Every minute a fertility doctor spends typing into the system is a minute not spent with the patient. VRepro includes native AI speech recognition, the doctor dictates freely during the consultation and the system automatically transcribes and structures the dictation into the correct fields of the IVF clinical record, without interrupting the flow of care.
Contents
The problem: how much time clinical documentation takes in fertility
A follow-up consultation at a fertility clinic lasts between 10 and 20 minutes. The clinical record for that consultation, the ultrasound findings, hormonal parameters, protocol adjustment, patient instructions, plan for the next control, can take an additional 5 to 10 minutes of typing.
With 20 to 25 consultations per day, that adds up to between 100 and 250 minutes of pure documentation. Nearly two hours that the doctor spends typing instead of caring, thinking or resting.
And in the embryology laboratory, where EU SoHO Regulation now requires immutable traceability records for every sample manipulation, thorough documentation is not optional, but it does not have to be manual.
How AI speech recognition works in VRepro
VRepro’s speech recognition is a native feature of the platform, not an external add-on or a third-party service integration. This means the transcription engine is designed specifically for the clinical vocabulary of assisted reproduction, and that data never leaves the secure VRepro environment.
Most medical dictation systems transcribe speech into a free-text field. The doctor then has to copy, paste and organise manually. VRepro goes a step further: the AI automatically identifies which clinical field each part of the dictation belongs to and structures it without manual intervention. It is not a transcriber, it is a clinical assistant that understands the context of assisted reproduction.
Before and after: what changes in daily clinical practice
- Doctor attends patient and takes handwritten notes
- Manually transcribes to the system after the consultation
- Navigates between screens to complete each field
- 10–12 min of documentation per consultation
- Risk of error from delayed transcription
- Incomplete records under high consultation pressure
- Accumulated fatigue by end of schedule
- Doctor dictates while attending the patient
- AI automatically structures into the correct fields
- Doctor reviews and validates in 1–2 minutes
- 2–4 min of documentation per consultation
- Complete, immediate record after each consultation
- More time for the patient in every appointment
Why native matters
Privacy and GDPR compliance
With an external integration, the audio from the medical dictation, which contains patient health data travels to a third party’s servers for processing. With VRepro’s native dictation, processing occurs within the platform’s secure environment, with no data transfer to external systems. In the context of the EU SoHO Regulation and GDPR, this eliminates a significant layer of compliance complexity.
Fertility-specific vocabulary
Generic dictation systems have high error rates with specialised clinical terminology, protocol names, stimulation drugs, embryo grading classifications, ultrasound terminology. VRepro’s engine is trained specifically on assisted reproduction vocabulary, resulting in significantly higher transcription accuracy in the real clinical context.
Direct integration with the clinical record
Being native, dictation does not generate an intermediate text that the doctor then has to review and organise, it generates directly the structured fields of the VRepro clinical record. That is what enables the real reduction in documentation time, not simply shifting the work from typing to reviewing.
- No transfer of voice data to external systems
- Vocabulary trained for assisted reproduction and embryology
- Automatic structuring into fields, not free text to organise
- Available without additional configuration for all VRepro users
- Structured data integrates in real time with traceability and laboratory modules
- Audio recordings are not stored, only the resulting structured text
Want to see VRepro’s AI dictation in a real consultation workflow?
We’ll show you how speech recognition works in your clinical team’s specific workflow, with a demo tailored to your protocols.
Request a VRepro demo
No commitment · Assisted reproduction specialists · Response within 24h
Frequently asked questions
Does VRepro’s speech recognition require prior voice training?
No. The system is pre-trained and works from the first use without any voice adaptation period. As the doctor uses dictation, the system can refine its accuracy for the specific vocabulary and speech patterns of each user over time.
What happens if the doctor makes a mistake during dictation?
The review step is designed precisely for this. Before the record is saved to the clinical history, the doctor sees the structured transcription and can correct any field. Correction is fast because the structure is already in place, only the content that needs adjustment is modified.
Are voice recordings stored?
No. VRepro processes the audio to generate the transcription and does not store voice recordings. Only the resulting structured text is retained in the clinical record. This ensures compliance with the GDPR data minimisation principle, no more information is retained than necessary for the clinical purpose.
How does AI dictation interact with EU SoHO traceability requirements?
AI dictation improves the quality of traceability records by reducing the friction of manual documentation, resulting in more complete and immediate records. Each transcription validated by the doctor is digitally signed with a timestamp in the system, meeting the immutable record requirements that EU SoHO Regulation establishes for reproductive SoHO manipulations.
Is VRepro’s AI speech recognition available for clinics outside Spain?
Yes. VRepro is used by fertility clinics in more than 28 countries across Europe, LATAM, the Middle East and Asia. The AI speech recognition feature is available across all markets where VRepro is deployed. For clinics in the UK, Germany, the Netherlands and Nordic countries, NaturalSoft’s local implementation teams provide onboarding and configuration support in the relevant languages and regulatory contexts.
