Partial ERA Import Completeness: Can File Success Hide Missing Lines?

Research question: Can source and import evidence identify ERA transactions that never reached the posting batch?

Scope: This September 9, 2026 study concerns administrative evidence available to an outsourced medical billing operation. The unit of analysis is one source 835 transaction and its local import disposition. It does not evaluate clinical care, individual staff merit, legal compliance, contract entitlement, or the correctness of a payer's final decision.

Methodology: Freeze the population, observation window, source report, filters, cutoff, timezone, duplicate rule, and exclusions before reviewing results. Reconcile file control values, transaction and service-line counts, import messages, batch rows, deposits, and unresolved exceptions without relying on a successful file-level status. Retain routine cases, exceptions, reopened work, missing evidence, and access failures in the reported denominator.

Measurement plan: Study 1 reports raw numerators and denominators with every rate. Event time, receipt time, system-entry time, retrieval time, and review time remain separate. Each statement is labeled observed, externally reported, calculated, owner-interpreted, or unknown.

Reliability check: Give a mixed sample to a second authorized reviewer who cannot see the first classification. Compare cohort membership, source selection, identifiers, dates, calculations, exclusions, and category. Preserve disagreements and document whether resolution required a clarified rule or owner judgment.

Analysis: Report source-confirmed, conflicting, missing, access-blocked, superseded, interpretation-required, and unresolved results. Keep sequence distinct from cause. A later event can complete a chronology without proving why an earlier event occurred.

Role boundary: Support staff may retrieve approved evidence, preserve source wording, compare fields, reproduce arithmetic, and prepare a neutral exception. Qualified owners retain coding, coverage, medical-necessity, credentialing, contract, privacy, accounting, submission, refund, write-off, and patient-communication decisions.

Inference limits: Findings describe only the frozen cohort, systems, definitions, and observation window reported here. They should not be generalized to another payer, team, or period without replication. Association, order, and missing records do not establish causation or responsibility.

Limitations: Vendor transformations, inaccessible raw files, split deposits, and delayed corrections can prevent complete reconstruction. Non-random selection, interface changes, incomplete history, access restrictions, and unresolved records can bias the observed distribution. Unknowns remain visible rather than being removed to improve the result.

Conclusion: The design can locate an evidence-chain break and frame a bounded question for an authorized owner. It cannot forecast acceptance, payment, appeal success, liability, or clinical correctness.

Source reviewed September 9, 2026: https://www.cms.gov/medicare/coding-billing/electronic-billing/electronic-remittance-advice

Source reviewed September 9, 2026: https://x12.org/products/health-care-claim-paymentadvice-835

Source reviewed September 9, 2026: https://www.caqh.org/core/operating-rules