do ‘what is best’ for the patient, and hail the ‘what would you do if you were me, doctor’ as the ultimate bar for ensuring good decision making. Training can help clinicians gain the skills to discuss the limitations of the evidence.

Additionally, many clinicians view regulation, in any form, as an invasion of the doctor-patient relationship and a further demand on their time. An evolving body of science shows that regulations that lead to increased administrative tasks and that are perceived as being low-value, threaten clinician well-being and increase burnout. While easy to dismiss, those of us in the field of quality improvement and outcomes must concede that in the effort to improve and standardize quality, physician burnout has climbed; whether these are directly related is unknown, but certainly plausible. As such, some have considered clinician well-being to be the 4^th aim for value-based care.⁸ With new policies to drive the implementation of shared decision making, we must be mindful of the potential for increased clinician burden. Strategies that re-conceptualize care as being delivered by a team (and not just by an individual) and that encourage system-wide investments that integrate decision aids into routine work-flow, can help to create a culture that supports shared decision making. Yet, evidence from dissemination and implementation studies are needed. As such, there is great interest in the forthcoming pilot of 2 shared decision making models (under the Centers for Medicare & Medicaid Innovation) to better define the role of the health system and team in engaging patients in shared decision making, and to illuminate best practices as well as ineffective or unfavorable strategies.

Measuring Shared Decision Making

Measures of shared decision making are needed to prevent it from becoming a check-box item for reimbursement programs. Yet it is not clear how best to measure shared decision making outside of the research setting. Elwyn, Frosch and Kobrin et al propose a conceptual model of the outcomes that are salient to the process of collaboration and deliberation; they are organized into proximal, distal and distant effects or consequences that can emerge from a shared decision making encounter.¹³ Currently, two patient-reported measures have been reviewed by the National Quality Forum: a measure of decisional quality (Healthwise) and a measure of patient engagement (CollaboRATE).¹⁴ However, some practical issues remain, including knowing when patients are in the decision making window. Additionally, while there is certainly value in patient reported experiences,¹⁵ they often do not correlate with direct observation of quality, which has raised concerns about whether patients can be a good judge of shared decision making.

An alternative approach is to start with measuring the quality of informed consent. Given that informed consent is a routine part of clinical care, yet fails to meet even basic standards to support patient autonomy and safety, there are efforts to use the informed consent milestone in clinical care as a starting point for shared decision making.¹⁶ CMS is supporting the development of a hospital-wide measure of the quality of informed consent, which is not a measure of shared decision making, but creates a set of standards to ensure that patients are given information about the procedure in a timely manner, as a basis for collaboration, deliberation, and shared decision making. In Washington State, shared decision making with a certified decision aid is the preferred alternative to traditional informed consent. The limitation of this work, however, is that, necessarily, the cohort of informed consent measures is limited to those who have decided on a procedure or treatment for which informed consent is standard practice. Additionally, the focus is on delivering content to patients, which is necessary though not sufficient for helping patients develop their preferences, values and goals to make high-quality decisions.

Next steps to ensure SDM advances a value-based system.

To date, most research on shared decision making has arguably been efficacy research – does it work under ideal (or near ideal) conditions? As we move further into implementation, we’re asking a different question. Does this still work in the real world? Which elements don’t happen reliably? (Perhaps because they’re not feasible in routine practice). And which elements are truly essential for us to be able to say that shared decision making happened? Health systems such as Massachusetts General Hospital, the Mayo Clinic, the Dartmouth-Hitchcock Health System, and the University of Colorado are using decision aids for a number of conditions. There are also many clinicians using decision support interventions, yet we do not know how extensively these tools are being used, nor the quality of the shared decision making process. The directional goals of PCORI and the Centers for Medicare & Medicaid Innovation to compare different strategies for implementation are aligned with this next-generation research. More research is also needed to understand how shared decision making impacts the broader goals of patient experience and outcomes, patient safety, clinician experience and burnout, and impact on health system (i.e., throughput; cost).

Future efforts to implement SDM might also consider integrating shared decision making as part of a broader goal of patient engagement. Clinicians, researchers and health systems are increasingly engaging patients in the coproduction of goals and strategies to achieve these goals. Shared decision making may be considered the highest, most sophisticated form or patient engagement, as it attempts to level the balance between clinicians and patients, making them equal players in the clinical care. This idea remains revolutionary and inconceivable to many, and is unlikely to gain traction unless there is a larger culture of patient engagement.¹⁷ Efforts that acknowledge patient well-being and experience (e.g., disruptions and quiet time in the hospital; wait times and e-communication in the outpatient setting); open notes; recording of medical interviews and such, may compliment efforts to implement shared decision making.

Shared decision making is finally getting recognition as an important component of value based care. Though we still have a long way to go to achieve a true shift in attitude that accepts the notion that patients should be active participants in their care. The community of shared decision making, which has mostly been comprised of researchers and increasingly clinicians, is necessarily needing to open its doors to designers, health system innovators, technology experts and patients – who may have broader goals, such as patient safety and experience. The future of shared decision making may very well rest of these other key players, yet it is up to us to ensure that the science to inform best practices stays current and rigorous, and that new incentives to promote shared decision making result in effective implementation that raises the quality of care for patients, clinicians and the larger health system.