Embedding a Choice Experiment in an Online Decision Aid or Tool: Scoping Review
Bibliographic record
Abstract
BACKGROUND: Decision aids empower patients to understand how treatment options match their preferences. Choice experiments, a method to clarify values used within decision aids, present patients with hypothetical scenarios to reveal their preferences for treatment characteristics. Given the rise in research embedding choice experiments in decision tools and the emergence of novel developments in embedding methodology, a scoping review is warranted. OBJECTIVE: This scoping review examines how choice experiments are embedded into decision tools and how these tools are evaluated, to identify best practices. METHODS: This scoping review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines. Searches were conducted on MEDLINE, PsycInfo, and Web of Science. The methodology, development and evaluation details of decision aids were extracted and summarized using narrative synthesis. RESULTS: Overall, 33 papers reporting 22 tools were included in the scoping review. These tools were developed for various health conditions, including musculoskeletal (7/22, 32%), oncological (8/22, 36%), and chronic conditions (7/22, 32%). Most decision tools (17/22, 77%) were developed in the United States, with the remaining tools originating in the Netherlands, United Kingdom, Canada, and Australia. The number of publications increased, with 73% (16/22) published since 2015, peaking at 4 publications in 2019. The primary purpose of these tools (20/22, 91%) was to help patients compare or choose treatments. Adaptive conjoint analysis was the most frequently used design type (10/22, 45%), followed by conjoint analysis and discrete choice experiments (DCEs; both 4/22, 18%), modified adaptive conjoint analysis (3/22, 14%), and adaptive best-worst conjoint analysis (1/22, 5%). The number of tasks varied depending on the design (6-12 for DCEs and adaptive conjoint vs 16-20 for conjoint analysis designs). Sawtooth software was commonly used (14/22, 64%) to embed choice tasks. Four proof-of-concept embedding methods were identified: scenario analysis, known preference phenotypes, Bayesian collaborative filtering, and penalized multinomial logit model. After completing the choice tasks patients received tailored information, 73% (16/22) of tools provided attribute importance scores, and 23% (5/22) presented a "best match" treatment ranking. To convey probabilistic attributes, most tools (13/22, 59%) used a combination of approaches, including percentages, natural frequencies, icon arrays, narratives, and videos. The tools were evaluated across diverse study designs (randomized controlled trials, mixed methods, and cohort studies), with sample sizes ranging from 23 to 743 participants. Over 40 different outcomes were included in the evaluations, with the decisional conflict scale being the most frequently used in 6 tools. CONCLUSIONS: This scoping review provides an overview of how choice experiments are embedded into decision tools. It highlights the lack of established best practices for embedding methods, with only 4 proof-of-concept methods identified. Furthermore, the review reveals a lack of consensus on outcome measures, emphasizing the need for standardized outcome selection for future evaluations.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.123 | 0.400 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.008 | 0.010 |
| Bibliometrics | 0.018 | 0.020 |
| Science and technology studies | 0.002 | 0.004 |
| Scholarly communication | 0.009 | 0.011 |
| Open science | 0.004 | 0.005 |
| Research integrity | 0.006 | 0.004 |
| Insufficient payload (model declined to judge) | 0.012 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".