Patient decision aids
A patient decision aid sets out the options for a health decision, with their benefits, harms and uncertainties, and helps the patient weigh them against their own values, so that patient and clinician can decide together. Building one well means knowing when a decision calls for an aid, what aids achieve and why health systems want them, what the international standards require of one, and how an aid is developed with patients and professionals, brought into a service, evaluated and reported.
What a decision aid is
Shared decision making is an underused communication process, designed for the clinical encounter, that helps patients navigate complex health choices and make informed decisions about their own care. It encourages doctors to personalise population-level evidence for individual patients whose socio-cultural contexts and personal preferences may be as varied as their literacy and their tolerance of risk. Decision aids — also called patient decision aids or decision tools, and made as brochures, videos, websites, cards or leaflets — have been developed across disciplines and decision points, from choosing a statin to hip replacement surgery to taking antibiotics for bronchitis, to support that two-way communication between practitioner and patient.
A patient decision aid can be defined as an aid that supports patients by making their decisions explicit, providing information about options and associated benefits and harms, and helping clarify congruence between decisions and personal values. At minimum, an aid should provide information on options, benefits and harms, and help patients clarify their values for the outcomes of options. Aids come as leaflets, videos and internet-based tools, and patients can use them before, during or after meetings with their health professionals, or on their own with no connection to a health professional.
Decision aids are evidence-based information packages outlining likely outcomes and uncertainties, used to support major or complex medical decisions. They range from brief summaries designed as conversation aids in a consultation to more detailed booklets, websites or videos for the patient to review at home, and they are available for many of the most common medical conditions.
The family of tools is wider than the decision aid itself. Some tools are explicitly designed to facilitate shared decision making, such as decision aids; others supply some of the information needed for parts of the process, such as risk calculators and evidence summaries; others provide ways of starting and structuring the conversation, such as communication frameworks and question prompt lists. For-profit and non-profit hospitals, research centres and private businesses had been developing decision aids for fifteen years by 2014, and by then the tools came in a variety of formats across print and digital media.
Decision support tools can help patients and their care team work out individual decision-making needs and plan next steps. An option grid compares the usual treatment or screening options, with patients' frequently asked questions in the rows, the options in the columns and the answers under each option, and health care professionals can use it during discussions with patients about treatment planning. A patient decision aid differs from usual patient education materials: its options and facts are tailored to the patient's own stage of disease, it presents each option's advantages and disadvantages, and patients consider which outcomes matter most to them; some aids also give the likelihood of complications, side effects or other consequences, and an aid may address a single health decision or decisions about a condition generally. Decision support tools complement the discussions between patients and practitioners and do not replace them.
What an aid shows a patient
A Mayo Clinic tool for people in a high-risk group weighing whether to take a statin shows what such a tool puts in front of a patient. It asks what the patient's risk of a heart attack in the next ten years is and what the downsides of taking statins are. The risk is shown for 100 people like the patient, with and without a statin, those who have a heart attack, those who avoid one and those who do not have one distinguished by colour. The downsides are listed plainly: a statin is taken every day for a long time, perhaps for ever; it costs money, to the patient or the drug plan; common side effects are nausea, diarrhoea and constipation, which most patients tolerate; muscle aching or stiffness affects 5 in 100 patients and a rise in a liver blood test, with no pain and no permanent liver damage, 2 in 100, and some patients need to stop statins because of either; and muscle and kidney damage affects 1 in 20,000 and requires stopping. The choices offered are to take or continue a statin, not to take it or to stop, or to decide at some other time.
Without a statin, 50 of the 100 have a heart attack and 50 do not; with one, 12 avoid a heart attack, 38 still have one and 50 still do not, and 88 of the 100 experience no benefit from taking it. Shared decision-making tools frequently use visual displays of quantitative information like this to help people interpret the benefits and risks of a course of action.
Decisions that call for an aid
Shared decision making rests on two complementary kinds of expertise: patients know their bodies, their health and what is important to them, and the health care team knows the latest treatments and options that could benefit them. Clinician and patient bring different but equally important expertise — the clinician's knowledge of the diagnosis, likely prognosis, treatment and support options and the range of possible outcomes, and the patient's knowledge of the condition's impact on their daily life and their own attitude to risk — and good-quality decisions draw on both. Practitioners, who may include physicians, nurses and social workers, provide information about the likely reasonable options, benefits and risks, patients bring what is most important and practical for them, and the goal is a decision based on the most up-to-date scientific evidence and in harmony with the patient's values. Defined as a conversation between two experts — the care provider as the expert on medicine and the patient as the expert on his or her own values, fears and lifestyle realities — shared decision making uses the best external evidence, the provider's clinical experience and the patient's knowledge of social influences, finances and the support available at home to identify the most suitable treatment options, and a decision aid facilitates that conversation.
In shared decision making, clinicians — doctors, nurses, therapists and other health professionals — and patients work together to select tests, treatments, prevention strategies or support packages based on clinical evidence and the patient's informed preferences. It involves asking patients about their experiences, listening actively, informing them about all feasible options, eliciting their preferences and jointly agreeing a plan of action, so that patients can understand, think about and weigh the likely benefits and harms of the different options. It draws together evidence-based medicine and person-centred care, two of the major goals of modern health care, into a pinnacle of excellence that many aspire to. A medical condition can usually be treated in more than one way, yet too often doctors tell patients what they have decided to do instead of laying out the options and asking which the patient would prefer. Clinicians trained to believe they are the only expert in the room respond this way, and so ignore the patient's knowledge of their own situation, their experiences and their values. The result is poorer-quality decisions, less adherence to recommendations and more unwanted, inappropriate care.
Some decisions are easy because the evidence is one-sided, such as taking iron replacement on dialysis, whose benefits clearly far outweigh its side effects and long-term complications, but many have no clear right or wrong choice, such as home against in-centre dialysis. A person with end-stage kidney disease may have as many as five options — a transplant from a living or deceased donor, no dialysis with care to manage symptoms until natural death, peritoneal dialysis at home, haemodialysis at home, or haemodialysis in a facility — and one decision often leads to others: most patients want a transplant, the treatment of choice because it prolongs survival, improves quality of life and costs less than dialysis, and must then choose between the deceased-donor waiting list and a living donation, which some prefer not to ask of relatives while others think a kidney from a known relative more important. Personal views about the importance of the options' features lead people to different decisions, none of them wrong. Shared decision making is widely recognised as relevant to major treatment decisions with more than one feasible option, screening tests and preventive strategies, diagnostic decisions, maternity care choices, goal setting and action planning in long-term conditions, advance care planning for mental health problems, and end-of-life care. A person newly diagnosed with a chronic disease such as end-stage renal disease, breast cancer or obstructive sleep apnoea may, under the stress of the diagnosis, have trouble taking in new information and choosing among numerous treatment options within a limited time.
The uncertainty such a choice produces is called decisional conflict. It can come from not having enough facts or knowledge, being unsure what to expect, being unclear what is most important to avoid or achieve among the possible consequences, lacking advice or resources, or feeling pressured by others, and decisions about newer treatments are often difficult because the evidence on their effects may be limited or inconclusive. In kidney care decisions like these, decision tools can help prepare patients for discussions with their health care teams. A patient with kidney disease does well to decide early because each option needs preparation: if the disease progresses faster than expected she may face an urgent choice while more acutely ill and anxious, and a delayed decision can mean starting a type of dialysis that would not be her first choice and fits neither her values nor her lifestyle. Early preparation can lower the risks associated with urgent hospitalisation or an unplanned start to dialysis.
In the 20 years to 2021, many health authorities and health care organisations worldwide encouraged care more centred on patients and their families — care that is respectful of and responsive to individual patient preferences, needs and values, with patient values guiding all clinical decisions. Patients' participation in decisions can be slowed by barriers: they need to know about their condition and the options and their outcomes, to know their own values and preferences, and to believe they can influence the decision — that they have permission to take part, are confident in their knowledge and are effective at applying their decision-making skills. Patient decision aids are designed to help patients overcome these barriers and take a more active part, by increasing knowledge of the condition, options and outcomes, clarifying patients' values and providing a structured approach to making decisions.
What decision aids achieve
Research has demonstrated that patient decision aids are effective, and high-quality evidence indicates that they are effective interventions without associated harm. A Cochrane review of 105 randomised controlled trials found that patients exposed to decision aids had more knowledge of their options, more realistic expectations, less decisional conflict related to feeling uninformed or unclear about their values, and more involvement in the decision than patients receiving standard care. As of 2021, a recent Cochrane review had shown that decision aids, patient decision aids and option grids can make people more knowledgeable, better informed and clearer about their values, and in all likelihood more active in decision making and more accurate in perceiving risk. Clinical trials had shown decision aids increasing knowledge and the accuracy of risk perception, improving communication between patient and clinician, and reducing decisional conflict, feeling uninformed, passivity in decision making and indecision about the choice made. Meta-analytic reviews confirm that patients whose decisional conflict is assessed by professionals trained in decision coaching reduce their conflict and are more likely to adhere to, and be satisfied with, their choices. Most patients find well-designed information of this kind helpful, including those from disadvantaged groups, who are especially likely to benefit.
Decision aids also reduced the number of patients choosing major elective invasive surgery over more conservative options. As of 2021, reliable, evidence-based information on options, benefits, harms and uncertainties that informs discussions between doctors and patients had been shown to produce more realistic expectations and greater congruence between patients' values and their choices; patients who better understand the trade-offs between benefits and harms often opt for less aggressive interventions, with reductions in elective surgery and in unnecessary antibiotics among the examples, and a large US trial found that supporting patient involvement in treatment decisions led to fewer hospital admissions and fewer elective procedures and lower overall medical costs. A collaborative approach to care planning — learning an individual's concerns, finding out what changes they feel able to make and supporting them in doing so — improves physical and emotional health and self-management in long-term conditions, and patients actively engaged in decisions about their condition feel more responsible and motivated to cope with it, which improves adherence and so treatment outcomes. Decision tools are not panaceas, but as of 2014 they could be valuable if regularly employed at decision points across the health care system, because of their demonstrated ability to improve patients' knowledge of treatment options and hospitals' bottom lines. A decision aid will not prevent an emergency visit for a perforated appendix, and how many preventable visits and costs aids used at each decision point for conditions such as diabetes or cardiovascular disease might avoid remained unknown, though the question was receiving increased scrutiny.
Why health systems want them
Patients and clinicians deciding together on the best course for that patient is not a new concept, and it is what anyone hopes for from a consultation. There are circumstances in which people would rather others acted in their best interests using the best available evidence, but in most interactions there is room to weigh the options and for patient and clinician to discuss the best one as equals. Shared decision making has been endorsed by doctors since 1959. In 1982 a US Presidential Commission on the ethics of medical decision making recommended it as the appropriate ideal for patient–professional relationships that a sound doctrine of informed consent should support, seeing it as a way to reform doctor–patient communication and make informed consent more meaningful. It has since, slowly, gained international recognition as a hallmark of good clinical practice, an ethical imperative and a way of enhancing patient engagement and activation. In 2014 the process had not yet reached a critical mass, though the transition from volume-based to value-based care was creating a business case for it as well as an ethical one.
Policy makers value shared decision making for its potential to reduce overuse of options not clearly associated with benefit for all, such as prostate cancer screening; to increase use of options clearly beneficial for the vast majority, such as managing cardiovascular risk factors; to reduce unwarranted variation in practice; to foster the sustainability of the health care system; and to promote patients' right to be involved in decisions about their health. Patients and clinicians typically overestimate the benefits of interventions and underestimate their harms, and shared decision making offers a way to resolve that mismatch and may reduce the inappropriate use of tests and treatments, giving it a role in reducing overdiagnosis and overtreatment. As the number of possible clinical options increases, so does the complexity of the decision.
In the US in 2014, high and increasing health care costs were arguably the single biggest threat to the long-term fiscal solvency of federal and state governments. A 2013 study found acute care spending on Medicare patients in 2009 and 2010 highly concentrated, with 10% of the Medicare population accounting for more than half of the programme's costs. For hospitals taking part in programmes under the Affordable Care Act, with its federal mandates, pay-for-performance incentives and public reporting of survey data on patients' perceptions of care, organisational success depended on patients' perceptions of and satisfaction with their care, and in 2013 the Centers for Medicare and Medicaid Services allotted $964 million in value-based payments for acute inpatient care for eligible hospitals, with payments set to rise.
Evidence-based medicine, and its application in health technology assessment as NICE practises it, produces recommendations on standardised care for broad populations; individual decision making, and the science of how people decide about their care, have not routinely been part of the approach. Processes for bringing treatments into health systems have typically stopped short of the science of decision making and of the tension between recommendations meant to realise population-level benefit and the individual's choices. People have different attitudes to risk and may make different choices from the same options, so they need access to standardised information about their options to reach a decision reflecting their own values and preferences and to discuss it with their clinicians. Tools such as decision aids, patient decision aids and option grids also support clinicians by giving them easy access to standardised information to share with patients. NICE's initiatives, as of 2021, included developing patient decision aids for a number of topics and identifying preference-sensitive decisions while developing recommendations — decisions that depend on a lack or uncertainty of evidence or, most importantly, on the individual's circumstances, experiences, values and preferences; guideline recommendations on good practice in shared decision making; an online learning package for clinicians developed with Keele University; a quality framework for users and developers of decision aids, commissioned from NICE by NHS England; and, in its five-year strategy, developing mechanisms to make shared decision making an integral part of its methods and processes. In this way NICE aims to help resolve the tension between population-level, evidence-based recommendations and individualised decision making. One author at NICE, writing in a personal capacity, hoped its work in this field would add to the tools supporting a change in culture in which shared decision making is part not only of routine care but also of developing evidence-based guidance and assessment recommendations.
Health systems support the tools directly: the UK's guidance on patient experience in adult NHS services recommends high-quality decision aids, the US Centers for Medicare and Medicaid Services reimburses a first lung cancer screening by low-dose computed tomography when a decision aid is used, and several countries have health policies recommending aids; the next priority is making their use part of routine clinical practice.
Before the COVID-19 pandemic the requirement to inform and involve patients was moving up policy agendas in many countries, and case studies from several countries underscored the importance of effective leadership, appropriate infrastructure, training and practical demonstrations. The directive public health response, rational and probably unavoidable, was a major setback for advocates of patient and public involvement, especially where autocratic decision making was replicated in the clinic or at the bedside, and shared decision making was conspicuously absent during the pandemic. Care for non-COVID conditions halted, delayed or moved online left a backlog that will take years to work through, and the pandemic exposed health inequalities; dealing with that unmet need will require a public consensus on priorities, eliminating unnecessary treatment and effective support for self-care. Embracing shared decision making will be more important than ever as health care organisations face that backlog. Much published health information, from leaflets to newspapers, has tended to present a biased, uncritical view of the benefits of medical care, leading patients and professionals to overestimate benefits and underplay harms and fuelling demand for unnecessary treatments, for interventions that put longevity before quality of life, and for screening that promises early diagnosis of conditions that cannot be cured or may not need treatment. After the pandemic, the aim should be to correct this imbalance so that a better-informed public is more critical of false promises and less tolerant of clinicians who fail to involve them or ignore their views, and redoubled efforts to share decisions between clinicians and patients should be central to building more resilient health services — the path to enhanced self-care and less dependence, fewer inappropriate interventions, more effective prevention, better targeting of resources on those with the greatest needs, and better health outcomes.
Why uptake has been slow
Involving patients in decisions attracts wide support in theory, but making it happen in practice has proven difficult, and despite research demonstrating the tools' effectiveness, their uptake in the clinic has been slow, and the uptake of shared decision making into mainstream care has been slow and highly variable. Most people want to be involved in decisions about their care, but clinicians are often unwilling or unable to cede control: doctors commonly do most of the talking, some believe they practise shared decision making when they do not, and some think patients do not want it when they do. Some patients used to a more paternalistic style may need preparation and encouragement, but the process should be shared rather than a delegation of responsibility to the patient. Many clinicians believe that informing patients about options, asking their preferences and deciding together takes too long for a standard consultation, but the evidence shows it need not be burdensome if well supported. As of 2021 shared decision making was still not embedded in routine clinical practice, and the reasons suggested for this include assumptions that patients are not interested in making decisions, that there are no tools to support it, and not knowing how to measure it.
The international standards
According to the International Patient Decision Aid Standards (IPDAS) Collaboration, a decision aid at a minimum needs to explicitly state the decision that needs to be considered; provide evidence-based information about the health condition, the options, their benefits, harms and probabilities, and the scientific uncertainties; and help patients recognise that their personal values are an important part of the process, which can be used to help guide their decision.
Decision aids can increase uptake of particular options, which is good when it follows from patients' understanding or from acknowledged values but not when it comes from bias, and concern about potential bias in aids intended to increase uptake of specific options led an international collaboration of researchers, patients, health care professionals and policy makers to develop the standards in 2003, as a set of criteria for the quality of the content, development, implementation and evaluation of patient decision aids. In 2005 the Collaboration — a panel of more than 100 researchers, practitioners, patients and policy makers from 14 countries — drew up a checklist for assessing the content, development process and effectiveness of print and digital tools, recommending that a tool state its funding source and the date it was created and present probabilities, information and options in a balanced way. The quality checklist includes plain language at an 8th-grade reading level or lower; visual graphics, with more than one way to view probabilities — words, numbers and diagrams; equal presentation of positive and negative features in fonts, order and display of statistics; discussion of chance, of uncertain probabilities, of probabilities as they relate to age, and of the context of other events; equal use of positive and negative frames, showing survival and death rates alike; suggestions for ways patients can share what matters most to them with others; and evidence from studies of patients similar to the target audience. Web-based tools have criteria of their own, such as step-by-step guidance on moving through the site and the ability to print the tool as a single document.
The evidence behind the standards was updated in 2013 and the original 74 criteria revised into a minimum set of 44: six qualifying criteria defining what a patient decision aid is, ten criteria to minimise the risk of bias, and 28 quality criteria whose omission would not present a high risk of harmful bias. The criteria were adopted by the Washington State Health Care Authority to certify aids and by the Norwegian Department of Health to approve them.
To qualify, the aid describes the health condition or problem for which the index decision is needed, states that decision explicitly, describes the options, describes each option's positive features and its negative ones, and describes what experiencing the consequences is like or includes an explicit values-clarification exercise. To minimise bias it shows positive and negative features in equal detail (similar fonts, order and presentation of statistics), cites the evidence used, gives a production or publication date and an update policy, gives the levels of uncertainty around outcome probabilities, and states the funding source; an aid for screening or diagnostic testing also says what the test measures, the usual next steps when the condition is or is not detected, and the consequences of detecting a condition that would never have caused problems. The quality criteria cover matters such as the natural course of the condition without action, comparing positive and negative features, comparing outcome probabilities over the same time period and denominator, more than one way of viewing probabilities, asking patients which features matter most to them, a step-by-step way to make the decision, worksheets or question lists for discussing options with a practitioner, a needs assessment and review with patients and professionals, field testing, how the evidence was selected and its quality, developers' credentials, readability levels, evidence that the aid improves patients' knowledge and the match between their informed preferences and the option chosen, and for screening the chances of true and false results.
Despite the standards, decision aids are not equally reliable. Their uncertain quality, including whether they suit low-literacy populations, is one of a few barriers to their widespread uptake. One research group found no systematic examination of how health literacy affects outcomes relevant to decision aid development, or of interventions that might mitigate the potential adverse effects of low health literacy in decision making, although in many developed countries only half the population has more than basic reading skills.
Hundreds of tools were available on the internet in 2014, and sites offering them included the Ottawa Hospital Research Institute, the US Agency for Healthcare Research and Quality, Med-Decs and the Mayo Clinic. The largest is the Ottawa Hospital Research Institute's A to Z Inventory of Patient Decision Aids, an international repository of publicly available aids, every one of which has been rated against the standards, and a place to view examples of decision aids.
Dialysis Choice, an aid designed to meet the standards, was developed in Aarhus, evaluated in four Danish hospitals and later implemented in three more, and is freely available in Danish, English and Arabic through the Ottawa inventory. It supports patients with chronic kidney disease choosing a dialysis option in shared decision-making meetings with a dialysis coordinator — a nurse specially trained to tailor the support using mirroring, active listening and value clarifying — and at home with family, through a decision map and an overview of symptoms, to show why a choice is being made and which options are available, and a values-clarification exercise rating the importance of option features on a five-point scale, whose answers the coordinator uses to tailor the meetings to the patient's needs, expectations and values. Rated against the standards it met all qualifying criteria, all but one bias criterion, and most quality criteria; the missing one was references to the evidence, omitted because the patients involved in development asked that the sources not be included. In evaluations it increased involvement, led to choices reflecting patients' values for outcomes, and made home-based treatment more often chosen, with those on home treatment becoming more involved in their treatment and care over time; the aid and the coordinator meetings were the two active mechanisms. One patient said the aid can help a lot when a patient sits there naive and knows nothing, and another, after starting home haemodialysis, that the meetings had left no big surprises.
Developing an aid with patients and professionals
The standards recommend that patients and professionals take part in each stage of development — telling developers what they need to prepare to discuss a decision, reviewing the draft as experts not involved in producing it, and field testing it with patients facing the decision and professionals who counsel them — and more recently they take part as partners on the research team designing it. A survey of 98 researchers who had evaluated 108 aids in randomised trials found co-design with professionals and patients important for making an aid fit clinical practice. According to a 2021 update of the standards, development should be an iterative process in three phases: understanding patients' and professionals' decision-making needs through interviews, surveys, observation and literature review; developing the aid collaboratively, for example in multidisciplinary workshops; and assessing how patients and professionals interact with it and experience it. Patient and public involvement in research can contribute to developing and evaluating aids; patients and professionals were involved throughout Dialysis Choice, and one patient was particularly proud of it and wanted every Danish hospital to use it.
In spring 2020, COVID-19 outbreaks in Canadian nursing homes left many families wondering whether to move older relatives home, a decision with benefits and harms to weigh. Experienced decision aid developers and professionals caring for older adults in Ottawa built an aid in two weeks from the Ottawa Decision Aid Template and Personal Decision Guide, on an umbrella systematic review showing that outcomes did not differ between private homes and nursing homes as long as personal care needs were met, with value statements drawn from public responses to news coverage of the outbreaks. Endorsed by the National Institute on Ageing, it was disseminated through the Ottawa inventory and traditional and social media and downloaded more than 25,000 times, and one user whose wife lives with dementia in a long-term care home found it most helpful. It was built so fast that only professionals were on the development team, and its developers acknowledged the limitation of not involving patients.
An effective decision aid elicits and clarifies the patient's values, fears and questions, is written with health literacy principles in mind, and presents the risks and benefits of each option clearly and equally, including the option of no treatment. A medical writer at Intermountain Healthcare recommends a repeatable process, such as the four steps used there. Ask patients, through advisory councils, focus groups, surveys or one-to-one questions, what they would have wanted to know when deciding. Ask care providers — doctors, nurses, home health workers, physical therapists, social workers — which hard conversations they have with patients, and when and where choices are discussed: one provider's hard conversation was telling patients their kidneys were failing when they might wait at least a month to see the nephrologist, and learning when that conversation happened led to a fact sheet for patients newly diagnosed with end-stage renal disease. Create the aid by arranging what was gathered in a meaningful way, aligned with the decision situation — assessment, diagnosis or treatment — replacing text with visuals, worksheets and self-quizzes where possible, guarding against inadvertent bias by giving each option equal colours, graphics and length of description, and distilling everything to what the reader needs to know right now to decide. Evaluate drafts with focus groups and surveys of patients and providers, and use the International Patient Decision Aid Standards checklist to check that the aid helps patients clarify and express their values and presents information in a balanced way.
Shared decision-making tools can educate, activate and engage patients in their care. One such tool was being developed for patients with atrial fibrillation. Decision analytic models generated patient-specific decision support templates, which were personalised from each patient's calculated risk of stroke and of major bleeding in order to gather the patient's own values for the relevant health states of stroke and gastrointestinal bleeding. The materials were pilot tested with experts in anticoagulation therapy and in health literacy, and feedback from patients, obtained through interviews and focus groups, was used to improve them iteratively.
Typical decision-making tools for patients choosing dialysis focus on the treatment methods and their risks and benefits, whereas the Agency for Healthcare Research and Quality defines shared decision making as also considering the patient's characteristics and values, and a search found no published tool for this population that adequately supported the patient–provider conversation about those concerns. Working with kidney disease experts, one developer reviewed the literature on health literacy in dialysis patients, surveyed patients and providers about information needs, treatment fears and barriers to shared decision making, and turned existing patient education into a pilot decision aid addressing cost, physical appearance, accountability for treatment, aversion to needles, caregiver availability and diet. The developer recommends using it in follow-up appointments with later-stage kidney disease patients to open conversations about possible dialysis decisions; as part of a three-step shared decision-making programme of patient education, self-evaluation of characteristics and values, and collaboration with providers; and to increase decision satisfaction, preference for home-based and lower-cost options, early decisions about access, and adherence.
Bringing aids into a care pathway
Embedding tools that support shared decision making, such as patient decision aids and decision coaching, into care pathways has the potential to promote open and early communication between patients and providers, facilitate compassionate and relevant treatment decisions, and promote high-quality care. In 2012 the Ottawa Hospital and the Ontario Renal Network ran a pilot, Shared End-Stage Renal Patients Decision Making (SHERPA-DM), in the pre-renal insufficiency clinics of the Ottawa and Renfrew Victoria hospitals, built with extensive patient and family engagement to address patients' decisional conflict through three resources: a renal treatment options grid answering patient-identified questions about the options for when kidneys stop working; a patient decision aid on whether to dialyse at home or in a facility, which takes the patient through information resources, health and functional status, values clarification, support needs and next steps; and skill-building workshops training providers to coach patients through decisional conflict. Providers came first — five interactive workgroup sessions with literature review, identification of barriers, problem solving for the local context and simulation exercises — and then every new patient at the clinics attended a decision-coaching visit with a nurse and a social worker who used the tools.
Patients and providers evaluated the tools for acceptability, usability and feasibility of integration into existing care pathways. More than 95% of patients polled would recommend the options grid to a fellow patient and all 17 would endorse the decision aid; patients and providers alike felt better prepared to identify a treatment option after reviewing the grid and agreed unanimously that it described the options clearly and relevantly. Providers reported that the aid increased the likelihood that patients' decisions reflect what matters most to them and endorsed its continued use, and 90% of them (n=21) found that it helped patients take an active part in the decision. Decision coaching with the tools became the standard of practice at the pilot sites and was being rolled out to the Ottawa Hospital's haemodialysis new-start programme, which gives patients beginning haemodialysis intensive education focused on self-management, and to another regional nephrology programme, and the Ontario Renal Network set out to disseminate the grid and the aid across the province's chronic kidney disease programmes. It is generally recognised that implementing such tools without building providers' skills is not effective, and coaching training delivered in person raises doubts about scaling across a province as geographically spread as Ontario, so, with the Canadian Foundation for Healthcare Improvement and patients and families, the team had a distributable training video made on decision coaching and on answering the common questions, fears and anxieties of patients and families in decisional conflict. Video teaching aids have long been used to scale up education, and in a randomised trial of an interactive skills-building workshop on decision coaching about place of care at the end of life, participants rated a scenario-based video the most effective learning aid. The video, coauthored by a patient and a provider, contrasts a decision-coaching approach with traditional patient teaching in two vignettes, interspersed with reflections from a patient and providers who have used both. Multipronged approaches are needed to spread the approach, and by customising implementation to local realities while keeping standardised teaching and carefully monitoring outcomes, population-based coverage is possible, with relief of unresolved decisional conflict and better quality of life for patients with chronic kidney disease as the final outcome.
Getting decision aids into routine clinical practice can be challenging. In the survey of 98 trial authors, 28% reported that their aid was implemented after the trial; barriers included outdated aids with no funding for updates, professionals' reluctance and lack of infrastructure support. A rapid realist review of 23 implementation studies found aids succeeded when content and application were developed with patients and professionals, the whole team was trained, patients were prepared and prompted to take part, management support was ensured, and the quality of decision making was monitored. The strategies for implementation run across four areas. For the intervention, keep aids as simple as possible and in plain language, and establish processes for using them in practice. In the clinical setting, help patients with a strong emotional response to a new diagnosis come to terms with it so they can better process an aid; assess each patient's decision-making needs to identify a suitable aid; help the whole team — senior leadership, administrative staff and professionals — understand the value of aids and their roles in decision processes; make sure aids are given to patients and discussed by patients and professionals alike; provide continuing education on aids and on supporting patients in decision making; and ensure that senior leadership supports and encourages their use. Among individuals, health professionals should be aware, trained and motivated to use aids and understand their intended use; engage them in selecting the aid and establishing how it is used; have them invite and encourage patients to use aids; and be aware of the potential for significant power imbalances between patients and professionals. In the process, embed aids early, when options are first communicated, establish delivery to every eligible patient, and use aids within a learning health system that monitors their outcomes and uses them to inform care and quality improvement.
Reporting an evaluation
Reporting on the characteristics of decision aids is suboptimal: in a review of 17 randomised trials only 59% of authors reported all the qualifying criteria, so readers could not tell whether the intervention was in fact a decision aid, and few trials described the aid well enough to show whether the bias criteria were met. The 2021 update on development does not include an adequate description of the development process, although authors can give details in appendices or other supporting documents. In 2018 the IPDAS collaboration developed SUNDAE, the Standards for Universal reporting of patient Decision Aid Evaluation studies, published in two papers — one on the standards and their development, one elaborating them with examples — and the guideline is in the EQUATOR Network's library, with journals encouraged to have authors follow it and acknowledge its use; some require the checklist as supplementary material. A Google Scholar search in May 2021 found the papers cited 59 and 9 times in their first three years, and some studies support its value in ensuring adequate reporting. Combining the IPDAS standards with SUNDAE lets authors of articles on decision aids report clearly, concisely and understandably.
What a medical writer brings
The growing need for high-quality decision tools across health care settings presents an opportunity for medical writers, and the opportunity for writers to help develop high-quality decision support tools has been inadequately leveraged. Evidence that professional medical writers were helping to develop decision tools was scant in 2014: one study, in a relatively new open-access journal, specifically cited a medical writer's role, while another reference described a multidisciplinary team of designers, investigators, clinicians, patient representatives and other stakeholders. On one medical writer's account, the IPDAS criteria demand the technical, linguistic and rhetorical skills that underpin professional writing and editing curricula across the United States. To meet the criteria, developers face the classical rhetorical questions of invention, arrangement, style and delivery — arguing cause and effect, creating a logical message, using language suited to the subject and the audience, and choosing the mode of communication, such as digital or print. On Marshall McLuhan's "the medium is the message", the message of a decision tool — in print, digital or as a set of memorised prompts — is the change in interpersonal dynamics it brings, because it lets both empirical evidence and the patient's values and preferences into the decision, with effects that reach beyond the encounter to families, practices, hospitals and communities; as health care providers adopt shared decision making, professional medical writers are equipped to develop high-quality decision tools that span disciplines and decision points and deliver important health messages to diverse populations.
Medical writers are often called on to communicate complex topics and specialised information clearly to lay audiences. Their skills in translating complex information such as probability ratios into accessible formats, in patient-centred language and in choosing the most relevant medium could round out decision support teams and contribute substantially to designing and testing decision support tools, and their aptitude for organising scientific data and text into accessible language and for formatting and presenting information can enhance a tool's conception and development. A writer could help tool creators keep the language clear and show the links between factual detail and complex relationships, such as by visualising outcome probabilities and presenting complex information in approachable formats. The SHERPA-DM pilot missed the contribution of a medical writer trained in usability tests, surveys, interviews and focus groups to its evaluation, where a writer, seen as a neutral party, could also have removed an element of bias. For the training video, a writer's expertise in script writing and tailoring information to select audiences could have informed the work, and promoting early involvement of medical writers could strengthen efforts to build sustainable, meaningful decision support tools and connections with decision aid creators, a way to enhance messaging and dissemination.
One senior medical writer holds that medical writers can create effective decision aids and advocate the shared decision-making process in health care. She encourages them to tell providers that shared decision making leads to better health outcomes, including improved quality of care, patient satisfaction and adherence to treatment recommendations, and she holds that it reduces medical errors and that an effective aid can save clinicians time by answering patients' questions up front. Toolkits and curated resources are available from the Dartmouth Center for Shared Decision Making, the Informed Medical Decisions Foundation, the Mayo Clinic Shared Decision Making National Resource Center and the Oregon Rural Practice-Based Research Network.