ThePattern#015EN — The Missing Links: From AuDHD Diagnostics to a European Learning System
How existing medical, scientific and technological infrastructures could be connected to improve neurodevelopmental assessments, accelerate mental health research and create new opportunities — with neurodivergent people helping design the system from the beginning.
What if Europe doesn’t lack the resources to improve neurodivergent lives nearly as much as it lacks the ability to connect the resources it already has?
That’s the question I’ve been exploring.
And the further I went, the more I realised that this isn’t just about autism or ADHD. It’s about the way we organise healthcare, scientific research, social support and even access to professional opportunities.
Somewhere in Germany, an autistic adult is waiting years for a specialist assessment. Somewhere else, a highly qualified diagnostician has expertise that people living hundreds of kilometres away can barely access. Meanwhile, university hospitals and research institutions are collecting valuable information, often within separate projects, databases and administrative structures.
All of them are doing something useful.
But they’re not necessarily working as parts of one larger system.
Europe has specialists. Europe has research centres. Europe has secure data infrastructure, high-speed networks, cloud computing and internationally recognised scientific institutions.
Europe also has a remarkable talent for keeping these things in separate administrative boxes.
And I keep wondering: What would happen if we connected them properly?
I spent many years working in enterprise IT, including advising on complex infrastructure architecture during the later part of my career. Designing systems from existing components, identifying unnecessary bottlenecks and finding ways to make separate resources work together were part of my professional background.
So instead of looking at this primarily as a medical problem, I started approaching it as a systems architecture problem.
What emerged was a concept with three interconnected, but deliberately separated areas: clinical assessment and care, scientific research, and social and professional participation.
But this isn’t supposed to be an architecture designed by technical experts for a passive group of patients.
I want it to become an open discussion involving autistic people, people with ADHD, AuDHDers and the broader neurodivergent community, including those who are still questioning or seeking assessment.
Researchers, clinicians, engineers and policymakers absolutely have a place in that discussion. But so do people who know these systems from the inside because they’ve spent years trying to navigate them.
And they shouldn’t have to wait until the architecture is finished before someone asks for their opinion.
Let’s take it apart, layer by layer.
Layer One: When one person’s judgement can change another person’s life
I received my autism diagnosis in 2016, when I was 41.
By that time, I’d already spent years investigating autism and understanding how my own mind worked. The diagnosis was a scientifically grounded professional confirmation of something I’d figured out about myself earlier.
ADHD was different.
My psychologist was highly experienced in autism diagnostics and even supervised other professionals. But ADHD wasn’t her specialist field, and she didn’t have the complete assessment instruments required for a comprehensive ADHD evaluation.
That wasn’t a failure of her expertise.
It was a limitation of specialisation.
And that distinction matters.
Autism and ADHD assessments rely on developmental histories, structured interviews, standardised questionnaires, behavioural observations and, where available, childhood records or information from family members and partners.
But none of these automatically produces a perfectly objective answer.
Professionals must interpret the evidence.
Imagine someone who appears highly organised, rarely misses appointments and maintains a carefully structured daily routine.
One diagnostician might see little evidence of ADHD-related organisational difficulties. Another might discover that the person maintains this apparent efficiency through relentless compensatory strategies, exhausting mental effort and a collection of reminders that would make an air traffic control centre look under-equipped.
The visible outcome is similar.
The underlying effort isn’t.
Good diagnostics must distinguish between the two.
This is particularly important for adults who have spent decades compensating, masking or adapting their behaviour to environments that don’t naturally suit them.
It also raises a scientific question.
How much of a diagnostic result depends on the person being examined, and how much depends on the particular professional conducting the examination?
We already have methods for investigating that question.
We just don’t apply them consistently across everyday clinical practice.
Layer Two: Five diagnosticians — but not five for everyone
My first idea was deliberately ambitious.
What if particularly difficult cases were assessed by five independent professionals?
Two specialising in autism, two specialising in ADHD, and a fifth with substantial expertise in both.
Each would initially conduct and document an independent assessment before seeing anyone else’s conclusions.
The first four could then compare their findings, challenge one another’s interpretations and identify disagreements. The fifth person’s assessment would remain sealed until that comparison was complete.
Only afterwards would the fifth assessment be revealed.
Why?
Because four professionals agreeing with one another doesn’t necessarily mean they’re correct. They might share similar assumptions, rely on the same diagnostic stereotypes or influence one another during discussion.
An independently documented fifth opinion could reveal something the others collectively overlooked.
The scientific principle behind this is related to interrater reliability: measuring the consistency of assessments made by different observers.
Structured expert judgement methods already exist. One example is the IDEA protocol — Investigate, Discuss, Estimate, Aggregate.
However, IDEA was developed for quantitative expert estimates and probabilistic judgements, not as a validated clinical diagnostic procedure. Its relevance here lies in the underlying method: independent judgements first, critical discussion afterwards, followed by carefully documented reassessment.
But there’s an obvious problem with my original five-person approach.
We already have too few specialists.
Making every patient undergo five complete assessments would be a rather spectacular way to make waiting lists even longer.
So the practical system should work differently.
A four-tier assessment model
Tier 1 — Standard clinical assessment
A properly qualified diagnostic service conducts a guideline-based assessment, drawing on additional professional expertise where necessary.
This remains the normal route. Good screening, structured history-taking and appropriate differential diagnosis are essential.
Tier 2 — Independent second opinion
Where significant uncertainty, conflicting evidence or difficult differential diagnoses remain, a second appropriately qualified professional conducts an independent review or supplementary assessment.
Tier 3 — Blinded adjudication
When the first two assessments disagree materially, a third specialist can provide an additional independent judgement before the earlier conclusions are discussed.
The disagreement, the reasons for it and the eventual outcome are documented.
Tier 4 — Five-person reference assessment
The original five-person model is reserved for carefully selected, highly complex cases and scientific validation cohorts.
Some research participants could be selected independently of case complexity, allowing researchers to investigate whether routine procedures systematically overlook particular profiles.
One important methodological distinction
There are actually two different things we need to investigate.
First, interpretation variability: different specialists reviewing the same information and reaching different conclusions.
Second, information-gathering variability: different specialists asking different questions and therefore discovering different things.
Using the same recorded interview for multiple independent assessments reduces the burden on participants, but it doesn’t reveal what another interviewer might have discovered.
Five entirely separate developmental interviews could reveal more about information-gathering differences, but would also introduce exhaustion, practice effects and unnecessary repetition.
So I’d favour a standardised initial assessment combined with targeted supplementary interviews. A smaller, voluntary research cohort could undergo more extensive independent assessments, with the effects of repeated interviewing explicitly studied.
The clinical system should use additional expertise where it adds demonstrable value. The research system should investigate how much value that actually is.
An invitation before we go any further
This is the first part of the concept where I’d particularly welcome input from the neurodivergent community.
If you’ve been assessed for autism, ADHD or both, what did the process overlook?
Were there questions you wished someone had asked? Did you feel that your compensation strategies were misunderstood? Were your experiences interpreted differently by different professionals?
And would you welcome an independent second opinion, or would another assessment feel like an additional burden?
I’m not looking exclusively for agreement. I’m looking for the perspectives that could make this model better.
If you have an idea for a better approach, share it. If you can identify a flaw, please challenge it.
This isn’t intended to be my finished blueprint handed down to everyone else. I’d much rather see it develop into something we design and improve together.
Layer Three: Screening should prioritise resources, not become another barrier
Before we even reach specialist assessment, we need a sensible way to determine who requires which kind of investigation.
Standardised screening questionnaires, structured interviews and existing medical or developmental records can help.
Someone with a well-documented history of neurodevelopmental difficulties may need a different diagnostic pathway from someone presenting with relatively nonspecific symptoms.
But screening must never become an automatic exclusion mechanism.
An adult who has spent thirty years developing sophisticated compensation strategies might not produce an obvious result on a short questionnaire.
That doesn’t mean there is nothing worth investigating.
Clinical judgement, developmental evidence and documented difficulties must still count.
The UK’s NICE guidance, for example, allows comprehensive autism assessment based on clinical suspicion even when a preliminary screening threshold isn’t met.
And here’s something scientifically important.
We shouldn’t only study the people who receive a positive diagnosis.
People whose assessments are negative, uncertain or inconclusive could also be invited to take part in voluntary follow-up research.
Otherwise, how would we investigate whether our screening methods systematically overlook certain people?
A good research system must be able to study its own failures.
Not merely count its successes.
Layer Four: Remote assessment — without turning people into laboratory specimens
Now we come to the technical possibilities.
Many parts of an autism or ADHD assessment can already be performed remotely: developmental interviews, questionnaires, document reviews, consultations with relatives and interdisciplinary case discussions.
A 2026 study involving 100 adults found 86% agreement between original autism assessments conducted through a structured telehealth approach and blinded expert classifications based on substantially overlapping material.
That’s encouraging.
But it measures agreement, not proof of diagnostic accuracy. Independent, prospective validation is still needed.
For some examinations, dedicated regional assessment centres could provide professional facilities with appropriate audio and video equipment, accessible workstations and qualified support staff.
The specialists wouldn’t necessarily have to be physically present.
They could participate remotely or review relevant material later.
Multiple camera perspectives might be useful for certain observational questions.
But here’s a problem I hadn’t fully considered when first developing this idea.
What if being observed by several cameras changes the very behaviour we’re trying to understand?
For someone who has spent their life masking, a room full of recording equipment might trigger even more masking, heightened vigilance or severe examination anxiety.
We could end up measuring the person’s reaction to surveillance rather than the characteristics we’re interested in.
That would be an impressive technological achievement in producing scientifically questionable data.
So the system must provide genuine choice.
Someone might prefer a remote assessment from home. Another person might prefer a quiet regional centre with discreet equipment. Someone else might need a traditional face-to-face assessment without any recording.
Audio and video recordings would require a defined purpose, appropriate safeguards and a suitable legal basis. Optional research recordings must never become a hidden condition for receiving healthcare.
And the setting itself could become a valuable research question.
How do assessments differ between home environments, dedicated centres and traditional clinical settings?
Which environments reduce stress? Which improve the quality of information? Does one setting systematically disadvantage particular groups?
That could be investigated scientifically, with voluntary participation.
Instead of pretending technology is neutral, we could examine how it influences the results.
Layer Five: We already have much of the infrastructure
This is where my former professional field enters the picture.
If I were designing the technical foundation, I wouldn’t begin by ordering a nationwide collection of new servers.
I’d start by examining what already exists.
Germany’s Medical Informatics Initiative has established data integration centres at university medical institutions. Their purpose includes improving the standardisation and exchange of medical data for research and healthcare.
University hospitals already operate professional networks, data centres, secure communications, identity management systems and clinical information platforms.
The German Center for Mental Health also brings together institutions conducting research across different locations.
We don’t need to pretend those foundations are missing.
A shared clinical assessment environment could build on existing capabilities using interoperable data standards, secure European cloud services, encrypted connections, appropriately separated administrative domains and Zero Trust access controls.
Whether particular connections use MPLS, ExpressRoute, private networking or encrypted internet-based transport is an implementation decision.
None of those technologies replaces proper authorisation or encryption.
And yes, European cloud infrastructure is technically capable of handling high-quality medical video, document exchange and coordinated assessment processes.
YouTube can deliver high-resolution video to millions of people. I’m fairly confident European research institutions can manage several authorised clinical video streams.
The challenge isn’t inventing video transmission.
It’s designing a trustworthy medical service around it.
Where the German ePA fits
Germany already has an electronic patient record, the elektronische Patientenakte or ePA.
That could serve as one component for patient-centred documentation and clinical information exchange.
But it isn’t currently a complete diagnostic research workspace.
The existing ePA doesn’t allow patients to make an individual document visible exclusively to a selected group of healthcare institutions. Documents can be hidden from all providers, but not shared with precisely chosen institutions on a document-by-document basis.
A dedicated clinical assessment workspace would therefore need additional access controls.
Video recordings, detailed test material and independent research assessments should be managed in an appropriately protected environment rather than simply placed inside the general patient record.
And the infrastructure must respect the reality of healthcare law.
Germany’s Section 393 of Social Code Book V already provides a framework for cloud computing in healthcare, subject to specified security, location and certification requirements.
So this isn’t a request to legalise cloud computing from scratch.
But existing cloud permissions don’t automatically authorise every proposed exchange of medical information.
The legal responsibilities and individual data-processing purposes must still be established.
Layer Six: Trust is not an optional software feature
There is a reason sensitive psychiatric and psychological information requires more than ordinary enterprise security.
A database containing developmental histories, psychological assessments, trauma experiences, personal relationships and intimate details about people’s lives can cause enormous harm if misused.
And a person who doesn’t trust the system may withhold information or refuse to participate entirely.
That affects both clinical care and the scientific quality of research.
In March 2026, the German Society for Psychiatry and Psychotherapy, Psychosomatics and Neurology (DGPPN) published a critical position on the ePA, raising concerns about medical confidentiality, potential discrimination and unresolved legal protections, including the need for explicit statutory safeguards against seizure of electronic patient records.
That’s not something a VPN can fix.
Nor can we solve it by simply announcing that the servers are located in Europe.
A trustworthy architecture needs clearly separated responsibilities, strictly limited access, independent oversight, transparent audit logs, defined retention periods and meaningful patient control.
It should also protect against unauthorised administrative access, including risks associated with cloud support operations and unnecessary linkage between datasets.
Where research participation requires consent, that consent must be informed, specific enough for the intended purposes and genuinely voluntary. Other research processing must also have a valid legal basis under the applicable rules.
Clinical care cannot depend on agreeing to additional research.
And if someone withdraws from future participation, that decision must be respected, subject to the legal rules governing information already processed in completed research.
Data protection isn’t the enemy of scientific progress. Trustworthy data protection is one of the conditions that makes long-term research possible.
Layer Seven: Turning diagnostic disagreement into scientific knowledge
Here’s what might become the most valuable part of the entire system.
Normally, a clinical assessment ends with a report.
It might explain which criteria were met, which weren’t, and how the final conclusion was reached.
But imagine systematically preserving the original, independent judgements of several specialists.
Not just whether they agreed.
What did each person notice? Which criteria did they interpret differently? What additional information changed their understanding? Which uncertainties remained unresolved?
We could begin studying diagnostic disagreement as a scientific phenomenon in its own right.
And with enough carefully standardised data, researchers could investigate whether particular assessment methods systematically produce different results.
Perhaps certain types of compensation are frequently missed. Perhaps some characteristics attributed to autism are better explained by ADHD, or vice versa. Perhaps existing assessment tools perform differently across age groups, genders, communication styles or cultural backgrounds.
These would be research questions, not conclusions we could simply assume beforehand.
But there’s another methodological problem.
Agreement is not the same as truth
If five diagnosticians agree, we have evidence of consistency.
We don’t automatically have evidence that the diagnosis is correct.
There is no single biological gold standard capable of independently confirming autism or ADHD in every case.
So how would we evaluate the validity of the proposed approach?
We would need multiple, independently collected indicators: developmental evidence, functional characteristics, careful evaluation of alternative explanations, longitudinal stability and changes in functioning or wellbeing over time.
We could also examine whether better assessment reduces inappropriate treatment or leads to more suitable support.
But the response to a particular medication or therapy cannot, by itself, prove a diagnosis.
And longitudinal stability is not infallible either. A stable diagnosis can still be wrong, while a person’s outward difficulties may change considerably as their environment changes.
The objective should therefore be stronger evidence from multiple sources, not the illusion of absolute certainty.
The research infrastructure should also retain disagreements and uncertainty rather than forcing every case into an artificial consensus.
Sometimes the scientifically correct answer is simply that the available evidence remains inconclusive.
That’s not a weakness.
That’s intellectual honesty.
Layer Eight: From individual assessments to lifelong research
Now imagine that people entering the diagnostic system are also invited to participate in long-term scientific studies.
Not automatically enrolled.
Invited.
Children, adolescents and adults could choose to participate in repeated assessments, interviews, questionnaires and other research activities throughout their lives.
Some might participate for five years. Others for twenty. Some could remain involved for much longer.
A follow-up every few years might examine changes in everyday functioning, mental health, support needs, relationships, education or employment.
It wouldn’t necessarily mean repeating the entire original diagnostic process.
Autism and ADHD are neurodevelopmental conditions. They don’t simply appear for the first time in adulthood because someone has a difficult year.
But their expression, their impact and their recognition can change. Previously overlooked co-occurring conditions may become apparent, and new research may justify reviewing earlier conclusions.
For children, appropriate parental permission and age-sensitive involvement would be necessary. Continued participation into adulthood would require the person’s own informed decision under the applicable rules.
And participation could be paused or ended.
The data architecture matters here, too
I would separate identifiable clinical records from protected research datasets.
Approved longitudinal projects could use pseudonymised data where participants need to be followed over time. Researchers would work within secure environments under tightly controlled permissions, while publications would use appropriately anonymised or aggregated results.
Raw video and audio data would receive additional restrictions because faces, voices and personal narratives cannot simply be made anonymous by removing someone’s name.
And there’s one more problem worth studying.
People who volunteer for long-term research may differ systematically from those who don’t. Participants who leave a study may also differ from those who remain.
Without accounting for those selection and attrition effects, we might end up developing better diagnostic methods for the people who are already easiest to reach.
That would rather defeat the purpose.
Layer Nine: Why limit this to autism and ADHD?
At this point, the architecture becomes much larger than the original question.
And that’s intentional.
The same foundational research and collaboration infrastructure could support projects in psychiatry, clinical psychology, psychotherapy, neuropsychology and mental health epidemiology.
The clinical methods would remain specific to each field.
The common foundation would handle secure collaboration, interoperable information, research governance and authorised access to relevant datasets.
That distinction matters.
I’m not proposing one universal diagnostic algorithm that somehow understands every psychiatric and neurodevelopmental condition.
I’m proposing shared infrastructure that allows different disciplines to work together without repeatedly building their own isolated technical environments.
Autism and ADHD are not simply psychiatric illnesses. But their study overlaps with research into depression, anxiety, trauma, sleep, cognition and other aspects of mental health.
Imagine researchers examining how particular environmental conditions, sleep patterns, chronic stress and neurodevelopmental characteristics interact over decades.
Or how different kinds of workplace adjustments affect mental health and long-term functioning.
Or which interventions improve someone’s quality of life, even when their underlying neurodevelopmental characteristics remain unchanged.
These questions require careful scientific methods, including appropriate comparison groups and independent validation.
More data alone doesn’t establish causation.
But well-structured, comparable and appropriately governed data can enable research that isolated institutions struggle to conduct.
And Europe already has important starting points.
The German Center for Mental Health, the Medical Informatics Initiative and the European Health Data Space all demonstrate movement towards more coordinated healthcare research.
The EHDS entered into force in March 2025, with major provisions on secondary data use scheduled to apply from March 2029.
It isn’t a ready-made psychiatric research platform.
But it provides an evolving European framework that such a system could eventually connect with.
European research isn’t the same as European clinical licensing
One distinction matters here: exchanging research data or collaborating on scientific methods across Europe is not the same as providing formal clinical assessments across national borders.
Cross-border telemedicine can be legally possible under European law, but professional licensing, liability, reimbursement and recognition of diagnoses for national social benefits remain subject to the relevant rules.
A legally performed assessment abroad doesn’t automatically qualify someone for disability-related benefits, rehabilitation or treatment in another country.
For an initial pilot, I would therefore keep formal clinical responsibility clearly assigned within the participating national healthcare systems while expanding European collaboration through research, shared methodology and specialist consultation.
That would make the first implementation manageable without ruling out broader cross-border clinical services in the future.
The opportunity isn’t to replace all existing research initiatives. It’s to make their capabilities more interoperable and their collaboration more effective.
Layer Ten: Healthcare, research and employment must not become one giant database
Now comes a particularly important architectural boundary.
A person who participates in psychiatric research should never have to worry that their clinical records will somehow become available to an employer.
And I don’t mean merely hiding them behind a different screen in the same application.
I mean genuine separation.
The clinical care environment, the research environment and a potential European professional participation network would have different purposes, legal responsibilities, administrative controls and access permissions.
The professional network should have its own identity-management domain and data storage, with no automatic linkage to clinical records.
That doesn’t prevent the systems from using compatible technical standards or even infrastructure technologies provided by the same vendors.
But it prevents a shared platform from quietly becoming a mechanism for combining sensitive medical and employment information.
What would connect the systems?
The individual.
Not an invisible database relationship.
Someone might voluntarily share evidence of a professional qualification, an agreed workplace adjustment or an officially recognised entitlement.
European digital identity technologies could eventually make certain user-controlled attestations easier to exchange.
But a functional workplace requirement is usually more relevant than a complete diagnostic history.
An employer might need to know that someone requires a quiet working environment, flexible hours or specific accessibility measures.
They don’t need a detailed psychiatric record to arrange that.
The European Health Data Space itself prohibits certain uses of health data obtained through its regulated secondary-use mechanisms, including decisions about job offers.
That boundary should be reflected directly in the system architecture.
And why connect the broader ideas at all?
Because better assessment and support can sometimes make professional participation more accessible.
People with strong abilities in research, software engineering, data analysis, AI development or defensive cybersecurity may struggle in conventional working environments for reasons that have little to do with their actual technical competence.
This doesn’t mean neurodivergent people automatically possess exceptional abilities in those fields.
It means people should be assessed by their individual skills rather than by their ability to perform conventional workplace behaviour.
Someone might work twelve hours a week on a research project from home. Someone else might contribute to a European AI initiative under a flexible contract. Another person might choose a multi-year position abroad if professional relocation assistance and continuing everyday support made the move manageable.
And some people won’t be able to work at all, regardless of support.
They still deserve healthcare, dignity and meaningful participation in society.
The purpose of support isn’t to make people economically productive enough to deserve it.
But where people want to contribute, we should stop treating avoidable organisational barriers as if they were permanent limitations of the individual.
The professional participation network would therefore be a separately governed, optional extension of the broader European vision.
Not part of the clinical database.
Not a recruitment tool disguised as healthcare.
And certainly not a way for institutions to monetise psychiatric information.
Layer Eleven: The economic argument — and the costs we keep ignoring
Eventually, someone in a ministry or funding organisation will ask the inevitable question.
How much would this cost?
Fair enough.
A coordinated assessment network requires funding. Independent specialist reviews consume professional time. Secure research infrastructure, longitudinal studies and individual support services are not free.
But evaluating the system solely by its immediate costs would miss much of the economic question.
What happens when someone spends years receiving inappropriate treatment because their underlying difficulties have not been recognised?
What happens when repeated overload leads to preventable mental health crises?
What happens when a highly qualified person cannot sustain employment because the available working conditions are unsuitable?
Or when family members reduce their own working hours because appropriate assistance is inaccessible?
Those situations also have costs.
Sometimes substantial ones.
But claiming that our proposed system would automatically save billions would be speculation.
We need a proper economic evaluation.
That would include the cost per completed assessment, specialist hours, waiting times, downstream healthcare use, quality of life, the value of additional support and long-term changes in employment or other forms of participation.
We would also have to compare the new system with realistic alternatives, including improved conventional services.
An expensive five-person assessment might provide enormous value for a small group of difficult cases while being completely unjustified for routine cases.
That’s precisely the kind of distinction the research should establish.
There’s also a structural problem.
The institution financing a better diagnostic assessment may not be the one benefiting from reduced healthcare use or improved employment years later.
Health insurance, pension systems, social services, employers and public budgets don’t necessarily share costs and benefits.
This is one of those situations where the accounting structure can make a potentially worthwhile investment look unattractive to every individual organisation.
From an enterprise architecture perspective, it’s a familiar problem: local optimisation can produce a very inefficient overall system.
You can have departments that all meet their individual budgets while the organisation as a whole performs badly.
Healthcare isn’t immune to that.
Layer Twelve: How I would actually start
I wouldn’t begin by demanding that Europe rebuild its entire mental healthcare system.
That would be an excellent way to produce several years of conferences, committees and presentation slides before anyone receives a better assessment.
I’d begin with a limited, properly designed pilot.
A small group of university hospitals and specialist centres could collaborate across different German regions, ideally incorporating existing data integration infrastructures.
The first objective would be to compare current diagnostic pathways with structured independent review.
Selected cases would receive additional assessments, while a research cohort would test the five-person reference model.
Some participants would be selected because their cases are particularly difficult. Others would be recruited independently of complexity, allowing the researchers to examine potential blind spots in routine assessment.
Different settings could be tested voluntarily: remote interviews from home, dedicated regional assessment rooms and traditional face-to-face examinations.
The methodology should be preregistered, with clearly defined inclusion criteria and outcome measures.
What would count as success?
Not simply more diagnoses.
Not simply higher agreement among specialists.
We would need evidence addressing several different questions.
Does the approach improve diagnostic consistency without sacrificing accuracy? Does it reduce unnecessary repeat assessments? Can patients access appropriate care sooner? Does it uncover information that changes treatment or support decisions? Does it reduce or increase stress for the people being assessed?
And perhaps most importantly: Does it improve outcomes sufficiently to justify the additional professional time and financial investment?
Longitudinal follow-up would examine diagnostic stability, functioning, wellbeing and the consequences of different support approaches.
Economic evaluation would compare the additional costs with demonstrated benefits.
Researchers would need to account for participant selection, possible shared diagnostic biases and the absence of a definitive biological gold standard.
Results would require replication beyond the original participating institutions.
And neurodivergent people should be directly involved in research design, governance and evaluation from the beginning.
Not merely invited to comment after the decisions have already been made.
If one part of the system doesn’t work, it should be redesigned or abandoned.
That’s not a problem.
That’s how serious research is supposed to function.
Where do we go from here? An open invitation
I started with a fairly narrow question about the quality and accessibility of autism and ADHD assessments.
That question gradually led to independent diagnostic reviews, research on diagnostic disagreement, lifelong studies, shared infrastructure, European scientific collaboration and a broader discussion about meaningful participation.
Not by proposing one enormous system that does everything, but by asking how existing capabilities could cooperate while respecting their necessary boundaries.
The result is an architectural concept, not a finished medical programme, and certainly not something that should be developed behind closed doors.
I want this to be a conversation among the people who would actually live with its consequences.
That means autistic people, ADHDers, AuDHDers, other neurodivergent people, parents, partners, carers, educators, researchers and clinicians. It includes people with a formal diagnosis, people still seeking one, and people whose experiences have repeatedly been dismissed because they didn’t fit someone else’s expectations.
Neurotypical expertise is welcome. Scientific expertise is essential.
But neither should replace the voices of the people whose lives are being studied.
And I’d like this discussion to go beyond simply agreeing that the idea sounds interesting.
If you’re experienced in clinical assessment, challenge the methodology.
If you work in research, tell me where the proposed studies would fail or how they could be improved.
If you understand European healthcare law, identify the regulatory obstacles.
If you’re an engineer or IT architect, question the infrastructure.
And if you’re neurodivergent, tell me what the system would need to do differently to earn your trust and actually make your life easier.
Your criticism is as welcome as your ideas. So is your willingness to collaborate.
Maybe you’ve already developed a useful approach in your own community. Maybe you know researchers or organisations doing something similar. Maybe you’ve experienced a failure that this architecture hasn’t even considered yet.
I would genuinely like to learn about it.
If enough people are interested, we could begin turning this discussion into an open, community-informed proposal that researchers, medical institutions and policymakers could examine, challenge and potentially help develop.
Not a promise that every idea will be adopted, and not an invitation to share sensitive medical records in public comments. We would need proper governance before any formal collaboration or research began.
But a place to start thinking together.
Perhaps Europe’s problem isn’t that we lack the technology, the expertise or the people.
Perhaps we’ve simply become too accustomed to keeping them apart.
And perhaps the next major breakthrough won’t come from inventing something entirely new, but from finally connecting what we’ve already built.
This time, I’d like neurodivergent people to help design that future, rather than spend another generation trying to adapt to systems that were never designed with us in mind.
So let’s hear your objections, your experiences and your ideas.
And if you think you could help build something better, I’d like to hear that too.
Sources and further reading
The following sources document existing scientific methods, clinical guidance, technical infrastructures and legal frameworks. They support the feasibility of individual components, not the proven effectiveness of the complete proposed system.
-
National Institute for Health and Care Excellence (NICE). Autism spectrum disorder in adults: Diagnosis and management (CG142). Clinical guidance on specialist assessment and differential diagnosis.
https://www.nice.org.uk/guidance/cg142 -
Hanea, A. M., et al. (2017). InvestigateDiscussEstimateAggregate for structured expert judgement. International Journal of Forecasting, 33(1), 267–279.
https://doi.org/10.1016/j.ijforecast.2016.02.008 -
Crehan, E. T., et al. (2026). Diagnostic Agreement of a Structured Telehealth Adult Autism Assessment. Journal of Autism and Developmental Disorders.
https://doi.org/10.1007/s10803-026-07571-2 -
Medical Informatics Initiative, Germany. Data Integration Centres. Existing infrastructure connecting medical research and clinical data.
https://www.medizininformatik-initiative.de/de/konsortien/datenintegrationszentren -
German Center for Mental Health (DZPG). Research Programme. Coordinated mental health research across participating institutions.
https://www.dzpg.org/en/research/research-programme -
gematik. Electronic Patient Record (ePA): Information and FAQ. Current capabilities and limitations of Germany’s electronic patient record.
https://www.gematik.de/anwendungen/epa-fuer-alle/faq -
Federal Republic of Germany. Social Code Book V, Section 393. Legal requirements for cloud computing in healthcare.
https://www.gesetze-im-internet.de/sgb_5/__393.html -
German Society for Psychiatry and Psychotherapy, Psychosomatics and Neurology (DGPPN), 2026. Position on the electronic patient record. Confidentiality, legal safeguards and trust in psychiatric care.
https://www.dgppn.de
-
European Commission. European Health Data Space. European health data framework and implementation schedule.
https://health.ec.europa.eu/ehealth-digital-health-and-care/european-health-data-space-regulation-ehds_en -
European Union (2025). Regulation (EU) 2025/327 on the European Health Data Space, particularly Article 54 on prohibited secondary uses.
https://eur-lex.europa.eu/eli/reg/2025/327/oj/eng -
European Union (2011). Directive 2011/24/EU on the application of patients’ rights in cross-border healthcare.
https://eur-lex.europa.eu/eli/dir/2011/24/oj/eng
Transparency: This article presents my independently developed conceptual proposal, not an existing or clinically validated programme. AI tools assisted with editorial development, language refinement, critical review and source research. The underlying ideas, architecture and editorial responsibility remain with me.


