A practice runs claims in athenahealth. When a payer rejects a charge, the agent that proposes a fix needs a place to record the rationale, the CPT or ICD-10 edit it suggests, and which biller approved the resubmission. Dock holds that row. athenahealth holds the claim. The two systems point at each other, and the audit trail survives staffing changes, payer disputes, and downstream questions from the healthcare operations team.
athenahealth and Epic stay the system of record for the claim itself, the encounter, the patient chart, and the remittance. Dock is the system of record for what the billing agent interprets from that data. Each Dock row carries athena_claim_id (or epic_claim_id for Epic-integrated practices), the agent's identity, the proposed correction, the human reviewer, and a timestamp. When the agent needs current claim state, it re-fetches from the athenahealth API rather than trusting a stale snapshot in Dock.
The Dock surface: claim_review_queue
| athena_claim_id | denial_code | agent_rationale | proposed_edit | reviewer | status |
|---|---|---|---|---|---|
| ATH-884217 | CO-11 (dx not consistent with procedure) | Encounter note supports 99214; original coding used 99213. AAPC guidance on E/M level selection points to MDM complexity here. | Resubmit as 99214 with modifier 25 | k.nguyen@practice | approved |
| ATH-884301 | CO-97 (bundled service) | 36415 is bundled into 80053 per payer policy. No separate reimbursement available. | Write off $18.00 | s.alvarez@practice | approved |
| ATH-884355 | CO-16 (missing referral) | Referral on file in athena under prior auth ATH-PA-44219. Not attached to claim line. | Attach PA-44219, resubmit | pending | awaiting_review |
Three rows. Each one points back at an athenahealth claim and forward at a human who took responsibility for the decision.
A worked workflow
The billing agent pulls the daily denial report from athenahealth via API. For each denial, it reads the encounter note, the payer's policy, and the relevant coding guidance. It writes one Dock row per denial with its proposed action and rationale. A biller opens the queue, reviews the row, and either approves the resubmission, requests a revision, or rejects it. Approval triggers the agent to post the corrected claim back to athenahealth. The Dock row records the resubmission ID and closes. No edits happen in athenahealth without an approved Dock row, which makes agent audit and compliance reviews straightforward.
Why this matters
Practice-management software treats a claim as a single object with a status field. That object cannot hold the reasoning that produced a resubmission, the alternate codes the agent considered, or the biller who signed off. When auditors or payers ask why a code changed, the answer lives in a free-text note or, more often, nowhere. Dock gives that reasoning a row.
It also fixes the attribution gap. A shared service account posting to athenahealth tells you nothing about which agent ran which logic on which claim. With agent identity attached to every row, the practice can answer the question "who decided this" for any line item, going back as far as the rows exist.
The same pattern carries to the practice's books. The write-offs and resubmitted balances flow into the same attributed surface the controller already uses for Dock for accounting work, so revenue cycle and ledger reconcile against the same evidence.
Start with the healthcare operations pillar and the compliance overview to see how the pieces connect.
FAQ
Does Dock change codes inside athenahealth automatically? No. The agent proposes the edit in a Dock row. A human biller approves it. Only then does the agent call the athenahealth API to resubmit. The approval is the consent gate.
What if athenahealth and Dock disagree on claim status? athenahealth is the source of truth for status. The agent re-fetches before acting. Dock rows record what the agent saw and decided at a given moment, not the current claim state.
Does this work for Epic-integrated practices too?
Yes. The row schema swaps athena_claim_id for epic_claim_id and the agent reads from the Epic API instead. The Dock layer, the rationale, the reviewer, and the audit trail behave identically. See agent identity for how the same agent can carry credentials across both.
What standards back the agent's coding decisions? AAPC publishes the CPC certification standards and coding guidance the agent cites in its rationale (AAPC certifications). Volume context for U.S. practices comes from the AHA's annual hospital survey, which tracks 6,100 hospitals and over 35 million admissions (AHA Fast Facts).
