TL;DR: A multi-physician primary care practice in Philadelphia spends six weeks every year assembling a 340-page evidence binder for accreditation review. The practice is compliant. It's been compliant for years. The problem is proving it, which requires pulling documentation from seven disconnected systems and formatting it to the auditors' preferred structure. One person (Linda, the practice manager) carries the institutional knowledge of what the auditors want and where to find it. We built a four-stage agent that maps every evidence source, monitors compliance status continuously instead of annually, assembles the binder automatically, and routes gaps to responsible staff the week they appear rather than the week before the audit. Binder assembly: six weeks down to three days. Audit findings: two in the past three years (both documentation gaps, not compliance failures) down to zero. Agent cost: $260/month.
The Scavenger Hunt
Every February, Linda clears her calendar. Not for a holiday. For the auditors.
The accreditation review is in April. Between now and then, Linda will assemble a 340-page evidence binder that proves the practice did what it was supposed to do over the past twelve months. Credentialing current for all physicians. Continuing education completed. Infection control protocols followed. Patient satisfaction surveys conducted. HIPAA training documented. Incident reports filed. Equipment maintenance logged.
The practice did all of these things. Linda knows this because she was involved in most of them.
The problem is not whether the clinic is compliant. It is. The problem is proving it. Which requires pulling documentation from seven systems: the EHR, the credentialing database, the training platform, the HR system, the incident reporting tool, the patient feedback portal, and a shared Google Drive full of PDFs that Linda maintains with the quiet desperation of someone who knows the folder structure makes sense only to her.
This is, when you think about it, a process whose fundamental purpose is to assemble evidence that something happened. Not to make it happen. Not to check whether it happened. Just to prove, to someone else's satisfaction, that it did. The work was done months ago. Linda is spending six weeks writing the proof.
I don't want to suggest that spending six weeks proving you did work you already did is an absurd use of a practice manager's time. But I will observe that Linda's salary during those six weeks costs the practice roughly $8,400, and that the physicians sign attestations and review documents at a combined cost of another $3,200, bringing the annual price of proving compliance to approximately $11,600. Which is the cost of being regulated, apparently. Nobody had questioned whether it could also be the cost of having seven disconnected systems.
The Practice
Multi-physician primary care practice in Philadelphia. Six physicians, four nurse practitioners, twelve clinical staff, four admin including Linda. Primary care, chronic disease management, preventive health, minor procedures. Accredited by a national body requiring annual compliance reviews.
Compliance areas span credentialing (physician licences, DEA registrations, board certifications), training (HIPAA, infection control, BLS/ACLS), patient safety (incident reporting, medication reconciliation), quality metrics (patient satisfaction, clinical outcomes), and facility standards (equipment maintenance, fire safety, ADA compliance). The documentation lives in seven places. Seven. I counted.
Linda's process: a 47-item checklist. For each item, identify the evidence source, retrieve the document, verify it's current, format it to the auditors' preferred structure, compile. Forty-seven items. Seven systems. Six weeks at roughly half her working day.
The practice's audit findings over three years: two. Both documentation gaps. A training record filed in the wrong system. A maintenance log missing one quarterly entry. Neither was a compliance failure. The training happened. The maintenance happened. The proof was just stored in the wrong place.
The practice failed not at compliance but at filing. And the entire six-week annual ritual exists because "stored correctly" and "findable by auditors" are, in a seven-system environment, not the same thing.
Why Nothing She Tried Worked
Linda evaluated a compliance management platform. $400 to $800 per month. Designed for larger health systems. Required migrating data from all seven systems. Linda's assessment: "I'd spend three months setting it up, and then I'd still need to pull from the systems it doesn't integrate with." The platform solved the problem of having multiple systems by adding an eighth system. Which is a solution in the same way that giving someone a ninth plate to spin solves the problem of spinning eight.
She tried starting earlier. November instead of February. Finished a week sooner. Same number of hours, just spread across more weeks. The problem wasn't the timeline. It was the retrieval.
She tried delegating sections to department leads. Two of four completed their assignments. The other two didn't, because treating patients was, understandably, a higher priority than assembling documentation for Linda's binder.
Her best idea: a "living audit binder" updated monthly. She maintained it for four months. Then flu season hit, two staff members quit, and the binder fell three months behind. By February, the full sprint anyway. The living binder died of the same disease that kills every maintenance-dependent system in a busy practice: the maintenance isn't anyone's primary job.
The design is next. The design, the build guide, the running costs and the take-away are free to subscribers. Enter your email below and the rest of this Blueprint opens. New ones arrive every Tuesday and Friday.
