The challenge
A 22-centre physiotherapy chain across South India was losing a third of patients before completing prescribed treatment plans. A typical post-surgical knee rehab is 18-24 sessions over 8 weeks; patients dropped at session 6 once acute pain eased. Recurring billing for session packages was manual — front desk reconciled cash, UPI and cards weekly. Therapists had no view of which patients were at risk of dropping until they had already missed two appointments.
How we deployed
- Triggered session-prep WhatsApp reminders 24h before each appointment with the prescribed exercises for that session.
- Built a between-session adherence check at midpoint — short video prompt, patient logs pain and exercise completion.
- Automated package billing on Razorpay with auto-debit mandates for multi-session treatment plans.
- Trained an AI dropout-risk model on past patient patterns — missed appointments, low pain reduction, low engagement.
- Surfaced at-risk patients to the lead therapist with a recommended re-engagement script per dropout reason.
- Logged outcome scoring (pain VAS, range of motion) into the EMR for cohort-level outcome reporting.
What changed
- Mid-treatment dropout fell from 34% to 12% across all 22 centres.
- Recurring billing operations time dropped 81% — front desk no longer chases payments.
- Patient NPS climbed 22 points across the network.
- Average session-package completion rate hit 87% versus 64% baseline.
- Therapists got a usable dashboard of cohort outcomes by injury type and protocol.
"Patients drop when they feel better, not when they are better. Catching them at session 5 with the right nudge is the entire game."
— Founder · 22-centre physio chain