Shared Decision Making

Contents

1. A defined method
2. Decision aids and the language of risk
3. Barriers in practice
4. What the research shows

Candace Cooley’s later years in practice turned more and more toward office gynecology and preventive care for women, work that is full of choices between reasonable options, and shared decision making is the name of a defined method for working through them. The method has a standard instrument. The Agency for Healthcare Research and Quality teaches it in a one day workshop called the SHARE Approach, built to help clinicians and patients reach the best possible decisions using patient centered outcomes research.² A summary prepared for clinical teams under a federal contract lays out the five steps behind the acronym: seek the patient’s participation, help her explore and compare options, assess her values and preferences, reach a decision together, and evaluate that decision afterward.⁸ Inside those steps are smaller instruments. Clinicians are told to check what the patient already knows, describe each option without jargon, include the option of no care, and summarize with a teach back to confirm understanding.⁸

A defined method

The agency’s own definition describes a collaborative process in which patients and clinicians make decisions informed by evidence, by the care team’s knowledge and experience, and by the patient’s values, goals, preferences and circumstances.¹ It adds that the process can happen in one visit or across several, with one clinician or more, and that family members and caregivers often take part.¹ A decision sciences center in Boston names three required components: clear, accurate and unbiased evidence about the reasonable options and the risks, benefits and burdens of each; clinician skill in communicating and tailoring that evidence; and the patient’s goals, informed preferences and concerns, including the burden of treatment.⁴

The method is defined partly by what it replaces. A review in the Journal of Brown Hospital Medicine contrasts it with a traditional model in which the physician decides alone and then presents the choice, which can reduce the patient’s role to agreement without a real discussion of her preferences.⁵ The same review notes that most patients prefer an active role but often perceive decisions being made against their wishes, and that there is rarely a perfect option, since nearly every treatment carries some uncertainty and side effects.⁵

Figure

Two ways a clinical decision gets made

	Traditional approach	Shared decision making
Who decides	The physician, then presents the choice	Patient and physician together
Evidence	Applied by the physician	Laid out for the patient with risks, benefits and alternatives
Patient's values	Often not explored	A required input
Patient's role	Agreement to the plan	Active part in the choice

Decision aids and the language of risk

Decision aids are the best known tools. The Boston center describes them as materials that help patients understand the evidence and identify their own preferences without steering them toward any one option.⁴ They range from pamphlets and videos to interactive programs. The AHRQ guide recommends building a library of evidence based aids and describes practices that mail a decision aid to patients before the visit.² A simpler device is the set of BRAN questions used in the United Kingdom: what are the benefits, what are the risks, what are the alternatives, and what happens if nothing is done.⁵

Numbers are the hard part. An NPR report described a survey in which only about 20 percent of primary care physicians said they were very comfortable explaining numbers and probabilities, and quoted an internist who noted that a phrase like small risk can mean very different things to a doctor and to a patient.⁹ One answer is a picture of 100 figures in a grid, with some colored to show how many people like the patient would be expected to have an outcome over ten years.⁹ Not every patient wants the picture. The same report followed a man weighing a hip replacement who had read a booklet, watched a video and spoken with friends who had the operation, and who found a plain figure of about half a percent for infection risk the most useful number of all.⁹ Another answer is fixed vocabulary. The report explained that a European scale defines a very common side effect as one affecting more than 10 percent of people and a very rare one as fewer than one in 10,000.⁹ Research funded by AHRQ continues to build digital aids, among them a web tool to help older adults understand their lab results and a collaborative tool for breast cancer decisions.³

Figure

The five SHARE steps

1. 01
   SeekInvite the patient to take part, and say that her input matters.
2. 02
   HelpExplore and compare the options, including no treatment, in plain words, with decision aids where they exist.
3. 03
   AssessAsk open questions about what matters most to her and acknowledge those values.
4. 04
   ReachCheck whether she is ready to decide or needs more information, then confirm the choice.
5. 05
   EvaluatePlan follow up on the decision and on how she is doing.

Barriers in practice

The agency acknowledges that clinicians often do not involve patients in this way or struggle to do it in a busy day.¹ Time is the barrier named most often. The Brown review lists it first and suggests visual aids, simpler language, sitting at the patient’s level, and longer or follow up appointments for complex decisions.⁵ The AHRQ guide answers the time concern directly, citing studies in which trained clinicians used the approach without lengthening the consultation, and advises starting small in one department before spreading it.² It also warns that a single training session is not enough and recommends ongoing coaching.² Measurement keeps the practice honest. The guide points to a validated 12 item observer scale, called OPTION, for rating how far a clinician involves the patient, to international criteria for judging the quality of decision aids, and to an AHRQ tool for scoring how understandable and actionable patient materials are.²

Other barriers sit with communication itself. Low health literacy limits the discussion, and undiagnosed cognitive impairment or hearing loss can be mistaken for one another.⁵ Observations on hospital rounds found that clinicians most often explained the clinical issue and matched their language to the patient, and least often checked understanding or asked whether the patient had questions.⁵ Choices that come back year after year, like whether to have a screening pelvic exam, also lean on a lasting relationship with one clinician, the subject of continuity of care.

Figure

One national study of cost

rise in shared decision making

3 percent

associated fall in health expenditures

about 10 percent

years of patient surveys analyzed

2003 to 2017

What the research shows

The Patient Centered Outcomes Research Institute summarizes the evidence as showing that shared decision making can raise satisfaction, change the care patients receive and improve health outcomes.⁶ One study it funded found that a decision aid used in emergency departments helped parents understand concussion symptoms, their child’s relative risk of brain injury and the trade offs of a head CT scan.⁶ AHRQ reports that its funded research shows greater patient engagement and a better experience for patients and clinicians.¹ The Brown review adds that the approach reduces decisional conflict and increases patient knowledge, and that live and online training improves clinicians’ skills.⁵

The cost evidence is newer. A national study led by a health economist at the UC Berkeley School of Public Health, using surveys of more than 60,000 patients from 2003 to 2017, found that a 3 percent rise in shared decision making went with about a 10 percent fall in health spending, with no measurable change in physical or mental health, drug spending or inpatient use.⁷ The effect was roughly double when Latino patients saw Latino physicians and roughly triple when Black patients saw Black physicians, which the authors linked to trust.⁷ They acknowledged that some effects might take longer to appear.⁷ AHRQ separately reports that the approach has improved the experience and trust of racial, gender and sexual minority patients.¹

The evidence is strongest closest to the conversation itself: patients who go through a structured process with good decision aids understand their options better and feel less conflicted about the choice. Effects on long term health and on spending rest on fewer and more varied studies. Which option fits one woman’s goals is, by design, settled between her and her own clinician, and many of those conversations start at the yearly preventive checkup.