Same Protocol, Different State, Different Answer: How Regional Coverage Breaks Multi-Site MCAs

THE BILLING GRID
A weekly column on research billing and coverage analysis, by Krishma Shah

A billing compliance officer at a four-site oncology network once described her problem this way: the MCA her team built designated a CA-125 draw as standard of care at every site. Three sites were in the same MAC jurisdiction. The fourth was not. That fourth site billed Medicare. Medicare denied. The denial did not come from a protocol error or a consent gap. It came from a Local Coverage Determination that the MAC governing that site had never issued, meaning no affirmative coverage policy existed for that indication in that region, and the fallback national policy did not close the gap.

That scenario repeats constantly across multi-site trials, and the fix is not complicated once you understand the architecture of how Medicare coverage actually works at the site level.

Why the Same Item Gets a Different Answer

Medicare coverage flows through two channels. National Coverage Determinations, or NCDs, are issued by CMS and apply everywhere. Local Coverage Determinations, or LCDs, are issued by the Medicare Administrative Contractor, known as a MAC, that administers claims for a specific geographic region. There are roughly a dozen MACs in the United States, each covering a defined set of states. When an LCD exists for a given item or service, it controls coverage within that jurisdiction. When no LCD exists, the MAC medical director has discretion to make coverage decisions on a claim-by-claim or policy basis, and that discretion varies.

This means the same CPT code for the same lab draw, under the same protocol, can be covered as a routine cost in a qualifying clinical trial in one state and denied in another, not because the trial changed, but because the MAC changed. The clinical trial coverage rules under NCD 310.1, which governs routine costs in trials that meet the qualifying criteria, establish a floor. They do not override the absence of an LCD in a jurisdiction that has not issued one for a particular item.

For a medical coverage analysis, the MCA, which is the document that assigns each protocol procedure at each visit a billing designation of standard of care, research-paid, or non-billable or invoiceable, this geographic variability is a structural problem that a single-column designation table cannot solve.

The Defensible Unit Is the Cell, Not the Row

A well-constructed MCA assigns coverage designations at the level of the visit-grid cell: the intersection of a specific procedure and a specific visit. The same CBC, which maps to NCD 190.15 for routine blood studies, might be standard of care at the screening visit and research-paid at an unscheduled safety visit where no independent clinical indication exists. That cell-level logic is the baseline expectation for a defensible MCA under current billing compliance standards.

Regional coverage adds a second dimension. When a trial enrolls at sites in multiple MAC jurisdictions, the honest MCA has a region column, or a separate coverage matrix per region, that reflects what the controlling LCD actually says in each geography. A designation of standard of care that is valid under Novitas Solutions guidance in one state is not automatically valid under WPS Government Health Administrators guidance in another.

The citation discipline here matters. Writing that a procedure is covered as a routine cost because it is a standard diagnostic test is not a citation. It is an assertion. A defensible citation names the NCD number, or the LCD identifier, which takes the format L followed by a five-digit number, and the MAC that issued it. An auditor following up on a denied claim will pull the LCD and check whether your site’s state falls within that MAC’s jurisdiction. If it does not, the citation is decorative.

How to Cite an LCD Correctly in an MCA

LCD identifiers are public and searchable through the CMS Coverage Database. The correct citation format in an MCA references the LCD identifier, the title of the determination, the MAC name, and the effective date. For items where no LCD exists in the relevant jurisdiction, the MCA should say exactly that and note whether the MAC has issued a billing article, a less formal guidance document, or whether coverage rests entirely on NCD authority or MAC discretion. Leaving that cell blank or copying the LCD citation from a different region is the audit weak point.

For lab procedures specifically, the citation chain is: NCD if one applies, for example NCD 190.15 for routine blood studies, then the applicable LCD if one exists in the site’s jurisdiction, then the ICD-10 codes that the LCD requires for covered billing. If the LCD in a given region restricts a lab draw to specific diagnoses and the protocol population does not uniformly carry those diagnoses, the standard-of-care designation for that cell needs a qualifier or needs to move to research-paid for that region’s sites.

A Worked Example: Imaging Across Two MAC Jurisdictions

Consider a solid tumor protocol that requires CT chest, abdomen, and pelvis every eight weeks for nine cycles. The MCA team correctly preserves the parent schedule rather than decomposing the scan set into individual visit rows, which is a common error that causes MCAs to designate roughly one visit instead of nine. They designate the scans as standard of care, citing an LCD from their primary MAC that covers CT imaging for cancer staging and response assessment.

The protocol adds two sites in a state covered by a different MAC. That MAC has an LCD covering CT for cancer staging but restricts coverage for response assessment scans beyond the initial staging workup to cases where results will change active treatment. The protocol does not require a treatment change decision at each imaging interval; it requires documentation of response. In that second MAC’s jurisdiction, cycles two through nine may not qualify as standard of care under that LCD’s restrictions, and billing them to Medicare as routine costs creates exactly the exposure that the Rush University Medical Center billing settlement made famous: charging Medicare for items that the sponsor is obligated to fund.

The fix is not to reclassify every scan to research-paid across all sites. The fix is to apply the correct designation by region in the MCA, budget the research-paid scans into the invoiceable section at those sites, and price them at the greater of the sponsor rate or the site’s charge master rate. The MCA and the budget reconcile. The audit trail holds.

What Multi-Site MCAs Need From the Start

Any MCA built at the protocol stage, before site selection is complete, should flag imaging and lab line items as jurisdiction-sensitive and note that final designations require LCD review once sites are confirmed. That is not a hedge; it is accurate practice. Protocol-only MCAs are real deliverables at feasibility and serve a purpose: they scope the trial, identify research-paid cost drivers, and support site budget negotiations. They are not final billing authorization documents, and they should not pretend to be.

Once sites are confirmed, the team that builds or reviews the MCA needs to pull the LCDs for every MAC jurisdiction represented in the site list, check each contested line item against the LCD’s covered diagnoses and clinical criteria, and update designations by region. That work takes time. It belongs on the site activation critical path alongside the budget and the IRB submission, not after activation, when billing has already started.