Home›Radiology & Imaging› LLM lay summaries improve patient satisfaction and subjective comprehension without increasing objective comprehension
LLM lay summaries improve patient satisfaction and subjective comprehension without increasing objective comprehensionAI Summaries Improve Patient Satisfaction After Brain MRI Reports
medRxivPublished August 8, 2026Study authors: Le Guellec, B.; Bentegeac, R.; Tran, V.-T.; El Homsi, M.; Amouyel, P.; Kuchcinski, G.; Hamroun, A.DOI ↗Editorial oversight: Dr. Lars van Dijk, PhD · Surgical, Procedural & Diagnostic
AI-generated summary of the cited source, checked by automated accuracy review.
How we work
Share
Key Takeaway
Note that LLM lay summaries improve patient satisfaction and subjective comprehension but do not improve objective comprehension.
This randomized controlled trial enrolled 2,727 adult participants from the ComPaRe e-cohort to evaluate the impact of LLM-generated lay summaries on brain MRI reports for patients with headaches. Participants were randomized to receive either a report in its native format or a report appended with a lay summary generated by an open-weights LLM (Mistral Small 3.2).
Primary outcomes showed no significant difference in overall objective comprehension between the intervention and control groups (59.4% vs 58.3%, OR 0.97, 95% CI: 0.90-1.06, P =.54). However, subjective comprehension was significantly higher in the intervention group (50.3% vs 24.0%, OR 3.17, 95% CI: 2.82-3.56, P <.001), and overall satisfaction also improved significantly (64.9% vs 36.7%, OR 3.26, 95% CI: 2.93-3.64, P <.001).
Secondary outcomes showed a modest reduction in high anxiety in the intervention arm (25.1% vs 26.6%, OR 0.92, P =.037). Notably, while objective comprehension of symptom-explaining reports improved (42.4% vs 37.4%, P <.001), objective comprehension of normal reports actually decreased in the intervention arm (72.5% vs 76.6%, P =.001).
Safety and tolerability data were not reported. A key limitation is the identified gap between perceived and actual understanding. Clinicians should note that while LLM summaries may improve patient experience, they do not necessarily ensure accurate comprehension of medical reports.
How this fits prior evidence
How this fits prior evidence: This finding does not relate to the previously covered information regarding ibuprofen for fever and minor aches in children. It addresses a gap in how patients process complex diagnostic imaging results through AI-assisted communication tools.
Researchers conducted a randomized controlled trial involving 2,727 adults who received brain MRI reports for headaches. The study compared standard medical reports against reports that included an extra summary written in plain language by an AI model.
The results showed that patients who received the AI-generated summaries reported much higher levels of satisfaction and felt they understood their results better than those who only saw the original report. These patients also reported slightly lower levels of high anxiety. However, while patients felt more confident in their understanding, the actual objective comprehension of the reports did not show a significant overall difference between the two groups.
It is important to note that there was a gap between how well patients felt they understood the information and what they actually comprehended objectively. While the AI summaries helped with patient comfort and feelings of clarity, they did not change the core facts of the medical report. Patients should still consult their doctors to discuss the specific details of their results.
What this means for you:
AI-generated summaries can improve patient satisfaction and confidence but do not necessarily improve objective understanding.
Common questions
Does an AI summary help patients understand their results better?
Patients who received AI-generated summaries reported significantly higher subjective comprehension, with 50.3% reporting they understood the report compared to 24.0% in the standard group. However, objective comprehension—the actual accuracy of understanding—did not show a significant difference between the two groups.
Does using AI summaries reduce patient anxiety?
The study found that patients who received the AI-generated summary reported slightly lower levels of high anxiety. Specifically, 25.1% of those with the AI summary reported high anxiety compared to 26.6% in the group receiving standard reports.
How did patient satisfaction change with AI summaries?
Patient satisfaction improved significantly in the group that received the AI-generated lay summary. In that group, 64.9% of participants reported being satisfied, compared to only 36.7% of those who received the standard report format.
Study Details
Study typeRct
Sample sizen = 1,401
EvidenceLevel 2
PublishedAug 2026
View Original Abstract ↓
Background: Large language models have been proposed to improve patient comprehension of radiology reports. However, whether they improve objective understanding remains unproven. Purpose: To evaluate the effect of appending an LLM-generated lay summary to brain MRI reports on objective and subjective patient comprehension in a randomized controlled trial. Materials and Methods: In this randomized controlled trial, 2,727 adult participants from the ComPaRe e-cohort were randomly assigned to interpret six standardized brain MRI reports for headache, presented either in their native format (control; n = 1,401) or appended with a lay summary generated by an open-weights LLM (Mistral Small 3.2) (intervention; n = 1,326). The primary outcome was objective comprehension, defined as the rate of correct classification of whether the report provided a probable explanation for the headache, with ground truth established by four-radiologist consensus. Secondary outcomes included satisfaction, subjective comprehension, anxiety, and willingness to contact a healthcare professional. Generalized estimating equations accounted for repeated within-participant observations. Results: A total of 2,727 participants (mean age, 52 years +/- 15; 75.2% women) were evaluated. Objective comprehension did not differ between arms (58.3% vs 59.4%; odds ratio (OR) 0.97; 95% CI: 0.90-1.06; P = .54). The intervention significantly improved overall satisfaction (64.9% vs 36.7%; OR 3.26; 95% CI: 2.93-3.64; P < .001) and subjective comprehension (50.3% vs 24.0%; OR 3.17; 95% CI: 2.82-3.56; P < .001). High anxiety was modestly reduced (25.1% vs 26.6%; OR 0.92; P = .037). The effect on objective comprehension varied by report type (P for interaction < .001): summaries improved comprehension of symptom-explaining reports (42.4% vs 37.4%; P < .001) but reduced it for normal reports (72.5% vs 76.6%; P = .001). Conclusion: LLM-generated lay summaries appended to brain MRI reports improved patient satisfaction and subjective comprehension but did not improve objective comprehension, indicating a gap between perceived and actual understanding that should be addressed before clinical integration.
