Safety & governance

Safety is a system requirement

Mental-health technology must be designed for uncertainty, vulnerability and the possibility of harm. BioTekk’s governance approach is intended to combine defined clinical boundaries, human oversight, evidence generation, information governance and continuous safety monitoring.

Crisis and safeguarding boundaries

BioTekk is not an
emergency service

If you need help now

If you or someone else is in immediate danger, contact the emergency services. Urgent mental-health need should be met by people, not by software. The routes below are available in the United Kingdom.

Emergency
999 Immediate danger to life
Urgent health advice
NHS 111 Select the mental-health option
Samaritans
116 123 Free, 24 hours, every day
Text support
SHOUT to 85258 Free 24/7 text service

In-product behaviour

Stated before use,
and repeated
during it.

BioTekk is not an emergency service. The platform should communicate this limitation clearly before and during use, so that no user forms the impression that submitting a message is equivalent to contacting help.

When an interaction suggests immediate danger or urgent mental-health need, the user should receive direct instructions to contact appropriate emergency or urgent-care services. Organisational deployments may additionally activate locally approved safeguarding and escalation procedures.

Escalation design is a joint decision with the deploying organisation. A notification route that no one is resourced to answer is not a safety control, so escalation capacity and out-of-hours arrangements are assessed before a configuration goes live.

01

Intended purpose

Documented before
anything is
deployed.

Each BioTekk product configuration should have a documented intended purpose specifying its users, use environment, supported activities, excluded uses and role within the wider care pathway. Clinical and regulatory requirements should be assessed against that intended purpose before deployment.

Users
Who the configuration is for, including any age, capacity or eligibility conditions, and which professional roles may access which information.
Use environment
The setting in which it operates — self-directed use, a university service, an employer programme or a clinical pathway — and the assumptions each setting brings.
Supported activities
The specific functions enabled: reflection, journalling, wellbeing exercises, trend summaries, prioritised review, aggregated reporting.
Excluded uses
What the configuration must not be used for, stated explicitly — including diagnosis, emergency response and any decision the intended purpose does not support.
Pathway role
Where the product sits relative to existing services, and what happens at each boundary between automated support and human care.
02

Clinical safety management

Accountability,
hazards, controls,
evidence.

Clinical safety management should include nominated accountability, hazard identification, risk controls, testing, residual-risk review, incident management and documented safety cases where applicable.

Deployment should also consider local workflows, staffing, escalation capacity and foreseeable misuse. A control that depends on a step no one has time to perform will not hold in live service.

01 / ACCOUNTABILITY

Nominated accountability

A named individual is responsible for clinical safety activity, with the authority to require changes or halt a release.

02 / HAZARDS

Hazard identification

Systematic identification of ways the system could contribute to harm, including missed indicators, over-escalation, misinterpretation and misuse.

03 / CONTROLS

Risk controls

Design, process and training controls proportionate to each hazard, with an explicit record of what each control depends on.

04 / TESTING

Testing

Verification that controls behave as specified, including safety-critical paths and behaviour under unusual or distressing input.

05 / RESIDUAL RISK

Residual-risk review

Documented judgement on the risk that remains after controls, reviewed by people with appropriate clinical expertise.

06 / INCIDENTS

Incident management

Defined routes for reporting, investigating and learning from incidents and near misses, with feedback into design.

03

Human-in-the-loop operation

Designed to augment human support, not displace judgement

Where automated analysis influences triage, escalation or prioritisation, qualified or appropriately trained personnel should retain responsibility for review and action within a defined service model.

Outputs are inputs to a decision
Indicators, summaries and priority ordering are presented as information for a professional to weigh, with their basis visible and their uncertainty stated.
Traceability to source
A professional can see what an output was derived from, so that they can judge it rather than accept it.
Defined service model
Roles, response times, coverage and record-keeping are agreed with the deploying organisation before automated analysis influences any pathway.
Override and suspension
Professionals can disagree with an automated indicator, and BioTekk can suspend or modify an automated function while a concern is investigated.
Escalation levels · illustrative
04

Continuous monitoring

Governance that
continues after
go-live

Safety assurance is not completed at launch. Model behaviour, service performance and organisational context all change, so monitoring, evidence generation and information governance are treated as continuing obligations.

Performance drift

Monitored against the baseline established at evaluation, with investigation triggered by change rather than by complaint.

Incidents & near misses

Recorded, investigated and reported through agreed routes, including near misses that caused no harm.

Information governance

Lawful basis, data-protection assessment, retention and access reviewed per deployment and re-reviewed on material change.

Evidence generation

Evaluation of defined functions in the populations and settings for which they are intended, reported with their limitations.

Development status

BioTekk is in active development. Statements on this page describe the governance approach the product is being built to satisfy. They are not claims of completed regulatory assessment, certification or clinical validation. Documentation relevant to a specific deployment is provided during pilot discussions.

Bring your
governance team.

Clinical safety officers, DPOs and information-governance leads are the right people to be in the room early. Tell us who they are and what they need to see.