GPs submitted referrals into a system that returned nothing until an appointment appeared weeks later or a rejection arrived with no reason. Referrals were re-submitted, duplicated and escalated out of pure uncertainty.
STUDY 24 / 30 · ANONYMISED · NDA · HEALTHCARE · NETHERLANDS · NATIONAL
HEALTHCARE · NETHERLANDS Referral triage that tells the GP what happens next.
RoleLead designer
Timeline13 weeks
Team2 designers, 1 researcher, 1 PM, 2 frontend, 1 clinical SME
VerticalCare coordination platform · primary to secondary care
Made triage state and expected wait visible from the moment of submission, and made rejection carry a reason and a corrected path rather than a terminal notice.
Duplicate referrals dropped sharply and rejection re-submission became targeted, which reduced the load on secondary care triage.
-64%DUPLICATE REFERRALSvs baseline
82%REJECTIONS CORRECTED ON FIRST RETRYfrom 21%
4.2 / 5GP-REPORTED CONFIDENCEfrom 2.3
THE ARGUMENT
Why the obvious solution was wrong.
The study matters because the product problem was reframed before the interface was polished.
A referral disappearing into a system with no acknowledgement produces exactly the behaviour you would predict: the GP, responsible for the patient and given no information, refers again elsewhere, chases by phone, or escalates. Every one of those actions loads the system further. The duplication was not GPs behaving badly. It was the only rational response to an interface that returned nothing.
Submission now returns a triage state immediately and updates it as the referral moves - received, triaged, prioritised at this level, expected wait in this band. Rejections carry the specific reason and the corrected path: this needs a particular test result first, this belongs with a different specialty, this patient meets criteria for a faster route. Eighty-two percent of rejections are now corrected on first retry against 21% before, because the GP is told what to fix rather than that they were wrong.
THE INTERFACE CRAFT
The interaction, rendered as a working product surface.
The specimen below is code-native and uses the study's own design logic. The client interface remains protected.
HEALTHCARE
ALARM PRIORITYNeeds action
Room 408 · 02:14
DESIGN DECISIONS
Positions we would defend.
Each decision names the principle and the product consequence, not a stylistic preference.
Silence generates load
An interface that returns nothing generates duplicate work as a rational response. Acknowledgement is a capacity intervention.
Honest bands beat false dates
Clinicians can work with a range and its basis. They cannot work with a date that slips, and neither can the patient.
Design the rejection
Rejection is the highest-information moment in the flow and was carrying the least information.
PRODUCT LEADER READOUT
What transfers, and what should remain specific to this product.
A case study is useful when its operating principle travels without turning the original interface into a template.
Read the operating condition
For Care coordination platform · primary to secondary care, the transferable lesson is not a copied screen. It is the condition the interface had to make legible: The referral's state and priority band are visible immediately and update as it moves. Silence is what produced the duplication. Rebuild that visibility for your own roles, risk, terminology, and operating cadence.
Protect the design rule
An interface that returns nothing generates duplicate work as a rational response. Acknowledgement is a capacity intervention. Keep that rule in the acceptance criteria, component states, and production QA record so later visual cleanup cannot erase why the interaction exists.
Measure behaviour after ship
The evidence record is -64% for duplicate referrals, vs baseline. Recreate the baseline and outcome window before rollout, segment the result by role and context, and state clearly what the measure cannot prove.
RESEARCH RECORD
The work behind the interface.
These artefacts connect the final interaction back to the evidence and product model that produced it.
Referral pathway trace
Followed 60 referrals end to end across four specialties, recording every duplicate, chase and escalation and its trigger.
GP interview series
18 interviews on what they tell the patient after referring, and what they do when the system tells them nothing.
Rejection cause taxonomy
Coded 900 rejections into 14 causes with secondary-care triage clinicians; 11 had a clear corrective action.
Wait band validation
Established which wait bands could be honestly published at what confidence with the capacity planning team.
“I can tell the patient something true. Before, I could only tell them I had sent it.”
GP partner, primary care network · under NDA
NEXT STUDY · 25 / 30 · HEALTHCARE