• Sidelining all-payer claims databases

    As I wrote yesterday, SAMHSA has proposed a rule that’ll go far to relieve Austin’s and my concerns about counterproductive restrictions on research into substance use disorders. But the rule has at least one glaring problem. As written, it would make the all-payer claims databases (APCDs) that now exist in about sixteen states all but useless when it comes to figuring out how to help those who struggle with addiction.

    Under the proposed rule, researchers can get patient identifying data about substance use disorders from anyone who’s a “lawful holder” of such data, subject to a whole bunch of rules to protect patient privacy. In SAMHSA’s view, a “lawful holder” is someone who receives identifiable data “as the result of a … patient consent … or as a result of one of the limited exceptions to the consent requirements specified in the regulations.”

    The problem is that APCDs don’t appear to be lawful holders. To my knowledge, patients don’t consent to sharing data on substance use disorders with APCDs. Nor is there an exception in the federal privacy rules that would enable such data sharing. State laws that try to require payers to disclose that information to APCDs would, I think, be preempted.

    Under the rule, then, APCDs will be ineligible to receive identifiable data on substance use disorders. That’ll blow an enormous hole in the databases. (Although it may not matter much: the Supreme Court is considering a case that could blow an even bigger hole in them.) APCDs in Massachusetts, Vermont, and New Hampshire, for example, will be scrubbed of data pertaining to the heroin epidemic. That’s completely loony.

    I’m not sure SAMHSA appreciates the problem. The proposed rule says nothing about APCDs. It’s as if they didn’t exist.

    That’s weird. As third party payers, state Medicaid agencies already hold substance use claims. Indeed, because they’re lawful holders, they can share their claims data with researchers. SAMHSA has even said that it’s open to letting state databases link identifiable data for researchers.

    If we trust state Medicaid agencies to compile and share addiction-related claims, why not trust APCDs to do the same? APCDs are unique among data sets because they offer a comprehensive view of a state’s health-care system—and they were designed in large part to facilitate research. If their security procedures are up to snuff, I don’t see the case for preventing them from doing their jobs.

    From my perspective, the best result would be for SAMHSA to clarify in the final rule (1) that APCDs are eligible to receive data on substance use disorders if they have adequate privacy protections in place and (2) that APCDs are “lawful holders” and can thus share such data with qualified researchers where appropriate.

    But the agency won’t do any of that unless APCDs and the research community make the case. SAMHSA is taking comments until April 1. Let’s not squander the opportunity.


    Comments closed