The empty 2pm chair
A patient forgets, the slot goes unused, and someone on the waitlist was never called. The clinic pays for both.
See the national pictureDieses Beispiel wird gerade überarbeitet. Der Text unten ist noch auf Englisch.
Example · Healthcare clinicsReminder and rescheduling agent
Fewer empty chairs, without more calls
An agent that sends the reminder, handles a reschedule inside your rules, offers the slot to the waitlist, and passes every clinical question to a clinician.
Illustrative scenario — not a real client result
Tomorrow's list · 17:40
- Answered
24 reminders delivered
17:41 · Three replies: two confirmations, one reschedule
- Answered
11:20 slot released
17:44 · Offered to the waitlist, first reply holds it
- Person
“Is this pain normal?”
17:46 · Clinical question · with a clinician
Clinical questions never go to the agent
Reminder and rescheduling agent · Healthcare clinics
- Sourced facts
- Clinicians decide
- A person, whenever asked
01 — A typical day
Three small failures that add up.
What healthcare clinics owners tell us goes wrong in an ordinary week.
Reminders depend on one person
The receptionist sends texts between calls. On a busy day they go out late, or not at all.
See the cost of manual remindersClinical questions in the inbox
“Is this normal?” arrives by text. A general agent must never answer it, and the person who can is with a patient.
See the handover rule01 — A typical day · the evidence
What the numbers actually say.
Each figure opens its source, year, sample and limits. Click a badge to check it.
01
Source
GOV.UK / Department of Health, missed outpatient appointments
2016
England outpatient data for 2014–15: about 5.6 million appointments missed, 9% of the total.
England only, historical administrative data, not a current rate.
Open the source9%
of NHS outpatient appointments were missed in 2014–15
About 5.6 million appointments in England, from national administrative data. Historical, not a current rate.
GOV.UK / Department of Health · 201602
Source
GOV.UK / Department of Health, missed outpatient appointments
2016
England outpatient data for 2014–15: about 5.6 million appointments missed, 9% of the total.
England only, historical administrative data, not a current rate.
Open the source5.6M
outpatient appointments missed in one year
The same England dataset. A national scale, not a clinic benchmark.
GOV.UK / Department of Health · 201603
Source
Cochrane review, mobile messaging reminders
2013
Cochrane review: seven trials, 5,841 participants; SMS reminders versus none, risk ratio 1.14 (95% CI 1.03–1.26).
Relative effect, older studies, mixed settings. Not a 14-point increase.
Open the source14%
higher attendance with text reminders
A Cochrane review of seven trials (5,841 participants) found a risk ratio of 1.14. Relative, not percentage points.
Cochrane review · 201302 — What the helper does
From your records to a decision you own.
The agent reads tomorrow's list and sends the reminder in the clinic's name, with the practical details and a clear way to change the time.
A reply is either a confirmation, a reschedule inside your rules, or a question. Clinical, urgent and distressed messages go straight to a person.
A released slot is offered to the waitlist in order. The record is updated and the day's list is reported to the clinic.
6 steps · 2 decisions you own
Reads from
- Practice management system
- SMS
- Phone notes
Text version of the flow
- 1. Read tomorrow's list — Appointments, rules, consent.
- 2. Send the reminder — Time, place, what to bring.
- 3. Handle the reply — Confirm, reschedule, or pass on. Human checkpoint: You set the rescheduling rules.
- 4. Offer the freed slot — Waitlist, in order.
- 5. Hand over anything clinical — Symptoms, medication, distress. Human checkpoint: A clinician answers clinical questions.
- 6. Report the day — Filled, missed, waiting.
02 — What the helper does · a simulated example
One simulated example, step by step.
Use the buttons or the left and right arrow keys. Nothing here is a real customer.
Simulated example
Riverside Physio
SMS · 17:40
A four-clinician physiotherapy clinic. The agent handles tomorrow's reminders and one released slot. Every message is simulated; no real patient data is used.
You are talking to Riverside Physio's AI assistant, not a clinician. Reply HUMAN at any time for a person.
03 — What you decide
The helper prepares. You decide.
The agent handles the diary. Clinicians handle the patient.
You set the rules
Notice period, allowed slots, and which appointment types always need reception.
You set the consent
Who may be contacted, on which channel, and how to opt out.
A clinician takes over
Symptoms, medication, results and distress go to the clinical queue, not to the agent.
You can pause it
Stop reminders for a clinician, a day or the whole clinic.
Without your approval · Licensed clinicians decide every clinical matter. The agent never gives medical advice, triages symptoms, or changes a clinical plan.
04 — What changes when a safeguard is missing
Documented cases, not scare stories.
Two documented findings and two labelled limits. None is a clinic-level result, and none is a claim that automation alone fixes attendance.
Missed first appointments cost up to £225m
The NAO estimated that missed first outpatient appointments cost the NHS up to £225 million in 2012–13. It is an “up to” estimate for first appointments, not current spending.
What the helper would do differently
The agent targets the recoverable part: reminding, rescheduling and refilling a released slot, with a daily count of what was actually recovered.
£225M
upper estimate, England 2012–13
Text reminders help, but the effect is modest
Across seven trials (5,841 participants), text reminders improved attendance with a risk ratio of 1.14. Two studies also reported text costs per attendance 55–65% lower than phone reminders.
What the helper would do differently
The workflow starts with the evidence-backed part — a timely reminder and an easy change — rather than promising to eliminate no-shows.
RR 1.14
attendance with SMS versus none
Wording matters, and one trust is not every clinic
In a London trial of 10,111 patients, mentioning the cost of a missed appointment reduced the DNA rate from 11.1% to 8.4%. Mobile numbers existed for only about 20% of eligible patients.
What the helper would do differently
The agent tests one clinic's wording against its own baseline, and reports the share of patients it could actually reach.
8.4% vs 11.1%
DNA rate with a cost message vs the existing one
A 2014–15 figure is not today's rate
The 9% and 5.6 million figures come from national administrative data for 2014–15. They are England-only and historical, and should not be read as a current clinic benchmark.
What the helper would do differently
The page labels the year and country, and the pilot measures the clinic's own no-show rate before and after.
9%
missed outpatient appointments, England 2014–15
04 — What changes · the evidence
The rest of the evidence.
Every figure resolves to a source with its sample and limits. No figure without one.
14%
higher attendance with text reminders
A relative risk ratio of 1.14 from seven trials; not a 14-point rise.
Source
Cochrane review, mobile messaging reminders
2013
Cochrane review: seven trials, 5,841 participants; SMS reminders versus none, risk ratio 1.14 (95% CI 1.03–1.26).
Relative effect, older studies, mixed settings. Not a 14-point increase.
Open the source9%
of NHS outpatient appointments missed in 2014–15
About 5.6 million appointments in England. Historical administrative data.
Source
GOV.UK / Department of Health, missed outpatient appointments
2016
England outpatient data for 2014–15: about 5.6 million appointments missed, 9% of the total.
England only, historical administrative data, not a current rate.
Open the source8.4%
DNA rate with a cost-framed reminder
Versus 11.1% with the existing message in a Barts Health trial. One trust, not a national benchmark.
Source
PLOS ONE, SMS wording and DNA rates at Barts Health
2015
Two randomised trials at Barts Health, London; 10,111 patients in Trial One. Cost-framed reminder: 8.4% DNA vs 11.1%.
One trust; mobile numbers available for only about 20% of eligible patients.
Open the source
04 — What changes · your week
The same week, two ways.
Drag the handle, or use the left and right arrow keys when it is focused.
Typical week, beforeTypical week, after
Mon
Reminders go out between calls, sometimes after 6pm.
Tue
Two patients arrive at the wrong time; one slot sits empty.
Wed
The waitlist is a paper note; nobody calls the third person.
Thu
A clinical question waits in the inbox until the clinician has a gap.
Mon 17:40
All reminders sent. Confirmations update the record automatically.
Mon 17:44
A released slot is offered to the waitlist in order, with a 20-minute hold.
Tue
The list is accurate; reception starts with the two patients who need a call.
Thu
The clinical question is in the clinician queue with the patient record, not the general inbox.
Illustrative example · not a measured result
05 — Your version
Four questions about your own week.
Keep your answers. They are the brief for the first version we would build for you.
How many appointments does the clinic book in a normal week?
What may be rescheduled without asking a clinician, and how close to the visit?
How is the waitlist managed when a slot opens?
Who answers a clinical question, and how does it reach them?
Next
Put your own figures into the calculator below. It runs in your browser, and nothing is sent or stored.
Put your week into the calculatorBring the four answers to the free one-hour call. We will use them to sketch the first version, and you keep the written plan either way.
Talk about your business05 — Your version · your numbers
Put your own week into it.
An estimate, not a promise. The model and its assumptions are printed below.
Your inputs
All clinicians, not just the pilot list.
count
Confirmations and simple changes inside your rules.
%
Sending, calling back, updating the record.
min
Pay plus employer costs.
per hour
Evenings and weekends.
%
From patient message to updated record.
hours
Example defaults are typical for a small business. Replace them with yours. Nothing is stored, sent or shared.
Estimated result
IllustrativeReception hours back each month
21–28
- Staff time released
- £392–£530
- After-hours replies handled the same day
- 24–28%
- Time to a reschedule confirmation
- Under 1 hour
How we calculated this
Routine work handled by the helper: 70% of the share you marked routine.
Time saved per item: 80% of the minutes you entered.
Monthly volume: weekly figures × 4.33; monthly figures used as entered.
Cost: hours saved × your staff cost per hour, in pounds as an illustration.
Reply time: routine items move within the hour while the helper runs. Complex work still waits for a person.
Questions
The things people ask next.
No. Symptoms, medication, results and distress go to a clinician. The agent acknowledges the message and says who will reply.
Yes. The first message says it is the clinic's AI assistant and that a person is available by replying HUMAN.
Urgent wording bypasses the rules and alerts reception. The agent does not assess urgency itself.
It reads and writes through the system's API where one exists. If not, we start with a daily export and a review queue.
Data stays in the clinic's systems, is encrypted, and is kept for the period the clinic sets. It is not used to train a model.
Two to three weeks, including consent checks, message testing and a two-week parallel run with reception.
Start here
Show us tomorrow's list.
One hour with a recent day's schedule and your rescheduling rules. We will show you where the gaps are and what the agent would handle.