Cognitive accessibility in diagnosis games
Table of contents
You open the case. On screen, three dense paragraphs, a red timer and five tabs blinking. Before you form your first hypothesis, your attention has already scattered. The difficulty is not just the clinical content: it is the interface asking for memory, speed and jargon at the same time. This is the territory of cognitive accessibility in diagnostic games and digital clinical cases.
Cognitive accessibility is about making study usable for people with attention difficulties, limited working memory, slower processing or high cognitive load caused by stress, short sleep or a second language. It applies to neurodivergent students and also to anyone under pressure. The focus of this text is the design of the game and of the case: language, pace, clues, feedback and noise. It is not a guide to screen readers nor a list of technical software standards. The topic of the neurodivergent student in study habits appears in neurodiversity and medical study.
Key points
- Cognitive barriers in games and clinical cases include dense language, clue overload, rigid timing, ambiguous feedback, jargon and interface distractions.
- Useful principles: clarity, chunking, consistency, flexible timing, explicit feedback, less noise and plain language in the clues.
- Working memory is limited; novices use this resource more when reasoning about cases, which makes the design of the task especially sensitive.
- Extended time and self-paced rhythm reduce assessment stress without replacing a clear design of the statement.
- In NEXO, fictional cases with a real clinical presentation, up to six progressive clues and a daily pace allow training with a more controllable load, without claiming effectiveness.
What cognitive accessibility is
Cognitive accessibility is the set of content and interface decisions that reduce barriers for people with cognitive or learning difficulties. The W3C COGA task force sums up practical goals: help people understand what each control does, find what they need, use clear text and media, avoid errors, keep focus, not rely on memory alone, offer support and allow personalization (W3C, Making Content Usable). These goals apply to web pages and also to games and interactive clinical cases.
Do not confuse it with visual or motor accessibility. A case can be readable by a screen reader and still impossible to process because the vignette mixes ten findings with no hierarchy, the feedback says only "incorrect" and the timer forces a guess. It is also not the same as lowering the clinical level: the goal is to remove friction that does not measure reasoning. The difficult content stays; the unnecessary confusion goes.
In medical education, cognitive load theory already shows that working memory is the bottleneck. Guidelines for health education describe how to reduce extraneous load (noise from the format), calibrate intrinsic load (complexity of the case) and protect germane load (schema building) (van Merriënboer and Sweller, 2010; Young et al., 2014). Cognitive accessibility uses this same logic with an inclusive angle: a design that eases overload helps those with attention or memory deficits and also those who are tired, anxious or studying in a second language.
For the student, the practical question is: "does this game ask me for clinical reasoning or does it ask me to decipher the interface?". For those who design cases, the question is: "does each element on screen justify the space it takes up in working memory?". The rest of this text translates that into barriers and concrete remedies, with a link to diagnostic games and to case-based learning.
Barriers in games and clinical cases
In serious games and case platforms, the most common cognitive barriers repeat. Dense language: long sentences, abbreviations without expansion, specialty jargon where it is not needed. Cue overload: ten findings on the same screen, all with the same visual weight. Timed play without pause: the clock becomes the main opponent. Ambiguous feedback: "try again" without saying what failed. Forced memory: game rules that change on every screen and must be remembered without a reminder. Distraction: animations, pop-ups, live ranking and sounds competing with the hypothesis.
Accessibility guidelines for games (for example the work of the Game Accessibility Guidelines community and notes from the IGDA Game Accessibility SIG) call for adjustable pace, consistent language, visible goals, less unnecessary stimulation and reversible errors. In educational games, recent rubrics add the learning dimension: accessibility must make the pedagogical content perceptible and understandable, not just the clickable button. The point for Medicine is the same: if the barrier is cognitive, the student "gets the case wrong" without having gotten the reasoning wrong.
An objective list of barriers to inspect in any playable case:
- Presentation text without hierarchy (everything in the same block).
- Cues that all arrive at once, with no progression.
- A mandatory timer with no option for your own pace.
- Feedback that only marks right or wrong, without explaining the criterion.
- Controls with different names for the same action on different screens.
- Decorative elements that compete with the clinically useful cue.
These barriers affect people unequally. Someone with ADHD can lose the thread when the interface shifts focus. Someone with dyslexia spends extra effort on jargon and on poor typography. Someone with reduced working memory (because of a condition, sleep or stress) cannot hold six hypotheses and three conflicting findings at the same time. Someone studying under anxiety suffers the same effect: rumination takes the mental space that should hold the differential list. Inclusive design does not "favor" a group; it widens who can train what the case is meant to measure. There is a parallel with the post on types of games for medical students: choosing the right format includes asking whether the format is cognitively usable.
A quick cognitive barrier test: ask a colleague who does not know the case to narrate, out loud, what the screen asks for in the first thirty seconds. If the narration stalls at "I don't know where to click" or "I didn't understand what counts as a cue", the interface is measuring literacy in the product, not reasoning. Fix labels, hierarchy and order before adding new clinical content.
Another sign: the student gets the differential right in an oral discussion and gets the same case wrong in the game. In that pattern, investigate feedback and button naming before concluding there is a knowledge gap. The post on knowledge gap versus reasoning gap in other texts on the blog covers the educational diagnosis; here the focus is the hypothesis that the gap is one of cognitive usability.
Working memory and cue overload
Neuroimaging studies in clinical reasoning show greater activation of regions linked to working memory in novices than in experts, in both easy and difficult cases (Hruska et al., 2016). In practical terms: the student is still building the schema; the expert already has it. Any interface noise takes from the novice the resource they need most.
Cases in virtual patient format interact with prior knowledge. In a trial with students, the format (whole case versus cues in series) alone did not decide the knowledge gain, but students with less prior knowledge reported greater extraneous load and learned less from the interaction (Hege et al., 2020). A useful message: progressive cues help when the design respects the capacity to process; piling up data without structure is not "realism", it is overload.
Simulation and complex tasks for novices also illustrate the risk of overload: too much complexity without support harms learning (Haji et al., 2016; Haji et al., 2015). Fraser and colleagues translate cognitive load theory into the design of medical simulations: reduce what is not essential to the task, sequence, keep a consistent format (Fraser et al., 2015). In a diagnosis game, this becomes a product rule: up to six clues, one at a time or in short blocks; a hypothesis requested at a clear moment; explanation afterward, not in the middle of the noise.
Training in differential diagnosis with clues already points to the value of weighing findings one by one. From the cognitive angle, "one by one" is also accessibility: each clue takes up a slot in working memory; six well-used slots beat twelve mixed findings. Schmidt and Mamede, in a narrative review on teaching clinical reasoning, reinforce design proposals that respect the student's cognitive limits (Schmidt and Mamede, 2015).
There is a useful distinction between desirable difficulty and useless barrier. Recalling a hypothesis from memory is desirable difficulty. Deciphering a hidden menu or remembering an unannounced game rule is a useless barrier. Cognitive accessibility removes the second without erasing the first. Keeping the clinical challenge and clearing interface friction is the balance that inclusive design seeks.
When the case requires integrating many elements (several systems, a long timeline, conflicting data), the answer is not "show everything at once to be realistic". The answer is to sequence: first the temporal picture, then the targeted examination, then the complementary test that tests the leading hypothesis. Clinical realism fits in the sequence; overload fits in the single block.
Chunking practices in the clinical case: separate demographics, complaint, examination and complementary tests into short blocks; deliver clues in order, without revealing laboratory, imaging and biopsy in the same paragraph; ask for an intermediate hypothesis before the next batch of data; keep the current clue visible and a minimal summary of what has already been seen, without forcing memorization of the previous screen; reserve alarm findings with simple typographic emphasis, without aggressive animation.
Clear language in the clues
Plain language is not "clinical simplism". It is choosing familiar words, short sentences and a predictable structure when the goal is comprehension. Health trials show that plain-language versions of instructions improve comprehension and task performance, for example in self-injection (Smith and Wallace, 2013). Plain-language guides and student training also improve communication with patients with lower health literacy (Sagi et al., 2021; van der Giessen et al., 2021).
In the diagnosis game, the clue is the equivalent of the instruction. If the clue says "the patient presents recurrent postprandial epigastric pain radiating to the back with partial relief from antacids", the student spends mental cycles just on parsing. A clear version keeps the clinical data: "Pain in the pit of the stomach after meals, which sometimes goes to the back and improves a little with antacid." The technical term can appear later, in the explanation, when the burden of the decision has passed.
Practical rules for clues:
- One main idea per sentence.
- Prefer concrete verbs to abstract nouns.
- Expand abbreviations on first occurrence.
- Avoid metaphors and humor that depend on local culture.
- Separate the finding from the interpretive comment ("fever of 39 °C" versus "worrying high fever").
This does not erase medical vocabulary. The vocabulary enters the feedback and the study texts, when the student has already stabilized the hypothesis. Mixing jargon teaching with demanding a decision under time is a double load. The post on useful feedback in clinical cases complements this: the explanation is the right place to go deeper into terms.
Term consistency is also cognitive accessibility. If the interface calls the same finding "dyspnea" in one clue and "shortness of breath" in another without saying they are synonyms, part of working memory goes to reconciling the vocabulary. Choose one term per concept in the game flow; present the synonym in the explanation. The same applies to button names: "submit guess", "confirm" and "close case" need to mean distinct, stable actions.
Flexible time and timers
Rigid time introduces variance that is not clinical competence: reading speed, anxiety, processing. In nursing, a program that offered extended time to all students reported less stress and maintained academic outcomes, with the argument that too short a limit injects noise into the score (Birkhead, 2018). In high-stakes assessments, changes in time alter the perception of sufficiency and the pace of response; the effect on the score is not always large, but the design of the clock matters (Swygert et al., 2004).
For games and case studies (not formal exams), the cognitive accessibility recommendation is direct: allow pausing, rereading and advancing at one's own pace; if there is a timed mode, offer it as an option, not as the only path. The timer can be an advanced training tool, after the student already understands the flow. Using the clock as an entry barrier excludes those who most need calm rehearsal.
Good time practices in a playable case: default mode without aggressive countdown; optional timed mode, announced before starting; pause that freezes the case without penalty; the possibility of rereading earlier clues at no artificial cost; avoid penalizing delay in a formative activity.
Those who design summative assessment follow institutional accommodation rules. Those who design daily training can be more generous: the goal is rehearsal of reasoning, not a typing ranking. There is an echo of discussions of gamification in medical education: time pressure and points only help when they do not become a cognitive barrier.
Explicit feedback and less noise
Ambiguous feedback is a classic cognitive barrier. "Wrong" with no criterion forces the person to guess what the system wanted. Explicit feedback names the deviation: "the guess ignores the fever and the lung focus; the next step would be to consider pneumonia before isolated reflux." This reduces the load of inferring the game's hidden rules and frees memory for the clinical content.
COGA emphasizes fast, direct feedback, error prevention and easy recovery. In games: confirm destructive actions, allow undo, keep the goal visible. In clinical cases: show where the student is in the flow (clue 2 of 5; guess requested; closing). Spatial orientation on the screen (headings, stable regions) keeps the person from "getting lost" and spending attention just to find their way back.
Reducing noise is the other side of the same coin. Celebration animations, confetti, loud sounds, live rankings and side messages compete with the clue. Motivating matters; saturating gets in the way. The rule of thumb: one stimulus channel at a time at the moment of decision. Save the celebration for after the closing. At the moment of the guess, relative visual silence.
Metacognition also comes in: after the case, a brief "what weighed on your decision?" helps the person review their own process. The post on metacognition in Medicine goes deeper into this habit. From the accessibility angle, the post-case question should be short and with clear options, not a mandatory free essay under fatigue.
Practical principles in a table
The table below summarizes the barrier and response mapping. Use it when reviewing a case or a diagnosis game before publishing.
| Cognitive barrier | What the student feels | Design response |
|---|---|---|
| Dense language | Spends effort just to decode | Short sentences; jargon in the explanation |
| Clue overload | Does not know what to weigh first | Clues in parts; up to six in the flow |
| Rigid time | Anxiety and superficial reading | Own pace; optional timer |
| Ambiguous feedback | Guesses the system's rule | Explicit criteria in the response |
| Interface inconsistency | Relearns the game on every screen | Same labels and stable layout |
| Visual or sound noise | Loses focus on the hypothesis | Less stimulation in the decision |
| Reliance on memory | Forgets an earlier clue | Visible summary of what has been seen |
These principles align with the COGA goals and with recommendations on cognitive load in health education. They do not replace testing with real users (including neurodivergent students), but they prevent the costliest mistakes before the first playtest.
How NEXO fits in
In NEXO, the flow of the case of the day is built for practicing reasoning with incomplete information: fictional cases with a realistic clinical presentation and up to six progressive clues. This cap on clues is itself a choice about load. The progression lets you form a hypothesis with what has already appeared, instead of processing a single block. The daily pace favors consistency with short sessions, which reduces the fatigue that degrades working memory.
Ecosystem features that connect with accessible study: NEXO Researcher to go deeper into topics; NEXO Closer when the focus is closing gaps; usage limits described under Economy. The site's Accessibility page gathers the platform's general guidelines. The text what NEXO is places the game on the map of clinical teaching.
How to use the case of the day with a cognitive lens (with no promise of results):
- Start without time pressure; only later try a faster pace if you want to.
- Read each clue aloud or note it in one line; externalizing reduces the load on memory.
- Write your hypothesis before opening the next clue; avoid piling up data without an intermediate decision.
- In the explanation, mark just one new vocabulary point per case; do not let the feedback become a new overload.
- If the interface or the discomfort gets in the way of focus, pause and seek the help you need when the distress is intense.
Honest comparisons: NEXO organizes clues and feedback in a single flow; banks of long questions and simulators with many tabs open at once ask more of working memory. The choice depends on the goal of the day. To train differential diagnosis with few clues, the lean format fits. To train searching a complex medical record, another environment may be needed. What does not change is the principle: measure the reasoning, not the tolerance for noise.
Those who create cases for classes (preceptors, monitors, study groups) can apply the table above even outside the platform: a vignette in plain language, numbered clues, flexible time in the formative activity and feedback with criteria. NEXO illustrates one path; the standard of cognitive accessibility holds for any playable case.
For preceptors who use the case of the day in a group: agree on one minute of silent reading of the clue, one minute of individual written hypothesis and only then the discussion. This external ritual reduces the social pressure that competes with working memory and gives processing time to those who read more slowly. The game stays the same; the protocol for using it becomes more inclusive.
If you adapt your own cases for a class, copy the structure: an opening in a short scene, numbered clues, an explicit request for a hypothesis, feedback with criteria, a short glossary at the end. Cognitive accessibility lives in this skeleton, regardless of the platform. NEXO is an example of a lean flow; the principle travels.
Frequently asked questions
Is cognitive accessibility the same as WCAG?
No. WCAG covers many technical criteria for web accessibility. The W3C's COGA document offers supplementary guidance focused on cognitive usability. This post uses that angle applied to games and cases, without repeating a normative software guide.
Does plain language make the clinical case poorer?
No, if the clinical data remains. You change the form, not the substance. The technical term can appear in the explanation, once the decision has already been made.
Does flexible time take away the value of training under pressure?
No. Offer your own pace by default and a timer as an optional mode. Pressure without prior understanding of the flow measures anxiety, not just reasoning.
How many clues are too many?
It depends on the student's level and on the interaction between the data. As a practical design rule, prefer a few progressive clues to a single block. In NEXO, the flow works with up to six clues.
Does this only matter for people diagnosed with ADHD or dyslexia?
No. High cognitive load from sleep, stress, a second language or on-call fatigue produces similar barriers. Clear design benefits most people.
How does NEXO fit in without becoming a promise of results?
As a rehearsal environment with limited clues and feedback after the guess. Use it to practice the habit, not as proof of competence or as treatment.
Final checklist
Before publishing or playing a case with cognitive accessibility in mind, go through this list:
- Does the presentation fit into short blocks with headings?
- Do the clues arrive in parts, with a clear cap?
- Is the clue language direct, with acronyms spelled out?
- Is there a mode without a punitive timer?
- Does the feedback name the criterion, not just right or wrong?
- Does the interface repeat the same labels and layout?
- Are visual and sound noise turned off at the moment of the guess?
- Is there a summary of what has already been seen, without forcing pure memory?
If any item fails, fix the design before attributing the error to the student. Then practice with a complete case in the real flow.
References
- W3C. Making Content Usable for People with Cognitive and Learning Disabilities (COGA). https://www.w3.org/TR/coga-usable/
- W3C WAI. Cognitive Accessibility at W3C. https://www.w3.org/WAI/cognitive/
- Hruska P, et al. Working memory, reasoning, and expertise in medicine. Adv Health Sci Educ Theory Pract. 2016. PMID 26537964.
- van Merriënboer JJG, Sweller J. Cognitive load theory in health professional education. Med Educ. 2010. PMID 20078759.
- Young JQ, et al. Cognitive Load Theory: implications for medical education: AMEE Guide No. 86. Med Teach. 2014. PMID 24593808.
- Hege I, et al. Learning clinical reasoning: how virtual patient case format and prior knowledge interact. BMC Med Educ. 2020. PMID 32171297.
- Schmidt HG, Mamede S. How to improve the teaching of clinical reasoning. Med Educ. 2015. PMID 26383068.
- Fraser KL, et al. Cognitive Load Theory for the Design of Medical Simulations. Simul Healthc. 2015. PMID 26154251.
- Haji FA, et al. Measuring cognitive load: performance, mental effort and simulation task complexity. Med Educ. 2015. PMID 26152493.
- Haji FA, et al. Thrive or overload? The effect of task complexity on novices' simulation-based learning. Med Educ. 2016. PMID 27562895.
- Smith MY, Wallace LS. Reducing drug self-injection errors: plain language instructions. Res Social Adm Pharm. 2013. PMID 23506652.
- Sagi D, et al. Teaching plain language to medical students. BMC Med Educ. 2021. PMID 34320965.
- van der Giessen JAM, et al. Plain-language guide for genetic counseling discussions. Support Care Cancer. 2021. PMID 33001269.
- Birkhead SF. Testing Off the Clock: Allowing Extended Time for All Students on Tests. J Nurs Educ. 2018. PMID 29505076.
- Swygert K, et al. The impact of timing changes on examinee pacing. Acad Med. 2004. PMID 15383389.
- Game Accessibility Guidelines. https://gameaccessibilityguidelines.com/