Goal Attainment Scaling Across Languages: A Practical Review Guide

Goal Attainment Scaling, often shortened to GAS, organizes a person’s goals into predefined descriptions of possible outcomes. A common approach uses five attainment levels, with the expected outcome assigned zero. Baseline placement and scoring conventions vary, so teams must follow their selected methodology rather than assume that a familiar-looking scale is implemented identically everywhere. The goal is personalized; the rules for setting and evaluating it need to be explicit.

ISPOR’s October 8, 2026 webinar listing places GAS within personalized clinical outcome assessment and highlights goal elicitation, scaling, rater training and analysis. It also identifies implementation variability as a source of bias and reduced reliability. This article builds on Doraid Matani’s related LinkedIn discussion to examine the multilingual workflow. It is not a report of the webinar proceedings or evidence that a particular GAS implementation has been validated.

Begin by separating the materials that will be common to everyone from the language that may be individualized. Shared content can include interviewer instructions, goal-selection prompts, explanations of the attainment levels and electronic navigation. Individual content may include a participant’s own activity, relevant circumstances and agreed outcome descriptions. These two streams need connected review rules, but they may not follow the same production schedule.

A practical project brief should identify who proposes each goal, who approves the wording, who assigns attainment and which language each person uses. Record the assessment time point, administration mode and intended respondent as well. If a bilingual investigator summarizes a local-language discussion in English, distinguish that research record from the wording actually shown or read to the patient. Otherwise, reviewers may approve a version the participant never encountered.

Consider the illustrative goal of walking to a nearby shop without another person’s help. A translation that turns help into encouragement may allow physical assistance that the original goal excluded. A translation that changes walking to going may permit transport. Both alternatives can sound natural while altering the activity being evaluated. Review the action, setting, assistance and time frame separately before deciding that the whole sentence is equivalent.

Cultural relevance also deserves explicit discussion. A nearby shop may be a meaningful destination for one participant and inaccessible or irrelevant for another. The clinical and methodological team should decide whether a different activity is acceptable within the study’s goal-setting rules. A translator should flag the issue and explain the linguistic implications, not silently substitute a culturally familiar activity. Keep the approved rationale alongside the affected goal version.

Reviewers should see all levels together rather than translate isolated cells. Ask whether neighboring descriptions overlap, leave an unintended gap or change more than one dimension at once. If one level concerns frequency and the next concerns independence, it may be unclear how to classify a participant who improves on only one dimension. Such ambiguity belongs with the methodological team; fluent wording cannot resolve the underlying measurement decision.

For a language-review exercise, imagine descriptions that differ only in how often an agreed activity can be completed within a defined week. The reviewer can check that the period, activity and assistance conditions stay constant as frequency changes. This is an example of a consistency check, not a ready-to-use GAS instrument. The actual levels and their clinical meaning must be developed and approved for the intended assessment.

Where the project methodology includes respondent testing, plan questions that reveal how participants interpret the wording. Asking someone to explain a goal in their own words can uncover a different understanding of effort, help or opportunity. A neutral follow-up can explore what they believe would count as achieving a described level. Avoid supplying the preferred interpretation inside the question and then recording agreement as evidence of understanding.

Separate problems in the translated instructions from problems in an individual goal. A recurring misunderstanding of a shared instruction may affect several participants and languages. A difficulty with one person’s activity may require a narrower review. Document the version tested, the relevant observation, the proposed response and who made the decision. Interview findings support language decisions; they do not, on their own, establish all measurement properties or justify a regulatory claim.

The final review should include the environment in which the assessment will be used. A correctly approved sentence may become confusing when a screen truncates its condition, places two levels far apart or hides the explanation behind an unclear control. Check small-screen display, text expansion, line breaks and the order of response options. Confirm that an exported report preserves the connection between the goal, selected level, visit and language version.

Site guidance should explain how to handle a participant who asks for clarification, proposes a change or describes an outcome that does not fit the available levels. The study team must define the permitted response and escalation route. Consistency does not require every interviewer to sound identical, but unrecorded explanations can introduce different meanings. Include realistic language questions in training so sites know which decisions they can make and which require central review.

For each issue, retain a stable identifier, the source wording, the affected locale, the proposed wording and the reason for the decision. Name the responsible reviewer and record approval dates. Link a changed shared instruction to the goals, interview materials and screens that may depend on it. A short, specific rationale is more useful than a generic statement that a translation was checked and accepted.

Plan how later changes will be handled before recruitment begins. An approved goal revision, a correction to a translation and a change in assessment methodology are different events. Record which event occurred and follow the study’s rules for its consequences. Preserve the previous version so a reviewer can understand what was in use at an earlier visit. Do not replace historical wording merely to make the current file look consistent.

A team may choose to evaluate AI for limited tasks such as flagging inconsistent terminology or preparing a comparison of approved versions. Give the system a defined task and approved inputs, then check its findings against the underlying text. An apparent inconsistency may reflect an intentional distinction between patients, and a polished suggestion may remove a clinically important condition. Neither should be accepted solely because the output sounds confident.

Keep identifiable patient information out of tools that have not been approved for that data. Retain the ability to reject suggestions, leave an issue unresolved and escalate it to a qualified specialist. Workflow technology can organize the record, but the project still needs named people responsible for clinical meaning, language suitability and methodological decisions. Evaluate any proposed tool against those responsibilities before introducing it into the study.

Before requesting linguistic validation support, assemble the selected GAS approach, study population, intended use, language-country pairs and available goal-setting materials. Identify which wording is standardized, which may be individualized and how local-language goals enter the central record. Add the expected review stages, interview plans, electronic platform and decision owners. These details let a language partner propose a scope that matches the actual assessment rather than a simple word count.

Agree in advance what evidence will be delivered and what questions remain outside the language team’s remit. FDA’s final October 2025 guidance provides the broader context of selecting, developing or modifying fit-for-purpose clinical outcome assessments; it should not be treated as blanket endorsement of a specific GAS project. For multilingual work, the practical objective is a traceable account of what each participant understood, which wording was approved and how language questions were resolved.