UK MHRA Guidance: Adverse Event Reporting in Digital Mental Health Technologies (DMHTs)

The proliferation of Digital Mental Health Technologies (DMHTs)—ranging from self-help apps and online therapy services to artificial intelligence–powered chatbots—has transformed how mental health care is delivered in the UK and globally. However, with the increased adoption of these platforms comes a growing need for robust mechanisms to monitor patient safety, especially in reporting and acting upon adverse events (AEs).

On 1 August 2025, the Medicines and Healthcare products Regulatory Agency (MHRA) published a landmark document Report on adverse event (AE) reporting in digital mental health technologies (DMHTs) and the development of updates to regulatory guidance.

An “adverse event” (AE) in the context of DMHTs encompasses any unwanted, unintended, or harmful outcome that may arise during use. These can range from software glitches (such as loss of crisis alerts), and misleading information, to unintended psychological harms (like exacerbated distress, anxiety, or self-harm triggers).

The use of digital mental health technologies (DMHTs) has expanded rapidly in recent years, offering scalable tools for depression, anxiety, PTSD, and wellbeing. Many DMHTs now use:

• Artificial intelligence and machine learning

• Behavioural nudges and gamification

• Self-guided or therapist-supported modules

These products promise increased access and personalisation, but they also introduce new safety risks — for example:

• Digital content triggering distress

• Algorithm errors giving harmful advice

• Data misuse impacting mental health

The MHRA recognised that traditional medical device reporting frameworks may not capture the unique nature of these risks.

The report points out several specific challenges:

• Dynamic content and updates: Apps may change daily, so static risk assessments are insufficient.

• User-driven pathways: Many DMHTs depend on self-report or voluntary engagement, making harm underreported.

• Non-traditional harms: Emotional distress, anxiety escalation, or social stigma are harder to detect than physical events.

• Complex causality: It can be unclear if harm stems from app design, user context, or external events.

Existing definitions of adverse events (AEs) were found to be too narrow for the digital mental health context.

The updated guidance includes illustrative examples to help manufacturers and clinicians identify reportable AEs:

• Clinical Outcome Failures:An app designed to guide users through crisis situations fails to trigger an emergency alert during a suicidal episode—an immediately reportable AE.

• Information and Content Errors:Misinformation generated by a chatbot—such as suggesting hazardous coping strategies, or providing out-of-date treatment guidance—must be logged and investigated.

• Data Handling Incidents:Unintentional disclosure of patient-sensitive information (e.g., a privacy breach) is now viewed as a safety AE with reporting obligations, not merely an IT issue.

• User Interface Flaws:Updates that inadvertently remove critical warnings or safety prompts, potentially leading to harm, are included as AEs.

For full details, read the official MHRA report and guidance:

👉 Report on adverse event (AE) reporting in digital mental health technologies (DMHTs) and the development of updates to regulatory guidance