Clinical Standards of Safety and Psychometric Quality: Why Doctors Need the Digital Validity of DIN 33430, MDR and GDPR

The mobile health sector (mHealth) is flooded with thousands of uncertified “child apps”, “quizzes from social media” and “development calculators”, built without any involvement of scientists or clinicians. A parent can install one in a minute, answer ten questions while the kettle boils, and receive a verdict that looks authoritative: “everything is fine” or “your child is at risk”.

Paediatricians and neurologists treat such services with deep suspicion, and for good reason. A questionnaire that has never been validated does one of two things. It frightens parents without cause, or it gives a false sense of security, and a real difficulty is put off until the school desk makes it impossible to ignore.

For a doctor, the question is therefore not whether an app is pleasant to use. The question is whether its result can be trusted, whether it is built within the rules of medical software, and whether the data of a child are safe. Three pillars answer these questions.

I. Why the Medical Community Rejects Everyday Apps

A clinician works with measurable things. A result is useful only if it can be reproduced: the same child, assessed on two different days, should produce a comparable picture. Everyday apps cannot offer this. Their questions are written “by feel”, their thresholds are invented, and nobody can say what a given score actually means.

There is a second problem. A consumer app is designed to hold attention, not to support a decision. Its output is a bright badge and a share button. A doctor needs the opposite: a sober, structured profile that shows where the child is doing well, where support is needed, and what to discuss at the appointment.

The third problem is the data. Information about a child aged one to six is among the most sensitive a family has. When it travels to unknown servers or is kept forever, no responsible clinician will recommend the tool, whatever it promises on the front page.

II. First Pillar: The Psychometric Standard DIN 33430

DIN 33430 is a German standard that regulates the quality of psychodiagnostic instruments. It does not care how a questionnaire looks. It asks whether the instrument measures what it claims to measure, and whether it does so consistently. OkiTalki strictly follows its requirements, which rest on three ideas doctors know well:

• Reliability: the result is stable. If the same child is observed under comparable conditions, the picture does not change from one attempt to the next.
• Validity: the tool measures the function it was built to measure, such as hearing, speech, fine and gross motor skills, vision or social skills, and not something nearby.
• Representativeness of the sample: the reference values rest on a sample that reflects the children to whom the tool is applied, so that normal variation is not mistaken for deviation.

These properties separate a screening instrument from a quiz. Consider a small example. A child of three “does not like to draw”. A parent may read this as character. A validated tool places the same observation next to age-appropriate norms, shows that it is one of several signals in the fine motor block, and asks the next question in a structured way. The difference is between an impression and a measurement.

III. Second Pillar: The Medical Devices Regulation MDR 2017/745

The European Medical Device Regulation (MDR 2017/745) defines which software counts as a medical device and what is required of it. Software that supports a clinical decision is called SaMD, Software as a Medical Device, and it is treated very differently from a lifestyle application.

The architecture of OkiTalki is built in accordance with the requirements for medical software of Class I / IIa. For a doctor this has a plain meaning: the system is designed from the start with documentation and risk in mind, not as a consumer toy that was later given a medical vocabulary. That is what allows OkiTalki reports to be used in official medical practice, for example as a structured basis for a conversation at a paediatric appointment.

It also changes the tone of the product. A tool built under this logic does not promise cures. It shows a profile, marks the zones of attention and leaves the clinical decision to the specialist. For a physician this is the right division of roles: the screening sees early, the doctor decides.

IV. Third Pillar: Uncompromising GDPR / DSGVO Compliance

The General Data Protection Regulation (GDPR, in German DSGVO) is the framework for personal data in Europe, and data about children is its most sensitive category. OkiTalki handles it with a short list of concrete rules:

• All personal data of children and parents are anonymised.
• Data are encrypted with the AES-256 protocol.
• Data are stored physically on protected servers inside the European Union, in Frankfurt.
• Data are never passed on to third parties.
• Data can be deleted immediately and irreversibly at the request of the user.

The last point matters more than it seems. A family must be able to leave. When a parent changes their mind or a child grows up, the information should disappear completely and not remain in somebody’s archive.

V. What This Means for Families, Kitas and Doctors

Taken together, the three pillars change who can use a screening result and for what. A parent receives an understandable profile without fear of leaks. A Kita, the German day-care and kindergarten setting, gets a shared language with families that is not built on opinions. A doctor receives a document that can be read in a minute and rests on recognised standards.

OkiTalki is built for this role. The screening takes about ten minutes and consists of 36 questions. The result is a profile of the child across the key areas of development, not a single frightening number. For zones where support is useful, the parent gets short two-minute games for home, which fit into an ordinary evening. Nothing here replaces a specialist. It builds a bridge between the family, the kindergarten and the doctor’s office.

Doctors’ distrust of everyday apps is not conservatism. It is a professional reflex that protects children from random conclusions. Trust is earned not by louder promises, but by standards anyone can check: a psychometric framework in line with DIN 33430, an architecture aligned with MDR 2017/745, and data protection under GDPR / DSGVO.