Controlled GenAI virtual-patient and communication studies reported short-term skill gains, but positive findings were not universal¶

Claim · All claims
Evidence · 1 study (2 entries) · review r3 · q2
Currency · newest study 2026 · checked against its sources 2026-10-16 (automated, 2 of 2 entries) · not reviewed by a person

Subclaims¶

q2 i? Several controlled studies reported short-term objective gains for history taking, communication, and OSCE practice with GenAI patients. → Zhao J 2026 q2 i? Human or conventional cases were sometimes rated as more educational or authentic than AI cases, and automated scoring agreed less well for communication than information-gathering domains. → Zhao J 2026 (2)

Evidence¶

Zhao J 2026¶

Zhao J, Yang K, Gao H and Jiang G (2026). Generative artificial intelligence and the transformation of medical education: a scoping review. Front. Educ. 11:1907589. https://doi.org/10.3389/feduc.2026.1907589

q2 · i? · review · r3

The review reports, citing Wang et al. and Borg et al., that controlled simulation studies showed short-term gains; "Several controlled studies reported short-term objective gains" in this 47-report category.

"A GPT history-taking trial reported a large post-test difference and lower anxiety (Wang et al., 2025), while other trials found improvements in gynecologic history taking, communication domains, or OSCE performance (Borg et al., 2026a; Gao et al., 2026b)."

Zhao J 2026 (2)¶

Zhao J, Yang K, Gao H and Jiang G (2026). Generative artificial intelligence and the transformation of medical education: a scoping review. Front. Educ. 11:1907589. https://doi.org/10.3389/feduc.2026.1907589

q2 · i? · review · r3

The review reports countervailing findings in the same category, including that automated total scores approximating faculty ratings showed weaker agreement for communication than information-gathering domains.

"However, positive findings were not universal. Some single-arm implementations measured confidence or usability without objective change. Human or conventional cases were sometimes rated as more educational or authentic than AI cases (Fąferek et al., 2025; Karampinis et al., 2025; Sheth et al., 2025)."

Discussion¶

Learner Variables¶

• Communication Skills — outcome: instruction changes it

Related Claims¶

• Simulation Based Education Improves Outcomes — the general claim this finding bears on, with the studies behind it
• Students treated the genAI avatar as a sophisticated question base rather than a virtual patient, using predominantly closed-ended, efficiency-focused questioning — related
• GenAI students collaboratively phrased and troubleshooted questions, with collaborative problem-solving varying substantially between the two genAI groups — related
• GenAI tutorials took longer than ePBLM tutorials, with approximately 10 additional minutes spent on history-taking — related
• Evidence clustered at reaction and short-term performance levels: only 13 reports contained follow-up signals and none evaluated patient outcomes — related
• Multi-agent AI standardized patient training improved final OSCE-aligned examination scores compared with a structured non-LLM control condition — a narrower finding that bears on this claim
• The largest OSCE-domain improvement from multi-agent scaffolding was in communication performance — a narrower finding that bears on this claim
• The largest checklist-item gain from multi-agent scaffolding was in observable empathic expression, with additional gains in selected history-taking items — a narrower finding that bears on this claim
• Multi-agent scaffolding showed no significant difference from the structured control in binary diagnostic accuracy or worksheet completion — related
• The genAI-enabled patient presented pertinent positive findings consistently with the ePBLM but deviated on non-essential content — related
• Overall 3E, Empower, and Empathize scores were higher in the second AI-patient encounter, but the uncontrolled design precludes attributing this to training — a narrower finding that bears on this claim
• One RCT found AI-assisted simulation inferior to standardized patient simulation for breast self-examination skills — related
• AI-powered simulations show short-term gains in communication confidence and structured knowledge in RCTs, with limited or inferior effects for tactile psychomotor skills — a broader claim this one bears on
• Virtual patient design-variant comparisons rarely show significant differences, with feedback and scaffolding exceptions — related
• AI-driven case simulations in ambulatory settings support diagnostic reasoning and anticipatory thinking in nursing education — related
• GenAI-condition students rated their PBL tutorial significantly higher after the tutorial than they rated their previous ePBLM-based experiences, especially clinical accuracy — related
• Patient history recall was near ceiling in both conditions, with significant decline over two weeks but no difference between genAI and ePBLM conditions — related
• CLT-aligned e-learning modules improved cognitive load profiles and OSCE performance in simulation-based medical training (review-attributed to Gutierrez et al., 2023) — related
• In an evaluative study, medical learners interpreted MedEasy as a connected consultation environment and used its components together to judge case coherence — related