Of paramount importance in the release of Public Use Files was the protection of beneficiary confidentiality. To that end all directly identifiable information was removed from each data set. Moreover, other potentially identifying variables, which might cause identification by themselves or in combination with other variables, were either removed from the files or their values re-coded. In selecting variables for inclusion in these PUFs priority was given to the measures most commonly reported/studied in health services research.
The PUFs were designed to provide researchers and analysts a convenient initial look at data drawn from CMS claims records. The PUFs assist users in hypothesis generation and testing as well as provide a very inexpensive and efficient data set that researchers will use to gain familiarity with CMS claims data.
The following Research Briefs were commissioned by CMS and are based on analyses performed using the PUFs.