The Practice Manager's Guide to Referral Tracking

It’s 4:30 on a Friday. A GP’s receptionist calls: Dr Chen sent a referral three weeks ago, and the patient has been ringing them asking why nobody has been in touch. You check the inbox, the fax tray, the shared drive. Eventually the letter turns up wedged between two pathology reports on the scanner.

Three weeks, no action, one unhappy patient, and a referrer now quietly wondering where else to send their work.

That is almost never a competence problem. The referral existed and the team was busy. Nothing in the practice was watching it move.

Why the spreadsheet stops working

Most practices start with the best of intentions and a spreadsheet: patient, referring GP, date received, status. It holds up fine while one person can still hold the whole list in their head.

Then volume grows, staff change, someone forgets a row, and a second copy appears on a different desktop. The failure isn’t that the spreadsheet is wrong; it’s that nobody can tell whether it’s right. Once staff stop trusting it, the real tracking system becomes “I think I remember seeing that one.” Spreadsheet vs software sets out where that line usually falls.

The Australian Institute of Health and Welfare reports millions of referred specialist attendances a year, every one of them beginning as a document somebody had to receive, review and action. Volume alone makes ad-hoc tracking untenable.

Five stages, five places it stalls

Every referral takes the same path whether you track it or not. Knowing the stage names matters far less than knowing the specific way each one fails and the single check that catches it.

Booked is the stage that deserves more than a line, because it is where referral leakage happens: the patient was referred and never seen. It’s easy to underestimate from inside a practice, because you see the appointments that happened, not the ones that quietly didn’t. A systematic review in the Journal of General Internal Medicine found only 35% of specialty referrals result in a documented visit. Not all of that gap is administrative, but the administrative share is the part you control. MDA National’s analysis of lost referrals goes further, noting that duty of care can extend to whether the patient actually attends, not merely whether an appointment was offered.

What “good” actually looks like

You don’t necessarily need software. You do need the same three properties, whatever you build them out of.

So the design brief for any system, paper or otherwise, is short: a single source of truth, stage visibility, and something that speaks up when a referral stops moving. A referral board delivers all three, which is why SimpleRef is built that way. Tell us how your practice triages and we build those stages into your board.

Two numbers worth keeping

Most practices either measure nothing or try to measure everything and abandon it by March. Two numbers do most of the work.

The first is time from receipt to first patient contact: the gap between a referral landing on your desk and the patient hearing from a human being. It is the single most sensitive indicator of whether your intake and triage are keeping up, because it lengthens before anything else visibly breaks.

The second is the count of referrals with no movement in fourteen days. Not a percentage, not an average, but a raw count, because a raw count is a worklist. Every item on it is a specific patient somebody can pick up this afternoon.

Both are useful only if the definition never changes. Agree what “first contact” means, write it down, and resist the urge to refine it each quarter; a mediocre metric measured consistently beats a perfect one measured three different ways. If you want a broader set to grow into, the practice manager’s KPI guide covers the rest.

Where practices get stuck

Two objections come up almost every time this framework is proposed internally.

“We already know which referrals are waiting.” Usually one person does: the one who has been there longest. That is not a tracking system. It is a single point of failure with annual leave. The test isn’t whether someone knows; it’s whether anyone else could find out in thirty seconds.

“We’ll fix it when things quiet down.” They don’t, and the backlog is worst precisely when the practice is busiest. Start with a fortnight of counting rather than a project plan. Counting costs nothing and tells you whether the problem is intake, triage or patient contact. That is the difference between fixing the right stage and buying something that addresses the wrong one.

Start with one question

If you’re currently running on spreadsheets and memory, don’t try to overhaul everything at once. Pick one number and watch it: how many referrals received in the past fourteen days have not been actioned?

Ask it at the same time each week. A fifteen-minute standing meeting is enough. The discipline is in asking consistently, not in the meeting itself. A number that moves in the wrong direction tells you where to look long before a GP’s receptionist does.

If the answer surprises you, work out what it’s worth before deciding what to spend fixing it. The referral leakage calculator does the arithmetic on your own volumes.

The goal isn’t perfection. It’s visibility. You can’t fix what you can’t see.