Framing the Issue

  • Shared decision-making is a collaborative process in which patients and providers make health care decisions together, weighing the medical evidence of various options and considering the patient's values and preferences.
  • SDM often involves the use of decision aids — usually written materials or videos — that help patients learn about and evaluate their treatment options.
  • Most clinicians support SDM in theory and may believe they use it. However, many structural barriers make the effective use of SDM challenging. Saying, "Here's my recommendation. How does that sound to you?" is not a shared decision.
  • The transition from volume-based to value-based care creates a business case for SDM. Research shows that, when SDM is used, many patients opt to avoid some procedures.

After years of quietly spreading from one physician practice to another, shared decision-making is leaping onto center stage. Hospitals hoping their institutions will thrive in a value-based environment would be well-advised not just to understand the movement, but also to support it and insist that clinicians buy into it.

Under SDM, patients and physicians collaborate on making health care decisions, taking into account the medical options and the patient's preferences.The federal government issued its second-largest innovation grant — $26 million — to push shared decision-making to a national collaborative that includes Dartmouth-Hitchcock, Mayo Clinic, Denver Health and 12 other leading health systems. Shared decision-making is a quality measure for accountable care organizations in the Medicare Shared Savings Program. The National Committee for Quality Assurance makes SDM a criteriaon for patient-centered medical home recognition. And at least four states have policies promoting SDM.

The reason: Shared decision-making supports the move to value in health care delivery. When patients choose what they want — rather than what the health care provider is selling — many opt for less-intensive, less costly treatments and they report higher satisfaction with their care.

In the fee-for-service world, investing in the infrastructure to support SDM meant a capital outlay that would likely would reduce demand for services. But, as value-based contracts emerge, SDM is increasingly is seen as a way to lower costs while improving patient satisfaction.

"In the past, we weren't necessarily rewarded for truly engaging patients and families — to the detriment of their care," says John G. O'Brien, president and CEO of UMass Memorial Health Care in Worcester. "I think every health care system can benefit from looking at this, not only from a patient perspective, but for health care systems that have significant financial risk for populations of people."

State of SDM

The SDM concept dates to 1989, when two Dartmouth-Hitchcock Medical Center physicians — John Wennberg, M.D., and Albert Mulley Jr., M.D. — started the Informed Medical Decisions Foundation. The next year, the foundation published its first decision aid — a video to help patients understand the pros and cons of prostate cancer treatments.

Since then, the foundation and others have produced hundreds of decision aids and SDM has gained traction steadily, mostly in primary care practices. For the last few years, the foundation has supported SDM demonstrations at Massachusetts General Hospital, Oregon Rural Practice-based Research Network and other sites around the country.

The foundation conducted patient surveys in 2008 and 2011 to examine the prevalence of SDM and the quality of medical decisions patients had made.

"We found that [typically] the discussion with providers involves the pros of doing something, like having the surgery or taking the medication or having a screening test, much more than the cons," says Richard Wexler, M.D., the foundation's chief medical officer. "Generally, patients are not asked to weigh in on the decision and not asked what they would like to do. So, we have a long way to go."

That said, Wexler believes patients increasingly are aware that tests and treatments may have a downside, and they want more information. Add that to the push SDM is getting from policymakers and health systems' newfound motivation to lower costs, and he believes SDM eventually will become standard medical practice.

"We are getting much closer to the tipping point than we've ever been," he says.

Research results

Updating research from the early 2000s, investigators from the school of nursing at the University of Ottawa in 2011 reviewed numerous studies on shared decision-making and found that when patients are presented with decision aids, they are less likely to choose to have an elective surgery. Those findings were underscored last year when Seattle-based Group Health Cooperative reported on its use of decision aids in a real-world setting — an orthopedic service line that includes 27 surgeons and 15 physician assistants working in five specialty clinics.

After introducing decision-aid videos for Group Health patients with osteoarthritis, researchers found a 38 percent reduction in knee replacement surgery over a six-month period and a 26 percent reduction in hip replacement. That translated into a 12 to 21 percent reduction in costs, depending on the location, for patients with knee and hip osteoarthritis.

"Patients don't always want the most expensive care," says Dominick L. Frosch, a Gordon and Betty Moore Foundation fellow who has been studying SDM for 15 years.

Barriers abound

Despite that, many studies show serious barriers that are preventing widespread adoption of SDM.

When Frosch and some colleagues reviewed SDM in five primary care clinics in northern California for patients suffering back pain or considering colorectal cancer screening, they found that decision aids reached only about 10 percent of eligible patients. Physicians cited time as a primary barrier to using decisions aids, but researchers also found that some physicians did not fully understand SDM despite many attempts to inform them; some disagreed that patients should have input on certain decisions; and some would not deviate from physician-directed decision-making even when they claimed to support the SDM concept.

"We don't have reason to believe that our experience is unique," Frosch says. "It is the same thing we hear from our colleagues around the country."

Indeed, Wexler and his colleagues found similar barriers when they evaluated SDM at eight primary care sites participating in the foundation's three-year demonstration.

In addition to overworked physicians and insufficient physician training, the evaluators found that clinical information systems, either paper or electronic, were inadequate to flag patients as candidates for SDM or track them through the process of receiving a decision aid, communicating their values and preferences to a clinician, and having an SDM shared decision-making conversation.

The researchers concluded that substantial investments in physician training, information systems and process reengineering are needed to successfully implement SDM.

Advancing SDM

The federal government believes those barriers can be overcome. That's why the Center for Medicare & Medicaid Innovation provided that $26 million grant to the Dartmouth Institute for Health Policy and Clinical Practice to implement SDM for patients at Dartmouth-Hitchcock and the other 14 large health systems.

James N. Weinstein, D.O., CEO and president of the Dartmouth-Hitchcock system, expects the project to save $64 million in three years, thanks to reduced utilization and costs stemming from patients' being actively involved in their own health care decisions.

The Innovation Center grant funding will be used to hire and train about 48 "patient and family activators" — typically nurses — who will engage patients and their families in SDM for a variety of medical situations. This avoids tasking physicians to take the lead.

"Most spine surgeons do what they're trained to do — spine surgery," says Weinstein, who started Dartmouth-Hitchcock's Center for Shared Decision-Making in 1999. "This patient activator will be somebody who understands how to present information in an unbiased way to the patient."

Meanwhile, Massachusetts General, another longtime leader in SDM, also is experimenting with new ideas. For many years, Mass General primary care physicians have been able to prescribe decision aids for patients via electronic health record technology. Every physician receives a quarterly report showing how his or her own use of decision aids compares with other PCPs in the system.

"What we find is that doctors are busy and they forget or they don't always have it top of mind, so many patients are not getting the programs," says Karen Sepucha, M.D., director of the Health Decision Sciences Center at Mass General.

One new initiative is informing patients about the availability of various decision aids before they see the physician and asking them which ones they would like to view. That approach increased decision-aid orders by tenfold.

Additionally, Mass General found that patients ordered decision aids about the management of chronic, symptomatic conditions much more frequently than physicians did. "The doctors tend to order PSA testing for prostate cancer, colon cancer screening, advance directives," she says. "When we opened it up to the patients, they asked, 'What are my choices for dealing with insomnia? Depression? Low back pain?'"

Mass General's newest initiative is to extend the use of decision aids into specialty care. Its ob-gyn group focused on quality improvement measures related to SDM in two subsequent quarters. In the first quarter, each physician had to view one of the decision aids relevant to his or her patient population; in the subsequent quarter, each physician had to use one in practice.

"That has been tied to the quality improvement bonus as a way to try to get this accepted and integrated into practice," Sepucha says.

Surgeons at Mass General are getting an even stronger nudge to use SDM. A protocol is being implemented to ascertain the appropriateness of elective surgery procedures. Before an operating room will be scheduled, a physician must document that the patient meets the clinical criteria for the procedure and that a shared decision-making process has been completed.

"This is starting to get rolled out across Mass General with the eventual goal of getting it rolled out across the entire Partners HealthCare System," Sepucha says.

Meanwhile, technology may help organizations adopt SDM. Holly Toomey, associate vice president, care management strategy, at McKesson, says "impact analysis" programs can help to identify patients whose medical situations might be appropriate for this type of intervention.

Need for support

After studying SDM for more than a decade, the Informed Medical Decisions Foundation's Wexler believes financial incentives will be needed to give SDM a boost into standard practice. "We've come quite far in being able to measure decision quality ([see sidebar)], and we would like to see financial incentives to help reward that when we see evidence of high-quality, decision-making at a population level," he says.

Frosch, meanwhile, thinks health system leaders must support a culture change by making SDM an expectation for preference-sensitive medical situations. "This isn't just a nice thing to do or the right thing to do," he says. "It must be 'We will work with you so that it is possible and we expect you to do this.' "

Lola Butcher is a freelance writer in Springfield, Mo.

Executive Corner

In a new report, "Engaging Health Care Users: A Framework for Healthy Individuals and Communities," the American Hospital Association Committee on Research found that shared decision-making is one of many patient-engagement strategies that are paying off for America's hospitals. "One of the things that we are seeing from the most successful systems is that not only are they getting better patient outcomes, but they had increased patient satisfaction," says John G. O'Brien, president and CEO of UMass Memorial Health Care. "And almost all of the efforts are positively benefitting the bottom line." O'Brien, who co-chaired the Committee on Research in 2012 with current AHA Chairman Benjamin K. Chu, M.D., Southern California regional president, Kaiser Foundation Hospitals, says the committee found four broad categories of patient engagement:

Community-level strategies, which include providing health education and physical education classes, and training patient navigators. Methodist Le Bonheur Healthcare in Memphis partnered with almost 400 churches to improve congregants' health and help patients make a smooth transition from hospital to home.

Organizational strategies such as patient and family advisory councils.

Health care team strategies ranging from patient "teach-back" communication to shared decision-making tools. At Helen DeVos Children's Hospital in Grand Rapids, Mich., family members are included in daily rounds so they can help with decision-making and the day's care plan.

Individual-level strategies that help patients be proactive about their health. Howard University Hospital in Washington, D.C., provides patients in its diabetes program with electronic personal health record technology that allows them to monitor their blood sugar and weight and report to their clinicians between visits.

What good decisions look like

Many physicians think they are using shared decision-making — and most of them are wrong. "When you explain what shared decision-making is and ask physicians about their use of it, often they will say, "'Well, I already do that,' " says Leigh Simmons, M.D., a physician fellow who studies shared decision-making at the John D. Stoeckle Center for Primary Care Innovation at Massachusetts General Hospital. "Research has shown that shared decision-making is usually not happening even if the physician thinks that it is or despite best intentions."

Her colleague, Karen Sepucha, director of MGHMass General's Health Decision Sciences Center, has led the development of decision quality survey tools to support the actual — rather than perceived — use of SDM.

Decision quality is defined as the extent to which patients are informed and receive tests or treatments that reflect their personal values and preferences. So far, survey instruments have been created for more than a dozen common medical decision points, including treatment of knee and hip osteoarthritis, breast cancer and prostate cancer screening. The goal of the tools is to answer three questions:

1 | Is the patient informed?

Sepucha's research has shown that patients may report feeling well-informed about a medical decision even when they are unable to answer basic questions that demonstrate an understanding of the issue. That suggests that the standard informed-consent process does not support SDM. "They just have to sign the consent form, but does that really mean that they have understood the options, benefits and harms?" Sepucha says.

2 | Did the patient get the treatment that best matched his or her goals?

If a patient has the primary goal of pain relief and is not too concerned about surgery, an operation might be best. For someone who is able to manage his or her current pain, but is very concerned about having surgery, nonsurgical options are a better bet.

3 | Was the patient meaningfully involved in the decision?

"You can't really have a shared decision if the doctor makes a recommendation and discusses the pros of that option and the cons of the other ones that they don't recommend and doesn't ask the patient what they wanthe or she wants to do," Sepucha says. "That's not a very balanced discussion and not a shared decision."

So far, the survey instruments have been used primarily in retrospective national surveys in which patients are surveyed within a year after having a certain procedure. Going forward, Sepucha and her colleagues are trying to integrate the tools into care pathways so patients are surveyed at the decision-making point or very shortly after it. That will allow an ongoing assessment of whether clinicians are using the SDM process effectively and identify opportunities for improvement. — Lola Butcher