Proficiency of artificial intelligence in addressing patient frequently asked questions on oral lichen planus

Objectives: To investigate the proficiency of artificial intelligence (ChatGPT-4 and DeepSeek-v3) in responding to frequently asked questions (FAQs) related to oral lichen planus (OLP). Study design: Twenty-three OLP-related FAQs were adapted from known sources. Responses generated by ChatGPT-4 and DeepSeek-v3 were assessed by expert panels across 5 domains: accuracy, comprehensibility, logic, fairness, and simplicity. Results: DeepSeek-v3 outperformed ChatGPT-4 in comprehensiveness (4.65 ± 0.37 vs 4.04 ± 0.54;…

By the Truvace deskOctober 16, 20261 min read1 sourcePeer-reviewedTRV-2026-1626· cite & verify

Proficiency of artificial intelligence in addressing patient frequently asked questions on oral lichen planus — Chatbot by James grills. CC BY-SA 4.0 · https://creativecommons.org/licenses/by-sa/4.0

In brief

Researchers took 23 frequently asked patient questions about oral lichen planus and had ChatGPT-4 and DeepSeek-v3 answer them. Expert panels then scored the responses on accuracy, comprehensibility, logic, fairness and simplicity.

The findings suggest that general-purpose chatbots can give patients useful information about a chronic oral condition, and that the models differ in style: one is broader and more detailed, the other more concise. The abstract does not report how patients actually use or understand the answers, so the benefit to patients is a suggestion, not a demonstrated result.

Main points

DeepSeek-v3 beat ChatGPT-4 on comprehensiveness (4.65 vs 4.04), logic (4.41 vs 4.03) and fairness (4.34 vs 4.03), all P < .001.
ChatGPT-4 scored higher on simplicity (4.25 vs 4.04; P = .034).
DeepSeek-v3 had a modest accuracy edge (4.49 vs 4.22; P = .025).
Differences varied by topic, including causes, symptoms, risks and complications, and treatment and management.

The gain

Expert panels rated both chatbots well on 23 oral lichen planus patient questions, with DeepSeek-v3 more comprehensive and logical and ChatGPT-4 simpler, suggesting a combination could aid patient understanding.

The rundown

The study used 23 questions adapted from known sources, and expert panels rated each model's answers across five domains. The two models were compared directly on the same question set.

DeepSeek-v3 did better on comprehensiveness and logic for questions about causes and symptoms, and on accuracy for risks and complications. ChatGPT-4's simplicity advantage was most evident for causes and symptoms.

In treatment and management questions, DeepSeek-v3 excelled in logic and comprehensiveness, while ChatGPT-4 showed greater simplicity and occasional accuracy. The authors conclude that combining both models may improve patient understanding.