Thanks to all who participated in NAHDO's POA discussion May 25. It is clear that there are issues to be resolved in both collection and use of POA indicator.
The issue raised yesterday about the potential for the misuse of POA on public use files warrants further discussion. We will explore how the release of POA has been managed in NY and CA--two states with experience in POA collection/release.
We identified the following concerns:
--POA data collection will pose challenges, especially in the first few years of statewide data collection of this data element. Providers will vary in their coding practices, making data quality an issue.
--POA can be misinterpreted, especially in the early phases of data collection.
--How restrictive can/should states be in releasing POA with public use files? In most states, the data element is not a protected element, like identifiable data. And states will want to encourage its appropriate use.
States have in place data release policies that may limit the use of POA for individual cases. Data use agreements that many states use for their public use files prohibits the identification of an individual patient.
In the early years of collection, states will want to encourage appropriate research uses. Research-level data for bona fide researchers who understand the data and the data element is encouraged. (this use should be encouraged to promote use and understanding of the data element).
We also can use Pricing differentials. A basic public use file, with a data use agreement for all appropriate uses/users. Then a more expensive and restrictive use file, that requires more stringent agreement and terms of use. (An enhanced file).
Next, NAHDO welcomes ongoing discussion to develop use guidelines for POA.
I hope this stimulates further discussion as we venture into the world of POA in administrative data.
Denise Love
Executive Director
NAHDO