A patient-intake flow that had grown one field at a time for four years. We rebuilt it around what the clinic actually needs before a first appointment, and cut abandonment by more than half.
Intake had been extended once a quarter for four years. Every new regulation, specialty and insurer added a field, and nothing was ever removed. By the time we saw it, a new patient answered sixty-one questions across nine screens before booking.
Staff had built workarounds. Front desk filled the form on the phone for anyone who called, which meant the digital flow was measuring the patients least able to use it.
We sat with intake staff across three sites and logged which answers actually changed what happened next. Twenty-two of the sixty-one fields were never read before the first appointment. Nine were duplicated from insurance records already on file.
The clinical team needed six answers before a patient arrived. Everything else could wait for the visit, or be pulled from a system the clinic already paid for.
Three screens: who you are, what brings you in, and how you are paying. Anything derivable is prefilled and shown for confirmation rather than asked. Conditional depth replaced the flat list, so a patient only sees a specialty question if their reason for visiting triggers it.
The deferred fields moved into a short form the patient completes in the waiting room, where staff can help. Nothing was deleted from the record, only relocated to where it gets answered accurately.
Abandonment fell 58% in the first full quarter. Median completion time went from eleven minutes to four. Phone-assisted intake dropped by a third, which returned roughly six hours a week to the front desk at each site.
The pattern library we built alongside it now carries the clinic's other four patient-facing flows.
βThe team did not ask us what we wanted the screen to look like. They asked what a nurse needs to know before the patient walks in, and designed backwards from that.β