Connect
PetConnect
Cortex
An AI-powered SOAP system prepares a draft that a veterinarian can review against the consultation. The useful question is not simply how quickly text appears. It is whether the clinician can find the source information, correct the draft and finish a record that accurately reflects the visit.
In Vetigen, the SOAP product walkthrough shows how Veti helps organize supplied information into reviewable fields. This guide explains a practical review process. It does not promise a fixed reduction in examination time or treat a generated draft as a completed clinical assessment.
Start with the correct patient and visit#
Before entering a note, confirm which patient and encounter it belongs to. Separate information from the current consultation from details in an earlier record. If a fact came from the owner, preserve that distinction instead of turning it into an observed finding.
These checks matter before and after drafting. A polished note attached to the wrong visit is still wrong. A missing date or an unclear source should become a question for review, not a detail the system is expected to invent.
Use SOAP as a review structure#
SOAP separates subjective information, objective findings, assessment and plan. The structure helps a reviewer ask different questions of each section. It does not make an unsupported statement clinically valid.
Section | Review question |
|---|---|
Subjective | Is this what the owner reported, with the relevant timing and context? |
Objective | Does it match the observations, measurements and results actually recorded? |
Assessment | Is this the responsible clinician's assessment, with uncertainty preserved? |
Plan | Does it distinguish a proposed action from something already completed? |
Consider a documentation-only example: the source says, “Owner reports a change since yesterday; examination findings are recorded separately; follow-up timing has not yet been agreed.” A draft should not turn that into a confirmed diagnosis, normal examination or booked appointment. The review catches changes in meaning, not just spelling.
See the product workflow before creating a clinical record#
See the SOAP workflow in Vetigen. The current product page includes the actual product screens and explains the voice or text input, draft fields and clinician review. Use that walkthrough to identify which part of your existing note process you want to evaluate.
For a first exercise, use an approved example without real client or patient identifiers. Supply a short note, review each field against the source, and note what needed correction. Do not add real records to an evaluation until your clinic has agreed the appropriate data handling and access arrangements.
Review meaning before accepting the draft#
Pay particular attention to negation, dates, units and the distinction between historical and current findings. Also check whether the draft added information that was not supplied or moved an uncertain statement into a definitive assessment.
The RCVS guidance on AI in practice explains that professional responsibility remains with the veterinary team and that AI-generated records need appropriate verification. It is UK professional guidance; local obligations still apply elsewhere.
Use two passes: first check identity and transcription, then review the clinical meaning. If a correction affects another section, reread both. Where information is missing, leave a clear question for the responsible clinician rather than filling the gap with an assumption.
Define when a draft is finished#
Agree who reviews a note, how unresolved questions are handled and what the team means by “complete.” A colleague should be able to tell whether a record is awaiting review without asking the author at the end of every shift.
Sharing information with a pet owner is a separate decision. Finishing a SOAP note does not itself authorize an automated message or a new clinical instruction. Check the content, recipient and responsible person before using a record in another communication.
Evaluate the whole task#
Measure the time from preparing the source information to completing the reviewed record. Keep drafting, correction and final review visible so a faster first draft does not hide more work later. Record examples of omissions, unsupported additions or confusing section placement.
Start with a small, consistent set of permitted examples. Discuss why each draft was accepted or returned for correction, then adjust the team's checklist. The AI note review checklist provides a practical next step; the SOAP introduction explains the underlying format.
To explore your clinic's documentation workflow, use Request a call on this page. You can also Create your free account to explore Connect; check current plan details before assuming a particular AI feature or usage allowance is included.
