How laboratory billing workflows actually work

The stages, in order

• Order — the test, the provider, the diagnosis and the patient
• Eligibility — whether coverage exists, with which payer, and what it covers
• Coding — CPT for the tests performed, ICD for the reason, modifiers where required
• Claim preparation — assembly and scrubbing against payer-specific rules
• Submission — through a clearinghouse to the payer
• Adjudication — the payer's determination
• ERA / remittance — the electronic explanation of what was paid and what was not
• Posting — payment applied against the claim and reconciled
• Patient responsibility — deductible, coinsurance and non-covered amounts
• Follow-up — denials, appeals and accounts receivable

Nearly every failure originates at stage one or two

Look at a laboratory's denial reasons and the pattern is consistent: coverage not in effect, patient not found, missing or invalid information, missing prior authorization, diagnosis not supporting medical necessity. Almost all of them are intake failures.

The information that would have prevented them was available when the order was placed. It was either not captured, not validated, or captured in a system billing could not read. The denial forty-five days later is a delayed notification of a problem that already existed.

If denials are being worked but denial reasons are not being fed back into intake, the same denials will keep arriving indefinitely.

Why the disconnect between LIS and billing is expensive

When billing runs in a separate system fed by re-keyed data, three costs compound. The re-entry itself is a duplicated intake process. Discrepancies appear between what was tested and what was billed. And the feedback loop from denial back to cause is broken, because the two systems do not share the record that would connect them.

When the claim is built from the same order the laboratory ran, the codes come from the catalog, the demographics and coverage come from the validated intake record, and a denial can be traced to the specific intake step that caused it.

Work queues beat worklists

A billing team working chronologically will spend its day on whatever is oldest rather than whatever is most recoverable. The alternative is queues organized by what is blocking payment: missing information, failed eligibility, denial by reason, aged AR by payer and balance.

That ordering also makes patterns visible. Fifty claims blocked on the same missing field is not fifty problems; it is one problem, upstream, that can be fixed once.

The measures worth watching

• Clean claim rate — the share submitted without rework
• Days in AR, and the shape of the aging distribution rather than only the average
• Denial rate by reason and by payer
• Time from result release to claim submission
• Collection against expected reimbursement, not against billed amount
• Eligibility exception rate — the leading indicator for everything above

A note on promised improvements

Any vendor quoting a specific collection improvement without seeing your data is quoting someone else's laboratory. Outcomes depend on payer mix, test menu, documentation quality, client behaviour and the state of the existing AR. The useful conversation is modelling against your actual figures during scoping.