Truace
Subscribe
70/100

Moderate evidence

Cite
More
Export CSVExport JSONEmailLinkedInX

Medical students like AI but know only moderate amounts, review of 43 studies finds

Study title: Undergraduate medical students and artificial intelligence: A global systematic review and meta-analysis using a knowledge–attitude–practice framework

Evidence type
Systematic review or meta-analysis
Sources
1 sourcePeer-reviewed
Published
2 min read
Truvace record
TRV-2026-2780cite or check it

Study title: Undergraduate medical students and artificial intelligence: A global systematic review and meta-analysis using a knowledge–attitude–practice framework

A systematic review and meta-analysis of 43 studies, 37 cross-sectional and 6 psychometric validation studies, examined how undergraduate medical students engage with AI. Pooling 11 studies that used the MAIRS-MS scale, it found moderate knowledge (0.55), positive attitudes (0.64) and practice levels of 0.63.

Read the study abstract

BACKGROUND: Artificial intelligence (AI), including deep learning and large language models, is increasingly reshaping healthcare and medical education. Since the public release of generative AI tools such as ChatGPT, medical students have become a particularly exposed group, often engaging with these technologies without standardized curricular guidance. Although previous reviews have explored readiness, acceptance, or anxiety, the relationship between students' knowledge, attitudes, and practical engagement wi…

The problem

A systematic review of 43 studies, with 11 pooled, found undergraduate medical students scored only moderate on AI knowledge (0.55; 95% CI 0.49-0.61) even though attitudes were more positive (0.64).

In brief

The authors conclude that medical students are receptive to AI but that their knowledge is only moderate and their practical engagement reflects limited preparation. They say this supports structured undergraduate AI curricula covering theoretical, applied and ethical-legal competencies.

Attitudes were the most studied dimension, followed by knowledge and practice. Heterogeneity across the pooled studies stayed high, so the estimates cover a varied set of student groups and settings.

Main points

  1. The review searched PubMed, Scopus and Web of Science up to September 30, 2025 and followed PRISMA 2020 reporting.
  2. Forty-three studies met the inclusion criteria: 37 cross-sectional and 6 psychometric validation studies.
  3. Only 11 studies using the MAIRS-MS scale with comparable percentage outcomes were pooled, because instruments and reporting varied widely.
  4. Pooled knowledge was 0.55 (95% CI 0.49-0.61), attitudes 0.64 (0.59-0.69) and practice 0.63 (0.59-0.67).
  5. Egger's test showed no significant small-study effects.
  6. Authors call for curricula combining theoretical, applied and ethical-legal AI competencies.

What the review found when it combined the studies

These are the review’s own pooled results, copied from its abstract. Truvace did not recalculate them.

  • Attitudes toward AI among undergraduate medical students (MAIRS-MS scale)

    Proportion 0.64 (95% CI 0.59 to 0.69)

    11 studies · search to September 30, 2025

    The review’s own words
    In the 11 studies included in the quantitative synthesis, pooled estimates showed moderate knowledge (0.55; 95% CI: 0.49-0.61), positive attitudes (0.64; 95% CI: 0.59-0.69), and comparable practice levels (0.63; 95% CI: 0.59-0.67).

  • Practice with AI among undergraduate medical students (MAIRS-MS scale)

    Proportion 0.63 (95% CI 0.59 to 0.67)

    11 studies · search to September 30, 2025

    The review’s own words
    In the 11 studies included in the quantitative synthesis, pooled estimates showed moderate knowledge (0.55; 95% CI: 0.49-0.61), positive attitudes (0.64; 95% CI: 0.59-0.69), and comparable practice levels (0.63; 95% CI: 0.59-0.67).

  • Knowledge of AI among undergraduate medical students (MAIRS-MS scale)

    Proportion 0.55 (95% CI 0.49 to 0.61)

    11 studies · search to September 30, 2025

    The review’s own words
    In the 11 studies included in the quantitative synthesis, pooled estimates showed moderate knowledge (0.55; 95% CI: 0.49-0.61), positive attitudes (0.64; 95% CI: 0.59-0.69), and comparable practice levels (0.63; 95% CI: 0.59-0.67).

Code checks every number above against the review’s text before it is shown. The short labels are an AI model’s reading of that text, so the review’s own sentence sits under each one. A pooled figure is only as strong as the studies behind it. How Truvace uses pooled results

The full story

The review used a Knowledge-Attitude-Practice framework to link what medical students know about AI, how they feel about it and how they use it. It was motivated by the spread of generative tools such as ChatGPT, which students often use without standardized curricular guidance.

Study quality was assessed with the AXIS tool and the research question was structured with the PCC framework. Because instruments and reporting differed substantially, only 11 of the 43 studies entered the quantitative synthesis, using a random-effects DerSimonian-Laird model with Freeman-Tukey double arcsine transformation.

Pooled results put knowledge lowest at 0.55, with attitudes at 0.64 and practice at 0.63. The authors read this as receptive students whose practical engagement reflects limited preparation, and recommend structured undergraduate AI teaching.

Sources

1 on file
  1. Medical Teacherdoi:10.1080/0142159x.2026.2663855
    Peer-reviewed2026-05-31

Talk it over

Four questions to start onWho carries this cost, and what would have to change to prevent it?

Start with a question

Medical students like AI but know only moderate amounts, review of 43 studies findshttps://truvace.com/item/undergraduate-medical-students-and-artificial-intelligence-a-global-systematic-r

  1. Who carries this cost, and what would have to change to prevent it?

    The problem A systematic review of 43 studies, with 11 pooled, found undergraduate medical students scored only moderate on AI knowledge (0.55; 95% CI 0.49-0.61) even though attitudes were more positive (0.64).

    Answer in the thread

  2. How do we know? What can this kind of evidence show, and what can’t it?

    The evidence This rests on a systematic review or meta-analysis, from Medical Teacher.

    Answer in the thread

  3. What does it leave open? What else would you want to know before acting on it?

    The record’s own caveat Only 11 of the 43 studies could be pooled because of differences in instruments and reporting, and heterogeneity among them remained high.

    Answer in the thread

  4. Have you seen AI used this way where you study, work or live? Did it go the same way?

    Answer in the thread

The questions are the same on every record. The details under them are taken from this record as written.

What readers say

More from Health