Somewhere between a quarter and a half of specialty referrals never result in a completed consultation, depending on which study and which specialty. The figure is quoted often enough to have lost its shock value, which is unfortunate, because it represents a very large volume of clinical intent evaporating.
Where referrals actually fail
Instrumenting the path from order to consultation note usually reveals that the technology handled its part. The failures cluster at the human steps: the patient was never contacted to book, the specialist’s office requested records that were not sent, or the consultation note arrived and nobody reviewed it.
- Ordered but never scheduled — the largest single bucket in most practices.
- Scheduled but not attended, with no follow-up.
- Attended, note returned, never read by the referring clinician.
- Attended, note never returned, never chased.
A low-technology fix that works
Assign the chase
One named person owns open referrals and reviews them on a fixed cadence. This is unglamorous and consistently effective. Most practices that improve referral completion did so by assigning the work, not by procuring software.
Close the inbound loop too
A returned consultation note that nobody reads is a failed referral with better paperwork. Route notes to a queue with a review expectation, and measure the time from receipt to acknowledgement.
When a platform does help
At scale, across many external specialists, with a genuine need for bidirectional status. Below that threshold, the platform mostly digitises a process that is failing for reasons the platform does not address.
