RESEARCH DOSSIER
Patient–Clinician Communication and Shared Decision-Making
Abstract. Patient–clinician communication and shared decision-making concern the interpersonal, cognitive, ethical, and organizational processes through which clinical knowledge becomes an intelligible and personally meaningful care choice. This fundamental research monograph reconstructs the history of patient-centered consultation and collaborative decision models; defines information exchange, epistemic asymmetry, preference-sensitive decisions, trust, listening, risk comprehension, agency, and decision quality; reviews observational consultation research, psychometric instruments, randomized decision-aid trials, and evidence syntheses; and evaluates communication inequalities, chronic illness, rehabilitation, digital care, and artificial decision support. Its integrative critical narrative synthesis finds consistent evidence for improvements in knowledge and several decision-process measures from well-designed decision aids, while long-term health outcomes, the causal role of relational processes, equity effects, and generalizability depend on context and measurement. The analysis identifies mechanistic and methodological priorities for longitudinal, participatory, implementation, and cross-cultural research.
Keywords: patient–clinician communication; shared decision-making; patient-centered care; decision aids; informed consent; health literacy; clinical trust; risk communication; preference elicitation; decisional conflict; OPTION; SDM-Q-9; CollaboRATE; rehabilitation; patient agency; health equity; health communication
Scientific Scope and Questions
Patient–clinician communication is the situated exchange of information, emotion, interpretation, and commitments between a person receiving health care and a clinician responsible for professional judgment. Shared decision-making (SDM) is the deliberative component of this interaction in which available evidence about reasonable options is combined with the person's values and circumstances to guide a choice. The objects are related yet analytically distinct. A respectful consultation may communicate a diagnosis effectively without facing a preference-sensitive choice; a technically well-described choice may leave unspoken an individual's priorities. The resulting research problem concerns how clinical encounters establish shared understanding and transform it into decisions that can be understood, owned, implemented, and revisited.
This monograph asks how the field defined collaboration, what conversational behaviors make a decision meaningfully shared, how knowledge and power shape participation, what instruments adequately measure these processes, and which patient outcomes can be causally attributed to communication interventions. Its unit of analysis extends from conversational turns to longitudinal relationships and care systems. The emphasis remains on actual clinical encounters; general persuasion, independent health-literacy assessment, pharmacological efficacy, and disability-rights adjudication receive treatment only where they directly shape a clinical exchange.
The methodological approach is an integrative critical narrative synthesis of foundational theory, recordings and observations of consultations, instrument-validation studies, randomized controlled trials, meta-analyses, qualitative syntheses, and practice guidance. Sources were selected purposively for conceptual importance, design diversity, and direct relevance; this is neither a prospectively registered systematic review nor an original experiment. Findings from a randomized comparison, observational association, qualitative interpretation, and professional recommendation are therefore identified as different kinds of evidence. The objective is a durable conceptual and empirical account, with claims tied to identifiable scholarly records.
The Historical Emergence of Clinical Partnership
The modern research tradition developed through several overlapping movements: attention to physician–patient conversation in medical sociology, empirical studies of clinical communication, bioethical accounts of consent and autonomy, and the rise of patient-centered health services research. Early investigations asked who speaks, who asks questions, whose concerns become clinically visible, and how perceived authority shapes the visit. In her Stewart, M. A (1995), Moira Stewart gathered evidence connecting communication with health outcomes and helped establish the encounter as an empirically investigable component of health care.
The decisive conceptual formulation by Charles, C., Gafni, A., & Whelan, T (1997)00221-3) described shared decisions as involving at least patient and clinician, reciprocal information exchange, steps toward a common preference, and agreement on the treatment to implement. Their Charles, C., Gafni, A., & Whelan, T (1999)00145-8) made room for movement among decision approaches during a consultation. The historical contribution was to distinguish disclosure, participation, deliberation, and agreement as observable processes, each of which can fail independently.
The subsequent research trajectory incorporated preference-sensitive clinical questions, decision aids, standardized observation, health literacy, implementation science, and patient-defined outcomes. Coulter, A., & Collins, A (2011) connected the clinical ideal with organizational design. Stiggelbout, A. M., Pieterse, A. H., & De Haes, J. C. J. M (2015) reconstructed the related intellectual history and examined difficulties translating the concept into everyday practice. Historical statements about the desirability of partnership must be evaluated separately from evidence about the effectiveness of particular training methods or technologies.
Decision Authority, Consent, and Delegation
Professional expertise and lived experience enter the encounter with different informational authority. Clinicians generally possess specialized knowledge about diagnostic possibilities, clinical evidence, therapeutic mechanisms, and harms. Patients hold first-person knowledge concerning experienced symptoms, daily demands, caregiving obligations, financial constraints, goals, beliefs, and acceptable trade-offs. Effective collaboration makes these knowledge sources mutually available in a form suitable for a choice, rather than pretending their contents are interchangeable.
Informed consent refers to authorization of a proposed intervention under relevant standards for information, capacity, and voluntariness. Shared decision-making is the broader process that supports deliberation, including choice among several reasonable pathways. A signed form does not demonstrate that the person recognized alternatives or appreciated consequences. Conversely, a clinician's ethical and professional responsibility continues throughout shared decision-making, including the duty to identify medically unsound options and communicate the reasons they fall outside a reasonable choice set.
An individual's desire to delegate details can itself be expressed autonomously. Research should distinguish preference for decisional control from opportunities for genuine participation: a person may wish the professional to recommend a treatment yet still expect accurate information and attention to personal concerns. Preferences for involvement may shift with illness severity, anxiety, prior experience, language access, and relational trust. This dynamic understanding draws on Charles, C., Gafni, A., & Whelan, T (1999)00145-8) and subsequent clinical models. The empirical question is whether the chosen distribution of responsibility was itself negotiated freely and knowledgeably.
Models of the Clinical Encounter
One influential practice model translated the shared-decision concept into conversational stages. Elwyn, G., Frosch, D., Thomson, R., Joseph-Williams, N., Llo (2012) proposed choice talk, option talk, and decision talk, connecting recognition of a choice to explanation and preference clarification. Their later Elwyn, G., Durand, M.-A., Song, J., Aarts, J., Barr, P. J (2017) used team talk, option talk, and decision talk, explicitly emphasizing a supportive partnership, deliberation about available approaches, and movement toward a decision. The revised terms help expose the relational work required before a patient can engage productively with probabilistic evidence.
These stages describe functions, not a rigid sequence of sentences. A consultation may revisit option discussion after the person reveals a previously unknown value; a clinician may need to return to team-building after a frightening finding; the decision may be deferred pending family consultation or further testing. The quality of decision-making is therefore better represented as a recursive process with moments of uncertainty, clarification, and reappraisal than as mechanical completion of a checklist.
The Agency for Healthcare Research and Quality (2014) offers another operational approach that explicitly combines comparison of options with discussion of what matters to the patient. National Institute for Health and Care Excellence (2021) similarly emphasizes an agreed agenda, accessible information, understanding checks, family participation when requested, joint planning, and review of the decision. Such guidance demonstrates codification of professional expectations; its existence does not itself establish the magnitude of real-world patient benefit.
Listening, Questions, and the Organization of Talk
Communication quality depends partly on the distribution and sequencing of conversational opportunities. Patients may present a symptom, express a fear, ask about causation, offer a hypothesis, or signal an implicit concern. The clinician's response can invite elaboration, redirect toward diagnostic categories, provide reassurance, or close a topic. Interruption, premature explanation, and unexplored cues may reduce the clinical information available, especially when concerns involve relationships, stigma, disability, or fear of treatment.
Active listening involves orienting to the person's account, recognizing uncertainty, encouraging questions, accurately reflecting meaning, and making space for correction. These behaviors should be assessed in relation to actual conversational content. A clinician who asks multiple formulaic questions but disregards the answers has not achieved the same process as one who integrates the answers into the diagnostic and therapeutic formulation. Likewise, high patient word counts can reflect confusion or distress; quantity of speech is at best an indirect indicator of participation.
The language of a consultation is jointly constructed under asymmetric expertise. A prompt such as asking what the person hopes a treatment will make possible elicits different information from asking only whether an adverse effect is acceptable. Recorded consultations, conversation analysis, and structured observer coding can identify these differences at the level of specific turns. Such methods show what was said, while interviews are often required to understand what participants believed, wanted, or left unspoken. These complementary methods are central to research on the encounter.
Trust, Therapeutic Alliance, and Relational Continuity
Trust in clinical relationships includes expectations of technical competence, honesty, confidentiality, responsiveness, and fidelity to the patient's interests. It can influence whether uncertainty is disclosed, whether recommendations are questioned, and whether information about difficulties following a plan is volunteered. Trust is therefore a plausible mediator between communication and continuing care, although its direction of causation is difficult to determine in cross-sectional surveys: positive outcomes can improve retrospective ratings of the relationship.
Street, R. L., Jr., Makoul, G., Arora, N. K., & Epstein, R (2009) proposed a set of indirect pathways from communication to health, including better understanding, therapeutic relationships, self-management, emotional regulation, and support. Their framework gives researchers a way to distinguish an immediate interpersonal effect from downstream behavior or physiological change. An intervention that improves trust does not automatically establish a clinically significant improvement in disease outcomes; intermediate steps must be tested.
Some randomized decision-aid research also considered relational consequences. A secondary study of the Statin Choice trial reported that a treatment decision aid Nannenga, M. R., Montori, V. M., Weymiller, A. J., Smith, S (2009). The wording matters: this is a context-specific finding rather than proof that every form of patient activation improves trust. Relational continuity introduces additional complexity because trust develops across visits, prior service experiences, and institutions. Studies limited to a single consultation can miss both cumulative confidence and cumulative harm.
Emotion, Vulnerability, and the Experience of Illness
A consultation is often conducted under pain, fatigue, fear, grief, uncertainty, or acute cognitive load. These states affect attention, memory, interpretation of probabilities, and the ability to formulate questions. Emotionally responsive communication may make choices more intelligible by reducing immediate distress and establishing a workable pace. It also provides information about the costs the person attaches to specific outcomes; fear of functional dependence, for example, may shape preference more strongly than the clinician's ranking of physiological risks.
Emotional expression should be treated as clinically meaningful data and as an experience requiring respectful engagement. Blanket reassurance can conceal a disagreement about risk, while detailed factual explanations offered before distress is acknowledged may be poorly retained. Conversely, an excessively reassuring manner can encourage unwarranted certainty. Effective communication combines emotional recognition with accurately calibrated information, permitting feelings and evidence to inform deliberation without making either the sole determinant.
The pathways framework of Street and colleagues (2009) identifies emotional management among several possible mechanisms connecting interaction with health. The evidence base is more developed for patient experience and short-term understanding than for isolating a unique causal pathway from a particular empathic utterance to long-term clinical endpoints. Future mechanisms research should obtain temporally ordered emotion, comprehension, behavior, and clinical measures, ideally alongside recordings rather than relying solely on retrospective satisfaction questionnaires.
Explaining Clinical Evidence Under Uncertainty
An accurate explanation needs a clear question, relevant baseline risk, plausible options, expected benefits and harms, uncertainty, and a connection to the patient's circumstances. Relative risk reductions can make a benefit appear dramatic when baseline risk is low; absolute risks and common denominators usually provide a firmer basis for comparing outcomes. Numerical frequency statements, time horizons, and distinctions between population averages and individual predictions should be consistent across options.
Comprehension includes the ability to explain what each option entails, how outcomes might differ, and which uncertainties remain. It does not require memorizing every numerical estimate. A meaningful decision process often involves repeated comparison between anticipated health outcomes and treatment burden: time, travel, financial cost, monitoring, adverse effects, social obligations, and changes to daily life. Describing only physiological efficacy leaves the decision informationally incomplete.
Guidance from National Institute for Health and Care Excellence (2021) encourages manageable information chunks, teach-back, and discussion of benefits, harms, and the option of no change when appropriate. The clinical validity of a risk representation depends on the source model, population, calibration, and update date. A well-designed visual aid cannot repair a misleading base estimate. Psychological research can contribute by testing whether framing alters choices, perceived understanding, or preference alignment while holding the underlying evidence constant.
Health Literacy and Comprehension Checks
Health literacy has consequences throughout the encounter: it affects a person's capacity to access, interpret, discuss, and act on health information within a particular service environment. Clinicians may overestimate understanding when patients nod, repeat technical terminology, or express satisfaction. Effective communication therefore treats comprehension as a property of the joint exchange rather than a fixed deficiency located solely in the person receiving care.
Teach-back asks a person to explain information or next steps in their own words, providing the clinician with an opportunity to clarify the explanation. The AHRQ SHARE teach-back resource presents this as a practical understanding check, especially alongside decision aids. Teach-back is not a test to catch a patient making a mistake. Its scientific question is whether a method changes comprehension, later recall, actual behavior, or outcomes under a specified implementation and comparison condition.
Accessible information requires attention to language, numeracy, vision, hearing, memory, attention, reading conventions, and digital access. Universal-precautions approaches reduce reliance on crude guesses about who needs simpler explanations. Such adaptations are connected to the neighboring dossier on health literacy and risk communication, while this monograph asks specifically how information becomes negotiated understanding within an interpersonal clinical decision. Measurement must distinguish what was presented, what was comprehended, and what the person could use when making a choice.
The Psychology of Preference Elicitation
Preferences are context-sensitive judgments about which consequences of available options are acceptable, desirable, or burdensome. They may depend on goals that are difficult to express as a stable numerical utility: maintaining independence, caring for a relative, avoiding hospitalization, preserving fertility, minimizing pain, or continuing work. The clinician's task is to help the person examine how such priorities relate to feasible alternatives without presuming that a preference elicited at the beginning of the conversation is complete or fixed.
A stated preference can change after new information, a different risk representation, or reflection about personal constraints. These changes are not necessarily inconsistency. Researchers should distinguish informed revision from framing susceptibility and from acquiescence to professional authority. An instrument that asks only whether the patient was invited to express preferences is less informative than one that also examines whether those preferences meaningfully influenced the plan.
In the Elwyn, G., Durand, M.-A., Song, J., Aarts, J., Barr, P. J (2017), preference elicitation and integration occur within deliberation rather than as an isolated survey question. The Montori, V. M., Gafni, A., & Charles, C (2006) applies the concept to chronic diabetes care, where the feasibility of a treatment is inseparable from repeated everyday activities. A comprehensive account requires measuring the person's understanding, actual opportunities, the elicited priorities, and the treatment plan together.
Decision Aids and Their Mechanisms
Patient decision aids are structured interventions intended to help people make specific treatment or screening decisions. At minimum, they make the decision explicit, present evidence-based options and their benefits and harms, and assist clarification of personal values. They may be delivered before, during, or across consultations as booklets, cards, videos, websites, or interactive tools. Their function is to support deliberation with a clinician, and their design must account for the actual decision and the available evidence.
The Stacey, D., et al (2024) assembled 209 randomized studies involving 107,698 participants across 71 decisions. It found high-certainty evidence that decision aids improve knowledge and accurate risk perceptions and reduce aspects of decisional conflict; the certainty and size of effects varied by outcome. The review's reported average knowledge difference was 11.90 points on a 100-point scale, with a 95% confidence interval of 10.60–13.19. This is a pooled effect on an information outcome, not a guaranteed change in health status.
Mechanisms include making alternatives visible, reducing informational asymmetry, providing comparably presented numerical estimates, prompting values clarification, and supplying a shared reference during conversation. These mechanisms can interact: improved knowledge without an opportunity to speak may leave preferences unexpressed, while supportive talk without accurate option data can produce emotionally satisfactory but poorly informed choices. Decision aids also differ in reliability, accessibility, conflicts of interest, update practices, and fidelity of use. Their favorable average effects should motivate careful design and local evaluation rather than indiscriminate deployment.
Evidence From Randomized Clinical Encounters
A trial directly tests a particular intervention in a particular context, rather than the entire concept of shared decision-making. The Weymiller, A. J., Montori, V. M., Jones, L. A., Gafni, A., G (2007) studied 98 patients with diabetes across specialist consultations, comparing a tailored statin decision aid with an educational pamphlet. Its scientific importance lies in embedding an aid within the actual clinician–patient exchange, making risk communication and conversational process observable.
Mullan, R. J., Montori, V. M., Shah, N. D., Christianson, T (2009) tested a treatment-choice aid for people with type 2 diabetes. Such trials make it possible to distinguish the effects of offering structured options from those of routine education, although outcome measures, encounter setting, professional behavior, and implementation fidelity determine what can be inferred. Better process ratings do not make long-term glycemic control or adherence effects inevitable.
The multicenter pragmatic Hess, E. P., Hollander, J. E., Schaffer, J. T., Kline, J. A (2016) tested shared decision-making for emergency-department patients with low-risk chest pain who faced alternatives concerning observation and further testing. This setting demonstrates that the feasibility of collaborative decisions extends beyond longitudinal primary care, while clinical safety criteria and patient selection are essential to interpreting the result. A trial focused on an eligible low-risk population cannot be generalized to every suspected acute coronary syndrome. Across trials, appropriate interpretation requires knowing the comparator, decision point, outcome timing, implementation, and clinically relevant exclusions.
What the Cochrane Evidence Establishes
The 2024 decision-aid synthesis reports moderate-certainty evidence of improved alignment between informed values and the resulting choice, with risk ratio 1.75 and a 95% confidence interval of 1.44–2.13 across the studies contributing to that outcome. It also finds improvements in several components of the decision process, including reduced feelings of being uninformed, increased accuracy of risk perception, and less passive decisional participation. These findings support the practical value of structured patient decision support.
At the same time, a favorable process profile does not imply that every health outcome will improve. The review found high-certainty evidence of little or no difference in measured decision regret. Its consultation-time findings depended on when an aid was used: preparation before a visit did not demonstrate a clear average duration difference, whereas use during the visit was associated with approximately 1.5 additional minutes across the contributing trials. The appropriate interpretation is contextual, without claiming all shared decisions are costless or that more time is invariably required.
The broad evidence base combines distinct clinical choices, baseline informational conditions, delivery methods, and instruments. Synthesis therefore estimates average contrasts over a heterogeneous set of interventions. Clinicians and organizations should match the decision aid to a real choice, verify that its risk information remains current, provide an opportunity for questions, and assess whether it supports the conversation as intended. Scientific evaluation should report knowledge, values-congruence, involvement, regret, clinical outcomes, and potential harms as different endpoints.
Observational Associations and Their Limits
Studies recording or surveying routine care can capture the ordinary complexity of conversation, including cases rarely represented in tightly controlled trials. They can examine continuity, clinician habits, institutional conditions, and patient characteristics, while distinguishing observer-rated behaviors from patient-experienced participation. Their major methodological difficulty is confounding: people with more resources or favorable prognoses may receive different communication and report different outcomes, and people with worse outcomes may retrospectively judge the encounter more negatively.
The Shay, L. A., & Lafata, J. E (2015) included 39 studies explicitly measuring shared decision-making and patient outcomes. The synthesis found that positive associations depended materially on whose perspective measured decision-making and which outcome was assessed. Patient-reported measures were more frequently associated with positive outcomes than observer or clinician ratings. This divergence cannot simply be dismissed as error: patient experience and demonstrable clinician behaviors are related but distinct constructs.
Correlation between communication and adherence is similarly consequential but requires careful interpretation. Haskard Zolnierek, K. B., & DiMatteo, M. R (2009) conducted a meta-analysis linking physician communication with adherence. The association supports mechanism research while leaving open the contributions of baseline motivation, illness severity, clinician selection, and supportive service conditions. A causal framework should specify temporal ordering and test whether changing communication changes understanding, trust, practical treatment fit, and subsequent behavior.
Patient-Reported Measures and Response Bias
Patient-reported measures describe whether a person felt informed, invited, supported, listened to, and involved. The Kriston, L., Scholl, I., Hölzel, L., Simon, D., Loh, A., & H (2010) was evaluated in a German primary-care sample of 2,351 patients, providing psychometric support for a concise experience-based instrument. Its utility depends on language adaptation, decision specificity, response-scale properties, and the characteristics of the health system in which it is used.
The Barr, P. J., Thompson, R., Walsh, T., Grande, S. W., Ozanne (2014) pursued a short patient-reported assessment of the collaborative process. Brevity supports routine implementation but can increase vulnerability to ceiling effects, social desirability, and conflation of professional kindness with genuine deliberation. Patients who trust a clinician may rate decision quality highly despite not hearing multiple options; those facing difficult outcomes may assign low ratings despite careful process. Both possibilities matter when interpreting feedback.
Questionnaires should specify the encounter and the decision, use appropriately adapted language, and be collected in conditions that minimize pressure to please the treating professional. Cross-national score comparison additionally requires evidence of measurement invariance. Treating different scales as interchangeable erases meaningful distinctions between perceived inclusion, observed option talk, satisfaction, and decisional quality. A good research design evaluates convergent and discriminant validity along with predictive value.
Decisional Conflict, Regret, and Decision Quality
Decisional conflict refers to uncertainty related to a choice, including perceived deficits in information, value clarity, and support. The O'Connor, A. M (1995) provided a structured way to assess these dimensions and has been used extensively in decision-aid research. Some uncertainty remains appropriate when consequences are genuinely uncertain or values are in tension. The goal of clinical communication is therefore informed, supported deliberation rather than an indiscriminate elimination of all doubt.
Decision regret is a later evaluation of an earlier choice, and it can arise because of adverse outcomes even when the original decision process was appropriate. Measures of perceived participation, decision satisfaction, value congruence, knowledge, and regret should not be collapsed into a single quality index. Each has a different causal position: knowledge can inform deliberation; a decision can reflect values; later experiences can reshape satisfaction and regret.
The 2024 Stacey, D., et al (2024) found important improvements in decision-process outcomes without a clear reduction in decision regret. This pattern illustrates why claims that shared decision-making improves every dimension of patient welfare are empirically excessive. The most appropriate evaluative framework reports the endpoint, assessment time, comparison group, and setting, recognizing that a defensible choice may still lead to an undesirable outcome because of uncertainty intrinsic to medicine.
Communication, Adherence, and Self-Management
Treatment adherence describes how medication taking, monitoring, appointments, or behavioral actions correspond to an agreed plan. It is influenced by comprehension, trust, habit, adverse effects, cost, treatment complexity, access, and competing responsibilities. Shared deliberation may improve the fit between a plan and everyday life, thereby creating a plausible pathway to sustained behavior. Yet a person who chooses an alternative plan after understanding evidence should not be classified as simply nonadherent to the clinician's initial preference.
The 2009 communication–adherence meta-analysis supports an association between communication quality and subsequent adherence, while the Street, R. L., Jr., Makoul, G., Arora, N. K., & Epstein, R (2009) makes explicit the intermediary roles of understanding, agency, and therapeutic alliance. For long-term conditions, outcome attribution also requires attention to changing medication access, family support, treatment adjustments, illness progression, and the feasibility of the plan.
High-quality follow-up should document what was agreed, what the person understood, whether barriers arose, and whether the plan was reconsidered. Conversation during one visit initiates a process that may require repeated adjustment over years. An assessment of whether the original recommendation was obeyed is narrower than an assessment of whether mutually chosen care was workable, clinically appropriate, and responsive to the person's circumstances. This distinction is especially important in multimorbidity and rehabilitation.
Chronic Illness, Multimorbidity, and Treatment Burden
Chronic illness exposes the limitations of isolated one-time treatment decisions. Patients may need to weigh small expected benefits against repeated appointments, monitoring, daily medication, fatigue, travel, and family responsibilities. Comorbidity can produce competing recommendations, contradictory self-care demands, and uncertainty about how evidence derived from single-condition trials applies. Care decisions accumulate, and so does the work required to enact them.
Montori, V. M., Gafni, A., & Charles, C (2006) argued for a shared approach responsive to chronic conditions, using diabetes as a model. A treatment that appears optimal under a narrow physiological endpoint may be unrealistic within a person's actual life. Deliberation therefore needs to connect medical evidence with the capacity to implement a plan without disproportionate disruption. This requires repeated assessment rather than a single declaration of preference.
Longitudinal designs should evaluate whether patients experience genuine flexibility in revising goals, whether clinicians recognize cumulative treatment workload, and whether multidisciplinary teams preserve coherent decisions. Meaningful outcomes include practical burden, patient-defined functioning, continuity, preventable crises, and sustainable self-management. Better information at baseline is only one input to these processes. Institutional scheduling, medication coverage, rehabilitation access, and social support can exert independent effects that communication alone cannot overcome.
Rehabilitation Goals and Functional Choice
Rehabilitation decisions often address mobility, communication, daily activities, assistive technology, vocational participation, and the trade-offs between intensity of therapy and ordinary life. The patient may be balancing physical improvement with independence, social identity, care responsibilities, fatigue, and acceptable risk. Collaborative goal setting therefore requires an explicit account of the functions that matter most to the person, the feasibility of alternative intervention plans, and the environmental supports needed for participation.
In this domain, clinician expertise about prognosis and methods is combined with the person's goals and experience of disability. Caregivers, therapists, interpreters, and advocates may be central to the conversation, but their involvement should be negotiated with the person wherever possible. The National Institute for Health and Care Excellence (2021) provides general principles for choosing whether to involve others and documenting what mattered in the decision. The rehabilitation setting adds repeated decisions and distributed professional responsibility.
The single canonical Research Dossier HPBM-16 belongs both to Health Psychology and Behavioral Medicine and to Rehabilitation Psychology because the same clinical decision process informs both areas. General questions of disability rights, supported legal decisions, and service autonomy remain analytically separate from this clinical exchange. Rehabilitation outcome research needs to examine whether actual goals emerge from dialogue and shape therapy, rather than inferring shared decisions from patient attendance or goal forms alone.
Family, Caregivers, and Triadic Deliberation
Many decisions involve people beyond the patient and a single clinician. Family members can help remember information, describe practical constraints, support future implementation, translate complex circumstances into questions, or introduce concerns of their own. They can also unintentionally dominate the encounter or speak for a person who would prefer more direct participation. The research object becomes triadic or multilateral communication, with multiple perspectives on understanding and authority.
A high-quality deliberative process establishes whom the patient wants involved, what support those individuals provide, and where the final decision authority rests under the applicable clinical and legal context. In some circumstances formal substitute or supported decision procedures are needed; the details vary by jurisdiction and decision-making capacity. Such cases require safeguards beyond the conversational technique of SDM.
Studying these encounters requires coding whose preferences are elicited, how conflicting priorities are resolved, and whether the person receiving care remains intelligibly represented in the plan. A single global satisfaction score may obscure tensions between caregiver convenience, clinical prognosis, and patient goals. Future studies should distinguish family assistance from family control and assess longer-term effects on care participation, burden, and relationship quality.
Development, Capacity, and Supported Understanding
Clinical decision-making changes across childhood, adolescence, adulthood, and older age because language, experience, legal authority, cognitive resources, dependency, and family participation vary. The underlying ideal of taking the person's understanding and preferences seriously can be adapted, but valid implementation depends on the individual's developmental and functional capacities and the applicable consent framework. Adolescents may participate meaningfully in discussions even when another person holds formal authorization; older adults may require accessible explanations without being presumed incapable.
Decision-making capacity is task- and situation-specific. A person can have difficulty processing a highly complex choice during acute illness yet participate effectively after pain control, explanation, sensory support, or additional time. Communication accommodations can reveal abilities otherwise hidden by testing conditions. Research must distinguish cognitive limitation from inaccessible communication and temporary overload.
Supported understanding can involve written summaries, visual diagrams, interpreters, trusted supporters, communication devices, and repeated conversations. These methods should be evaluated for actual comprehension and preference expression rather than their mere availability. A clinical partnership is most defensible when the environment is adapted to the person's functional abilities and the limits of any delegation are clear. Complex capacity determinations demand qualified clinical and legal judgment beyond the scope of an encyclopedia monograph.
Culture, Language, and Institutional Inequality
Communication practices occur within cultural conventions about expertise, questioning, family roles, uncertainty, and the expression of disagreement. Language discordance introduces additional risks of incomplete histories, misunderstood risk estimates, and exclusion from decision-making. Access to qualified interpretation, accessible materials, and sufficient consultation time influences the actual opportunity to participate. Cultural responsiveness requires curiosity about the individual rather than assuming preferences from group membership.
Structural inequalities also influence appointment access, professional continuity, diagnostic trajectories, and perceived safety in raising concerns. A person who has repeatedly experienced dismissal may appear passive during a consultation even when they have strong preferences. Patient activation initiatives can overlook these conditions if they treat participation as solely the individual's responsibility. Research should measure institutional behavior and resource constraints alongside individual knowledge and confidence.
A Miller, A. N., Duvuuri, V. N. S., Vishanagra, K., Damarla, A (2024) examined race/ethnicity concordance and clinical communication across 33 studies and found inconsistent associations after accounting for covariates. The finding cautions against reducing communication equity to demographic matching alone. Specific language access, clinician behaviors, organizational practices, and experiences of discrimination remain important potential determinants. Cross-cultural research needs both adapted measurement and attention to meaningful within-group differences.
Patient-Reported Barriers to Participation
A person may understand that a decision is available yet remain reluctant to question a clinician, reveal a conflicting value, or request a different option. Social expectations, fear of being labeled difficult, limited consultation time, uncertainty about the legitimacy of questions, and histories of exclusion can shape behavior. These are relational and institutional phenomena, not simply failures of information.
Joseph-Williams, N., Elwyn, G., & Edwards, A (2014) synthesized patient-reported barriers and facilitators to shared decision-making. Their work helps explain why increasing knowledge alone may be insufficient to secure participation. An effective intervention must sometimes modify the clinician's invitation, the surrounding institutional environment, and the perceived legitimacy of expressing a preference, not merely distribute more patient leaflets.
Qualitative evidence is particularly useful for studying intentions left unspoken and the meanings attached to professional authority. However, thematic accounts do not directly estimate how often a barrier appears in a population or the causal effect of changing it. Mixed-methods designs can connect interview themes with recorded consultations, standardized involvement measures, and intervention contrasts. Such a design makes possible a more robust explanation of why a well-intended patient-support tool succeeds in one setting and fails in another.
Time, Workflows, and Implementation
Short appointments, fragmented records, competing incentives, staff turnover, and rigid templates can constrain an otherwise willing clinician. Communication must compete for time with examination, tests, documentation, safety checks, and coordination. The assumption that collaborative choices necessarily require long consultations is too simple: poor communication can generate repeated misunderstandings, while well-prepared decision support can focus discussion. Effects on duration must be assessed in specific workflows.
The Mauksch, L. B., Dugdale, D. C., Dodson, S., & Epstein, R (2008) examined relationships among communication, clinician–patient relationships, and efficiency. The Stacey, D., et al (2024) also distinguishes use before and during the consultation in its time findings. Together, these sources justify investigating preparation, task sequencing, information design, and documentation practices as implementation variables.
Implementation science should record uptake, reach, fidelity, adaptation, clinician training, patient access, and sustainability. Repeated measurement is essential because a program may initially improve reported intentions but fail to alter routine conversations. Institutional measures should complement clinical endpoints: the availability of aids, interpreter services, and decision documentation does not establish their meaningful use. Organizations need to make effective communication possible under ordinary clinical conditions rather than relying exclusively on individual enthusiasm.
Training Clinicians for Deliberation
Communication skills can be learned and assessed through observation, feedback, simulated encounters, reflective practice, and supervised clinical work. Training should address the ability to explain that a choice exists, present evidence accurately, recognize emotional and social cues, elicit priorities, check comprehension, and document decisions. Practice across different kinds of clinical encounters is necessary because a specialty consultation, chronic-care follow-up, and urgent-care decision have different constraints.
The Berkhof, M., van Rijssen, H. J., Schellart, A. J. M., Anema (2011) examined training strategies for physician communication skills. The outcome of training should not be limited to learner confidence or satisfaction. Direct observation and patient experience provide more informative measures, while durability and transfer to real practice require follow-up studies. Coaching that changes simulated behavior may or may not survive institutional workload.
The Légaré, F., Adekpedjou, R., Stacey, D., Turcotte, S., Krywor (2018) addressed interventions to increase healthcare professionals' use of SDM. Its focus on practice adoption complements trials of patient decision aids. A robust training evaluation should separate knowledge acquisition, observable behavior, patient-experienced partnership, and decision outcomes. Organizations should also test whether training has equitable benefits across language, age, disability, and health-literacy groups.
Clinical Safety and the Limits of Choice
Shared decision-making operates within the boundaries of professionally defensible care and context-specific constraints. The presence of clinically relevant alternatives is essential; an emergent threat requiring immediate action may limit time for reflection, while an elective screening decision may permit multiple conversations. Clinicians remain responsible for explaining uncertainty and identifying options supported by the available evidence. Patient preferences cannot alone establish safety or efficacy of an unvalidated intervention.
Disagreement may be ethically and clinically meaningful. A person may prioritize avoidance of a particular adverse effect or decline an intervention after understanding its expected consequences. The communication task is to ensure clarity, evaluate comprehension and capacity where relevant, consider alternatives, and preserve a respectful continuing relationship. Conflict should be documented without treating every difference in preference as evidence of irrationality.
In high-risk contexts, the quality of a conversational process is inseparable from timely escalation, safeguarding, and appropriate clinical evaluation. This monograph analyzes psychological and methodological issues, and it does not offer individualized diagnosis or treatment selection. The National Institute for Health and Care Excellence (2021) describes general collaborative principles, which must be implemented together with the relevant condition-specific clinical standards and local consent requirements.
Remote Care, Portals, and Digital Decision Support
Telephone, video, patient portals, and asynchronous messaging redistribute the timing and cues of clinical communication. Some people can formulate questions more carefully in writing or review options with family members before a visit. Others experience barriers from poor connectivity, low digital literacy, limited privacy, disability-related access difficulties, or reduced availability of nonverbal cues. Digital delivery therefore changes both opportunities for deliberation and the inequalities through which those opportunities are distributed.
A patient decision aid presented through a portal may improve preparation without generating a meaningful conversation unless the clinician recognizes its use and invites discussion. In-session digital aids can make risk estimates easier to inspect together but also divert attention toward the interface. The AHRQ SHARE resources and National Institute for Health and Care Excellence (2021) accommodate multiple modalities, while evaluation must identify the channel, usability, actual engagement, and quality of subsequent clinical interaction.
Research should distinguish a click, page view, or portal acknowledgment from understanding and agency. Digital trace data can complement survey and observer measures but does not reveal whether the person felt safe disagreeing or correctly interpreted the estimates. Hybrid longitudinal designs can examine how information encountered before a consultation shapes questions raised during it, decisions made afterward, and the feasibility of follow-up care.
Artificial Systems in Clinical Deliberation
Artificial systems can support the preparation and presentation of information, summarize questions, translate content, generate alternative explanations, or identify when preferences and available options have not been documented. These uses introduce a new object of investigation within the Artificial Era: how clinical agency and epistemic authority are organized when patient, clinician, and computational system interact. The relevant empirical constructs include accuracy, uncertainty communication, calibrated trust, user control, source transparency, and the influence of an artificial intermediary on shared decisions.
The safety problem is not resolved by fluent language. A system may omit reasonable options, present unwarranted confidence, depend on outdated evidence, infer preferences the patient has never expressed, or create a false impression that a clinical recommendation is individualized and authoritative. Researchers should examine how outputs are checked against current sources, how errors are corrected, who retains authority to change a decision, and whether the system supports rather than suppresses direct patient–clinician dialogue.
Claims about beneficial health outcomes from artificial clinical decision tools require dedicated comparative trials, usability studies, equity analyses, prospective safety monitoring, and evidence of reliable integration into clinical practice. The mature human communication and SDM literature provides constructs and methods for such work; it does not itself prove that present or future artificial systems satisfy them. The distinctive scientific opportunity is to observe whether artificial participation changes who explains, who understands, who raises alternatives, and whose preferences govern the resulting plan.
The Research Design Problem
Causal evaluation requires a clearly defined intervention, a meaningful comparison, a specified unit of randomization, and outcomes measured at suitable intervals. In communication research, treatment can be a clinician training package, a decision aid, a structured checklist, a patient-coaching intervention, a redesigned workflow, or a combination. Cluster randomization can reduce contamination between clinicians, but analytic methods must account for clustering and unequal implementation across settings.
Observation studies can estimate natural variation in practices; longitudinal cohorts can examine changes in trust and care participation; conversation analysis can reveal the sequential organization of talk; qualitative interviews identify tacit meanings, power relations, and perceived constraints. An integrative research program uses the method suited to the question. A thematic finding about why a person remained silent is evidence about reported experience, while a randomized change in observer scores supports a more direct causal inference about the intervention's effects on observed behavior.
Researchers should preregister confirmatory primary outcomes when appropriate, report attrition and missing data, specify intention-to-treat and fidelity analyses, and account for multiple outcomes. Process evaluations should identify whether a null effect reflects implementation failure, inadequate contrast, or an ineffective mechanism. Evidence about an instrument's internal consistency does not establish its ability to detect meaningful change. Explicitly separating these claims protects the interpretability of a diverse literature.
The Measurement Perspective Problem
Patient, clinician, and observer perspectives often yield different estimates of shared decision-making. A clinician may remember having mentioned options; an observer may verify only a brief passing reference; the patient may still report feeling engaged and respected. These reports concern overlapping but different dimensions. Disagreement can reveal gaps between professional intent, observable action, and experienced participation. It should not be resolved automatically by privileging the most favorable perspective.
The Shay, L. A., & Lafata, J. E (2015) made this problem especially clear: associations between measured SDM and patient outcomes varied sharply by measurement source. This result limits any claim that a high score on one instrument establishes a universal shared-decision construct. Measurement studies need to assess how ratings change with decision complexity, language, visit length, interpersonal trust, patient vulnerability, and cultural expectations.
A defensible evaluation specifies whether its target is clinician behavior, patient experience, information exchange, preference congruence, or later health. Combining instruments can provide triangulation, but simply averaging dissimilar scores can obscure mechanisms. Clinical quality dashboards should therefore identify the construct and time point rather than display a context-free SDM number. Better construct validity is a prerequisite for accurate estimates of intervention effects.
Replication, Heterogeneity, and Evidence Certainty
The field contains multiple research traditions with uneven methods and populations. Randomized decision-aid evidence is comparatively extensive, whereas studies of long-term disease outcomes, nonverbal interaction, cross-cultural measurement, and complex multilateral decisions are less uniform. Pooled effects can hide variation among conditions, delivery modes, baseline education, usual care, and outcome instruments. The evidence must be interpreted at the level supported by study designs.
The Stacey, D., et al (2024) documents both substantial benefits on selected process outcomes and methodological limitations in included trials, including risk-of-bias concerns in some studies. The Légaré, F., Adekpedjou, R., Stacey, D., Turcotte, S., Krywor (2018) examines a different intervention target. Combining their conclusions as if they measured one intervention would conceal the distinction between improving patient decision support and changing clinician practice.
Replication should test whether benefits persist in diverse clinics, resource-constrained services, rehabilitation programs, translated instruments, and digitally mediated encounters. Independent teams, transparent protocols, and accessible materials can help isolate reproducible intervention components. Future evidence synthesis should stratify by the decision, intervention function, measurement perspective, study quality, and population, rather than assume that every positive SDM study evaluates the same phenomenon.
Theoretical Synthesis of Communication Pathways
Patient–clinician communication can be represented as a causal sequence whose nodes are observable but whose links must be established empirically: an accessible invitation to participate; accurate representation of alternatives; patient interpretation and expressed priorities; shared deliberation; a chosen plan; workable implementation; reassessment; and subsequent health or functional outcomes. Relational trust, emotional regulation, professional expertise, material resources, and service continuity influence several of these links simultaneously.
The framework of Street, R. L., Jr., Makoul, G., Arora, N. K., & Epstein, R (2009) supplies a useful account of pathways, while Elwyn, G., Durand, M.-A., Song, J., Aarts, J., Barr, P. J (2017) clarify the conversational organization of deliberation. Hoffmann, T. C., Montori, V. M., & Del Mar, C (2014) connect evidence-based medicine with the integration of patient values. Combining these strands explains why factual accuracy, relational quality, and practical feasibility are jointly relevant without treating them as a single unmeasured entity.
The synthesis also clarifies the limits of what can be claimed. Evidence that an aid increases knowledge does not establish the complete pathway to health; an association between clinician warmth and adherence does not identify the causal contribution of any particular utterance; and a patient's satisfaction does not prove alternatives were understood. Each link requires its own valid measure and, where possible, independent manipulation or prospective observation. The resulting model is an organizing scientific synthesis of established literature, not a newly validated psychological theory.
Unresolved Problems and Research Priorities
The first unresolved problem is construct integration. Scholars still use shared decision-making to denote clinician behavior, patient experience, decisional control, mutual understanding, or agreement. Instruments often capture one dimension, and evidence about their comparability is incomplete. Research should establish explicitly which components are necessary for a particular evaluative purpose and whether measurement properties hold across language, specialty, age, disability, and socioeconomic context.
The second problem is causal mediation. Decision aids reliably change important informational outcomes, but the contribution of trust, emotion, preference clarification, care burden, and continued clinician contact to clinical endpoints remains difficult to isolate. Longitudinal and factorial designs could test specific combinations of information support, relational training, and workflow redesign. Studies should include adverse effects, such as overload, unwarranted reassurance, and unequal accessibility, alongside positive outcomes.
The third problem is implementation at scale. Routine services must serve diverse populations without reducing collaboration to paperwork, patient responsibility, or a digital checkbox. Patient and caregiver participation in intervention design, pragmatic trials, cross-cultural psychometric work, and transparent evaluation of artificial communication tools can strengthen generalization. A more mature field will identify when particular communication mechanisms matter, to whom, under what institutional conditions, and through which observable pathways they affect meaningful outcomes.
Conclusion
Patient–clinician communication and shared decision-making constitute an interconnected research field organized around the transformation of professional evidence and personal priorities into clinically responsible, intelligible, and revisable choices. Foundational models established the conceptual requirements of reciprocal information exchange and deliberation; subsequent research made portions of the process observable through recordings, questionnaires, decision aids, and randomized interventions.
The strongest aggregated intervention evidence supports improvements in patient knowledge, risk understanding, and several dimensions of decision quality. Effects on regret, adherence, use of services, and long-term clinical outcomes require more differentiated interpretation, and a substantial part of the research problem concerns how actual communication processes mediate those effects. Studies of power, cultural context, disability, multimorbidity, and institutional workflows reveal why individual knowledge and willingness alone cannot determine the opportunity for participation.
The scientific objective is a traceable account of what was communicated, understood, valued, chosen, and subsequently possible. Such an account supports more defensible clinical practice and a rigorous basis for evaluating new mediators of communication, including artificial systems. Its continuing progress depends on precise constructs, transparent source attribution, sound comparative design, patient participation in research, and empirical assessment of outcomes that matter over time.
References
- Agency for Healthcare Research and Quality. (2014). *The SHARE Approach: Shared decisionmaking tools and training*. https://www.ahrq.gov/evidencenow/tools/share-approach.html
- Barr, P. J., Thompson, R., Walsh, T., Grande, S. W., Ozanne, E. M., & Elwyn, G. (2014). The psychometric properties of CollaboRATE: A fast and frugal patient-reported measure of the shared decision-making process. *Journal of Medical Internet Research, 16*(1), e2. https://www.jmir.org/2014/1/e2
- Berkhof, M., van Rijssen, H. J., Schellart, A. J. M., Anema, J. R., & van der Beek, A. J. (2011). Effective training strategies for teaching communication skills to physicians: An overview of systematic reviews. *Patient Education and Counseling, 84*(2), 152–162. https://doi.org/10.1016/j.pec.2010.06.010
- Charles, C., Gafni, A., & Whelan, T. (1997). Shared decision-making in the medical encounter: What does it mean? (Or it takes at least two to tango). *Social Science & Medicine, 44*(5), 681–692. )00221-3 https://doi.org/10.1016/S0277-9536(96
- Charles, C., Gafni, A., & Whelan, T. (1999). Decision-making in the physician–patient encounter: Revisiting the shared treatment decision-making model. *Social Science & Medicine, 49*(5), 651–661. )00145-8 https://doi.org/10.1016/S0277-9536(99
- Coulter, A., & Collins, A. (2011). *Making shared decision-making a reality: No decision about me, without me*. The King's Fund. https://www.kingsfund.org.uk/insight-and-analysis/reports/making-shared-decision-making-reality
- Deveugele, M., Derese, A., van den Brink-Muinen, A., Bensing, J., & De Maeseneer, J. (2002). Consultation length in general practice: Cross sectional study in six European countries. *BMJ, 325*, 472. https://doi.org/10.1136/bmj.325.7362.472
- Elwyn, G., Durand, M.-A., Song, J., Aarts, J., Barr, P. J., Berger, Z., et al. (2017). A three-talk model for shared decision making: Multistage consultation process. *BMJ, 359*, j4891. https://doi.org/10.1136/bmj.j4891
- Elwyn, G., Frosch, D., Thomson, R., Joseph-Williams, N., Lloyd, A., Kinnersley, P., Cording, E., Tomson, D., Dodd, C., Rollnick, S., Edwards, A., & Barry, M. (2012). Shared decision making: A model for clinical practice. *Journal of General Internal Medicine, 27*(10), 1361–1367. https://pubmed.ncbi.nlm.nih.gov/22618581/
- Elwyn, G., Hutchings, H., Edwards, A., Rapport, F., Wensing, M., Cheung, W.-Y., & Grol, R. (2005). The OPTION scale: Measuring the extent that clinicians involve patients in decision-making tasks. *Health Expectations, 8*(1), 34–42. https://doi.org/10.1111/j.1369-7625.2004.00311.x
- Elwyn, G., Tsulukidze, M., Edwards, A., Légaré, F., & Newcombe, R. (2013). Using a 'talk' model of shared decision making to propose an observation-based measure: Observer OPTION 5 Item. *Patient Education and Counseling, 93*(2), 265–271. https://doi.org/10.1016/j.pec.2013.08.005
- Epstein, R. M., Fiscella, K., Lesser, C. S., & Stange, K. C. (2010). Why the nation needs a policy push on patient-centered health care. *Health Affairs, 29*(8), 1489–1495. https://doi.org/10.1377/hlthaff.2009.0888
- Haskard Zolnierek, K. B., & DiMatteo, M. R. (2009). Physician communication and patient adherence to treatment: A meta-analysis. *Medical Care, 47*(8), 826–834. https://doi.org/10.1097/MLR.0b013e31819a5acc
- Hess, E. P., Hollander, J. E., Schaffer, J. T., Kline, J. A., Torres, C. A., Diercks, D. B., et al. (2016). Shared decision making in patients with low risk chest pain: Prospective randomized pragmatic trial. *BMJ, 355*, i6165. https://doi.org/10.1136/bmj.i6165
- Hoffmann, T. C., Montori, V. M., & Del Mar, C. (2014). The connection between evidence-based medicine and shared decision making. *JAMA, 312*(13), 1295–1296. https://doi.org/10.1001/jama.2014.10186
- Joseph-Williams, N., Elwyn, G., & Edwards, A. (2014). Knowledge is not power for patients: A systematic review and thematic synthesis of patient-reported barriers and facilitators to shared decision making. *Patient Education and Counseling, 94*(3), 291–309. https://doi.org/10.1016/j.pec.2013.10.031
- Kriston, L., Scholl, I., Hölzel, L., Simon, D., Loh, A., & Härter, M. (2010). The 9-item Shared Decision Making Questionnaire (SDM-Q-9): Development and psychometric properties in a primary care sample. *Patient Education and Counseling, 80*(1), 94–99. https://doi.org/10.1016/j.pec.2009.09.034
- Légaré, F., Adekpedjou, R., Stacey, D., Turcotte, S., Kryworuchko, J., Graham, I. D., et al. (2018). Interventions for increasing the use of shared decision making by healthcare professionals. *Cochrane Database of Systematic Reviews, 2018*(7), CD006732. https://doi.org/10.1002/14651858.CD006732.pub4
- Makoul, G., & Clayman, M. L. (2006). An integrative model of shared decision making in medical encounters. *Patient Education and Counseling, 60*(3), 301–312. https://doi.org/10.1016/j.pec.2005.06.010
- Mauksch, L. B., Dugdale, D. C., Dodson, S., & Epstein, R. (2008). Relationship, communication, and efficiency in the medical encounter: Creating a clinical model from a literature review. *Archives of Internal Medicine, 168*(13), 1387–1395. https://doi.org/10.1001/archinte.168.13.1387
- Miller, A. N., Duvuuri, V. N. S., Vishanagra, K., Damarla, A., Hsiao, D., Todd, A., & Toledo, R. (2024). The relationship of race/ethnicity concordance to physician–patient communication: A mixed-methods systematic review. *Health Communication, 39*(8), 1543–1557. https://doi.org/10.1080/10410236.2023.2223402
- Montori, V. M., Gafni, A., & Charles, C. (2006). A shared treatment decision-making approach between patients with chronic conditions and their clinicians: The case of diabetes. *Health Expectations, 9*(1), 25–36. https://doi.org/10.1111/j.1369-7625.2006.00359.x
- Mullan, R. J., Montori, V. M., Shah, N. D., Christianson, T. J. H., Bryant, S. C., Guyatt, G. H., et al. (2009). The diabetes mellitus medication choice decision aid: A randomized trial. *Archives of Internal Medicine, 169*(17), 1560–1568. https://doi.org/10.1001/archinternmed.2009.293
- Nannenga, M. R., Montori, V. M., Weymiller, A. J., Smith, S. A., Christianson, T. J. H., Bryant, S. C., et al. (2009). A treatment decision aid may increase patient trust in the diabetes specialist: The Statin Choice randomized trial. *Health Expectations, 12*(1), 38–44. https://doi.org/10.1111/j.1369-7625.2008.00521.x
- National Institute for Health and Care Excellence. (2021). *Shared decision making* (NICE guideline NG197). https://www.nice.org.uk/guidance/ng197/chapter/Recommendations
- O'Connor, A. M. (1995). Validation of a decisional conflict scale. *Medical Decision Making, 15*(1), 25–30. https://doi.org/10.1177/0272989X9501500105
- Roter, D. L., Hall, J. A., & Aoki, Y. (2002). Physician gender effects in medical communication: A meta-analytic review. *JAMA, 288*(6), 756–764. https://doi.org/10.1001/jama.288.6.756
- Scholl, I., Zill, J. M., Härter, M., & Dirmaier, J. (2014). An integrative model of patient-centeredness: A systematic review and concept analysis. *PLOS ONE, 9*(9), e107828. https://doi.org/10.1371/journal.pone.0107828
- Shay, L. A., & Lafata, J. E. (2015). Where is the evidence? A systematic review of shared decision making and patient outcomes. *Medical Decision Making, 35*(1), 114–131. https://doi.org/10.1177/0272989X14551638
- Stacey, D., et al. (2024). Decision aids for people facing health treatment or screening decisions. *Cochrane Database of Systematic Reviews, 2024*(1), CD001431. https://doi.org/10.1002/14651858.CD001431.pub6
- Stewart, M. A. (1995). Effective physician–patient communication and health outcomes: A review. *Canadian Medical Association Journal, 152*(9), 1423–1433. https://pubmed.ncbi.nlm.nih.gov/7728691/
- Stiggelbout, A. M., Pieterse, A. H., & De Haes, J. C. J. M. (2015). Shared decision making: Concepts, evidence, and practice. *Patient Education and Counseling, 98*(10), 1172–1179. https://doi.org/10.1016/j.pec.2015.06.022
- Street, R. L., Jr., Makoul, G., Arora, N. K., & Epstein, R. M. (2009). How does communication heal? Pathways linking clinician–patient communication to health outcomes. *Patient Education and Counseling, 74*(3), 295–301. https://doi.org/10.1016/j.pec.2008.11.015
- Weymiller, A. J., Montori, V. M., Jones, L. A., Gafni, A., Guyatt, G. H., Bryant, S. C., et al. (2007). Helping patients with type 2 diabetes mellitus make treatment decisions: Statin Choice randomized trial. *Archives of Internal Medicine, 167*(10), 1076–1082. https://doi.org/10.1001/archinte.167.10.1076