Values in Care

    Methodology

    How Values in Care keeps values separate from clinical evidence, preparation tools separate from validated decision aids, and worldview context from becoming an assumption about a person.

    What the Decision Guide is built on

    The shared decision-making and decision-aid literature shaped how the guide is structured. Those sources do not validate Values in Care. That would take its own study.

    Why every tool carries a label

    An interactive page is not automatically a decision aid.

    Educational reference

    Explains a topic or evidence base. It does not personalize a recommendation.

    Values clarification tool

    Helps a person surface priorities and context. It does not infer preferences from identity or produce a clinical recommendation.

    Conversation preparation

    Helps organize concerns and questions for a visit, family meeting, or care-team discussion.

    Decision guide

    Structures a choice, priorities, uncertainties, and questions. Generic guides do not substitute for condition-specific evidence.

    Patient decision aid

    Reserved for a specific healthcare choice only when options, benefits, harms, probabilities where appropriate, evidence provenance, values clarification, and review meet the relevant quality standard.

    What a module must meet before it is published

    1. 1.Name the decision or care question, and who it is for.
    2. 2.State each module's scope, exclusions, and urgent situations.
    3. 3.Describe the reasonable options, including waiting or no change when that is clinically appropriate.
    4. 4.Tie benefits, harms, burdens, probabilities, and uncertainties to evidence, never to a generic values template.
    5. 5.Keep clinical facts separate from the person's values, worries, assumptions, and questions.
    6. 6.Offer values prompts without implying one value is better, or that a score picks the right option.
    7. 7.Record the source, the date it was searched or verified, the review date, the reviewer's role once review has happened, and what stays uncertain.
    8. 8.Use plain language, accessible design, keyboard support, and clear risk communication.
    9. 9.When evidence or review details are missing, omit the content or label it. Never fill the gap with plausible language.

    The source has to match the claim

    Different kinds of claim take different kinds of source. They are not ranked on one ladder.

    Clinical facts and outcomes

    We prefer current clinical guidelines, systematic reviews, peer-reviewed research, government or regulatory sources, and established professional bodies. The source has to be as strong as the claim.

    Shared decision-making structure

    Established shared decision-making and decision-aid frameworks shape how the guide is structured.

    Faith and worldview

    Claims about a tradition need qualified sources from that tradition. Interfaith bodies can inform dialogue and shared ethics, but they are not the clinical authority for any one tradition.

    Culture, disability, family, and lived context

    Group-level evidence never becomes an assumption about one person. What the person says comes first.

    Faith, culture, and worldview

    You are asked whether this context matters to you before any is shown.

    • Ask whether faith, spirituality, culture, philosophy, or ethical commitments matter before showing tradition-specific material.
    • Never infer a treatment preference from a religious, cultural, racial, ethnic, disability, gender, or other identity label.
    • Use a tradition's own authoritative or appropriately qualified sources for tradition-specific claims; umbrella interfaith bodies may provide pluralism or shared-ethic context, not clinical rulings for individual traditions.
    • Listing, citing, or discussing an organization does not imply affiliation, endorsement, review, sponsorship, or partnership.
    • Represent meaningful intra-tradition diversity and uncertainty rather than flattening a tradition into a single rule.

    Interfaith bodies, including the Parliament of the World's Religions, are cited only as public sources for pluralism and dialogue; no affiliation, endorsement, sponsorship, or review is implied.

    How a claim gets published

    1. Finding candidates

    Search tools and large language models help find candidate sources and organize the structure. Nothing they find is evidence until a person has checked it.

    2. Checking the source

    Every factual claim is traced to its source. What cannot be substantiated is cut, narrowed, or labeled as interpretation.

    3. Automated checks

    Routes, citation format, identifiers, links, contrast, and exported data are checked by machine. Whether a claim fairly represents its source is not.

    4. Human review

    An editor reads everything before it goes up. Sensitive pages also need review from the relevant expert or community, and the page says whether that has happened yet. Review is never implied before it occurs.

    What a review means

    The review record says exactly what it covered.

    Bound the scope

    Record exactly what was reviewed: a claim set, source set, page, flow, domain, translation, or usability task. Review of one item is never presented as review of the whole product.

    Match expertise to the question

    Clinical accuracy, lived experience, disability access, cultural and worldview representation, decision science, privacy, and usability are different questions. One reviewer is not assumed to cover domains outside their stated role.

    Preserve disagreement

    A reviewer may validate, correct, narrow, challenge, or leave a question open. Real disagreement is recorded, not smoothed into consensus.

    Separate review from endorsement

    Credit or acknowledgement means only that the recorded material was reviewed within the stated scope.

    Credit only with permission

    Reviewer names, organizations, and public credit are opt-in. Anonymous review is still recorded by role and scope, without inventing identity or institutional backing.

    Review status

    The resource as a whole is not yet independently clinician-reviewed. Sensitive modules need review by relevant clinicians, specialists, chaplains, tradition scholars, and patients or caregivers before their claims are treated as mature.

    The review log lists completed reviews only. A citation is not external validation.

    Citations you can download

    The citation dataset is versioned and exported in machine-readable formats. Current dataset version: 1.24.2; last verified: 2026-09-09; 254 sources; license: CC BY-NC 4.0.

    Automated checks confirm a link resolves. They cannot confirm a sentence represents its source fairly. That is a separate human job.

    What we do not do

    • We do not infer a person's preferences from diagnosis, religion, culture, disability, race, ethnicity, gender, family role, or another identity label.
    • We do not label a generic values or conversation tool as a validated patient decision aid merely because it is interactive.
    • We do not invent option probabilities, clinical outcomes, citations, reviewer names, affiliations, partnerships, or endorsements.
    • We do not present a Decision Brief as consent, an advance directive, a clinical order, or a substitute for individualized professional judgment.
    • We do not hide missing evidence by broadening language until a claim becomes impossible to verify.

    Last updated: 2026-10-13. Read the limitations, integrity report, and evidence index for additional detail.

    Cited frameworks belong to their owners. No certification or endorsement by AHRQ, IPDAS, NICE, OHRI, Cochrane, or any other cited body is claimed.