Care Clinics: 468 patient conversations read by hand, none left unanswered
A case study. Care Clinics in Cairo takes patient bookings on WhatsApp with Helm. We read all 468 conversations: none was left unanswered, the median reply took 12 seconds, and 86 bookings reached the front desk's sheet with no field out of place.
The short answer. Care Clinics, a clinic group in Cairo, takes patient bookings on WhatsApp with Helm. Patients book in the chat, and each booking lands in the front desk's sheet with the visit type and insurance already filled in. To 4 September 2026 we read all 468 conversations one by one. None was left unanswered, the median reply took 12 seconds, and 86 bookings reached the sheet with no field out of place.
How it works
- A patient writes on WhatsApp, at any hour.
- Helm answers, finds a slot and takes the booking.
- The booking lands in reception's Google Sheet, visit type and insurance included.
- Patients who go quiet get one follow-up.
The numbers
To 4 September 2026:
- 468 conversations, every one read by a person, with no sampling.
- 0 left unanswered.
- 12 seconds median reply, and 90% within 44 seconds.
- 86 bookings, all present in reception's sheet: none missing, none extra, none duplicated.
- 0 mismatches across the 516 booking fields checked against the sheet.
What this shows, and what it does not
Every conversation was read by a person, so the counts are not a classifier's guess. The period ends on 4 September 2026. The follow-up for patients who go quiet started later, and its results are not in these numbers.
If you run a clinic
How Helm works for clinics and hospitals · Helm for health · How a pilot works
Helm builds the workflow that does this work in production, in your channels and your systems.
Start a workflow ↗All posts ↗