{"paper_id":"19c4e115-d68a-46af-b7fd-408cbd75ed01","body_text":"Dynamic CBME in Action: AI-Assisted Digital Case-Based Learning to Improve Antibiotic-Stewardship Decisions in MBBS Learners | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Dynamic CBME in Action: AI-Assisted Digital Case-Based Learning to Improve Antibiotic-Stewardship Decisions in MBBS Learners Sunil Kumar D, Sucheta Dandekar, G Hari Prakash This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9384886/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Background Antimicrobial resistance is a major global public health threat, necessitating improved antibiotic stewardship. The introduction of Competency-Based Medical Education (CBME) in India requires robust methods to assess and cultivate clinical reasoning. This study evaluates an AI-assisted Digital Case-Based Learning (DCBL) module intended to improve antibiotic stewardship decisions among medical undergraduates. Methods A quasi-experimental, single-group pre-post study evaluated the impact of three AI-assisted DCBL micro-cases (identifying upper respiratory tract infection [URTI], urinary tract infection [UTI], and acute watery diarrhoea) on Script Concordance Test (SCT) performance in 271 MBBS learners. Usability and workload were measured using the System Usability Scale (SUS) and NASA Task Load Index. Results Overall SCT scores improved significantly from 5.65 ± 4.73 to 6.47 ± 4.48 (p = 0.04), indicating enhanced clinical reasoning. The UTI case showed the greatest improvement (p = 0.003). Over 76% of learners rated usability as Good/Average, with moderate cognitive effort. Conclusion A single-session AI-assisted DCBL intervention improved SCT-measured clinical reasoning for antibiotic stewardship. Targeted domain-specific design is needed to address heterogeneous intervention effects. Script Concordance Test Antimicrobial Stewardship Medical Education Artificial Intelligence Case-Based Learning Figures Figure 1 Figure 2 Figure 3 Introduction Clinical reasoning, the cognitive process through which clinicians integrate clinical data, formulate diagnoses, and arrive at treatment decisions under conditions of uncertainty, is widely recognised as the cornerstone of medical competence[ 1 , 2 ]. Unlike factual recall, clinical reasoning demands that learners navigate ambiguity, weigh competing hypotheses, and apply context-dependent judgement, skills that are difficult to cultivate through didactic lectures or conventional multiple-choice assessments alone [ 3 ]. This gap has become particularly pressing in the context of antibiotic prescribing, where reasoning errors translate directly into patient harm and population-level antimicrobial resistance (AMR). Antimicrobial resistance is now among the foremost global public health threats. The Indian Council of Medical Research (ICMR) Antimicrobial Resistance Surveillance Network has documented alarming declines in the effectiveness of commonly used antibiotics against gram-negative organisms, with key drugs such as ciprofloxacin and ceftriaxone showing susceptibility rates of less than 20% against Escherichia coli isolates in several outpatient settings [ 4 ]. In India, where 45–80% of patients presenting with upper respiratory tract infections (URTI) and diarrhoea receive antibiotics despite predominantly viral aetiologies, the outpatient setting remains a critical frontier for stewardship[ 5 ]. Urinary tract infections (UTI), another high-volume outpatient condition, carry their own stewardship challenges, with guideline-discordant prescribing documented across multiple countries [ 6 ]. Yet, despite the magnitude of the problem, antibiotic stewardship education in undergraduate medical curricula remains inconsistent and largely didactic [ 7 , 8 ]. The introduction of Competency-Based Medical Education (CBME) by India's National Medical Commission has fundamentally reoriented the goals of undergraduate training toward demonstrable competencies rather than mere content acquisition [ 9 ]. The CBME framework explicitly demands that the Indian Medical Graduate (IMG) function as a clinician capable of first-contact decision-making, a communicator, a leader, a professional, and a lifelong learner [ 10 ]. However, the transition from curriculum intent to classroom reality has been challenging, with faculty reporting difficulties in assessing higher-order cognitive skills such as clinical reasoning [ 11 ]. This underscores an urgent need for educational interventions that go beyond knowledge transmission and actively engage learners in authentic clinical decision-making within the CBME framework. Digital case-based learning (DCBL) has emerged as a promising pedagogical strategy for bridging this gap. Unlike traditional case discussions, digital platforms can embed branching clinical logic, simulate realistic decision pathways, and crucially provide immediate, individualised formative feedback at each decision node [ 12 , 13 ]. When augmented with artificial intelligence (AI), such systems can deliver contextualised rationale aligned to stewardship principles, helping learners understand not merely what the correct answer is, but why a particular clinical decision is appropriate in a given context. Recent evidence suggests that AI-enhanced educational tools can improve clinical reasoning and knowledge retention, particularly when embedded within structured pedagogical frameworks rather than deployed as standalone resources [ 14 , 15 ]. The Script Concordance Test (SCT), originally described by Charlin and colleagues, offers a robust method for assessing clinical reasoning under uncertainty [ 2 , 16 ]. Unlike conventional assessments that test factual knowledge, the SCT presents learners with clinical scenarios that require probabilistic reasoning, evaluating how new information modifies the likelihood of a particular hypothesis or the appropriateness of a management decision. Responses are scored against an expert reference panel using the aggregate-credit method, thus accommodating the legitimate variability inherent in expert clinical judgement [ 17 ]. The SCT has been validated across diverse medical disciplines and training levels, including emergency medicine, paediatrics, neurology, and paramedicine, and has been recommended as both a summative assessment tool and a formative learning instrument[ 18 – 20 ]. Despite growing interest in both digital learning and SCT-based assessment, evidence linking AI-assisted DCBL specifically to improvements in SCT-measured clinical reasoning remains scarce, particularly in the context of antibiotic stewardship education for undergraduate medical learners in low- and middle-income countries (LMICs). Most existing studies on antibiotic stewardship education have focused on knowledge gains measured through conventional tests, with limited attention to higher-order reasoning outcomes or the cognitive demands imposed on learners [ 21 , 22 ]. Furthermore, the usability and cognitive workload of technology-enhanced interventions, critical determinants of sustained adoption, are infrequently reported. This study, therefore, aimed to determine whether a single-session AI-assisted DCBL intervention comprising digital micro-cases on three common outpatient conditions (URTI, UTI, and acute watery diarrhoea) with AI-coach formative feedback could improve overall SCT performance in MBBS learners. Secondary objectives included evaluating case-specific effects across clinical domains, characterising individual-level heterogeneity of response, and assessing platform usability and cognitive workload using the System Usability Scale (SUS) and NASA Task Load Index (NASA-TLX) items. METHODS Study Design and Setting This was a quasi-experimental, single-group pre–post study embedded within routine Competency-Based Medical Education (CBME) teaching at a tertiary-care teaching hospital in South India. The study evaluated the effect of an artificial intelligence (AI)-assisted Digital Case-Based Learning (DCBL) intervention on clinical reasoning in antibiotic stewardship for common outpatient conditions. The study was conducted as an educational intervention study within the existing CBME framework and did not involve patient clinical data. Voluntary participation was ensured, and all data were de-identified before analysis. The study adhered to the ethical principles of the Declaration of Helsinki for educational research. Participants A convenience sample of undergraduate medical (MBBS) learners enrolled in the CBME curriculum was recruited. Eligibility criteria included current enrolment in the MBBS programme and willingness to participate in the complete study session. A total of 271 participants provided complete paired (pre–post) data and constituted the analytic sample for the primary analysis. Participation was voluntary; coded identifiers were assigned to enable data pairing while preserving anonymity. Intervention: AI-Assisted Digital Case-Based Learning Module The intervention comprised three online digital micro-cases addressing high-volume, error-prone outpatient antibiotic stewardship scenarios: (i) upper respiratory tract infection (URTI), (ii) urinary tract infection (UTI), and (iii) acute watery diarrhoea. Each micro-case incorporated branching clinical logic with two to three decision nodes, simulating authentic clinical decision-making under uncertainty. At each decision node, an AI-style formative feedback mechanism (\"AI-coach\") provided immediate, contextualised rationale addressing antibiotic stewardship principles, clinical decision thresholds, and safety-netting strategies. The micro-cases were designed to activate illness-script formation and refinement, consistent with the script theory underpinning the development of clinical reasoning. Study Procedure and Data Collection The study was conducted in a single sitting, lasting approximately 45–60 minutes. The intervention sequence followed a structured protocol: (1) Administration of the pre-intervention Script Concordance Test (pre-SCT); (2) Completion of three DCBL micro-cases (URTI, UTI, acute diarrhoea), each requiring two to three clinical decisions with AI-coach feedback at every node; (3) Administration of the post-intervention SCT (post-SCT), using parallel items to minimise testing effects; (4) Completion of the System Usability Scale (SUS) and two NASA Task Load Index (NASA-TLX) items for usability and workload evaluation. Outcome Measures and Instruments Primary outcome. The primary outcome was overall SCT performance, operationalised as both total SCT score and mean item score. The SCT comprised 12 items, designed in accordance with established guidelines, each presenting a clinical vignette and a 5-point Likert-type scale assessing the likelihood or appropriateness of the clinical decision. Aggregate credit scoring was employed, with scores weighted according to the response distribution of a reference expert panel, thereby capturing the inherent uncertainty of clinical reasoning. Secondary outcomes. Domain-specific (case-level) SCT scores for URTI, UTI, and acute diarrhoea cases were analysed as secondary outcomes. Additionally, case-level decision accuracy was evaluated at Node-2 (the treatment decision node) for each case, serving as the primary case-level outcome: correct or appropriate antibiotic decision-making. Decisions at Node-1 and Node-3 were retained for descriptive triangulation. Usability and workload. The System Usability Scale (SUS) was administered to evaluate platform usability, yielding a composite score ranging from 0 to 100, categorised as Poor (≤ 50), Average (51–70), or Good (> 70). Cognitive workload was assessed using two items adapted from the NASA Task Load Index (NASA-TLX): self-reported mental effort and frustration, each rated on a 1–7 Likert scale. Statistical Analysis All analyses were performed on paired complete cases, matched on unique learner identifiers. Where duplicate entries existed, the most recent response was retained. Continuous variables were summarised as mean ± standard deviation (SD), and categorical variables as frequencies and percentages. For the primary analysis, paired t -tests were used to compare pre- and post-intervention total SCT scores and mean item scores. Mean change, t -statistic, p -value, and Cohen's d were reported as measures of effect size, with d = 0.2, 0.5, and 0.8 interpreted as small, medium, and large effects, respectively. For case-level analyses, paired pre–post comparisons were conducted for each of the three case domains (URTI, UTI, diarrhoea) using paired t -tests. To account for multiple comparisons across three case-wise tests, the Holm–Bonferroni sequential correction was applied; however, overall inference was anchored to the primary SCT result. Individual-level heterogeneity of response was characterised by computing the proportions of learners who improved, declined, or showed no change per case. Usability outcomes (SUS) were reported as counts and percentages across predefined categories. Workload metrics (mental effort and frustration) were reported as medians with interquartile ranges (IQR), given their ordinal nature. A two-tailed p -value < 0.05 was considered statistically significant for the primary outcome. Statistical analyses were conducted using appropriate software. The study is reported in accordance with the Transparent Reporting of Evaluations with Nonrandomised Designs (TREND) statement guidelines. Ethical approval was obtained from the Institutional Ethics Committee, JSS Medical College, JSS Academy of Higher Education and Research, Mysuru, Karnataka, India (NABH Accredited) (Approval No. JSSMC/IEC/972025/01 NCT/2024-25; dated 10 July 2025). Informed consent was obtained from all participants prior to data collection. RESULTS Participant Characteristics Of the learners approached, 271 provided complete paired pre–post data and were included in the primary analysis. All 271 included participants completed the full intervention sequence, including the three DCBL micro-cases, both SCT administrations, and the usability and workload assessments, in a single sitting of approximately 45–60 minutes. Primary Outcome: Overall SCT Performance The AI-assisted DCBL intervention was associated with a statistically significant improvement in overall SCT performance (Table 1 ). The mean total SCT score increased from 5.65 ± 4.73 at pre-test to 6.47 ± 4.48 at post-test, yielding a mean change of 0.82 ± 6.77 ( t = 1.98, p = 0.04). Correspondingly, the mean item score increased from 0.471 ± 0.395 to 0.539 ± 0.373 (mean change = 0.068 ± 0.564; t = 1.98, p = 0.04). While statistically significant, the observed effect was small in magnitude, consistent with a brief, single-session educational intervention. Table 1 Overall SCT Performance: Pre–Post Intervention Comparison (N = 271) Metric Pre-test (Mean ± SD) Post-test (Mean ± SD) Change (Mean ± SD) t-statistic p-value Test Used Total Score 5.65 ± 4.73 6.47 ± 4.48 0.82 ± 6.77 1.98 0.04* Paired t-test Mean Item Score 0.471 ± 0.395 0.539 ± 0.373 0.068 ± 0.564 1.98 0.04* Paired t-test *p < 0.05 Secondary Outcome: Case-Specific SCT Performance Case-level analyses revealed heterogeneous intervention effects across the three clinical domains (Table 2 ; Fig. 1 ). The UTI case demonstrated the most pronounced improvement, with a mean score increase from 0.79 ± 2.07 to 1.37 ± 2.35 (mean change = 0.58 ± 3.18; t = 3.00, p = 0.003, Cohen's d = 0.280). This small-to-medium effect remained significant after Holm–Bonferroni correction and was accompanied by the highest proportion of individual improvers (49.8%) across all three cases. In contrast, the URTI case showed a non-significant decline in mean score from 1.80 ± 2.11 to 1.70 ± 1.97 (mean change = − 0.10 ± 2.70; t = − 0.63, p = 0.529, Cohen's d = − 0.049), with a greater proportion of learners declining (43.2%) than improving (38.0%). Similarly, the diarrhoea case demonstrated stable performance with a negligible, non-significant change (mean change = − 0.08 ± 3.20; t = − 0.40, p = 0.690, Cohen's d = − 0.032), with nearly equal proportions of improvers (39.1%) and decliners (38.7%). Table 2 Case-Specific SCT Performance: Pre–Post Intervention Comparison (N = 271) Case Pre-test (Mean ± SD) Post-test (Mean ± SD) Change (Mean ± SD) t p d Improved n (%) Declined n (%) No Change n (%) URTI 1.80 ± 2.11 1.70 ± 1.97 −0.10 ± 2.70 −0.63 0.529 −0.049 103 (38.0) 117 (43.2) 51 (18.8) UTI 0.79 ± 2.07 1.37 ± 2.35 0.58 ± 3.18 3.00 0.003** 0.280 135 (49.8) 92 (34.0) 44 (16.2) Diarrhoea 3.06 ± 2.41 2.98 ± 2.17 −0.08 ± 3.20 −0.40 0.690 −0.032 106 (39.1) 105 (38.7) 60 (22.1) **p < 0.01. d = Cohen's d. Proportions of individual learners who improved, declined, or remained unchanged. Individual-Level Heterogeneity of Response Analysis of individual-level response patterns revealed substantial heterogeneity across all three case domains. Across cases, the proportion of learners showing improvement ranged from 38.0% (URTI) to 49.8% (UTI), while the proportion of decliners ranged from 34.0% (UTI) to 43.2% (URTI). The UTI case exhibited the most favourable individual response profile, with nearly half of all learners demonstrating improvement. Notably, 16.2–22.1% of learners showed no change across the three domains, potentially reflecting floor or ceiling effects in pre-existing knowledge (Fig. 2 ). Case-Wise Improvement in Treatment Decisions The forest plot depicting risk differences in appropriate treatment decisions (Node-2) across the three cases (Fig. 3 ) demonstrated a consistent pattern with the SCT findings. The UTI case showed the largest positive risk difference (+ 0.059; 95% CI: −0.004 to 0.122), suggesting a trend toward improved treatment decision-making, although the confidence interval crossed zero. The URTI case showed a small negative risk difference (− 0.018; 95% CI: −0.072 to 0.035), and the diarrhoea case showed an essentially null risk difference (− 0.011; 95% CI: −0.064 to 0.042). Usability Assessment Evaluation of platform usability using the System Usability Scale (Table 3 ) revealed that most learners rated the DCBL platform as Average to Good. Approximately 39.1% (n = 106) rated usability as Good (SUS > 70), 36.9% (n = 100) as Average (SUS 51–70), and 24.0% (n = 65) as Poor (SUS ≤ 50). Taken together, over three-quarters (76.0%) of the cohort found the platform's usability acceptable or better, meeting the conventional threshold for adequate system usability. Table 3 Usability Assessment: System Usability Scale (SUS) Categories (N = 271) SUS Category n % Poor (≤ 50) 65 24.0 Average (51–70) 100 36.9 Good (> 70) 106 39.1 Cognitive Workload Self-reported cognitive workload was moderate overall (Table 4 ). The median mental effort score was 4.0 (IQR: 3.0–5.0) on a 7-point scale, indicating moderate cognitive engagement. The median frustration score was 3.0 (IQR: 2.0–4.0), suggesting relatively low levels of frustration. These findings indicate that the DCBL module imposed an acceptable cognitive load on learners without inducing excessive frustration, an important consideration for the sustained implementation of technology-enhanced learning interventions. Table 4 Self-Reported Cognitive Workload: Short NASA Task Load Index (N = 271) Metric Mental Effort (1–7) Frustration (1–7) Median 4.0 3.0 Q1 (25th percentile) 3.0 2.0 Q3 (75th percentile) 5.0 4.0 IQR (Q3–Q1) 2.0 2.0 DISCUSSION This study examined the impact of a single-session, AI-assisted Digital Case-Based Learning intervention on SCT-measured clinical reasoning in antibiotic stewardship among MBBS learners. The principal finding was a statistically significant, albeit small, improvement in overall SCT performance following the DCBL exposure (mean change = 0.82, p = 0.04). This improvement, while modest in magnitude indicates a modest but potentially educationally relevant improvement given the brevity of the intervention (a single 45–60-minute session), the challenging nature of the outcome being measured (clinical reasoning under uncertainty rather than factual recall), and the well-documented resistance of higher-order cognitive skills to short-duration educational exposures. Contextualising the Overall Improvement The small effect size observed in this study is consistent with findings from analogous interventions in medical education. Cook and colleagues, in their meta-analysis of internet-based learning in the health professions, reported that technology-enhanced learning interventions typically produce small-to-moderate effects on knowledge and clinical reasoning, with larger gains accruing from repeated, spaced exposures rather than single encounters [ 23 ]. The SCT, by design, assesses a particularly elusive competence, the capacity to modulate clinical decisions in response to new, uncertain information, which is unlikely to shift dramatically after a brief educational exposure. Naylor et al., evaluating SCT performance among paramedic students, similarly observed that the formative value of the SCT often exceeds its sensitivity to short-term interventions, and recommended regular exposure to facilitate familiarity and deeper script development[ 19 ]. Our finding that overall SCT scores improved significantly, even from a single session, thus provides encouraging preliminary evidence that AI-assisted DCBL can begin to activate the cognitive processes that underpin clinical reasoning in antibiotic stewardship. Differential Case-Specific Effects Perhaps the most instructive finding was the marked heterogeneity in intervention effects across the three clinical domains. The UTI case demonstrated a statistically significant improvement ( p = 0.003, Cohen's d = 0.280) with the highest proportion of individual improvers (49.8%), while URTI and diarrhoea cases showed no significant change. Several factors may explain this differential response. First, the UTI case may have offered the greatest scope for learning because baseline performance was the lowest among the three domains (pre-test mean = 0.79), leaving more room for improvement. This is consistent with the well-recognised floor-and-ceiling phenomenon in educational measurement, where learners with the lowest initial performance tend to benefit most from targeted instruction [ 24 ]. In contrast, the diarrhoea case had the highest baseline score (3.06), potentially limiting upward movement and creating a ceiling effect. Second, the clinical content itself likely contributed to the differential effect. UTI management involves relatively discrete, guideline-amenable decision points (choice of empirical antibiotic, duration of therapy, criteria for culture), which may be more amenable to the branching-logic and feedback structure of the DCBL format. In contrast, URTI management, where the primary stewardship challenge is to withhold rather than prescribe antibiotics, requires a different kind of reasoning that involves resisting prescribing pressure and managing diagnostic uncertainty about bacterial versus viral aetiology. This \"withholding\" decision may be less responsive to the type of structured feedback provided in the micro-cases. The challenge of improving antibiotic stewardship for URTIs has been documented in other settings as well, where even multi-faceted interventions required sustained effort to reduce URTI antibiotic prescribing from 62.6% to 7.2% [ 25 ]. Similarly, diarrhoea management, while conceptually straightforward (oral rehydration as first-line), involves nuanced decisions about the role of antibiotics in dysenteric versus non-dysenteric presentations that may require deeper clinical exposure than a single digital session can provide. Individual-Level Heterogeneity The substantial heterogeneity observed at the individual level, with 38–49.8% of learners improving, 34–43.2% declining, and 16–22.1% showing no change across cases, warrants careful consideration. This pattern is not unusual in medical education research; clinical reasoning development is highly individual, shaped by prior knowledge structures, motivation, and metacognitive capacity [ 26 ]. The proportion of \"decliners\" may partly reflect regression to the mean, particularly among learners whose pre-test scores were inflated by guessing. It may also indicate that the DCBL intervention, while activating new scripts for some learners, may have introduced cognitive interference in others, a phenomenon where new information temporarily disrupts existing knowledge structures before reorganisation occurs [ 27 ]. From a practical standpoint, the mixed individual response profile argues strongly for iterative, spaced DCBL exposure rather than a one-off session. Educational research consistently demonstrates that distributed practice produces more durable learning gains than massed practice, particularly for complex cognitive skills [ 28 ]. Incorporating DCBL micro-cases into a longitudinal curriculum, perhaps revisiting the same clinical domains at increasing levels of complexity over successive weeks, would likely amplify the effects observed in this study and provide opportunities for targeted scaffolding for learners who do not respond to initial exposure. Usability and Cognitive Workload The usability findings provide important context for interpreting the learning outcomes and for planning future implementation. With 76% of learners rating the platform as Average to Good on the SUS, the DCBL module meets the generally accepted threshold for adequate usability in educational technology [ 29 , 30 ]. This is noteworthy because poor usability can confound educational outcomes if learners struggle with the interface, diverting cognitive resources from the learning task to system navigation and potentially undermining the intended educational effect [ 31 ]. The moderate mental effort (median = 4.0/7) and low frustration (median = 3.0/7) reported on the NASA-TLX items further suggest that the DCBL module achieved an appropriate balance between cognitive challenge and accessibility, a critical consideration under cognitive load theory [ 32 ]. However, the 24% of learners who rated usability as Poor cannot be dismissed. Digital literacy varies considerably among Indian medical students, and interface design choices that seem intuitive to one subgroup may create barriers for another. Future iterations should incorporate user experience testing with low-performing subgroups to identify specific usability pain points and ensure equitable access to the learning experience. Implications for CBME and Antibiotic Stewardship Education This study carries several implications for the evolving CBME landscape in India. First, it demonstrates that digital, AI-enhanced educational tools can be successfully integrated into routine CBME teaching to address competencies that are difficult to teach through lectures alone, specifically, clinical decision-making under uncertainty in antibiotic stewardship. The DCBL format aligns naturally with several CBME principles: it activates authentic clinical reasoning, provides just-in-time formative feedback, yields individual-level analytics for programmatic assessment, and can be delivered at scale across institutions [ 33 , 34 ]. Second, the content-specific findings point to the need for domain-tailored instructional design. The success of the UTI case, relative to URTI and diarrhoea, suggests that the branching-logic, feedback-rich DCBL format is best suited to clinical scenarios with clear decision trees and guideline-concordant endpoints. For scenarios like URTI, where the stewardship challenge primarily involves withholding treatment, complementary strategies, such as communication skills training, shared decision-making simulations, or reflective debriefing, may be needed alongside the DCBL module. Third, from an antimicrobial stewardship perspective, the study reinforces the importance of engaging learners early in their training. The formation of prescribing habits begins in medical school and is shaped by clinical interactions and educational experiences [ 22 ]. Embedding stewardship reasoning exercises into the undergraduate curriculum, rather than deferring to postgraduate training, can help establish appropriate prescribing patterns before suboptimal practices become entrenched. Strengths and Limitations The strengths of this study include a substantial sample size (N = 271), the use of a validated clinical reasoning instrument (SCT) with expert-panel scoring, the reporting of both statistical significance and effect sizes, the evaluation of individual-level heterogeneity, and the inclusion of usability and cognitive workload assessments that are often omitted from educational intervention studies. Several limitations must be acknowledged. The single-group, pre–post design without a control group precludes causal inference, as observed improvements may reflect testing effects, maturation, or regression to the mean, despite the use of parallel SCT items. The single-session design limits conclusions about the durability of learning gains. The convenience sample from a single institution may constrain generalisability, and the absence of follow-up assessment means we cannot comment on retention of improvements over time. The URTI and diarrhoea cases showed no significant change, suggesting that the UTI domain largely drove the overall SCT improvement. Finally, the study was conducted within a specific CBME context in South India, and findings should be replicated in diverse institutional and geographical settings with varying baseline knowledge levels and curricular structures. Future Directions Future research should employ randomised controlled designs with active comparator groups to isolate the specific contribution of AI-assisted feedback from the case-based learning itself. Longitudinal studies with spaced DCBL sessions and delayed post-tests would clarify whether the observed improvements are durable and whether repeated exposure amplifies the effect. Adaptive AI algorithms that tailor feedback intensity and content to individual learner performance profiles could address the heterogeneity observed in this study. Additionally, aligning case content to local antibiograms and resistance patterns would enhance ecological validity and clinical relevance. Multisite studies across diverse Indian medical colleges would strengthen the evidence base for broader CBME integration. Declarations Ethics approval and consent to participate: Ethical approval for this study was obtained from the Institutional Ethics Committee, JSS Medical College, JSS Academy of Higher Education and Research, Mysuru, Karnataka, India (NABH Accredited) (Approval No. JSSMC/IEC/972025/01 NCT/2024-25; dated 10 July 2025). Participation was voluntary. Informed consent was obtained from all participants prior to inclusion in the study. Consent for publication : Not applicable. Data availability : The datasets generated and/or analysed during the current study are not publicly available because they form part of an ongoing educational research dataset, but de-identified data may be made available from the corresponding author on reasonable request. Competing interests : The authors declare that they have no competing interests. Funding : This research received no external funding. Authors’ contributions: Sunil Kumar D conceptualised the study, conducted the study, developed the methodology, and performed the results analysis. Sucheta Dandekar supervised the work and critically reviewed the manuscript. G Hari Prakash contributed to manuscript preparation and interpretation of the results. All authors read and approved the final manuscript. Acknowledgements: The authors thank the participating students for their engagement in the study. References Kassirer JP. Teaching clinical reasoning: Case-based and coached. Acad Med. 2010;85:1118–24. https://doi.org/10.1097/ACM.0b013e3181d5dd0d . Charlin B, Roy L, Brailovsky C, Goulet F, van der Vleuten C. The script concordance test: A tool to assess the reflective clinician. Teach Learn Med. 2000;12:189–95. https://doi.org/10.1207/S15328015TLM1204_5 . Lubarsky S, Dory V, Duggan P, Gagnon R, Charlin B. Script concordance testing: From theory to practice: AMEE Guide 75. Med Teach. 2013;35:184–93. https://doi.org/10.3109/0142159X.2013.760036 . Indian Council of Medical Research. Annual report: Antimicrobial Resistance Surveillance and Research Network (AMRSN). 2023. ICMR; 2024. Laxminarayan R, Chaudhury RR. Antibiotic resistance in India: Drivers and opportunities for action. PLoS Med. 2016;13:e1001974. https://doi.org/10.1371/journal.pmed.1001974 . Goebel MC, Trautner BW, Grigoryan L. The five Ds of outpatient antibiotic stewardship for urinary tract infections. Clin Microbiol Rev. 2021;34:e00003–20. https://doi.org/10.1128/CMR.00003-20 . Beck AP, Baubie K, Knobloch MJ, Safdar N. Promoting antimicrobial stewardship by incorporating it in undergraduate medical education curricula. Wis Med J. 2018;117:224–8. Wang EY, Siddiqi H, Engmann CN, Kang YS, Yin HS. Development of a multifaceted antimicrobial stewardship curriculum for undergraduate medical education. Open Forum Infect Dis. 2021;8:ofab154. https://doi.org/10.1093/ofid/ofab154 . National Medical Commission. Competency-based medical education (CBME) curriculum guidelines for the Indian medical graduate. Government of India; 2023. Soundariya K, Nishanthi A, Mahendran R, Vimal M. Evaluation of competency-based medical education (CBME) curriculum implementation for Phase II medical undergraduates: A qualitative study. J Adv Med Educ Professionalism. 2025;13:36–48. https://doi.org/10.30476/jamp.2024.104312.2052 . Singh K, Gupta P, Dhaliwal U, Kochhar K. Competency-based medical education in India: Clinicians’ multifaceted challenges. Indian J Otolaryngol Head Neck Surg. 2024;76:3741–9. https://doi.org/10.1007/s12070-024-04751-x . Cook DA, Levinson AJ, Garside S, Dupras DM, Erwin PJ, Montori VM. Instructional design variations in internet-based learning for health professions education: A systematic review and meta-analysis. Acad Med. 2010;85:909–22. https://doi.org/10.1097/ACM.0b013e3181d6c319 . Stevens NT, Bruen C, Boland F, Pawlikowska T, Fitzpatrick F, Humphreys H. Is online case-based learning effective in helping undergraduate medical students choose the appropriate antibiotics to treat important infections? JAC-Antimicrobial Resist. 2019;1:dlz081. https://doi.org/10.1093/jacamr/dlz081 . Garg T, Shrigiriwar A, Garg A. A scoping review of artificial intelligence in medical education: BEME Guide 84. Med Teach. 2024;46:446–70. https://doi.org/10.1080/0142159X.2024.2314198 . Masters K. Ethical use of artificial intelligence in health professions education: AMEE Guide 158. Med Teach. 2023;45:574–84. https://doi.org/10.1080/0142159X.2023.2186203 . Fournier JP, Demeester A, Charlin B. Script concordance tests: Guidelines for construction. BMC Med Inf Decis Mak. 2008;8:18. https://doi.org/10.1186/1472-6947-8-18 . Lubarsky S, Charlin B, Cook DA, Chalk C, van der Vleuten CP. Script concordance testing: A review of published validity evidence. Med Educ. 2011;45:329–38. https://doi.org/10.1111/j.1365-2923.2010.03863.x . Boulouffe C, Doucet B, Muschart X, Charlin B, Vanpee D. Assessing clinical reasoning using a script concordance test with electrocardiogram in an emergency medicine clerkship rotation. Emerg Med J. 2013;31:313–6. https://doi.org/10.1136/emermed-2012-201737 . Naylor K, Hislop J, Torres K, Mani ZA, Goniewicz K. The impact of script concordance testing on clinical decision-making in paramedic education. Healthcare. 2024;12:282. https://doi.org/10.3390/healthcare12020282 . Ross L, Semaan E, Gosling CM, Fisk B, Shannon B. Clinical reasoning in undergraduate paramedicine: Utilisation of a script concordance test. BMC Med Educ. 2023;23:39. https://doi.org/10.1186/s12909-023-04020-x . Malli IA, Mohamud MS, Al-Nasser SS. Enhancing medical students’ confidence and knowledge in antibiotic prescription and administration through virtual education: A quasi-experimental study. Antibiotics. 2023;12:1546. https://doi.org/10.3390/antibiotics12101546 . Silverberg SL, Zannella VE, Engel S, Engel AS. Implementing an antimicrobial stewardship medical student elective to impart good antibiotic prescribing habits early in medical training. Antimicrob Stewardship Healthc Epidemiol. 2023;3:e225. https://doi.org/10.1017/ash.2023.491 . Cook DA, Levinson AJ, Garside S. Internet-based learning in the health professions: A meta-analysis. JAMA. 2008;300:1181–96. https://doi.org/10.1001/jama.300.10.1181 . Cook DA, Artino AR. Motivation to learn: An overview of contemporary theories. Med Educ. 2016;50:997–1014. https://doi.org/10.1111/medu.13074 . Gulliford MC, Dregan A, Moore MV, Ashworth M, van Staa T, McCann G, et al. Continued high rates of antibiotic prescribing to adults with respiratory tract infection: Survey of 568 UK general practices. BMJ Open. 2014;4:e006245. https://doi.org/10.1136/bmjopen-2014-006245 . Eva KW. What every teacher needs to know about clinical reasoning. Med Educ. 2005;39:98–106. https://doi.org/10.1111/j.1365-2929.2004.01972.x . Schmidt HG, Rikers RM. How expertise develops in medicine: Knowledge encapsulation and illness script formation. Med Educ. 2007;41:1133–9. https://doi.org/10.1111/j.1365-2923.2007.02915.x . Dunlosky J, Rawson KA, Marsh EJ, Nathan MJ, Willingham DT. Improving students’ learning with effective learning techniques. Psychol Sci Public Interest. 2013;14:4–58. https://doi.org/10.1177/1529100612453266 . Brooke J. SUS: A retrospective. J Usability Stud. 2013;8:29–40. Vlachogianni P, Tselios N. Perceived usability evaluation of educational technology using the System Usability Scale (SUS): A systematic review. J Res Technol Educ. 2022;54:392–409. https://doi.org/10.1080/15391523.2020.1867938 . Sweller J, van Merriënboer JJG, Paas F. Cognitive architecture and instructional design: 20 years later. Educational Psychol Rev. 2019;31:261–92. https://doi.org/10.1007/s10648-019-09465-5 . Sweller J. Cognitive load theory. In: Mestre JP, Ross BH, editors. Psychology of learning and motivation. Academic; 2011. pp. 37–76. Ten Cate O. Nuts and bolts of entrustable professional activities. J Graduate Med Educ. 2013;5:157–8. https://doi.org/10.4300/JGME-D-12-00380.1 . van der Vleuten CPM, Schuwirth LWT, Driessen EW, Dijkstra J, Tigelaar D, Baartman LKJ, et al. A model for programmatic assessment fit for purpose. Med Teach. 2012;34:205–14. https://doi.org/10.3109/0142159X.2012.652239 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 16 May, 2026 Reviews received at journal 04 May, 2026 Reviewers agreed at journal 03 May, 2026 Reviewers agreed at journal 28 Apr, 2026 Reviewers invited by journal 28 Apr, 2026 Editor invited by journal 20 Apr, 2026 Editor assigned by journal 17 Apr, 2026 Submission checks completed at journal 17 Apr, 2026 First submitted to journal 11 Apr, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-9384886\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":633824551,\"identity\":\"a33a1b44-1e8e-49ff-8d4f-324d217bff28\",\"order_by\":0,\"name\":\"Sunil Kumar D\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABBklEQVRIiWNgGAWjYJCCAw8OgGnDBxIgihmIecAIj5YEiBZjA6K1MEC1mEnARfCpl5+Re/BAwpnD0fyzD2+rsPhTl7idnYHxwds2BhlzHFoMbuQlHEi4cTh3xrm0shuSbYcTdzYzMBvObWPgsWzAoUUix+BAwofDuQ1neMxuSDYcSNxwmIFNmheoxeAALodBtcwHaimQADoMqIX9Nz4tDDdAWoAO2wDUwiDBxgy2hRmfFoMzb4BazqTnbjzDViwB9IvxhsOMzZJzzkngdlh7jvGHD8esc+edYd74Gegw2Q3nDx/88KbMxh6nw5ABMyRiGBuAhARelXDA+IE4daNgFIyCUTDCAADsumKARaf8jQAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"JSS Medical College, JSS Academy of Higher Education and Research\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Sunil\",\"middleName\":\"Kumar\",\"lastName\":\"D\",\"suffix\":\"\"},{\"id\":633824552,\"identity\":\"9c13f51b-13c1-4a08-b577-0c97027bd243\",\"order_by\":1,\"name\":\"Sucheta Dandekar\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Era's Lucknow Medical College and Hospital\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Sucheta\",\"middleName\":\"\",\"lastName\":\"Dandekar\",\"suffix\":\"\"},{\"id\":633824553,\"identity\":\"1dd8b896-30a1-46ac-85d9-6d792e4c16a5\",\"order_by\":2,\"name\":\"G Hari Prakash\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"M S Ramaiah University of Applied Sciences\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"G\",\"middleName\":\"Hari\",\"lastName\":\"Prakash\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2026-04-11 05:55:17\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-9384886/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-9384886/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":108944277,\"identity\":\"daaba4f6-8838-4145-a8a1-8035263b2ee5\",\"added_by\":\"auto\",\"created_at\":\"2026-05-11 05:58:09\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":178469,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eIndividual case analysis showing pre- and post–SCT score distributions for URTI, UTI, and diarrhoea cases\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-9384886/v1/ef5a02ad72ca6dd4f8db5c6b.png\"},{\"id\":108944239,\"identity\":\"22530d0a-5dc0-4cd9-84ac-e0056367079f\",\"added_by\":\"auto\",\"created_at\":\"2026-05-11 05:57:51\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":268035,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003eDepicting case-wise risk differences in appropriate treatment decisions (Node-2) between pre- and post-intervention assessments with 95% confidence intervals.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-9384886/v1/3fed117efbbb74fc0eb151a8.png\"},{\"id\":108944278,\"identity\":\"c699ca3b-0a88-4b99-94ea-52685e1576b4\",\"added_by\":\"auto\",\"created_at\":\"2026-05-11 05:58:10\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":212360,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003ePaired comparison of overall SCT scores at pre-test and post-test time points (N = 271). Error bars represent 95% confidence intervals.\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"floatimage3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-9384886/v1/878bb2142cd62015db3ea654.png\"},{\"id\":108944449,\"identity\":\"d9279efe-698f-43f0-b231-a7ef3c777ca8\",\"added_by\":\"auto\",\"created_at\":\"2026-05-11 05:58:58\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":956772,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-9384886/v1/813244a4-b13c-411f-85b6-836296047896.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Dynamic CBME in Action: AI-Assisted Digital Case-Based Learning to Improve Antibiotic-Stewardship Decisions in MBBS Learners\",\"fulltext\":[{\"header\":\"Introduction\",\"content\":\"\\u003cp\\u003eClinical reasoning, the cognitive process through which clinicians integrate clinical data, formulate diagnoses, and arrive at treatment decisions under conditions of uncertainty, is widely recognised as the cornerstone of medical competence[\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e]. Unlike factual recall, clinical reasoning demands that learners navigate ambiguity, weigh competing hypotheses, and apply context-dependent judgement, skills that are difficult to cultivate through didactic lectures or conventional multiple-choice assessments alone [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e]. This gap has become particularly pressing in the context of antibiotic prescribing, where reasoning errors translate directly into patient harm and population-level antimicrobial resistance (AMR).\\u003c/p\\u003e \\u003cp\\u003eAntimicrobial resistance is now among the foremost global public health threats. The Indian Council of Medical Research (ICMR) Antimicrobial Resistance Surveillance Network has documented alarming declines in the effectiveness of commonly used antibiotics against gram-negative organisms, with key drugs such as ciprofloxacin and ceftriaxone showing susceptibility rates of less than 20% against \\u003cem\\u003eEscherichia coli\\u003c/em\\u003e isolates in several outpatient settings [\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e]. In India, where 45\\u0026ndash;80% of patients presenting with upper respiratory tract infections (URTI) and diarrhoea receive antibiotics despite predominantly viral aetiologies, the outpatient setting remains a critical frontier for stewardship[\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e]. Urinary tract infections (UTI), another high-volume outpatient condition, carry their own stewardship challenges, with guideline-discordant prescribing documented across multiple countries [\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e]. Yet, despite the magnitude of the problem, antibiotic stewardship education in undergraduate medical curricula remains inconsistent and largely didactic [\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eThe introduction of Competency-Based Medical Education (CBME) by India's National Medical Commission has fundamentally reoriented the goals of undergraduate training toward demonstrable competencies rather than mere content acquisition [\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e]. The CBME framework explicitly demands that the Indian Medical Graduate (IMG) function as a clinician capable of first-contact decision-making, a communicator, a leader, a professional, and a lifelong learner [\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e]. However, the transition from curriculum intent to classroom reality has been challenging, with faculty reporting difficulties in assessing higher-order cognitive skills such as clinical reasoning [\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e]. This underscores an urgent need for educational interventions that go beyond knowledge transmission and actively engage learners in authentic clinical decision-making within the CBME framework.\\u003c/p\\u003e \\u003cp\\u003eDigital case-based learning (DCBL) has emerged as a promising pedagogical strategy for bridging this gap. Unlike traditional case discussions, digital platforms can embed branching clinical logic, simulate realistic decision pathways, and crucially provide immediate, individualised formative feedback at each decision node [\\u003cspan citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e]. When augmented with artificial intelligence (AI), such systems can deliver contextualised rationale aligned to stewardship principles, helping learners understand not merely what the correct answer is, but why a particular clinical decision is appropriate in a given context. Recent evidence suggests that AI-enhanced educational tools can improve clinical reasoning and knowledge retention, particularly when embedded within structured pedagogical frameworks rather than deployed as standalone resources [\\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eThe Script Concordance Test (SCT), originally described by Charlin and colleagues, offers a robust method for assessing clinical reasoning under uncertainty [\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e]. Unlike conventional assessments that test factual knowledge, the SCT presents learners with clinical scenarios that require probabilistic reasoning, evaluating how new information modifies the likelihood of a particular hypothesis or the appropriateness of a management decision. Responses are scored against an expert reference panel using the aggregate-credit method, thus accommodating the legitimate variability inherent in expert clinical judgement [\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e]. The SCT has been validated across diverse medical disciplines and training levels, including emergency medicine, paediatrics, neurology, and paramedicine, and has been recommended as both a summative assessment tool and a formative learning instrument[\\u003cspan additionalcitationids=\\\"CR19\\\" citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eDespite growing interest in both digital learning and SCT-based assessment, evidence linking AI-assisted DCBL specifically to improvements in SCT-measured clinical reasoning remains scarce, particularly in the context of antibiotic stewardship education for undergraduate medical learners in low- and middle-income countries (LMICs). Most existing studies on antibiotic stewardship education have focused on knowledge gains measured through conventional tests, with limited attention to higher-order reasoning outcomes or the cognitive demands imposed on learners [\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e]. Furthermore, the usability and cognitive workload of technology-enhanced interventions, critical determinants of sustained adoption, are infrequently reported.\\u003c/p\\u003e \\u003cp\\u003eThis study, therefore, aimed to determine whether a single-session AI-assisted DCBL intervention comprising digital micro-cases on three common outpatient conditions (URTI, UTI, and acute watery diarrhoea) with AI-coach formative feedback could improve overall SCT performance in MBBS learners. Secondary objectives included evaluating case-specific effects across clinical domains, characterising individual-level heterogeneity of response, and assessing platform usability and cognitive workload using the System Usability Scale (SUS) and NASA Task Load Index (NASA-TLX) items.\\u003c/p\\u003e\"},{\"header\":\"METHODS\",\"content\":\"\\u003cdiv id=\\\"Sec3\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eStudy Design and Setting\\u003c/h2\\u003e \\u003cp\\u003eThis was a quasi-experimental, single-group pre\\u0026ndash;post study embedded within routine Competency-Based Medical Education (CBME) teaching at a tertiary-care teaching hospital in South India. The study evaluated the effect of an artificial intelligence (AI)-assisted Digital Case-Based Learning (DCBL) intervention on clinical reasoning in antibiotic stewardship for common outpatient conditions. The study was conducted as an educational intervention study within the existing CBME framework and did not involve patient clinical data. Voluntary participation was ensured, and all data were de-identified before analysis. The study adhered to the ethical principles of the Declaration of Helsinki for educational research.\\u003c/p\\u003e \\u003c/div\\u003e\\n\\u003ch3\\u003eParticipants\\u003c/h3\\u003e\\n\\u003cp\\u003eA convenience sample of undergraduate medical (MBBS) learners enrolled in the CBME curriculum was recruited. Eligibility criteria included current enrolment in the MBBS programme and willingness to participate in the complete study session. A total of 271 participants provided complete paired (pre\\u0026ndash;post) data and constituted the analytic sample for the primary analysis. Participation was voluntary; coded identifiers were assigned to enable data pairing while preserving anonymity.\\u003c/p\\u003e\\n\\u003ch3\\u003eIntervention: AI-Assisted Digital Case-Based Learning Module\\u003c/h3\\u003e\\n\\u003cp\\u003eThe intervention comprised three online digital micro-cases addressing high-volume, error-prone outpatient antibiotic stewardship scenarios: (i) upper respiratory tract infection (URTI), (ii) urinary tract infection (UTI), and (iii) acute watery diarrhoea. Each micro-case incorporated branching clinical logic with two to three decision nodes, simulating authentic clinical decision-making under uncertainty. At each decision node, an AI-style formative feedback mechanism (\\\"AI-coach\\\") provided immediate, contextualised rationale addressing antibiotic stewardship principles, clinical decision thresholds, and safety-netting strategies. The micro-cases were designed to activate illness-script formation and refinement, consistent with the script theory underpinning the development of clinical reasoning.\\u003c/p\\u003e\\n\\u003ch3\\u003eStudy Procedure and Data Collection\\u003c/h3\\u003e\\n\\u003cp\\u003eThe study was conducted in a single sitting, lasting approximately 45\\u0026ndash;60 minutes. The intervention sequence followed a structured protocol:\\u003cdiv class=\\\"BlockQuote\\\"\\u003e\\u003cp\\u003e(1) Administration of the pre-intervention Script Concordance Test (pre-SCT);\\u003c/p\\u003e\\u003cp\\u003e(2) Completion of three DCBL micro-cases (URTI, UTI, acute diarrhoea), each requiring two to three clinical decisions with AI-coach feedback at every node;\\u003c/p\\u003e\\u003cp\\u003e(3) Administration of the post-intervention SCT (post-SCT), using parallel items to minimise testing effects;\\u003c/p\\u003e\\u003cp\\u003e(4) Completion of the System Usability Scale (SUS) and two NASA Task Load Index (NASA-TLX) items for usability and workload evaluation.\\u003c/p\\u003e\\u003c/div\\u003e\\u003c/p\\u003e\\n\\u003ch3\\u003eOutcome Measures and Instruments\\u003c/h3\\u003e\\n\\u003cp\\u003e\\u003cb\\u003ePrimary outcome.\\u003c/b\\u003e The primary outcome was overall SCT performance, operationalised as both total SCT score and mean item score. The SCT comprised 12 items, designed in accordance with established guidelines, each presenting a clinical vignette and a 5-point Likert-type scale assessing the likelihood or appropriateness of the clinical decision. Aggregate credit scoring was employed, with scores weighted according to the response distribution of a reference expert panel, thereby capturing the inherent uncertainty of clinical reasoning.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eSecondary outcomes.\\u003c/b\\u003e Domain-specific (case-level) SCT scores for URTI, UTI, and acute diarrhoea cases were analysed as secondary outcomes. Additionally, case-level decision accuracy was evaluated at Node-2 (the treatment decision node) for each case, serving as the primary case-level outcome: correct or appropriate antibiotic decision-making. Decisions at Node-1 and Node-3 were retained for descriptive triangulation.\\u003c/p\\u003e \\u003cp\\u003e \\u003cb\\u003eUsability and workload.\\u003c/b\\u003e The System Usability Scale (SUS) was administered to evaluate platform usability, yielding a composite score ranging from 0 to 100, categorised as Poor (\\u0026le;\\u0026thinsp;50), Average (51\\u0026ndash;70), or Good (\\u0026gt;\\u0026thinsp;70). Cognitive workload was assessed using two items adapted from the NASA Task Load Index (NASA-TLX): self-reported mental effort and frustration, each rated on a 1\\u0026ndash;7 Likert scale.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec8\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eStatistical Analysis\\u003c/h2\\u003e \\u003cp\\u003eAll analyses were performed on paired complete cases, matched on unique learner identifiers. Where duplicate entries existed, the most recent response was retained. Continuous variables were summarised as mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;standard deviation (SD), and categorical variables as frequencies and percentages.\\u003c/p\\u003e \\u003cp\\u003eFor the primary analysis, paired \\u003cem\\u003et\\u003c/em\\u003e-tests were used to compare pre- and post-intervention total SCT scores and mean item scores. Mean change, \\u003cem\\u003et\\u003c/em\\u003e-statistic, \\u003cem\\u003ep\\u003c/em\\u003e-value, and Cohen's \\u003cem\\u003ed\\u003c/em\\u003e were reported as measures of effect size, with \\u003cem\\u003ed\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.2, 0.5, and 0.8 interpreted as small, medium, and large effects, respectively.\\u003c/p\\u003e \\u003cp\\u003eFor case-level analyses, paired pre\\u0026ndash;post comparisons were conducted for each of the three case domains (URTI, UTI, diarrhoea) using paired \\u003cem\\u003et\\u003c/em\\u003e-tests. To account for multiple comparisons across three case-wise tests, the Holm\\u0026ndash;Bonferroni sequential correction was applied; however, overall inference was anchored to the primary SCT result. Individual-level heterogeneity of response was characterised by computing the proportions of learners who improved, declined, or showed no change per case.\\u003c/p\\u003e \\u003cp\\u003eUsability outcomes (SUS) were reported as counts and percentages across predefined categories. Workload metrics (mental effort and frustration) were reported as medians with interquartile ranges (IQR), given their ordinal nature.\\u003c/p\\u003e \\u003cp\\u003eA two-tailed \\u003cem\\u003ep\\u003c/em\\u003e-value\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05 was considered statistically significant for the primary outcome. Statistical analyses were conducted using appropriate software. The study is reported in accordance with the Transparent Reporting of Evaluations with Nonrandomised Designs (TREND) statement guidelines.\\u003c/p\\u003e \\u003cp\\u003e \\u003cstrong\\u003eEthical approval\\u003c/strong\\u003e \\u003cp\\u003e was obtained from the Institutional Ethics Committee, JSS Medical College, JSS Academy of Higher Education and Research, Mysuru, Karnataka, India (NABH Accredited) (Approval No. JSSMC/IEC/972025/01 NCT/2024-25; dated 10 July 2025). Informed consent was obtained from all participants prior to data collection.\\u003c/p\\u003e \\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"RESULTS\",\"content\":\"\\u003cdiv id=\\\"Sec10\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eParticipant Characteristics\\u003c/h2\\u003e \\u003cp\\u003eOf the learners approached, 271 provided complete paired pre\\u0026ndash;post data and were included in the primary analysis. All 271 included participants completed the full intervention sequence, including the three DCBL micro-cases, both SCT administrations, and the usability and workload assessments, in a single sitting of approximately 45\\u0026ndash;60 minutes.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec11\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003ePrimary Outcome: Overall SCT Performance\\u003c/h2\\u003e \\u003cp\\u003eThe AI-assisted DCBL intervention was associated with a statistically significant improvement in overall SCT performance (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). The mean total SCT score increased from 5.65\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.73 at pre-test to 6.47\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.48 at post-test, yielding a mean change of 0.82\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;6.77 (\\u003cem\\u003et\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;1.98, \\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.04). Correspondingly, the mean item score increased from 0.471\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;0.395 to 0.539\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;0.373 (mean change\\u0026thinsp;=\\u0026thinsp;0.068\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;0.564; \\u003cem\\u003et\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;1.98, \\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.04). While statistically significant, the observed effect was small in magnitude, consistent with a brief, single-session educational intervention.\\u003c/p\\u003e \\u003cp\\u003e \\u003cdiv class=\\\"gridtable\\\"\\u003e\\u003ctable float=\\\"Yes\\\" id=\\\"Tab1\\\" border=\\\"1\\\"\\u003e \\u003ccaption language=\\\"En\\\"\\u003e \\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 1\\u003c/div\\u003e \\u003cdiv class=\\\"CaptionContent\\\"\\u003e \\u003cp\\u003e\\u003cem\\u003eOverall SCT Performance: Pre\\u0026ndash;Post Intervention Comparison (N\\u0026thinsp;=\\u0026thinsp;271)\\u003c/em\\u003e\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/caption\\u003e \\u003ccolgroup cols=\\\"7\\\"\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c1\\\" colnum=\\\"1\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" class=\\\"colspec\\\" colname=\\\"c2\\\" colnum=\\\"2\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" class=\\\"colspec\\\" colname=\\\"c4\\\" colnum=\\\"4\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c5\\\" colnum=\\\"5\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c6\\\" colnum=\\\"6\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c7\\\" colnum=\\\"7\\\"\\u003e\\u003c/div\\u003e \\u003cthead\\u003e \\u003ctr\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eMetric\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003ePre-test (Mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003ePost-test (Mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eChange (Mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003et-statistic\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003ep-value\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003eTest Used\\u003c/p\\u003e \\u003c/th\\u003e \\u003c/tr\\u003e \\u003c/thead\\u003e \\u003ctbody\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eTotal Score\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e5.65\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.73\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e6.47\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.48\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003e0.82\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;6.77\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003e1.98\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003e0.04*\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003ePaired t-test\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eMean Item Score\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e0.471\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;0.395\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e0.539\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;0.373\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003e0.068\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;0.564\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003e1.98\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003e0.04*\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003ePaired t-test\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003c/tbody\\u003e \\u003c/colgroup\\u003e \\u003c/table\\u003e\\u003c/div\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec12\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003e*p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.05\\u003c/h2\\u003e \\u003cdiv id=\\\"Sec13\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eSecondary Outcome: Case-Specific SCT Performance\\u003c/h2\\u003e \\u003cp\\u003eCase-level analyses revealed heterogeneous intervention effects across the three clinical domains (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e; Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig1\\\" class=\\\"InternalRef\\\"\\u003e1\\u003c/span\\u003e). The UTI case demonstrated the most pronounced improvement, with a mean score increase from 0.79\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.07 to 1.37\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.35 (mean change\\u0026thinsp;=\\u0026thinsp;0.58\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;3.18; \\u003cem\\u003et\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;3.00, \\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.003, Cohen's \\u003cem\\u003ed\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.280). This small-to-medium effect remained significant after Holm\\u0026ndash;Bonferroni correction and was accompanied by the highest proportion of individual improvers (49.8%) across all three cases.\\u003c/p\\u003e \\u003cp\\u003eIn contrast, the URTI case showed a non-significant decline in mean score from 1.80\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.11 to 1.70\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;1.97 (mean change\\u0026thinsp;=\\u0026thinsp;\\u0026minus;\\u0026thinsp;0.10\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.70; \\u003cem\\u003et\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;\\u0026minus;\\u0026thinsp;0.63, \\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.529, Cohen's \\u003cem\\u003ed\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;\\u0026minus;\\u0026thinsp;0.049), with a greater proportion of learners declining (43.2%) than improving (38.0%). Similarly, the diarrhoea case demonstrated stable performance with a negligible, non-significant change (mean change\\u0026thinsp;=\\u0026thinsp;\\u0026minus;\\u0026thinsp;0.08\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;3.20; \\u003cem\\u003et\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;\\u0026minus;\\u0026thinsp;0.40, \\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.690, Cohen's \\u003cem\\u003ed\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;\\u0026minus;\\u0026thinsp;0.032), with nearly equal proportions of improvers (39.1%) and decliners (38.7%).\\u003c/p\\u003e \\u003cp\\u003e \\u003cdiv class=\\\"gridtable\\\"\\u003e\\u003ctable float=\\\"Yes\\\" id=\\\"Tab2\\\" border=\\\"1\\\"\\u003e \\u003ccaption language=\\\"En\\\"\\u003e \\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 2\\u003c/div\\u003e \\u003cdiv class=\\\"CaptionContent\\\"\\u003e \\u003cp\\u003e\\u003cem\\u003eCase-Specific SCT Performance: Pre\\u0026ndash;Post Intervention Comparison (N\\u0026thinsp;=\\u0026thinsp;271)\\u003c/em\\u003e\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/caption\\u003e \\u003ccolgroup cols=\\\"10\\\"\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c1\\\" colnum=\\\"1\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" class=\\\"colspec\\\" colname=\\\"c2\\\" colnum=\\\"2\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" class=\\\"colspec\\\" colname=\\\"c4\\\" colnum=\\\"4\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c5\\\" colnum=\\\"5\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c6\\\" colnum=\\\"6\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c7\\\" colnum=\\\"7\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c8\\\" colnum=\\\"8\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c9\\\" colnum=\\\"9\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c10\\\" colnum=\\\"10\\\"\\u003e\\u003c/div\\u003e \\u003cthead\\u003e \\u003ctr\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eCase\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003ePre-test (Mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003ePost-test (Mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003eChange (Mean\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;SD)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003et\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003ep\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003ed\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c8\\\"\\u003e \\u003cp\\u003eImproved n (%)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c9\\\"\\u003e \\u003cp\\u003eDeclined n (%)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c10\\\"\\u003e \\u003cp\\u003eNo Change n (%)\\u003c/p\\u003e \\u003c/th\\u003e \\u003c/tr\\u003e \\u003c/thead\\u003e \\u003ctbody\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eURTI\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e1.80\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.11\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e1.70\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;1.97\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003e\\u0026minus;0.10\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.70\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003e\\u0026minus;0.63\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003e0.529\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003e\\u0026minus;0.049\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c8\\\"\\u003e \\u003cp\\u003e103 (38.0)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c9\\\"\\u003e \\u003cp\\u003e117 (43.2)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c10\\\"\\u003e \\u003cp\\u003e51 (18.8)\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eUTI\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e0.79\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.07\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e1.37\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.35\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003e0.58\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;3.18\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003e3.00\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003e0.003**\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003e0.280\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c8\\\"\\u003e \\u003cp\\u003e135 (49.8)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c9\\\"\\u003e \\u003cp\\u003e92 (34.0)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c10\\\"\\u003e \\u003cp\\u003e44 (16.2)\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eDiarrhoea\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e3.06\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.41\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e2.98\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;2.17\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\"\\u0026plusmn;\\\" colname=\\\"c4\\\"\\u003e \\u003cp\\u003e\\u0026minus;0.08\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;3.20\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c5\\\"\\u003e \\u003cp\\u003e\\u0026minus;0.40\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c6\\\"\\u003e \\u003cp\\u003e0.690\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c7\\\"\\u003e \\u003cp\\u003e\\u0026minus;0.032\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c8\\\"\\u003e \\u003cp\\u003e106 (39.1)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c9\\\"\\u003e \\u003cp\\u003e105 (38.7)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c10\\\"\\u003e \\u003cp\\u003e60 (22.1)\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003c/tbody\\u003e \\u003c/colgroup\\u003e \\u003c/table\\u003e\\u003c/div\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003cem\\u003e**p\\u0026thinsp;\\u0026lt;\\u0026thinsp;0.01. d\\u0026thinsp;=\\u0026thinsp;Cohen's d. Proportions of individual learners who improved, declined, or remained unchanged.\\u003c/em\\u003e \\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec14\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eIndividual-Level Heterogeneity of Response\\u003c/h2\\u003e \\u003cp\\u003eAnalysis of individual-level response patterns revealed substantial heterogeneity across all three case domains. Across cases, the proportion of learners showing improvement ranged from 38.0% (URTI) to 49.8% (UTI), while the proportion of decliners ranged from 34.0% (UTI) to 43.2% (URTI). The UTI case exhibited the most favourable individual response profile, with nearly half of all learners demonstrating improvement. Notably, 16.2\\u0026ndash;22.1% of learners showed no change across the three domains, potentially reflecting floor or ceiling effects in pre-existing knowledge (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig2\\\" class=\\\"InternalRef\\\"\\u003e2\\u003c/span\\u003e).\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec15\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCase-Wise Improvement in Treatment Decisions\\u003c/h2\\u003e \\u003cp\\u003eThe forest plot depicting risk differences in appropriate treatment decisions (Node-2) across the three cases (Fig.\\u0026nbsp;\\u003cspan refid=\\\"Fig3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003e) demonstrated a consistent pattern with the SCT findings. The UTI case showed the largest positive risk difference (+\\u0026thinsp;0.059; 95% CI: \\u0026minus;0.004 to 0.122), suggesting a trend toward improved treatment decision-making, although the confidence interval crossed zero. The URTI case showed a small negative risk difference (\\u0026minus;\\u0026thinsp;0.018; 95% CI: \\u0026minus;0.072 to 0.035), and the diarrhoea case showed an essentially null risk difference (\\u0026minus;\\u0026thinsp;0.011; 95% CI: \\u0026minus;0.064 to 0.042).\\u003c/p\\u003e \\u003cp\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec16\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eUsability Assessment\\u003c/h2\\u003e \\u003cp\\u003eEvaluation of platform usability using the System Usability Scale (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab3\\\" class=\\\"InternalRef\\\"\\u003e3\\u003c/span\\u003e) revealed that most learners rated the DCBL platform as Average to Good. Approximately 39.1% (n\\u0026thinsp;=\\u0026thinsp;106) rated usability as Good (SUS\\u0026thinsp;\\u0026gt;\\u0026thinsp;70), 36.9% (n\\u0026thinsp;=\\u0026thinsp;100) as Average (SUS 51\\u0026ndash;70), and 24.0% (n\\u0026thinsp;=\\u0026thinsp;65) as Poor (SUS\\u0026thinsp;\\u0026le;\\u0026thinsp;50). Taken together, over three-quarters (76.0%) of the cohort found the platform's usability acceptable or better, meeting the conventional threshold for adequate system usability.\\u003c/p\\u003e \\u003cp\\u003e \\u003cdiv class=\\\"gridtable\\\"\\u003e\\u003ctable float=\\\"Yes\\\" id=\\\"Tab3\\\" border=\\\"1\\\"\\u003e \\u003ccaption language=\\\"En\\\"\\u003e \\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 3\\u003c/div\\u003e \\u003cdiv class=\\\"CaptionContent\\\"\\u003e \\u003cp\\u003e\\u003cem\\u003eUsability Assessment: System Usability Scale (SUS) Categories (N\\u0026thinsp;=\\u0026thinsp;271)\\u003c/em\\u003e\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/caption\\u003e \\u003ccolgroup cols=\\\"3\\\"\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c1\\\" colnum=\\\"1\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c2\\\" colnum=\\\"2\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e \\u003cthead\\u003e \\u003ctr\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eSUS Category\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003en\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e%\\u003c/p\\u003e \\u003c/th\\u003e \\u003c/tr\\u003e \\u003c/thead\\u003e \\u003ctbody\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003ePoor (\\u0026le;\\u0026thinsp;50)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e65\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e24.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eAverage (51\\u0026ndash;70)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e100\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e36.9\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eGood (\\u0026gt;\\u0026thinsp;70)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e106\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e39.1\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003c/tbody\\u003e \\u003c/colgroup\\u003e \\u003c/table\\u003e\\u003c/div\\u003e \\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec17\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eCognitive Workload\\u003c/h2\\u003e \\u003cp\\u003eSelf-reported cognitive workload was moderate overall (Table\\u0026nbsp;\\u003cspan refid=\\\"Tab4\\\" class=\\\"InternalRef\\\"\\u003e4\\u003c/span\\u003e). The median mental effort score was 4.0 (IQR: 3.0\\u0026ndash;5.0) on a 7-point scale, indicating moderate cognitive engagement. The median frustration score was 3.0 (IQR: 2.0\\u0026ndash;4.0), suggesting relatively low levels of frustration. These findings indicate that the DCBL module imposed an acceptable cognitive load on learners without inducing excessive frustration, an important consideration for the sustained implementation of technology-enhanced learning interventions.\\u003c/p\\u003e \\u003cp\\u003e \\u003cdiv class=\\\"gridtable\\\"\\u003e\\u003ctable float=\\\"Yes\\\" id=\\\"Tab4\\\" border=\\\"1\\\"\\u003e \\u003ccaption language=\\\"En\\\"\\u003e \\u003cdiv class=\\\"CaptionNumber\\\"\\u003eTable 4\\u003c/div\\u003e \\u003cdiv class=\\\"CaptionContent\\\"\\u003e \\u003cp\\u003e\\u003cem\\u003eSelf-Reported Cognitive Workload: Short NASA Task Load Index (N\\u0026thinsp;=\\u0026thinsp;271)\\u003c/em\\u003e\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/caption\\u003e \\u003ccolgroup cols=\\\"3\\\"\\u003e \\u003cdiv align=\\\"left\\\" class=\\\"colspec\\\" colname=\\\"c1\\\" colnum=\\\"1\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c2\\\" colnum=\\\"2\\\"\\u003e\\u003c/div\\u003e \\u003cdiv align=\\\"char\\\" char=\\\".\\\" class=\\\"colspec\\\" colname=\\\"c3\\\" colnum=\\\"3\\\"\\u003e\\u003c/div\\u003e \\u003cthead\\u003e \\u003ctr\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eMetric\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003eMental Effort (1\\u0026ndash;7)\\u003c/p\\u003e \\u003c/th\\u003e \\u003cth align=\\\"left\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003eFrustration (1\\u0026ndash;7)\\u003c/p\\u003e \\u003c/th\\u003e \\u003c/tr\\u003e \\u003c/thead\\u003e \\u003ctbody\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eMedian\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e4.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e3.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eQ1 (25th percentile)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e3.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e2.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eQ3 (75th percentile)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e5.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e4.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003ctr\\u003e \\u003ctd align=\\\"left\\\" colname=\\\"c1\\\"\\u003e \\u003cp\\u003eIQR (Q3\\u0026ndash;Q1)\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c2\\\"\\u003e \\u003cp\\u003e2.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003ctd align=\\\"char\\\" char=\\\".\\\" colname=\\\"c3\\\"\\u003e \\u003cp\\u003e2.0\\u003c/p\\u003e \\u003c/td\\u003e \\u003c/tr\\u003e \\u003c/tbody\\u003e \\u003c/colgroup\\u003e \\u003c/table\\u003e\\u003c/div\\u003e \\u003c/p\\u003e \\u003c/div\\u003e\"},{\"header\":\"DISCUSSION\",\"content\":\"\\u003cp\\u003eThis study examined the impact of a single-session, AI-assisted Digital Case-Based Learning intervention on SCT-measured clinical reasoning in antibiotic stewardship among MBBS learners. The principal finding was a statistically significant, albeit small, improvement in overall SCT performance following the DCBL exposure (mean change\\u0026thinsp;=\\u0026thinsp;0.82, \\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.04). This improvement, while modest in magnitude indicates a modest but potentially educationally relevant improvement given the brevity of the intervention (a single 45\\u0026ndash;60-minute session), the challenging nature of the outcome being measured (clinical reasoning under uncertainty rather than factual recall), and the well-documented resistance of higher-order cognitive skills to short-duration educational exposures.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec19\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eContextualising the Overall Improvement\\u003c/h2\\u003e \\u003cp\\u003eThe small effect size observed in this study is consistent with findings from analogous interventions in medical education. Cook and colleagues, in their meta-analysis of internet-based learning in the health professions, reported that technology-enhanced learning interventions typically produce small-to-moderate effects on knowledge and clinical reasoning, with larger gains accruing from repeated, spaced exposures rather than single encounters [\\u003cspan citationid=\\\"CR23\\\" class=\\\"CitationRef\\\"\\u003e23\\u003c/span\\u003e]. The SCT, by design, assesses a particularly elusive competence, the capacity to modulate clinical decisions in response to new, uncertain information, which is unlikely to shift dramatically after a brief educational exposure. Naylor et al., evaluating SCT performance among paramedic students, similarly observed that the formative value of the SCT often exceeds its sensitivity to short-term interventions, and recommended regular exposure to facilitate familiarity and deeper script development[\\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e]. Our finding that overall SCT scores improved significantly, even from a single session, thus provides encouraging preliminary evidence that AI-assisted DCBL can begin to activate the cognitive processes that underpin clinical reasoning in antibiotic stewardship.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec20\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eDifferential Case-Specific Effects\\u003c/h2\\u003e \\u003cp\\u003ePerhaps the most instructive finding was the marked heterogeneity in intervention effects across the three clinical domains. The UTI case demonstrated a statistically significant improvement (\\u003cem\\u003ep\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.003, Cohen's \\u003cem\\u003ed\\u003c/em\\u003e\\u0026thinsp;=\\u0026thinsp;0.280) with the highest proportion of individual improvers (49.8%), while URTI and diarrhoea cases showed no significant change. Several factors may explain this differential response.\\u003c/p\\u003e \\u003cp\\u003eFirst, the UTI case may have offered the greatest scope for learning because baseline performance was the lowest among the three domains (pre-test mean\\u0026thinsp;=\\u0026thinsp;0.79), leaving more room for improvement. This is consistent with the well-recognised floor-and-ceiling phenomenon in educational measurement, where learners with the lowest initial performance tend to benefit most from targeted instruction [\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e]. In contrast, the diarrhoea case had the highest baseline score (3.06), potentially limiting upward movement and creating a ceiling effect.\\u003c/p\\u003e \\u003cp\\u003eSecond, the clinical content itself likely contributed to the differential effect. UTI management involves relatively discrete, guideline-amenable decision points (choice of empirical antibiotic, duration of therapy, criteria for culture), which may be more amenable to the branching-logic and feedback structure of the DCBL format. In contrast, URTI management, where the primary stewardship challenge is to withhold rather than prescribe antibiotics, requires a different kind of reasoning that involves resisting prescribing pressure and managing diagnostic uncertainty about bacterial versus viral aetiology. This \\\"withholding\\\" decision may be less responsive to the type of structured feedback provided in the micro-cases. The challenge of improving antibiotic stewardship for URTIs has been documented in other settings as well, where even multi-faceted interventions required sustained effort to reduce URTI antibiotic prescribing from 62.6% to 7.2% [\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e]. Similarly, diarrhoea management, while conceptually straightforward (oral rehydration as first-line), involves nuanced decisions about the role of antibiotics in dysenteric versus non-dysenteric presentations that may require deeper clinical exposure than a single digital session can provide.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec21\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eIndividual-Level Heterogeneity\\u003c/h2\\u003e \\u003cp\\u003eThe substantial heterogeneity observed at the individual level, with 38\\u0026ndash;49.8% of learners improving, 34\\u0026ndash;43.2% declining, and 16\\u0026ndash;22.1% showing no change across cases, warrants careful consideration. This pattern is not unusual in medical education research; clinical reasoning development is highly individual, shaped by prior knowledge structures, motivation, and metacognitive capacity [\\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e]. The proportion of \\\"decliners\\\" may partly reflect regression to the mean, particularly among learners whose pre-test scores were inflated by guessing. It may also indicate that the DCBL intervention, while activating new scripts for some learners, may have introduced cognitive interference in others, a phenomenon where new information temporarily disrupts existing knowledge structures before reorganisation occurs [\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eFrom a practical standpoint, the mixed individual response profile argues strongly for iterative, spaced DCBL exposure rather than a one-off session. Educational research consistently demonstrates that distributed practice produces more durable learning gains than massed practice, particularly for complex cognitive skills [\\u003cspan citationid=\\\"CR28\\\" class=\\\"CitationRef\\\"\\u003e28\\u003c/span\\u003e]. Incorporating DCBL micro-cases into a longitudinal curriculum, perhaps revisiting the same clinical domains at increasing levels of complexity over successive weeks, would likely amplify the effects observed in this study and provide opportunities for targeted scaffolding for learners who do not respond to initial exposure.\\u003c/p\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec22\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eUsability and Cognitive Workload\\u003c/h2\\u003e \\u003cp\\u003eThe usability findings provide important context for interpreting the learning outcomes and for planning future implementation. With 76% of learners rating the platform as Average to Good on the SUS, the DCBL module meets the generally accepted threshold for adequate usability in educational technology [\\u003cspan citationid=\\\"CR29\\\" class=\\\"CitationRef\\\"\\u003e29\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR30\\\" class=\\\"CitationRef\\\"\\u003e30\\u003c/span\\u003e]. This is noteworthy because poor usability can confound educational outcomes if learners struggle with the interface, diverting cognitive resources from the learning task to system navigation and potentially undermining the intended educational effect [\\u003cspan citationid=\\\"CR31\\\" class=\\\"CitationRef\\\"\\u003e31\\u003c/span\\u003e]. The moderate mental effort (median\\u0026thinsp;=\\u0026thinsp;4.0/7) and low frustration (median\\u0026thinsp;=\\u0026thinsp;3.0/7) reported on the NASA-TLX items further suggest that the DCBL module achieved an appropriate balance between cognitive challenge and accessibility, a critical consideration under cognitive load theory [\\u003cspan citationid=\\\"CR32\\\" class=\\\"CitationRef\\\"\\u003e32\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eHowever, the 24% of learners who rated usability as Poor cannot be dismissed. Digital literacy varies considerably among Indian medical students, and interface design choices that seem intuitive to one subgroup may create barriers for another. Future iterations should incorporate user experience testing with low-performing subgroups to identify specific usability pain points and ensure equitable access to the learning experience.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec23\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eImplications for CBME and Antibiotic Stewardship Education\\u003c/h2\\u003e \\u003cp\\u003eThis study carries several implications for the evolving CBME landscape in India. First, it demonstrates that digital, AI-enhanced educational tools can be successfully integrated into routine CBME teaching to address competencies that are difficult to teach through lectures alone, specifically, clinical decision-making under uncertainty in antibiotic stewardship. The DCBL format aligns naturally with several CBME principles: it activates authentic clinical reasoning, provides just-in-time formative feedback, yields individual-level analytics for programmatic assessment, and can be delivered at scale across institutions [\\u003cspan citationid=\\\"CR33\\\" class=\\\"CitationRef\\\"\\u003e33\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR34\\\" class=\\\"CitationRef\\\"\\u003e34\\u003c/span\\u003e].\\u003c/p\\u003e \\u003cp\\u003eSecond, the content-specific findings point to the need for domain-tailored instructional design. The success of the UTI case, relative to URTI and diarrhoea, suggests that the branching-logic, feedback-rich DCBL format is best suited to clinical scenarios with clear decision trees and guideline-concordant endpoints. For scenarios like URTI, where the stewardship challenge primarily involves withholding treatment, complementary strategies, such as communication skills training, shared decision-making simulations, or reflective debriefing, may be needed alongside the DCBL module.\\u003c/p\\u003e \\u003cp\\u003eThird, from an antimicrobial stewardship perspective, the study reinforces the importance of engaging learners early in their training. The formation of prescribing habits begins in medical school and is shaped by clinical interactions and educational experiences [\\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e]. Embedding stewardship reasoning exercises into the undergraduate curriculum, rather than deferring to postgraduate training, can help establish appropriate prescribing patterns before suboptimal practices become entrenched.\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e \\u003cdiv id=\\\"Sec24\\\" class=\\\"Section2\\\"\\u003e \\u003ch2\\u003eStrengths and Limitations\\u003c/h2\\u003e \\u003cp\\u003eThe strengths of this study include a substantial sample size (N\\u0026thinsp;=\\u0026thinsp;271), the use of a validated clinical reasoning instrument (SCT) with expert-panel scoring, the reporting of both statistical significance and effect sizes, the evaluation of individual-level heterogeneity, and the inclusion of usability and cognitive workload assessments that are often omitted from educational intervention studies.\\u003c/p\\u003e \\u003cp\\u003eSeveral limitations must be acknowledged. The single-group, pre\\u0026ndash;post design without a control group precludes causal inference, as observed improvements may reflect testing effects, maturation, or regression to the mean, despite the use of parallel SCT items. The single-session design limits conclusions about the durability of learning gains. The convenience sample from a single institution may constrain generalisability, and the absence of follow-up assessment means we cannot comment on retention of improvements over time. The URTI and diarrhoea cases showed no significant change, suggesting that the UTI domain largely drove the overall SCT improvement. Finally, the study was conducted within a specific CBME context in South India, and findings should be replicated in diverse institutional and geographical settings with varying baseline knowledge levels and curricular structures.\\u003c/p\\u003e \\u003cdiv id=\\\"Sec25\\\" class=\\\"Section3\\\"\\u003e \\u003ch2\\u003eFuture Directions\\u003c/h2\\u003e \\u003cp\\u003eFuture research should employ randomised controlled designs with active comparator groups to isolate the specific contribution of AI-assisted feedback from the case-based learning itself. Longitudinal studies with spaced DCBL sessions and delayed post-tests would clarify whether the observed improvements are durable and whether repeated exposure amplifies the effect. Adaptive AI algorithms that tailor feedback intensity and content to individual learner performance profiles could address the heterogeneity observed in this study. Additionally, aligning case content to local antibiograms and resistance patterns would enhance ecological validity and clinical relevance. Multisite studies across diverse Indian medical colleges would strengthen the evidence base for broader CBME integration.\\u003c/p\\u003e \\u003c/div\\u003e \\u003c/div\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eEthics approval and consent to participate:\\u003c/strong\\u003e Ethical approval for this study was obtained from the Institutional Ethics Committee, JSS Medical College, JSS Academy of Higher Education and Research, Mysuru, Karnataka, India (NABH Accredited) (Approval No. JSSMC/IEC/972025/01 NCT/2024-25; dated 10 July 2025). Participation was voluntary. Informed consent was obtained from all participants prior to inclusion in the study.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eConsent for publication\\u003c/strong\\u003e: Not applicable.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eData availability\\u003c/strong\\u003e: The datasets generated and/or analysed during the current study are not publicly available because they form part of an ongoing educational research dataset, but de-identified data may be made available from the corresponding author on reasonable request.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCompeting interests\\u003c/strong\\u003e: The authors declare that they have no competing interests.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eFunding\\u003c/strong\\u003e: This research received no external funding.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAuthors\\u0026rsquo; contributions:\\u003c/strong\\u003e Sunil Kumar D conceptualised the study, conducted the study, developed the methodology, and performed the results analysis. Sucheta Dandekar supervised the work and critically reviewed the manuscript. G Hari Prakash contributed to manuscript preparation and interpretation of the results. All authors read and approved the final manuscript.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAcknowledgements:\\u003c/strong\\u003e The authors thank the participating students for their engagement in the study.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\u003cli\\u003e\\u003cspan\\u003eKassirer JP. Teaching clinical reasoning: Case-based and coached. Acad Med. 2010;85:1118\\u0026ndash;24. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1097/ACM.0b013e3181d5dd0d\\u003c/span\\u003e\\u003cspan address=\\\"10.1097/ACM.0b013e3181d5dd0d\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eCharlin B, Roy L, Brailovsky C, Goulet F, van der Vleuten C. The script concordance test: A tool to assess the reflective clinician. Teach Learn Med. 2000;12:189\\u0026ndash;95. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1207/S15328015TLM1204_5\\u003c/span\\u003e\\u003cspan address=\\\"10.1207/S15328015TLM1204_5\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eLubarsky S, Dory V, Duggan P, Gagnon R, Charlin B. Script concordance testing: From theory to practice: AMEE Guide 75. Med Teach. 2013;35:184\\u0026ndash;93. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.3109/0142159X.2013.760036\\u003c/span\\u003e\\u003cspan address=\\\"10.3109/0142159X.2013.760036\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eIndian Council of Medical Research. Annual report: Antimicrobial Resistance Surveillance and Research Network (AMRSN). 2023. ICMR; 2024.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eLaxminarayan R, Chaudhury RR. Antibiotic resistance in India: Drivers and opportunities for action. PLoS Med. 2016;13:e1001974. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1371/journal.pmed.1001974\\u003c/span\\u003e\\u003cspan address=\\\"10.1371/journal.pmed.1001974\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eGoebel MC, Trautner BW, Grigoryan L. The five Ds of outpatient antibiotic stewardship for urinary tract infections. Clin Microbiol Rev. 2021;34:e00003\\u0026ndash;20. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1128/CMR.00003-20\\u003c/span\\u003e\\u003cspan address=\\\"10.1128/CMR.00003-20\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBeck AP, Baubie K, Knobloch MJ, Safdar N. Promoting antimicrobial stewardship by incorporating it in undergraduate medical education curricula. Wis Med J. 2018;117:224\\u0026ndash;8.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eWang EY, Siddiqi H, Engmann CN, Kang YS, Yin HS. Development of a multifaceted antimicrobial stewardship curriculum for undergraduate medical education. Open Forum Infect Dis. 2021;8:ofab154. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1093/ofid/ofab154\\u003c/span\\u003e\\u003cspan address=\\\"10.1093/ofid/ofab154\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eNational Medical Commission. Competency-based medical education (CBME) curriculum guidelines for the Indian medical graduate. Government of India; 2023.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSoundariya K, Nishanthi A, Mahendran R, Vimal M. Evaluation of competency-based medical education (CBME) curriculum implementation for Phase II medical undergraduates: A qualitative study. J Adv Med Educ Professionalism. 2025;13:36\\u0026ndash;48. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.30476/jamp.2024.104312.2052\\u003c/span\\u003e\\u003cspan address=\\\"10.30476/jamp.2024.104312.2052\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSingh K, Gupta P, Dhaliwal U, Kochhar K. Competency-based medical education in India: Clinicians\\u0026rsquo; multifaceted challenges. Indian J Otolaryngol Head Neck Surg. 2024;76:3741\\u0026ndash;9. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1007/s12070-024-04751-x\\u003c/span\\u003e\\u003cspan address=\\\"10.1007/s12070-024-04751-x\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eCook DA, Levinson AJ, Garside S, Dupras DM, Erwin PJ, Montori VM. Instructional design variations in internet-based learning for health professions education: A systematic review and meta-analysis. Acad Med. 2010;85:909\\u0026ndash;22. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1097/ACM.0b013e3181d6c319\\u003c/span\\u003e\\u003cspan address=\\\"10.1097/ACM.0b013e3181d6c319\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eStevens NT, Bruen C, Boland F, Pawlikowska T, Fitzpatrick F, Humphreys H. Is online case-based learning effective in helping undergraduate medical students choose the appropriate antibiotics to treat important infections? JAC-Antimicrobial Resist. 2019;1:dlz081. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1093/jacamr/dlz081\\u003c/span\\u003e\\u003cspan address=\\\"10.1093/jacamr/dlz081\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eGarg T, Shrigiriwar A, Garg A. A scoping review of artificial intelligence in medical education: BEME Guide 84. Med Teach. 2024;46:446\\u0026ndash;70. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1080/0142159X.2024.2314198\\u003c/span\\u003e\\u003cspan address=\\\"10.1080/0142159X.2024.2314198\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMasters K. Ethical use of artificial intelligence in health professions education: AMEE Guide 158. Med Teach. 2023;45:574\\u0026ndash;84. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1080/0142159X.2023.2186203\\u003c/span\\u003e\\u003cspan address=\\\"10.1080/0142159X.2023.2186203\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eFournier JP, Demeester A, Charlin B. Script concordance tests: Guidelines for construction. BMC Med Inf Decis Mak. 2008;8:18. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1186/1472-6947-8-18\\u003c/span\\u003e\\u003cspan address=\\\"10.1186/1472-6947-8-18\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eLubarsky S, Charlin B, Cook DA, Chalk C, van der Vleuten CP. Script concordance testing: A review of published validity evidence. Med Educ. 2011;45:329\\u0026ndash;38. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1111/j.1365-2923.2010.03863.x\\u003c/span\\u003e\\u003cspan address=\\\"10.1111/j.1365-2923.2010.03863.x\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBoulouffe C, Doucet B, Muschart X, Charlin B, Vanpee D. Assessing clinical reasoning using a script concordance test with electrocardiogram in an emergency medicine clerkship rotation. Emerg Med J. 2013;31:313\\u0026ndash;6. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1136/emermed-2012-201737\\u003c/span\\u003e\\u003cspan address=\\\"10.1136/emermed-2012-201737\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eNaylor K, Hislop J, Torres K, Mani ZA, Goniewicz K. The impact of script concordance testing on clinical decision-making in paramedic education. Healthcare. 2024;12:282. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.3390/healthcare12020282\\u003c/span\\u003e\\u003cspan address=\\\"10.3390/healthcare12020282\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eRoss L, Semaan E, Gosling CM, Fisk B, Shannon B. Clinical reasoning in undergraduate paramedicine: Utilisation of a script concordance test. BMC Med Educ. 2023;23:39. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1186/s12909-023-04020-x\\u003c/span\\u003e\\u003cspan address=\\\"10.1186/s12909-023-04020-x\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eMalli IA, Mohamud MS, Al-Nasser SS. Enhancing medical students\\u0026rsquo; confidence and knowledge in antibiotic prescription and administration through virtual education: A quasi-experimental study. Antibiotics. 2023;12:1546. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.3390/antibiotics12101546\\u003c/span\\u003e\\u003cspan address=\\\"10.3390/antibiotics12101546\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSilverberg SL, Zannella VE, Engel S, Engel AS. Implementing an antimicrobial stewardship medical student elective to impart good antibiotic prescribing habits early in medical training. Antimicrob Stewardship Healthc Epidemiol. 2023;3:e225. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1017/ash.2023.491\\u003c/span\\u003e\\u003cspan address=\\\"10.1017/ash.2023.491\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eCook DA, Levinson AJ, Garside S. Internet-based learning in the health professions: A meta-analysis. JAMA. 2008;300:1181\\u0026ndash;96. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1001/jama.300.10.1181\\u003c/span\\u003e\\u003cspan address=\\\"10.1001/jama.300.10.1181\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eCook DA, Artino AR. Motivation to learn: An overview of contemporary theories. Med Educ. 2016;50:997\\u0026ndash;1014. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1111/medu.13074\\u003c/span\\u003e\\u003cspan address=\\\"10.1111/medu.13074\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eGulliford MC, Dregan A, Moore MV, Ashworth M, van Staa T, McCann G, et al. Continued high rates of antibiotic prescribing to adults with respiratory tract infection: Survey of 568 UK general practices. BMJ Open. 2014;4:e006245. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1136/bmjopen-2014-006245\\u003c/span\\u003e\\u003cspan address=\\\"10.1136/bmjopen-2014-006245\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eEva KW. What every teacher needs to know about clinical reasoning. Med Educ. 2005;39:98\\u0026ndash;106. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1111/j.1365-2929.2004.01972.x\\u003c/span\\u003e\\u003cspan address=\\\"10.1111/j.1365-2929.2004.01972.x\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSchmidt HG, Rikers RM. How expertise develops in medicine: Knowledge encapsulation and illness script formation. Med Educ. 2007;41:1133\\u0026ndash;9. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1111/j.1365-2923.2007.02915.x\\u003c/span\\u003e\\u003cspan address=\\\"10.1111/j.1365-2923.2007.02915.x\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eDunlosky J, Rawson KA, Marsh EJ, Nathan MJ, Willingham DT. Improving students\\u0026rsquo; learning with effective learning techniques. Psychol Sci Public Interest. 2013;14:4\\u0026ndash;58. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1177/1529100612453266\\u003c/span\\u003e\\u003cspan address=\\\"10.1177/1529100612453266\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eBrooke J. SUS: A retrospective. J Usability Stud. 2013;8:29\\u0026ndash;40.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eVlachogianni P, Tselios N. Perceived usability evaluation of educational technology using the System Usability Scale (SUS): A systematic review. J Res Technol Educ. 2022;54:392\\u0026ndash;409. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1080/15391523.2020.1867938\\u003c/span\\u003e\\u003cspan address=\\\"10.1080/15391523.2020.1867938\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSweller J, van Merri\\u0026euml;nboer JJG, Paas F. Cognitive architecture and instructional design: 20 years later. Educational Psychol Rev. 2019;31:261\\u0026ndash;92. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.1007/s10648-019-09465-5\\u003c/span\\u003e\\u003cspan address=\\\"10.1007/s10648-019-09465-5\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eSweller J. Cognitive load theory. In: Mestre JP, Ross BH, editors. Psychology of learning and motivation. Academic; 2011. pp. 37\\u0026ndash;76.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003eTen Cate O. Nuts and bolts of entrustable professional activities. J Graduate Med Educ. 2013;5:157\\u0026ndash;8. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.4300/JGME-D-12-00380.1\\u003c/span\\u003e\\u003cspan address=\\\"10.4300/JGME-D-12-00380.1\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e \\u003cli\\u003e\\u003cspan\\u003evan der Vleuten CPM, Schuwirth LWT, Driessen EW, Dijkstra J, Tigelaar D, Baartman LKJ, et al. A model for programmatic assessment fit for purpose. Med Teach. 2012;34:205\\u0026ndash;14. \\u003cspan class=\\\"ExternalRef\\\"\\u003e\\u003cspan class=\\\"RefSource\\\"\\u003ehttps://doi.org/10.3109/0142159X.2012.652239\\u003c/span\\u003e\\u003cspan address=\\\"10.3109/0142159X.2012.652239\\\" targettype=\\\"DOI\\\" class=\\\"RefTarget\\\"\\u003e\\u003c/span\\u003e\\u003c/span\\u003e.\\u003c/span\\u003e\\u003c/li\\u003e\\u003c/ol\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"bmc-medical-education\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"meed\",\"sideBox\":\"Learn more about [BMC Medical Education](http://bmcmededuc.biomedcentral.com/)\",\"snPcode\":\"\",\"submissionUrl\":\"https://www.editorialmanager.com/meed/default.aspx\",\"title\":\"BMC Medical Education\",\"twitterHandle\":\"BMC_series\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"em\",\"reportingPortfolio\":\"BMC Series\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Script Concordance Test, Antimicrobial Stewardship, Medical Education, Artificial Intelligence, Case-Based Learning\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-9384886/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-9384886/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003ch2\\u003eBackground\\u003c/h2\\u003e \\u003cp\\u003eAntimicrobial resistance is a major global public health threat, necessitating improved antibiotic stewardship. The introduction of Competency-Based Medical Education (CBME) in India requires robust methods to assess and cultivate clinical reasoning. This study evaluates an AI-assisted Digital Case-Based Learning (DCBL) module intended to improve antibiotic stewardship decisions among medical undergraduates.\\u003c/p\\u003e\\u003ch2\\u003eMethods\\u003c/h2\\u003e \\u003cp\\u003eA quasi-experimental, single-group pre-post study evaluated the impact of three AI-assisted DCBL micro-cases (identifying upper respiratory tract infection [URTI], urinary tract infection [UTI], and acute watery diarrhoea) on Script Concordance Test (SCT) performance in 271 MBBS learners. Usability and workload were measured using the System Usability Scale (SUS) and NASA Task Load Index.\\u003c/p\\u003e\\u003ch2\\u003eResults\\u003c/h2\\u003e \\u003cp\\u003eOverall SCT scores improved significantly from 5.65\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.73 to 6.47\\u0026thinsp;\\u0026plusmn;\\u0026thinsp;4.48 (p\\u0026thinsp;=\\u0026thinsp;0.04), indicating enhanced clinical reasoning. The UTI case showed the greatest improvement (p\\u0026thinsp;=\\u0026thinsp;0.003). Over 76% of learners rated usability as Good/Average, with moderate cognitive effort.\\u003c/p\\u003e\\u003ch2\\u003eConclusion\\u003c/h2\\u003e \\u003cp\\u003eA single-session AI-assisted DCBL intervention improved SCT-measured clinical reasoning for antibiotic stewardship. Targeted domain-specific design is needed to address heterogeneous intervention effects.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Dynamic CBME in Action: AI-Assisted Digital Case-Based Learning to Improve Antibiotic-Stewardship Decisions in MBBS Learners\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2026-05-11 05:55:40\",\"doi\":\"10.21203/rs.3.rs-9384886/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0},{\"type\":\"reviewerAgreed\",\"content\":\"327290674154646850284307955706431614258\",\"date\":\"2026-05-17T03:22:43+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2026-05-04T04:16:48+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"257048767017993210812961201029800330695\",\"date\":\"2026-05-04T03:42:26+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"46971518456256746939114364853182357309\",\"date\":\"2026-04-28T14:39:32+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewersInvited\",\"content\":\"\",\"date\":\"2026-04-28T13:23:00+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorInvited\",\"content\":\"\",\"date\":\"2026-04-20T18:39:02+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorAssigned\",\"content\":\"\",\"date\":\"2026-04-18T01:45:58+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"checksComplete\",\"content\":\"\",\"date\":\"2026-04-18T01:45:29+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"submitted\",\"content\":\"BMC Medical Education\",\"date\":\"2026-04-11T05:38:36+00:00\",\"index\":\"\",\"fulltext\":\"\"}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"bmc-medical-education\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"meed\",\"sideBox\":\"Learn more about [BMC Medical Education](http://bmcmededuc.biomedcentral.com/)\",\"snPcode\":\"\",\"submissionUrl\":\"https://www.editorialmanager.com/meed/default.aspx\",\"title\":\"BMC Medical Education\",\"twitterHandle\":\"BMC_series\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"em\",\"reportingPortfolio\":\"BMC Series\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"ccd6d267-3705-4d1e-a6e6-31ad2f6aa2ef\",\"owner\":[],\"postedDate\":\"May 11th, 2026\",\"published\":true,\"recentEditorialEvents\":[{\"type\":\"reviewerAgreed\",\"content\":\"327290674154646850284307955706431614258\",\"date\":\"2026-05-17T03:22:43+00:00\",\"index\":57,\"fulltext\":\"\"},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2026-05-04T04:16:48+00:00\",\"index\":48,\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"257048767017993210812961201029800330695\",\"date\":\"2026-05-04T03:42:26+00:00\",\"index\":47,\"fulltext\":\"\"}],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"under-review\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2026-05-11T05:55:41+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2026-05-11 05:55:40\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-9384886\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-9384886\",\"identity\":\"rs-9384886\",\"version\":[\"v1\"]},\"buildId\":\"XKTyCvWXoU3ODBz1xrDgd\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}