Practice Operations
Pathology Referral Intake From Practices That Send Specimens
Pathology referral intake is client-practice intake, not patient intake. How AI works requisitions, referral sources and status calls inside athenaOne.
Pathology referral intake looks nothing like intake at a clinic. Almost nobody calling your lab is a patient. The work arrives as requisitions from other practices, and the person on the phone is a medical assistant at an office twenty miles away who needs to know whether the specimen you received yesterday has been accessioned yet.
A pathology group is a business-to-business front office wearing a clinical coat. Every case starts with a form somebody else filled out, in a hurry, between patients.
So the failure modes are all upstream of you. The ordering provider box has a group name instead of a person. The identifier that tells billing what to do is missing. The specimen site is written on the container but not on the paperwork. The practice that sent it changed management companies last quarter and its old account is still the one on file.
None of that stops the case from arriving. It stops the case from moving, and the delay lands on your lab even though the gap came in with the form.
The caller is a practice, and that changes the whole script
Front-office automation built for patient calls does not fit here. There is no appointment to book and no reminder to send. What there is instead is a small number of call types, repeated constantly, from a known set of accounts.
Status on a case. An add-on request. A correction to a form already sent. A question about which container or fixative a given order needs. A new office asking how to start sending work.
Each one is answerable from data, and each one currently costs a person several minutes because the caller has to be identified, the case has to be found, and the account has to be confirmed before anything can be said. Practices report that phones remain a persistent bottleneck on staff time, and a lab feels that in a concentrated form because its call mix is narrow and repetitive.
That narrowness is the opportunity. A short, well-defined call catalog is exactly what automation handles well.
An incomplete requisition is nobody’s assigned job
The single most expensive thing in this workflow is a form that is 90 percent complete.
The case gets accessioned. The work gets done. Then it stalls in billing because the ordering provider cannot be resolved to a real person with a real identifier, or because the reason for the order never made it onto the paper. Somebody has to call the sending office and ask. Usually nobody does, for days.
Automating that call is straightforward and it is worth more than it sounds. The system compares the incoming requisition against what a complete one needs for that account, flags exactly which fields are missing, calls the sending practice, reaches the person who handles their lab paperwork, collects the missing values, and attaches the corrected information to the case.
athenaOne carries the reference data this runs against. Referral sources, referral order types, and the order record on the encounter are structured configuration rather than free text, which means the completeness check is comparing a form against a definition instead of against somebody’s memory.
The handoff is clean and it is early. Anything about the specimen itself goes to your staff. If the question is whether the sample is adequate, whether a different block should be cut, or which test the case actually needs, the call is routed to the lab with the context already gathered. The automation chases paperwork. It never chooses a test.
Onboarding a sending practice is the intake workflow nobody documented
New client setup is where the compounding errors get introduced, because it happens rarely enough that it never becomes routine.
Start with the roster problem. The provider list a practice publishes on its website and the provider list inside its system routinely disagree, so working out who actually orders from there, and who is still there at all, is step one rather than cleanup. Get it wrong and every result you produce for that account routes to a person who left.
Then the coverage problem. Which plans your lab is in network for, and which of them a given practice’s patients actually carry, usually lives in a spreadsheet outside any clinical system. Practices often cannot hand over a clean grid because theirs is a mess too, so the workable artifact is the inverse: a short list of the plans that will not pay. That list is what turns a surprise write-off six weeks later into a conversation at the moment the requisition arrives.
An onboarding call sequence handles the mechanical part of this well. Confirm the ordering roster and identifiers, confirm the delivery channel and the number or interface that actually gets read, confirm who at the office owns lab paperwork, confirm the courier arrangement, and write all of it into the account record rather than into an email thread.
The difference shows up in month two, when the account either runs quietly or generates a phone call a day.
Add-on requests and status calls are the daily volume
The add-on call is the one that most rewards being answered immediately. A physician wants an additional stain or an additional test on material you already hold, and the answer depends on your own policy: how old the accession is, what material remains, and whether the account is set up for it.
That is a lookup and a policy check, not a judgment. The automation confirms the account, finds the accession, checks it against your stated window, and either records the add-on request against the case or routes it to the lab with everything already assembled. Where your policy says a person decides, a person decides, and the call arrives at that person with the case already open rather than as a voicemail.
Status calls collapse the same way. Received, accessioned, in process, released, delivered. Answering that on the first call removes a callback and, more usefully, surfaces the cases that are genuinely stuck to somebody who can act on them.
Inbound paperwork deserves the same treatment. Requisitions still arrive by fax at most labs, and a digital fax that lands in a shared inbox is less paper rather than more automation. The gap between a fax that becomes an indexed document on the right account and a fax that sits in a folder is the gap between a lab that knows its backlog and one that discovers it.
The line, and why it is easy to hold here
Pathology gives you an unusually clean boundary, which makes this one of the safer places to put automation in front of the phone.
The report belongs to the pathologist. Under the federal test report standard a laboratory releases results only to authorized recipients, and a referring laboratory may not revise results or the information tied to their interpretation. That is a rule about who receives, not about who explains, and it maps well onto what the automation is allowed to do.
So the automation moves forms, confirms accounts, chases missing fields, reports status, and routes exceptions. It does not read the report, characterize a finding, tell an office what a case shows, or decide which test to run. Every one of those goes to your staff with the case in front of them.
Patient calls follow the same rule and they are rare. Identity is verified, the caller is told what was released and when, and the conversation is routed to the ordering practice, which is where it belongs.
Key Takeaways
- Treat the sending practice as the caller, and build the automation around a short catalog of account-level call types rather than around patient intake.
- Check every incoming requisition against what a complete one needs for that account, and automate the call that collects the missing fields the same day.
- Reconcile the ordering roster at onboarding, because published rosters and system rosters disagree and every stale name misroutes results later.
- Capture the plans that will not pay for that account as an inverse list, since a clean enrollment grid rarely exists at either end.
- Keep specimen adequacy, test selection, and anything touching the report itself with your staff, and route those calls with the case already assembled.
A pathology lab’s front office is an account management desk that happens to sit next to a bench. The calls are few in kind and many in number, the forms arrive imperfect, and the accounts drift. Automate the completeness check, the onboarding sequence, the status answer, and the add-on intake, and what changes is not the work your pathologists do. What changes is how much of the day gets spent finding out what a form was supposed to say.
Related reading
- pathology results routing back to the ordering practice
- building a provider referral portal in cardiology
- the results callback queue at an imaging center
Sources
Ready to See It in Action?
See how PGA works incoming requisitions and client-practice calls inside athenaOne
Schedule a Demo →Written by Kevin Henrikson