A private hospital: four department diaries unified into one list
A private hospital in the South West was scheduling theatres across four departmental diaries, with the matron reconciling by phone. The CEO called it the hospital's most expensive manual process.
The reading
Six theatres, four surgical specialties, roughly 180 procedures a week. Each department kept its own diary: orthopaedics in a shared Outlook calendar, general surgery in a specialist bookings system, ophthalmology in an Excel workbook with macros, and plastics in a paper diary the consultant preferred.
The matron spent her mornings on the phone, reconciling. She was good at it; she had been doing it for twenty-two years. She was also sixty-one and had mentioned retirement twice in the preceding six months.
The CEO's view was that the hospital was one retirement away from a very public operational failure. He was correct.
Four diaries, six theatres
The conflicts were not primarily between departments but between resources: anaesthetists, specialist nurses, particular items of equipment. A theatre could be free but a required specialist nurse could be in another theatre. The departmental diaries did not know about each other, and did not know about shared resources at all.
The matron held the resource map in her head. When a department asked for a slot, she thought for a moment and said yes or no. The answer was correct about 94% of the time; the 6% caused the day's crises.
- 4 departmental diaries, 3 technologies, 1 paper
- 180 procedures per week with shared resources
- Matron spending 3.5 hours a day on reconciliation
- Shared resource conflicts discovered on average 2.3 hours before the procedure
- No queryable record of theatre utilisation by specialty
What we built
We built a unified scheduling layer that represented theatres, anaesthetists, specialist nurses, and specialist equipment as first-class bookable resources. Each department continued to work in its own system; we read from each and wrote a consolidated view that showed conflicts in real time.
The decision we spent the longest on was whether to replace the four diaries or layer above them. We layered. Replacing them would have taken eighteen months and would have failed; the plastics consultant was not going to give up his paper diary and the ophthalmology macros represented a decade of accumulated local knowledge. We photographed the paper diary each morning and OCR'd it into the shared view. It worked.
The stack: a Node service with a Postgres resource graph, scheduled polls of the three electronic diaries, and a morning intake process for the paper one.
The numbers at ninety days
| What | Before | After |
|---|---|---|
| Matron time on reconciliation | 3.5 hr/day | 40 min/day |
| Conflicts caught before procedure day | 60% | 97% |
| Procedures delayed for resource reasons | 11/week | 2/week |
| Theatre utilisation reporting | None | Weekly |
| Days of lead-time for anaesthetist booking | 2 | 7 |
The person and the system
The matron is still at the hospital. She retired in month seven as planned, but accepted a two-day-a-week consulting role because the hospital could now afford to use her for judgement rather than reconciliation.
That is the correct outcome of a system-first engagement: the experienced person stays, but their time is spent on the things only they can do. The reconciliation is now the system's job.