Practice Operations
How Pain Management Records Requests Stop Delaying Procedures
Pain management records requests, attorney releases, and prior conservative care documentation stall procedures. How AI clears them inside athenahealth.
Pain management records requests do not behave like anyone else’s. An attorney wants two years of chart notes for a personal injury case. A workers’ compensation adjuster wants the same file on a different form with a different signature block. A disability carrier wants a questionnaire back by Friday. And underneath all of it, a payer will not approve next month’s injection until somebody proves the patient already tried six weeks of physical therapy. That last one is not a records request in anybody’s workflow diagram, but it is the same work, done by the same person, against the same clock.
Most specialties treat records as a back office task that runs a week behind without much consequence. In interventional pain, records sit directly on the revenue path. The documentation packet is what turns a scheduled procedure into an approved one, and it gets assembled by whoever has a free minute between phone calls.
The volume also arrives in a shape no small team absorbs well. A two to ten physician pain practice generates litigation records requests, workers’ comp requests, disability paperwork, and payer documentation pulls at the same time, each with its own requester, its own authorization rule, and its own destination. Nobody owns the category, so it gets owned last.
And the cost of being late is not a complaint. It is a procedure slot that goes empty, or worse, a procedure that happens and gets denied on the back end because the conservative care documentation was never attached.
Four requesters, four rules, one phone line
Sort what actually arrives and it splits four ways. The patient asking for their own copy. A third party asking under a signed release, meaning an attorney, an insurer, or another practice. A form that needs completing and signing, which is not a records request at all. And a payer or an internal auth coordinator asking for documentation to support a pending authorization.
Each carries a different authorization rule, a different turnaround expectation, and a different owner. The first decision on every call is which of the four you are in, and it gets made by whoever answered the phone with a patient standing at the window.
The delivery format splits too. Information-blocking rules under the 21st Century Cures Act now require practices to provide electronic access when a patient asks, yet attorney, workers’ comp, and payer requests still arrive by fax and mail. Most practices are running a paper pipeline and a portal pipeline at the same time, and the caller decides which one they are in without knowing they are deciding anything.
An AI front office layer sorts at intake rather than downstream. It asks the questions that separate the four paths, captures requester, patient, date range, destination, and the reason for the request, then files it against the right athenahealth document class and department bucket before the call ends. Sorting at the front is what keeps an auth documentation pull from sitting in the same undifferentiated pile as a school form.
The clock starts when the request arrives, not when someone opens it
Under 45 CFR 164.524, a practice must act on a request for access within 30 days of receiving it, with one 30 day extension available if the patient is notified in writing. Hitting that deadline is not usually the hard part. Knowing when day zero was is.
Day zero is a Friday voicemail, a fax that printed at 4:50pm, or a portal message nobody opened until Monday. By the time somebody logs it, four of the thirty days are already spent and no one can say with confidence when it landed.
Automating intake fixes the timestamp before it fixes anything else. Every request, whether it came by phone, portal, or fax, gets captured at the moment it arrives, logged against the chart, and routed to the correct athenahealth department bucket with a due date attached. The caller hears what happens next and when, which is the single thing that prevents the second call.
The authorization decision stays with people. Identity verification, the scope of what a signed release actually covers, and anything touching sensitive record categories go to staff. The AI moves the paperwork and reports where it stands. It does not decide what is releasable.
The authorization packet is a records job wearing a billing costume
Interventional pain runs on prior authorization more than almost any other outpatient specialty, and payer rules differ by site of service, so the same injection needs a different packet depending on whether it happens in the office or the surgery center. What the payer wants is almost always documentation that already exists somewhere: prior conservative therapy notes, imaging reports, the last procedure’s outcome, a current medication list.
The work is finding it, not deciding it. Somebody has to pull the physical therapy notes from an outside practice, confirm the imaging report is the read and not just the order, and attach the whole set to the authorization before submission. That is records work, staffed as billing work, and it is why so many packets go out incomplete.
The timeline is tightening from the payer side too. Under the CMS Interoperability and Prior Authorization Final Rule, impacted payers must send prior authorization decisions within 72 hours for expedited requests and seven calendar days for standard ones, and must give a specific reason for denials. Faster payer turnaround only helps a practice that can assemble a complete packet on day one. If your documentation takes nine days to gather, a seven day payer decision window is not a gift.
This is where the scheduling side bites. At a multi-site pain practice, a meaningful share of procedures were getting scheduled outside the prior authorization window by human schedulers, which meant the appointment existed before the paperwork could possibly support it. The fix was not a faster typist. It was making the booking rule and the documentation chase the same workflow: a visit needing no authorization gets booked same or next day, one that does gets its earliest offer set out far enough that the packet can be assembled and submitted first.
An AI layer runs the gathering half. It calls the outside physical therapy office for the notes, confirms the imaging report has been received, opens the follow up task in athenahealth against the right document class, and calls the patient when a signature is the missing piece. The clinical content of the packet, and the decision about what supports medical necessity, belongs to the practice.
Workers’ comp and litigation records are a separate operational animal
A personal injury attorney’s office and a workers’ compensation adjuster do not behave like a patient. They call repeatedly, they escalate, they have their own deadlines, and they will call a fourth time rather than wait. In a pain practice this traffic is permanent, not seasonal.
They also have requirements a general records workflow does not anticipate. Comp requests are often scoped to a date of injury rather than a date range. Attorney requests frequently arrive with a release the practice cannot accept as written. Both types generate a status call cycle that runs for weeks.
The automatable part is the whole wrapper. Confirming which release is on file and whether it covers what was asked for, collecting the correct return fax or address, logging the date of injury scope, chasing a missing signature, and answering status honestly at seven in the evening when the requester is finally free to call.
What stays with staff is every judgment call: whether a release is legally sufficient, what a comp scope actually includes, and anything involving sensitive record categories. The handoff point is explicit, and the person picking it up gets the full request history attached rather than a sticky note.
The status call is the volume, not the request
Count the calls attached to a single records request and it is almost never one. There is the request, then did you get it, then did you send it, then the receiving office calling to say nothing arrived. The first call is short. The rest are not, and they all land during clinic hours.
MGMA Stat polling on where phone time actually goes puts eligibility and prior authorization at 45% of the most time consuming call work, scheduling at 31%, and intake at 9%. Intake reads small until you look at what makes the other two long: missing paperwork, incomplete demographics, and prior records that never showed up. A lot of what gets counted as authorization time is really documentation chasing wearing a different label.
Status is the easiest thing to automate and the least often automated, because answering it honestly requires knowing the real state of the request instead of repeating what was said last time. Once intake is captured against the athenahealth document record, status becomes a lookup. The AI answers it on the first ring at 7pm, and calls the requester when the record actually ships rather than waiting to be asked again.
Anything ambiguous still routes to a person. Identity mismatches, releases that do not cover the request, and anything needing a clinician to sign go to staff with the whole history attached.
Key takeaways
- Sort records traffic into four paths at intake, patient copy, third party release, forms completion, and authorization documentation, because each has a different rule and a different owner.
- Timestamp every request the moment it arrives across phone, portal, and fax. The 30 day clock in 45 CFR 164.524 starts on receipt, not when somebody opens the queue.
- Treat the prior authorization packet as records work, not billing work. Gathering outside physical therapy notes and confirming imaging reads is the bottleneck, not the submission.
- Set the booking rule and the documentation chase as one workflow, so a procedure needing authorization is never offered a slot the packet cannot reach.
- Give workers’ comp and attorney requests their own intake path with date of injury scoping and release sufficiency checks, and route every judgment call to a person.
- Track repeat status calls per request. It is the cleanest read on whether your intake captured enough on first contact.
Records work in a pain practice does not show up on a productivity report and it does not bill. It shows up as a procedure slot that went empty because the authorization packet was three documents short, an attorney’s office calling for the fifth time, and a front desk that cannot get to the phones. None of that requires touching what a clinician does. It requires capturing the request the moment it lands, chasing the paperwork on a schedule, and telling people the truth about where it stands.
If your practice runs on athenahealth, the pieces are already there: document classes, department buckets, the follow up task queue, the patient portal. What is missing is somebody working them at seven on a Tuesday evening.
Related reading: prior authorization automation for physical therapy visit limits, how family practice medical records requests clear faster, and orthopedic referral intake and the fax queue.
Sources:
- MGMA, Practices split regarding patient requests for electronic access to their medical record
- eCFR, 45 CFR 164.524 Access of individuals to protected health information
- CMS, CMS Interoperability and Prior Authorization Final Rule CMS-0057-F
- MGMA, Phones are still a bottleneck costing medical practices time
Ready to See It in Action?
See how Pretty Good AI captures records requests, chases release forms, and assembles authorization documentation inside athenahealth.
Schedule a Demo →Written by Kevin Henrikson