The Referrer Base Is Your Department's Capital

The orthopedist handing over his practice after 28 years was never an employee of the hospital. Yet a substantial share of the department's elective occupancy hung on him. His successor is 34, does not know the chief physician, and so far refers to the hospital where he himself trained as a resident. No controlling report captures this event. In the occupancy statistics, it will become visible in a year.

Someone keeps asset accounts for every ultrasound machine in the department. For the relationships that carry the occupancy, hardly anyone keeps books. Yet referrer relationships may be a department's most valuable capital — tied to individual people, built over years, damaged in weeks.

How concentrated the capital is

A few orders of magnitude from anonymized operational data of rehabilitation sites we work with: one site records around 23,500 referral inquiries per year — about 94 per working day — of which roughly 11 percent lead to an admission. Another facility counts around 9,000 inquiries for about 900 admissions, ten to one. And across the sites, one pattern repeats: 20 to 30 practices generate over 80 percent of the referral volume. The figures come from rehabilitation; the orders of magnitude shift by specialty, but in our observation the concentration on a few practices recurs wherever referrals are elective.

The consequence of this concentration: the loss of a single one of these practices — through retirement, practice succession, a disgruntled colleague, or simply a faster competing offer — has a measurable occupancy effect. A department that cannot name its 20 to 30 load-bearing practices is managing its most important capital blind. How ten inquiries become one admission (in German), we have described in detail elsewhere.

Why nobody keeps books

The hospital information system knows cases, not relationships. The referring practice appears as a field in the admission dataset; it is rarely analyzed. The real relationship knowledge — who sends what, who expects which callback, whose practice is about to be handed over — lives in two heads: the chief physician's and the front office's. Both are on vacation at some point, and at some point one of the two leaves. What is missing then never stood in any system.

Added to that is an asymmetry of perception. The hospital experiences the referral as one transaction among many that day. The practice experiences it as its one case, on which it must be able to answer to the patient. The relationship is decided at this asymmetry — long before anyone talks about quality.

The hardest currency: response time

What does a hospital earn a practice's loyalty with? Not the summer party. The practice judges the hospital at the points where it is the one waiting: how quickly after a registration does a usable response arrive — a confirmation with a date, a precise request for specific documents, or a substantiated message that the case does not fit? And how quickly after discharge does the letter arrive with which the practice can continue treatment? The benchmark for that is set: the German framework agreement on discharge management requires the provisional discharge letter on the day of discharge [1].

The present looks different at many sites. At one pilot site, 1,000 to 2,000 faxes arrive per month, five to forty pages, often incomplete. Every missing document costs a follow-up request loop, every loop takes days — and for the practice waiting for an answer, every one of these loops feels like silence. A practice copes better with a substantiated no on the same day than with a maybe that takes three weeks. The no is a response; the maybe is an imposition the practice remembers.

What a department can do

Five measures. None of them needs a budget; all of them need someone who feels responsible.

1. Keep the list. Name the 20 to 30 practices that carry the volume — with case numbers, contact persons and upcoming handovers. A practice succession in the referrer base deserves the same attention as a senior physician succession in the hospital. Where the list is missing, an afternoon with controlling helps: the referrer identifier from the last two years of admission data is enough for a first draft.
2. Measure the response time. Time from receipt of the inquiry to the first usable response — that is the metric the practice actually feels. What is not measured does not improve.
3. Bundle follow-up requests. Request once, completely and precisely, with a concrete document list, instead of three times in stages. Every avoided loop saves both sides days.
4. Send the letter on time. The same-day provisional discharge letter is not just a contractual duty — it is the document by which the practice re-evaluates the hospital after every single treatment.
5. Respond even on non-admission. Fast, substantiated, where possible with an alternative path. The practice does not remember the hospital's admission rate; it remembers whether it got an answer.

None of this needs software. All of it becomes easier when inquiries, documents and response status live in one place rather than in fax piles — what matters when selecting such tools is described here (in German). For the practice side, aiomics is currently preparing a referrer portal that bundles submission, completeness feedback and status updates (in pilot preparation).

Three questions for the next leadership meeting

1. Can we name our 20 to 30 most important referring practices — and do we know which of them will be handed over in the next three years?
2. Do we know our median time from receipt of an inquiry to the first usable response?
3. Does a practice whose patient is not admitted receive a substantiated response — and how fast?

If you want to walk through these questions for your department, write to us. Or subscribe to our weekly briefing Visite (German; English edition Grand Rounds is in preparation), our letter on documentation, economics and AI in German healthcare.

Sources

1. Rahmenvertrag Entlassmanagement nach § 39 Abs. 1a SGB V (GKV-Spitzenverband, Kassenärztliche Bundesvereinigung, Deutsche Krankenhausgesellschaft). https://www.gkv-spitzenverband.de
2. Anonymisierte Betriebs- und Volumendaten von Rehabilitationsstandorten und Interessenten (2025/2026), aiomics; Erfahrungswerte, keine kontrollierte Erhebung.