ArticleEuropean geriatric medicine2026
Evaluating how different versions of ChatGPT align with expert opinions on geriatric script concordance tests.
Article in European geriatric medicine, 2026. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Not yet cited in PubMed.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
0 citing papers in PubMed.
No citing paper in PubMed yet.
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
10 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
purposeLarge Language Models (LLMs), such as ChatGPT, are increasingly used in medical education, particularly for training clinical reasoning. Script Concordance Tests (SCTs) are valuable tools for assessing reasoning under uncertainty. Our objective was to evaluate the performance of four ChatGPT versions (3.5, 4, 4o, and 5) on SCTs compared to Geriatric Medicine experts.
methodsTwenty validated SCT questions covering diverse geriatric topics were administered to ChatGPT-3.5, 4, 4o, and 5, as well as a panel of 43 experts. SCTs assess responses on a five-point Likert scale, and scores were calculated using aggregate expert answers as the reference. Performance was analyzed based on alignment with expert majority responses.
resultsChatGPT-3.5 achieved a score of 8/20, aligning with the majority of experts on 15% of responses. ChatGPT-4 and -5 demonstrated significantly improved performance, scoring 15.6/20 and 15.2/20, respectively, and aligning with the majority on 45% of responses. ChatGPT 4o scored 13.6/20, slightly underperforming GPT-4, potentially due to differences in its architecture or training data. While GPT-4 and 5 excelled in reasoning and coherence, GPT-3.5 struggled with nuanced clinical scenarios.
conclusionChatGPT-4 and -5 performed comparably to geriatric experts in SCT-based reasoning, highlighting its potential for medical education. It can simulate case discussions, enhance autonomous learning, and support formative assessments. However, limitations such as variability across versions and question-specific challenges emphasize the need for further evaluation of LLMs across broader clinical domains. These findings support the integration of advanced LLMs into medical education to complement traditional teaching methods.
Indexed as
Identifiers
41114883What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.