What This Means for Your Practice and Patients
Cleaner data, better triage
We expect significantly fewer general questionnaires and more condition-specific submissions, with RAG ratings that consider the entire request — demographics, free text, and answers all together.
Patients completing condition-specific questionnaires has gone from 79.58% of submissions to 96.08%. This means fewer generic/'condition not listed' submissions by patients to ensure more accurate and relevant information provided.
Shorter time to complete for your patients
Across the board we are seeing patients have an improved, faster experience. Patients feel like the questions asked are more relevant to their conditions and easier to complete.
The average time for a patient to complete a form has gone from 7 minutes 21 seconds to 3 minutes 50 seconds.
Less patient friction, fewer mis-submissions
Patients describe their problem in their own words, then answer ~7-10 targeted questions. No more wading through a generic form and reduced chance of picking the wrong questionnaire.
Even with shortened questionnaires, the number of requests which require practices to request more details has gone from 9.75% to 8.46%.
Unified patient experience
The same condition-specific questionnaires are provided to patients whether they submit a request via your website link, the NHS App, or Andi the AI Voice Receptionist.
Key Considerations
Some changes to be aware of.
Shortened Generic Questionnaire
We've unified the generic patient questionnaire to provide a shortened form. We know many Reception teams use this to complete on behalf of patients. However, we also encourage teams to trial the condition-specific questionnaires while on the phone with patients. This can be enabled via Settings > Org Settings > Manage Content > Clinical Proxy request access. 'Toggle all' and save changes to make them active.
Routine requests are being updated
While still available and although most practices no longer use routine mode, the April contract changes alongside NHS App compatibility requires update to how practices use routine mode.
Assignment routes
Most auto-assignment rules are based on request type (medical vs admin) and will keep working unchanged. A handful of questionnaire-specific routes may need to be recreated against the new questionnaire set.
Auto appointments may require migration
If your practice leverages auto-appointments, please be mindful that these may not be routed if patients use new questionnaires. We’ll provide additional guidance to practices leveraging auto-appointments closer to ensure a smooth migration.
The Patient Flow, Step by Step
A walkthrough of what the patient experiences end-to-end.
1. Medical or Admin
Same starting point as today. Patients still choose between a medical request and an admin request at the top of the flow.
2. Standard Emergency Signposting
The standard NHS red flag screen runs first. Anything ticked here still escalates appropriately.
3. One open text box
Instead of selecting or searching for a questionnaire, patients describe their problem in their own words. “I’ve had a sore throat and a dry cough for two days.”
4. One problem at a time
If a patient describes multiple problems, we ask them to pick one and complete a questionnaire for it. After they submit, we loop them back to the remaining problems so they can submit each as a separate request. No more single submissions hiding three problems in free text.
5. The router picks the questionnaire
We match the description to the most appropriate condition-specific questionnaire. Before the questionnaire begins, patients see: “Sounds like you’ve got a cough or sore throat — press start to continue.” If we’ve got it wrong, they can go back and add detail.
If we can’t find a good match, we ask the patient to provide further detail or drop the patient into the general questionnaire. Spam or completely non-medical submissions will also be filtered out.
6. Improved Condition-specific questions
Targeted, sequenced, and tailored to the patient. A patient’s age and demographics will be considered in questionnaires and specific questions (i.e. pregnancy questions only shared with those with the appropriate demographics). Our questionnaires are now more targeted, averaging 6-7 clinically relevant questions per form.
7. Contextual autofill
If a patient already mentioned “severe left-sided chest pain for three weeks” in the opening text, the relevant questions come pre-selected. They just confirm and move on — keeping a human in the loop without making them repeat themselves.
8. Condition-specific red flags
We’ve heard the feedback. Most condition-specific red flags now flag for the GP rather than auto-escalating to A&E. Genuine emergencies still escalate, but patients will still have capacity to submit symptoms especially if they warrant a same-day GP review.
What Lands in Your Dashboard
The requests that land in your dashboard remain largely the same, meaning operationally there is minimal change for your team. We are just providing added benefit of better questionnaires with more relevant and accurate content, reducing the submission of general questionnaires and submissions with insufficient detail.
Full-context RAG ratings
The red-amber-green scoring that appears in the dashboard considers the entire submission: patient demographics, free text, and questionnaire responses. Your adjustment/correction of the RAG rating will have an active learning loop - meaning that every time you correct the RAG scoring Annie our AI will remember that scoring for next time.
Better, structured information
Condition-specific questionnaires mean positives and negatives are captured cleanly, not buried in free text. Easier to read, easier to triage.