Healthcare · Hong Kong

Patient record cleaning for a Hong Kong multi-branch dental clinic

Three messy systems, one clean patient list.

-100%duplicate records
+30%reachable customers
HK$27,000saved / month

Three branches, three booking systems, and the same patients appearing in each with clashing details. We merged everything into one trustworthy, de-duplicated list — and keep it clean automatically.

The challenge

Around 4,000 duplicate records meant reminders went to dead numbers, recalls were missed, and no one trusted the reports. Front-desk staff across the three branches lost around 245 hours a month fixing records, rebuilding recall lists and calling dead numbers. Cleaning it by hand had been tried twice and abandoned twice.

The solution

A matching engine that compares name, phone, email and visit history across all three systems, merges confident matches automatically, queues uncertain ones for a quick human yes/no, and syncs the clean master list back.

The workflow we built

  1. Extract everything. All three systems exported into one working dataset, untouched originals kept.
  2. Match & merge. Confident duplicates merge automatically with a full audit trail of every change.
  3. Human check. Borderline matches go to a simple review queue — a few minutes a week, not a project.
  4. Stay clean. New records are checked at entry, so duplicates never build up again.

A week before, a week after

TopicBeforeAfter
Patient recordsThe same patient could sit in all three branch systems, with a different phone number in each.One clean master list, synced back to all three systems.
RemindersReminders went to dead numbers, and recalls were missed.Recalls and reminders reach the patient.
Front-desk timeFront desks spent hours every day fixing records, rebuilding recall lists and calling dead numbers.Borderline matches wait in a review queue for a quick yes or no — a few minutes a week.
ReportsNobody trusted the reports, because the counts included duplicates.Confident duplicates merge on their own, each one recorded in the audit trail.
Staying cleanTwo attempts to clean it up by hand had been abandoned.New records are checked at entry, so duplicates don't build up again.

How the saving adds up

  • Each branch's front desk lost roughly 3¾ hours a working day to fixing records, rebuilding recall lists and calling dead numbers: about 82 hours a month.
  • 3 branches × about 82 hours ≈ 245 hours a month.
  • 245 hours × HK$110 an hour ≈ HK$27,000 a month.
  • 245 ÷ 173 hours (one full-time month) ≈ 1.4 full-time staff.
  • HK$110 an hour is about HK$19,000 a month for one full-time admin staff, MPF included.

These are the working assumptions behind the figure, not an audited result.

How the project ran

  1. Free audit. We looked at all three booking systems, and at how each branch's front desk fixed records and built recall lists.
  2. Built on their own files. Matching was built and tuned on a sample export from their own systems, with the originals left untouched.
  3. Side by side. For a period the front desks worked as normal while the matcher's results were checked against staff judgement on the same records.
  4. Live, then monitored. The full merge then ran with its audit trail, and new records are now checked at entry. We monitor it and fix anything that comes up after go-live.

Questions owners usually ask

  • Did anyone lose their job?
    The front-desk hours went back to patients, at the desk and on the phone. What to do with freed-up time is always the owner's decision, not ours.
  • What happens when it gets something wrong?
    Only confident matches merge on their own. Borderline ones wait in the review queue for a person's yes or no, every merge is in the audit trail, and the untouched originals are kept.
  • Where does our data live?
    In their own systems: the clean master list is synced back into them, and the original exports are kept untouched. Who has access to what is agreed with you during the free audit, before anything is built.

Tools we used

PythonGoogle WorkspaceCRM

The results

  • Roughly 4,000 duplicates went to zero, and the audit trail means every merge can be traced.
  • 30% more patients are actually reachable — recalls and reminders now land.
  • About HK$27,000 a month of admin time saved across the three branches — the work of about 1.4 full-time staff.

Representative example — figures illustrative, client anonymised.

Have an operation that looks like this?

Tell us the process and we'll show you exactly how to automate it — free, and no obligation.

Get a Free Automation Audit →