The problem the researchers set out to address is deceptively simple to state but difficult to solve. Community Medicine, known in many countries as preventive and social medicine, asks students to do something that traditional clinical teaching rarely demands: to hold clinical reasoning, epidemiological interpretation, the social determinants of health, national health-programme linkage and preventive decision-making in mind simultaneously. A student might need to manage a child with pneumonia while also calculating disease burden, tracing immunisation coverage through a national programme and deciding what preventive action the health system should take. Digital case-based learning, in which students work through interactive, realistic scenarios on screen, is widely seen as a promising way to teach this integration. But the literature offered little guidance on how such scenarios should actually be designed, delivered and assessed in this specific discipline, leaving individual faculty to improvise.

The methodological choice was the Delphi technique, a structured consensus method in which experts answer questionnaires in successive rounds, with feedback designed to move the group toward agreement. In its classic form, the Delphi process involves multiple rating rounds with controlled feedback between them, allowing participants to see the group’s responses and revise their own. The Mysuru team used a modified variant that compresses this into two distinct phases. In the first round, conducted between October and December 2025, six purposively selected Community Medicine faculty members answered an open-ended questionnaire designed purely to generate ideas. The researchers then applied inductive thematic analysis to those free-text responses, coding them without a predetermined framework and allowing themes to emerge from the data itself. That analysis produced a rating instrument of twenty-three items organised across seven domains, covering everything from which topics deserve priority to which digital formats and assessment methods belong in a well-designed digital case.

Round 2 shifted from generation to evaluation. Ten experts rated each of the twenty-three items once on a five-point relevance scale. Crucially, the researchers did not undertake controlled feedback or re-rating; each expert judged the items independently in a single pass. Consensus was defined in advance using three strict criteria: the median rating had to be at least 4, the interquartile range had to be no greater than 1, and at least 80 percent of experts had to rate the item 4 or 5. Setting these thresholds a priori is methodologically important, because consensus definitions chosen after the data are seen can be unconsciously tailored to flatter the results. By committing to all three criteria before ratings were collected, the team made their agreement standard transparent and reproducible. Response rates were 100 percent in both rounds, a notable strength given that attrition is a common weakness of Delphi studies.

The headline finding is that nineteen of the twenty-three items, or 82.6 percent, met all three predefined consensus criteria. Six of those nineteen met the agreement criterion exactly at the boundary, with eight of ten experts rating them 4 or 5, meaning the margin of agreement was narrow for roughly a third of the endorsed elements. Consensus was complete, meaning every item in the domain passed, for four of the seven domains: priority topics, case-design features, learning outcomes and reasoning steps. This suggests that experts agree most strongly on the pedagogical substance of digital cases, what they should teach and how the reasoning within them should unfold, and less uniformly on peripheral questions of format and assessment.

The pattern of disagreement is as informative as the pattern of agreement. On digital formats, interactive branching scenarios, in which the case unfolds differently depending on the learner’s decisions, reached consensus, as did embedded video or simulated patient interviews and decision trees. Plain text-based narratives did not. The experts appear to be saying that the digital medium should be used for what it uniquely offers: interactivity, simulation and consequence, rather than as a repository for electronic handouts. On assessment, the split was equally revealing. Clinical decision justification, reflective writing or case synthesis, and integration with objective structured practical or clinical examinations all reached consensus, while multiple-choice questions and short-answer formats did not. The endorsed assessment methods share a common property: they require the learner to expose the reasoning behind a decision, not merely select a correct option. This aligns with the study’s central aim of teaching reasoning rather than recall, and it echoes a broader movement in competency-based medical education away from knowledge tests toward performance-based assessment.

A distinctive strength of the study is its explicit mapping of consensus elements to India’s national undergraduate competencies for Community Medicine, the competency codes that govern what medical students must demonstrate under the country’s competency-based medical education framework. The researchers classified each link as either explicit, where the connection to a competency is direct and stated, or interpretative, where the connection requires inference. Content elements and most reasoning elements mapped explicitly to the national competencies, meaning that a digital case built on this framework would directly serve mandated learning outcomes. Design and digital-format elements, by contrast, mapped only interpretatively, which is unsurprising since national competency documents specify what students should learn rather than the technological means of teaching it. This mapping matters practically: any institution adopting the framework can trace each design decision back to an accreditation requirement, easing the path to curricular approval.

The authors are candid about the limits of what they have produced. Because the experts rated items only once and never revisited their judgments after seeing group feedback, the stability of the agreement is untested. In a full Delphi, a second rating round after feedback would reveal whether dissenting experts moved toward the group or held firm; without that step, the nineteen consensus items represent a snapshot of independent judgment rather than a settled group position. The panel was also small, six item generators and ten raters, drawn from a single medical college, which raises questions about whether the findings generalise across India’s diverse medical education landscape. The framework is therefore described as provisional and consensus-informed rather than definitive, and the authors state plainly that it requires confirmation in a larger, multi-institutional panel and evaluation of its feasibility and its effects on learner reasoning. Reporting followed the ACCORD and CREDES guidance, the recognised standards for transparently reporting consensus studies and Delphi research, and the study received ethics approval from the Institutional Ethics Committee of JSS Medical College with informed consent from all participants.

What makes the study worth wider attention is its disciplined approach to a question that educational technology often answers by enthusiasm rather than evidence. Digital learning tools proliferate faster than the pedagogy underpinning them, and medical education is no exception. By asking experienced faculty to specify, item by item, what a digital case in Community Medicine must contain, and by holding every proposed element to a pre-registered statistical standard of agreement, the researchers have converted opinion into a structured, testable framework. The emphasis on authentic community-based cases, progressive disclosure of information, explicit clinical and public health reasoning, linkage to national health programmes, interactive formats and assessment of decision justification amounts to a design specification that developers and educators can act on immediately, while researchers test it rigorously.

The next phase of this work will determine whether the framework survives contact with reality. Multi-institutional Delphi confirmation would test whether faculty in different regions and institutional cultures converge on the same essential elements, and empirical studies would need to measure whether students taught with consensus-designed digital cases actually reason better across the clinical-public health divide than students taught conventionally. Those studies will take time. But the Mysuru team has supplied something the field lacked: a concrete, expert-vetted answer to the question of what digital case-based learning in Community Medicine should look like, complete with honest caveats about how much confidence the evidence currently supports. For a discipline that trains the physicians who will manage epidemics, run immunisation programmes and judge the health of populations, that is a meaningful contribution.

Subject of Research: Expert consensus on the design of digital case-based learning for clinical and public health reasoning in undergraduate Community Medicine

Article Title: Consensus-informed design of digital case-based learning for clinical and public health reasoning in undergraduate Community Medicine: a modified Delphi study

Article References: Kumar D, S., Dandekar, S., Shabadi, N., & Prakash, G. H. (2026). Consensus-informed design of digital case-based learning for clinical and public health reasoning in undergraduate Community Medicine: a modified Delphi study. BMC Medical Education. https://doi.org/10.1186/s12909-026-10557-4

Image Credits: AI Generated

DOI: 10.1186/s12909-026-10557-4

Keywords: Delphi technique, digital case-based learning, clinical reasoning, public health reasoning, Community Medicine, competency-based medical education, undergraduate medical education, expert consensus, medical education, curriculum design, assessment methods, India