A specialist practice: three clerks returned to clinical hours
A regional specialist practice was keeping its most sensitive records in spreadsheets and third-party survey tools. Six weeks later, the records live where the work is done, three clerks are back on c
The reading
Twelve practitioners. Three clerks. Patient records held across a desktop spreadsheet, a shared survey tool used as the record of truth, and a filing cabinet that nobody opened after the migration four years ago.
The practice knew what the problem was in general terms — "we need proper records" — and had been quoted three times for software that would have taken ninety days to install and nine months to adopt. The quotations opened at a figure that would have obliged them to recoup it over five years, during which neither the regulator nor the industry would stand still.
We were asked to read the operation for a week and come back with the three things worth doing. We came back with one.
What we found
The practitioners were doing their work in the room with the patient. The clerks were doing the practitioners’ work after the fact — transcribing paper notes, chasing missing fields, resolving where a survey export disagreed with a spreadsheet. Roughly 12 hours a week per practitioner was being spent on tasks the clerks could not finish because the record did not yet exist in a form that could be finished.
Two compliance regimes applied. Both had been passed on the strength of best-effort narratives rather than demonstrable record. The next inspection would have gone the other way.
What we built
A tablet record at the point of care. The practitioner enters the finding once, in the room, on the device. The record writes itself to the central ledger when the tablet is in range of a signal; when it is not — which is most home visits — it writes to the device and reconciles when the clinician is back on the practice’s network. There is no second transcription anywhere in the workflow.
Six weeks from enquiry to first ledger. The build shipped with its own account: three figures the practice would see move within the first quarter.
The numbers at ninety days
| What | Before | Ninety days in |
|---|---|---|
| Clerical hours per practitioner / week | 12 | < 1 |
| Spreadsheets of record | 7 | 0 |
| Compliance regimes with demonstrable audit trail | 0 | 2 |
| Average record completion time (minutes) | 18 | 4 |
| Records with missing mandatory fields | 1 in 4 | 1 in 180 |
Three clerks moved to clinical hours. One of them returned to the qualification she had set aside when the paperwork took her off the floor in 2019.
What we would have done differently
Nothing in the build. One thing in the sequencing: we would have retired the survey tool before shipping the tablet, not after. The two months of overlap gave practitioners a legitimate excuse to keep entering findings in both places — which doubled the data-entry burden they were being freed from. A clean cut, with the old tool switched off on day one, would have been uncomfortable for a week and better for a quarter.
The standard
A new clerk can keep the books in a week. That is the test. The record exists at the point of work, it reconciles itself, and the compliance story is told by the ledger rather than by a narrator.
When the record is where the work is done, the clerks are freed to do the work.