OIG Audit Flags $15.2M in Sacroiliac Injection Coding Errors

·

OIG Audit Flags $15.2M in Sacroiliac Injection Coding Errors

A federal audit released this week puts a number on a coding problem that compliance
teams have suspected for years: physicians and facilities are struggling to bill
sacroiliac joint injections correctly, and the errors are expensive. In a report
published August 3, 2026 and posted August 5, the HHS Office of Inspector General found
that Medicare made an estimated $15.2 million in improper payments for
sacroiliac joint injection sessions performed between October 1, 2023 and September 30,
2024 — and the root cause traces directly back to how the procedures were coded.

What the audit found

OIG sampled 100 sacroiliac joint injection sessions covered under Medicare Part B and
found that 72 of them did not comply with Medicare billing requirements. Extrapolated
across the full audit period, that translates to 134,526 of 186,842 total sessions with
improper payments. The report,
“Medicare Improperly Paid Physicians an Estimated $15.2 Million for Sacroiliac Joint
Injections”
(report OAS-25-09-021), is the latest entry in OIG’s multi-year audit
series of Medicare spinal pain management services — a series that has already
turned up $30 million in improper payments for spinal facet-joint interventions and
$3.6 million for epidural steroid injection sessions in earlier phases.

A diagnostic-versus-therapeutic problem

The single largest driver of the error rate wasn’t a frequency violation or a missing
signature. OIG found that 46,711 sessions were billed as therapeutic sacroiliac joint
injections when the clinical circumstances indicated they should have been billed as
diagnostic. That distinction isn’t a technicality — Medicare coverage and
documentation requirements differ for diagnostic injections (typically used to confirm
the joint as a pain generator before further treatment) versus therapeutic injections
(intended to provide symptom relief), and CPT/HCPCS reporting depends on which one
actually occurred.

Where oversight broke down

OIG’s own conclusion was blunt: “oversight provided by CMS and the Medicare
Administrative Contractors (MACs) was not sufficient to ensure compliance with Medicare
requirements and guidance.” In other words, the documentation submitted with these claims
often didn’t clearly establish intent, and the claims-processing systems downstream
weren’t catching the mismatch before payment went out.

Why this is a CDI story, not just a billing story

It’s tempting to read an OIG audit like this as a payer-integrity issue that lives
downstream of coding. It isn’t. The diagnostic/therapeutic distinction has to be
established in the clinical note — the physician’s stated intent, the patient’s
prior treatment history, and whether the injection is being used to localize a pain
source or to treat one. If that intent isn’t documented clearly and consistently, no
downstream coding process can recover it, and the claim goes out wrong regardless of how
carefully the code itself is selected.

For coding and CDI teams working spinal pain management claims, a few patterns are
worth checking against this audit’s findings:

  • Whether documentation explicitly states the injection’s intent (diagnostic
    localization vs. therapeutic relief) rather than leaving it to be inferred from the
    order.
  • Whether repeat injections at the same site are supported by documented response to
    a prior diagnostic injection, not just a standing treatment plan.
  • Whether image-guidance documentation (fluoroscopic or CT) supports the specific code
    billed, consistent with the applicable local coverage determination.
  • Whether providers are using templated or copy-forward notes that default to
    “therapeutic” language regardless of the actual clinical scenario.
  • Whether claims edits flag sacroiliac joint injection codes for a documentation
    cross-check before submission, rather than relying on post-payment audit to catch the
    error.

CMS’s response

OIG recommended that CMS work with the MACs to develop coding and documentation
education specific to sacroiliac joint injection requirements, and to implement controls
that prevent diagnostic-versus-therapeutic misclassification going forward. CMS concurred
with two of the three recommendations but disagreed with the third, which suggests the
agency’s own edits and prepayment review process for this code set will likely tighten
in the coming months rather than staying static.

The broader pattern

Sacroiliac joint injections are a relatively small line item in the Medicare Part B
budget, but the underlying problem — clinical intent that isn’t captured cleanly
enough in the note to support the code that gets billed — shows up across the
entire spinal pain management audit series, and in plenty of other procedure families
beyond it. Every prior report in this OIG series has landed on some version of the same
root cause: documentation that describes what was done but not clearly enough why it was
done, or under what coverage circumstance. That’s a documentation integrity gap first
and a coding gap second, and it’s the kind of gap that’s very hard to catch by reviewing
claims data alone after the fact.

Coding teams that want to get ahead of the next audit in this series — rather
than read about it after the fact — are the ones building diagnostic-versus-
therapeutic intent checks, and coverage-specific documentation prompts, directly into
their workflow instead of relying on periodic manual review. That’s precisely the kind
of case-by-case, guideline-aware reasoning that
Medikode’s automated medical coding platform is
built to apply consistently across every claim, not just the ones that happen to get
sampled.