39 New PLA Codes Test Rules-Based Medical Coding

·

On October 1, 2026, the American Medical Association’s quarterly refresh of Proprietary Laboratory Analyses (PLA) codes went live: 39 new codes, three deletions, and three revisions, according to the AHA Coding Clinic for HCPCS’ Q3 2026 issue, printed October 28. For a code set most coders only think about when a denial forces them to, that is a lot of change to track four times a year — and the rule governing PLA codes makes getting it wrong more consequential than it looks.

A Code Set Built Around One Rule

PLA codes exist to identify a single proprietary test from a single laboratory. The AMA’s own guidance is blunt about what that means in practice: “When a PLA code is available for a proprietary laboratory service, it must be used exclusively for reporting that service.” A coder cannot substitute a generic Category I CPT code once a PLA code exists for that exact test, even if the generic code describes the underlying methodology well enough. Get the mapping wrong — use the old general code after a PLA code has gone live, or miss that a code was deleted — and the claim is coded incorrectly even though the clinical story is accurate.

That exclusivity rule is what turns a routine quarterly code list into a compliance exposure. It is not a judgment call coders can reason around with general coding knowledge; it is a lookup problem that has to be current.

What Changed in the October 2026 Release

New Codes Concentrate in Genomics and Transplant Testing

Several of the October additions illustrate how narrow these codes are. Code 0660U covers cell-free DNA testing for HPV genotypes 18, 31, 33, and 35 in whole blood via multiplex digital droplet PCR; 0661U covers the same methodology isolated to HPV genotype 16. Code 0662U covers a kidney allograft rejection test that profiles mRNA expression across 12 genes by real-time quantitative PCR — a transplant-medicine application that did not have a dedicated code before this cycle. Each of those three codes maps to exactly one lab’s test and exactly one clinical question.

Deletions and Revisions Carry Their Own Risk

The same release deleted three codes in infectious disease and oncology testing and revised three more spanning obstetrics, transplantation medicine, and oncology. A deleted PLA code is arguably the higher-risk event of the two: if a lab or coding team keeps billing a deleted code out of habit, or a charge description master was never updated, every claim behind it is wrong until someone notices.

Multiply this by four releases a year, going back to when PLA codes were introduced, and the result is a code set that has grown far faster — and far more unevenly — than the rest of the CPT book coders are trained against.

That pace creates a specific set of failure modes for coding operations built around static reference tables:

  • Charge description masters and encoder databases that refresh annually, or whenever IT has bandwidth, routinely lag a full quarter or more behind the live PLA list.
  • Generic CPT codes for a test’s underlying methodology remain technically billable in the system even after a PLA code supersedes them for that specific test, so nothing stops a coder from using the wrong one.
  • PLA test descriptions are written in highly specific lab and methodology language that doesn’t match the broader terminology coders are trained to search for.
  • Deleted codes rarely get flagged in claim-scrubbing logic unless someone has manually built a rule for that exact code.
  • Payer coverage policies and local coverage determinations for new PLA codes frequently take longer to publish than the code itself, leaving coders with a valid code and no clear payment rule to check it against.

Where Agentic Reasoning Changes the Equation

This is the kind of problem that pattern-matching against a static table is poorly suited for, because the table itself is the thing going stale. An agentic approach — one that reasons from the clinical documentation and the test’s actual methodology rather than matching keywords against a fixed list — has a better shot at catching that a kidney transplant rejection panel now has its own code, or that a previously valid infectious-disease PLA code no longer exists, because it can be built to check the current AMA and CMS code sets directly rather than trusting a quarterly data refresh someone forgot to run.

That is not a guarantee against error. An agent still needs to be grounded in the actual, current PLA list and the payer’s coverage policy, not just a general sense of how lab coding works — the exclusivity rule leaves no room for a plausible-sounding guess. But the core advantage is real: a system that can reason about “what test is this, and is there now a specific code for it” is structurally better matched to a code set that changes every three months than one built to apply the same lookup table until the next scheduled refresh.

What Coding and Compliance Teams Should Do Before Year-End

With Q4 underway, three practical steps matter most: reconcile the charge description master against the full October PLA release, not just the headline new-code count; audit recent claims for any of the three deleted codes still being billed; and confirm that coding staff — human or automated — are treating the exclusivity rule as mandatory rather than advisory when a generic code would otherwise “work.” None of that requires waiting for the January release to catch up; the exposure sits in claims going out right now.

Coding teams that want to stop re-running this exercise every quarter are increasingly turning to Medikode’s automated medical coding platform, which is built to reason about current code sets rather than depend on a static reference table that’s only as current as its last manual update.