Free up the data, Commissioner

The Governor of Massachusetts has made an excellent choice for the new Commissioner of the Division of Health Care Finance and Policy.  He is attorney Áron Boros, who comes to DHCFP after serving as director of federal finance for the state’s Medicaid office since 2008.  Prior to working for MassHealth, Boros was an associate at the Boston law firm Foley Hoag LLP, where he worked with health care clients on a wide variety of issues.  His work included initiatives related to chronic disease management, health information technology, and evidence-based medicine.  His first day at DHCFP will be Sept. 15.

Let's hope that Mr. Boros makes good on the promises of his predecessor.  Under state legislation, the DHCFP has been assembling an all-payer claims database (APCD). It comprises medical claims, dental claims, pharmacy claims, and information from member eligibility files, provider files, and product files. It includes fully-insured, self-insured, Medicare, and Medicaid data and clear definitions of insurance coverage (covered services, group size, premiums, co-pays, deductibles) and carrier-supplied provider directories.

The existence of this database offers the potential for all parties to study the actual transactions taking place in the Massachusetts health care system.  It permits testing of hypotheses with regard to payment models, clinical affiliations, and a variety of other pertinent matters with regard to the state's health care policy agenda.  Because Massachusetts took the national lead with regard to health care reform, it also offers potential value to the country as a whole.

But transparency only works if the data are made available to all and easily so.  The previous commissioner made this promise over a year ago:  "That the database would be widely accessible, so that independent researchers, policy analysts, advocates, market participants, and others would be able to manipulate it to test hypotheses and assumptions."

Here is what the current regulations (114.5 CMR 22.00: Health Care Claims Data Release) say about getting this data:

22.03: Procedures for Data Requests
(1) Public and Restricted Use Files. The Division will create Claims Data Public Use Files and Restricted Use Files to which Applicants may request access in accordance with 114.5 CMR 22.03(2).

(2) Application Review Procedures. 

(a) Applications for Data.
1. All Applicants must submit a written application. Each Applicant shall:
a. specify the data requested, including Public Use Files and any restricted data elements requested; 

b. specify the purpose and intended use of the data requested, including a detailed project description that describes any other data sources to be used for the project;
c. specify security and privacy measures that will be taken in order to safeguard patient privacy and to prevent unauthorized access to or use of such data; 

d. specify the Applicant's methodology for maintaining data integrity and accuracy;
e. describe how the results of the Applicant’s analysis will be published;

f. agree to provide the results of all analyses, research, or other product of the data requested to the Division for the Division’s own use;
g. agree to the data disclosure restrictions in 114.5 CMR 22.04; and 

h. obtain prior approval from the Division to release any reports that used restricted use files prior to publication or other release.

Obviously, protection of patient privacy is important and required by statute, but several of the other requirements here are not.  Truly, this is "old style" government.  Say I am an academic researcher, exploring multiple hypotheses related to payments or clinical care management.  Why should I have to tell the state what other data sources I plan to use?  Why should I describe how the results will be published?  (Indeed, I probably cannot, in that my work may have to go through levels of peer review at multiple journals.)  For public use files, rather than restricted files, why should there be a need for any application at all?  I fear that these requirements act as an effective bar on public inquiry, well beyond that intended by the Legislature.

But maybe I am over-reacting.  I would love to be proven wrong.  Here's a test to see whether my concerns are valid.  This is the most comprehensive data set in the state on this topic, in a state filled with public policy researchers at multiple academic institutions and NGOs.  How many applications for data have been received since the regulations were adopted in July of 2010?  How many have been approved?  For the ones not approved, what reasons were given?