The Blue Cross Blue Shield Association (BCBSA) recently submitted a comprehensive letter to the Centers for Medicare & Medicaid Services (CMS) on March 30. This letter responds to CMS’s request for information regarding its Comprehensive Regulations to Uncover Suspicious Healthcare (CRUSH) initiative. The BCBSA’s recommendations aim to enhance fraud prevention measures within Medicare Advantage (MA) plans and ensure the responsible use of artificial intelligence in healthcare oversight.
Seven BCBSA Recommendations:
1. CMS should promptly inform Medicare Advantage plans when it suspends payments to a provider due to suspected fraud. This measure is crucial, as fraudulent actors are currently exploiting delays by shifting billing from traditional Medicare to MA after CMS has acted on fraud within fee-for-service.
2. Any contractual or policy language requiring MA plans to continue paying claims when fraud is suspected should be eliminated. BCBSA also suggests tagging suspect claims with a unique code or pricing them at zero member liability when CMS suspends payment. This would enable MA plans to identify these claims beforehand.
3. Regarding whether MA plans should have mandatory authority to suspend payments akin to that of original Medicare, BCBSA advises a cautious approach. They emphasize that MA plans currently lack the necessary infrastructure for quick and precise payment suspensions. The association believes that simply applying the original Medicare framework could be problematic given the unique capitated payment structures of MA plans. A phased pilot program is recommended instead of an immediate implementation.
4. CMS should mandate rigorous testing and verification of any AI tools employed in MA coding oversight prior to and following their deployment. BCBSA also recommends that organizations using these tools establish internal AI governance programs. Moreover, developers should be obliged to disclose information regarding tool design, training datasets, and known limitations through a standardized framework.
5. If CMS chooses to use AI for Risk Adjustment Data Validation (RADV) audits, it should be used strictly as decision support and not as a substitute for qualified clinical and coding professionals. The association advises that AI-enhanced RADV workflows should be designed to identify both overpayments and underpayments, with performance metrics published by CMS. Insurers should retain the ability to review and contest AI-supported findings.
6. BCBSA asserts that the independent dispute resolution process established by the No Surprises Act is ineffective and requires significant structural improvements. They recommend that CMS swiftly implement the IDR Gateway, incorporate baseline eligibility screening prior to payment or review, establish an upfront eligibility fee to discourage bad-faith submissions, and create performance metrics to assess effectiveness.
7. CMS should enforce that non-participating durable medical equipment suppliers meet the accreditation and enrollment standards of original Medicare as a condition for billing MA plans. Additionally, BCBSA recommends establishing a central registry of non-participating DMEPOS suppliers, which MA plans can consult before processing claims.
