A browser-based serious simulation game designed to reproduce clinical workflows inside hospital emergency departments has produced substantial immediate improvements in acute care knowledge alongside high usability ratings among practicing medical personnel, according to a pilot study published in Frontiers in Medicine. Conducted at the Consorci Sanitari de Terrassa, a high-complexity university hospital in Spain, the investigation evaluated the novel educational platform known as the GEMAS project, an acronym for Gamificación en EnfermerÃa y Medicina para el Aprendizaje por Simulación. Across a single supervised training session lasting up to three hours, twenty-two healthcare professionals with no prior background in serious games or high-fidelity simulation recorded statistically significant knowledge gains within their respective professional disciplines. Evaluated through the lens of Levels I and II of the Kirkpatrick educational framework, participants demonstrated an overall median System Usability Scale score of 90.0 and reported comprehensive satisfaction across all operational domains, suggesting that narrative-driven virtual simulations can offer an accessible, resource-efficient pedagogical complement to conventional hospital training.
The conceptual motivation for the GEMAS platform stems from enduring structural limitations inherent to traditional lecture-based healthcare education, which frequently struggles to sustain learner engagement, promote long-term knowledge retention, and facilitate the effective transfer of complex procedural guidelines into urgent practice. In demanding emergency care environments, newly appointed nursing staff, incoming resident physicians, and junior clinicians are expected to rapidly master intricate, protocol-driven diagnostic and therapeutic pathways while managing severe workloads and acute time pressures. Although high-fidelity physical simulation represents the recognized gold standard for experiential emergency training, its broader adoption is consistently restricted by extensive financial costs, specialized spatial requirements, and limited institutional scalability. Conversely, low-fidelity and virtual simulation modalities, such as structured clinical scenarios and digital serious games, have emerged as viable alternatives capable of reinforcing clinical reasoning and cognitive decision-making without requiring expensive physical facilities or complex manikins, thereby providing equitable, flexible learning opportunities that function independently of geographical or logistical constraints.
Architecturally, GEMAS was developed as a two-dimensional, narrative-driven role-playing simulation using the commercial engine RPG Maker MV version 1.6.3 and exported as an HTML5 web package hosted on a dedicated server. By executing entirely through modern web browsers such as Google Chrome, Safari, Mozilla Firefox, and Microsoft Edge, the software requires no client-side installation and supports multiplatform deployment across desktop computers, tablets, and mobile smartphones. Over a twelve-month development period, the lead author programmed the game environment and mechanics using JavaScript, while visual assets representing clinical spaces, diagnostic tools, and non-player personnel were drafted using generative artificial intelligence via Midjourney version 5.2 and refined through image-editing software including GIMP 2.10. Clinical narratives and procedural branching logic were formulated and iteratively validated through face-to-face consensus meetings with a multidisciplinary panel of ten emergency specialists, comprising five physicians and five nurses with substantial experience managing complex acute pathology.
The simulated clinical environment closely mirrors the layout and operational zones of a tertiary hospital emergency department, integrating dedicated admission and triage stations, waiting halls, pre-duty staff lounges, and functional clinical bays categorized into fast-track, general medical assessment, infectious disease, and resuscitation rooms. Learners navigate the virtual hospital by selecting one of two distinct, non-overlapping professional trajectories designed to reinforce specialized scope-of-practice competencies. In the medical pathway, players assume the role of a first-year emergency resident—represented by character avatars Dr. David Salazar or Dr. Lucy Raiola—tasked with evaluating nineteen distinct patient scenarios that escalate in clinical complexity from localized conditions such as pneumonia, atelectasis, and ventricular extrasystoles to acute crises including hyponatremia, acute myocardial infarction, and severe chronic obstructive pulmonary disease exacerbations. In the nursing pathway, players embody novice nurses Nadia Kaddar or Juan Manuel Cordero, commencing with systematic patient triage using the Manchester triage system before managing critical nursing emergencies including acute pulmonary edema, acute coronary syndromes, and stroke.
To promote deliberate clinical reasoning rather than erratic trial-and-error mechanics, GEMAS incorporates structured gamified rules and consequences directly tied to evidence-based medical guidelines. Player avatars possess a finite pool of 500 health points, with each diagnostic omission, incorrect test selection, or inappropriate pharmacological intervention exacting a 50-point penalty. Accumulating ten clinical errors exhausts the health pool entirely, triggering an automated virtual dismissal that terminates the simulation session. To facilitate active learning during complex scenarios, players can access embedded reference materials, including electrocardiogram interpretation cards, prognostic calculators, and clinical score criteria such as Light criteria for pleural effusion analysis, Charcot triad in cholangitis, and the Pneumonia Severity Index. Additionally, players can invoke a specialized on-call colleague consultation mechanism known as CoR, which supplies contextual, just-in-time pedagogical advice without interrupting the simulation narrative.
The empirical evaluation of GEMAS utilized a single-group pre-post pilot design approved by the local institutional research ethics committee and registered under ClinicalTrials.gov identifier NCT06516250. Recruitment took place between June and August 2024, followed by experimental testing in September and October and analytical processing in November 2024. The final cohort comprised twenty-two active clinicians, including seven physicians representing 31.8 percent of the sample and fifteen nurses accounting for 68.2 percent. The overall participant cohort exhibited a mean age of 31.2 ± 6.8 years, a median age of 29.0 years with an interquartile range of 26.8 to 35.5 years, and a female representation of 59.1 percent. Participants had accumulated a median professional clinical experience of 4.0 years, with an interquartile range of 4.0 to 6.0 years. Sessions were held in the hospital computer laboratory in groups of six to seven participants seated at individual workstations under researcher supervision to prevent peer consultation, with each clinician independently completing a pre-test, the serious game, and an immediate post-test.
Assessments of user satisfaction and perceived usability under Kirkpatrick Level I demonstrated exceptional platform reception among both professional cohorts. On a ten-item Likert questionnaire combining metrics from Kirkpatrick’s model and the Questionnaire for User Interaction Satisfaction across a zero-to-ten scale, overall item means ranged between 9.3 and 9.7, with no statistically significant differences between doctors and nurses. High scores were recorded for clinical relevance, material interest, graphic representation, applicability to daily duties, and utility as a complement to traditional lectures. Usability evaluated through the Spanish-validated System Usability Scale revealed an overall median score of 90.0 with an interquartile range of 81.3 to 93.1. Physicians recorded a median score of 92.5 while nurses recorded 85.0, reflecting no significant disparity between groups according to Mann–Whitney U testing. Crucially, 77.3 percent of all participants awarded GEMAS an absolute score exceeding 78.9, corresponding to a Sauro–Lewis grade A rating and placing the platform within the top fifteenth percentile of usability benchmarks.
Kirkpatrick Level II cognitive outcomes demonstrated pronounced, statistically significant within-group improvements across both professions, as measured by twenty-five-item, expert-reviewed knowledge instruments scored from zero to one hundred points. Among medical practitioners, mean test scores escalated from a baseline of 48.6 ± 13.4 points to 80.9 ± 13.9 points following gameplay, representing a mean gain of 32.3 points with a 95 percent confidence interval of 24.1 to 40.5, a paired t-test value of p < 0.001, and an effect size of Cohen’s d_z = 3.63 with a 95 percent confidence interval of 1.49 to 5.75. Among nursing personnel, mean knowledge scores advanced from 23.7 ± 8.2 points pre-intervention to 43.3 ± 14.9 points post-intervention, producing a mean improvement of 19.6 points with a 95 percent confidence interval of 14.3 to 24.9, p < 0.001, and an effect size of Cohen’s d_z = 2.05 with a 95 percent confidence interval of 1.14 to 2.95. Bivariate correlation analyses revealed no statistically significant association between knowledge gains and participant age or clinical experience, indicating that educational benefits occurred across varying seniority levels.
Granular item-level analyses revealed that the most pronounced educational improvements corresponded to intricate diagnostic criteria and procedural standards where participants exhibited low initial competencies. Physicians registered marked score increases in items evaluating the Alvarado score for acute appendicitis, rising from a pre-test mean of 0.57 to a post-test mean of 4.00, the recognition of lucid intervals in intracranial hemorrhage, advancing from 0.00 to 3.14, and contraindications to prasugrel administration, improving from 1.14 to 3.43 out of four possible points. Nurses achieved substantial increases in mastering maximum electrocardiogram acquisition times in acute chest pain, rising from 0.80 to 4.00, neurological assessment using the National Institutes of Health Stroke Scale, increasing from 0.40 to 2.27, and quick Sequential Organ Failure Assessment scoring in febrile patients, rising from 0.00 to 1.87 points. Conversely, nursing performance on the Manchester triage classification item showed no post-intervention progress, shifting non-significantly from 1.07 to 0.93 points, indicating that complex multi-variable sorting algorithms require more explicit pedagogical feedback and repeated practice than was afforded within a single gameplay encounter.
Despite these encouraging preliminary results, the study authors emphasized several substantial methodological limitations that necessitate caution in interpreting the findings. Because the trial operated as a small, single-center pilot involving twenty-two self-selected volunteers without an external control group, the observed outcomes cannot establish definitive causal superiority over alternative educational methods. Administering identical evaluation tests immediately prior to and after gameplay introduces the possibility of a testing effect, wherein the pre-test prime directed participant attention toward specific in-game concepts. Furthermore, the protocol did not record individual elapsed gameplay duration, lacked a longitudinal follow-up to measure durable knowledge retention over time, and did not evaluate behavioral translation into actual clinical practice under Kirkpatrick Levels III or IV. The absence of a formalized post-game structured debriefing phase and the potential for recruitment bias linked to the lead author’s dual role as software developer and investigator also represent procedural constraints, which the research team intends to address in forthcoming multicenter controlled trials featuring equated testing instruments and larger professional cohorts.
Subject of Research: Evaluation of a browser-based emergency medicine simulation game for clinician training
Article Title: Proof of concept for a virtual serious game for the training of doctors and nurses: design and development of the GEMAS project
Article References: Mohamed Mimun, M., Campayo, M., Padró Vellvé, J., Jiménez Jiménez, S., Gozalbo Nieto, L., Romero Rascón, C., & Jovell-Fernández, E. (2026). Proof of concept for a virtual serious game for the training of doctors and nurses: design and development of the GEMAS project. Frontiers in Medicine, 13, Article 1949716. https://doi.org/10.3389/fmed.2026.1949716
Image Credits: AI Generated
DOI: 10.3389/fmed.2026.1949716
Keywords: emergency medicine education, serious games in healthcare, virtual clinical simulation, Kirkpatrick evaluation model, System Usability Scale, nursing clinical competencies, medical resident training, digital learning interventions, acute care decision-making, RPG Maker healthcare software, clinical protocol education, pre-post pilot study, formative educational feedback, diagnostic reasoning tools, Consorci Sanitari de Terrassa, emergency triage training, healthcare gamification
News Source: Ophelia Keating. (November 18, 2026). Virtual Emergency Simulation Game Yields Significant Short-Term Clinical Knowledge Gains. Scienmag.



