This randomized controlled trial evaluated whether an AI-assisted, rule-based adaptive screen-based simulation module could improve physiology learning outcomes among undergraduate health science students compared with conventional instruction. A total of 672 students from Physiotherapy, Occupational Therapy, Nursing, and Allied Health Sciences were randomly assigned in a 1:1 ratio to either the adaptive simulation group or the conventional teaching group. The intervention used web-based clinical physiology cases with algorithm-supported case sequencing, automated formative feedback, and structured faculty-led debriefing, while the control group received standard lectures, textbook reading, tutorial sessions, and laboratory practicals. The primary outcomes were physiological knowledge and reasoning ability, and the secondary outcomes were conceptual understanding, engagement, cognitive load, and academic self-efficacy. Assessments were performed at baseline, immediately after the 12-week intervention, and again at four-week follow-up.

This study was designed as a prospective, two-arm, parallel-group randomized controlled trial with repeated-measures assessment at three time points: baseline, immediately post-intervention, and four weeks after the intervention. It was conducted at Saveetha Institute of Basic Medical Sciences, India, between August 2025 and January 2026, and received institutional ethical approval before enrollment. Participants were undergraduate health science students aged 18 to 25 years who were enrolled in a Human Physiology course and had access to an internet-enabled personal device. Students with prior formal exposure to simulation-based physiology instruction or adaptive digital learning platforms were excluded. After baseline assessment, participants were randomized in a 1:1 ratio to the intervention or control group, with allocation concealment and blinded outcome assessment.

The intervention group received physiology instruction through a screen-based adaptive simulation environment over 12 weeks. The module was intentionally designed as a bundled educational strategy integrating adaptive case sequencing, automated formative feedback, and faculty-led debriefing. The adaptive component used predefined rule-based logic to personalize learning by adjusting case difficulty and feedback pathways according to learner performance; it did not use autonomous generative artificial intelligence or clinical decision-making. Participants completed structured simulation sessions for two hours per week, including pre-briefing, individual case-based simulation, and facilitated debriefing. The control group received conventional curriculum-based physiology instruction over the same 12-week period, including didactic lectures, prescribed textbook readings, tutorial sessions, and laboratory practicals.

The study prioritized objective learning outcomes. Physiological knowledge was measured using a 40-item multiple-choice test, physiological reasoning ability using a scenario-based rubric-scored assessment, and conceptual understanding using a physiology concept inventory. Secondary outcomes included student engagement measured with the USEI, cognitive load measured with NASA-TLX, and academic self-efficacy measured with an adapted CASES scale. Outcomes were collected at baseline, post-intervention, and follow-up using the same instruments across all time points.

Inclusion Criteria:

• Undergraduate students enrolled in Health Science programs including Physiotherapy, Occupational Therapy, Nursing, and Allied Health Sciences
• Registered for a Human Physiology course during the study period
• Age between 18 and 25 years
• Proficiency in English language
• Access to an internet-enabled personal device capable of supporting web-based educational applications
• Willingness to provide written informed consent for participation

Exclusion Criteria:

• Prior formal exposure to structured simulation-based physiology instruction
• Prior exposure to adaptive digital learning platforms related to physiology education
• Inability to access or use internet-enabled educational applications required for the intervention
• Declined or withdrew informed consent for participation

Individual participant data (IPD) will not be publicly shared because the dataset contains institution-linked educational performance information and participant-level academic assessment data. De-identified data may be considered for academic collaboration upon reasonable request to the corresponding author, subject to institutional ethical approval and data-sharing regulations.

Outcome assessors and data analysts were blinded to group allocation throughout the study. Allocation concealment was maintained using sequentially numbered, opaque, sealed envelopes prepared by an independent researcher not involved in recruitment or assessment.