Referral Loops That Never Close

Referrals fail quietly in both directions. Neither failure produces a message, a queue or a complaint, which is why a practice tends to learn about one from a patient months later, or from a referring office that has stopped sending.

A referral is two events and most practices record only the first

Placing it is an order. It gets entered, it leaves the building by fax or portal or in a letter the patient carries out, and in the chart that is the end of the trail. The second event is a consultation note or a report coming back and reaching the person who asked the question, and nothing in the ordinary course of a day prompts anyone to notice when it does not arrive.

That asymmetry is why this stays invisible longer than most operational problems. A denial argues with you. It appears in a work queue with a dollar amount and a deadline attached. A referral that went nowhere produces nothing at all.

The inbound side is the one with revenue attached

Referrals arriving at the practice are appointments somebody has already decided to send. They come through a fax line, a payer portal, a direct message into the EHR, and sometimes a phone call answered by whoever was closest. Each route lands somewhere different, and in a practice of about fifty people at least one of those destinations is not a work queue with a named owner reading it daily.

What becomes of an unclaimed one is undramatic. It sits a week, the patient calls the next name on their list or goes back to the referring office, and the visit happens elsewhere. The practice's own data holds no record of a referral that never became a patient, so this gets discovered sideways, by a referring office asking why their patients are not getting in. Access decides part of it before anyone touches the fax: where the third next available appointment is weeks out, some share of inbound referrals resolves itself somewhere else regardless of how fast the queue moves.

In the chart it looks like an order with nothing after it

Read later, by a covering physician or in response to a records request, an order with no result behind it is a question that was raised and not answered. The record does not say which of several things happened: the patient declined, or could not get an appointment anywhere close, or went and the report never came back, or was never called in the first place. Four situations with one appearance, and separating them afterwards is a phone call per patient.

A document arriving is not a person reading it

Exchange has improved a great deal, and in Ohio a fair amount now moves through the state health information partnership rather than a fax machine. What arrives is a document landing in a pool, routed by rules somebody configured during an implementation years ago and has not looked at since.

So the failure point moved rather than closing. A loop that used to break at a fax machine now breaks inside a results pool holding several thousand documents, which is harder to see and feels more like a solved problem. The ordering provider's inbox is the test. If the report came back into the building but not in front of the person who asked, the loop is open in every sense that matters clinically.

The gap belongs to two practices, so it belongs to neither

Every side of it holds a reasonable assumption. The ordering office assumes the receiving one will schedule the patient and send a note back, while the receiving office assumes a patient who wants the appointment will call. The patient assumes somebody would be in touch if it mattered. Nobody is wrong and the referral still sits there.

Around Cincinnati a growing share of these go to physicians employed by one of the systems, which changes the shape of the problem rather than the problem. The consultation happens and the note goes into that system's record, where it is genuinely closed, and that is not the same as closed in yours. The patient is now also inside an organization able to schedule everything else they need. A referral relationship is how a practice stays visible to the physicians it depends on, which is one reason this gets watched closely in some practices and is uncomfortable to look at in others.

What closes loops is a list and a fixed day

The operational form is dull and has not changed much: referrals placed with no result after a defined interval, inbound referrals not yet scheduled, and a named person reading both on a set day. The interval is a judgment about which kinds of referral can wait and which cannot, and that judgment is specific to a practice rather than general.

The first run of the list is the worst one. It surfaces a backlog built up over however long nobody was looking, and every line on it is a call to a patient about something months old. That cost is paid once. After it the list is short, and the loops on it are open now rather than open since last spring.