Why CMS never studies addiction data | Behavioral Healthcare Executive Skip to content Skip to navigation

Why CMS never studies addiction data

April 17, 2015
by Julie Miller, Editor in Chief
| Reprints

A perspective piece in the New England Journal of Medicine this week reveals the little known fact that the Centers for Medicare and Medicaid Services (CMS) in recent years has hampered research projects because it has withheld from the study data any claims that contain any substance use disorder (SUD) diagnosis or related procedure code. CMS cites the limitations of 42 CFR Part 2 as the reason.

It’s important to note SUD data can be disclosed for research purposes under 42 CFR Part 2, however, the privilege is limited to providers alone—not payers, such as Medicare and Medicaid. Therefore, CMS must follow the rule and withhold that data.

Authors Austin B. Frakt and Nicholas Bagley say prior to 2013 when the withholding policy began, studies were conducted using the SUD data and no harm came to any patients whose data was used.

Of broader concern is the fact that the data holes carry over to research on disorders outside of behavioral health because some claims chart co-occurring SUD, which would then cause the data to be excluded, for example, in research on hepatitis C or HIV/AIDS. The authors go on to encourage CMS and SAMHSA to rethink the policy and restore access to the SUD data for research.

My thoughts: Children are dying while we clamor for a landmark federal policy to address the crisis of addiction, and the last thing we need now is to entomb the very data that would support it. As an industry, we can argue our ideology and anecdotes in Congressional committee hearings all day long, but until we have quantified evidence to back our recommendations, very little will happen.


Julie Miller

Editor in Chief

Julie Miller


Julie Miller has more than 14 years of experience observing, analyzing and reporting on various...

The opinions expressed by Behavioral Healthcare Executive bloggers and those providing comments are theirs alone and are not meant to reflect the opinions of the publication.