AI Science Tracker
EN

Simplified medical reports improve comprehension in a randomized reader study

University Medical Center Mainz · 26 Sep 2026 · medicine

Discovery

Simplified medical reports improve comprehension in a randomized reader study

Source check: Source checked. Not independent scientific confirmation.

Finding: A Mainz study randomly assigned 576 adult participants to an original or GPT-4-simplified version of one of five German medical documents. Among 439 completed comprehension questionnaires, simplified texts generally produced more correct answers, with the pathology comparison not statistically significant. This measures immediate understanding of selected, researcher-checked texts, not improved treatment adherence or health outcomes. Single-site recruitment, incomplete questionnaires and only one text pair per document type limit generalization.

Technology: GPT-4 text simplification

Family GPT-4: profile · models page

Publication: Peer-reviewed paper · Test context: direct test. Catalog categories, not an evidence grade.

AI contribution
GPT-4 rewrites physician-authored medical reports into lay language; those generated texts form the experimental intervention used to test readers’ immediate comprehension.
Test and result
Randomized assignment to original or pre-generated simplified German medical reports, followed by immediate multiple-choice comprehension questions; 576 participants, with 439 completed comprehension questionnaires.
Limitation / still unknown
Generalization beyond five selected text pairs and a single German clinical department, long-term retention, adherence and clinical outcomes remain unestablished. Education/health-literacy measures were not collected; results include incomplete questionnaires. Output generation settings are described inconsistently as configured parameters and default web sessions; the represented result concerns the actual selected texts.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC13614978/

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 21 Nov 2026. Review scope: Primary model operation, randomized reader-study Methods, results and limitations; exact front-matter identity/date and current PubMed/PMC correction/retraction metadata. Supplementary questionnaire and independent semantic validation were not inspected..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 3/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.

Related catalog entries

Links use documented relationships where available, plus shared catalog context. A shared topic or model family does not establish a scientific dependency.