This Article  • References  • Full text PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Citation map  • Citing articles on HighWire  • Contact me when this article is cited  Related Content  • Similar articles in JAMA  Social Bookmarking   What's this?

Philip J. Held, PhD; Friedrich K. Port, MD, MS; Caitlin Carroll
University of Michigan Ann Arbor

Kenneth Petronis, MS, MPH
University of Maryland Baltimore

JAMA. 1994;272(23):1819.

	Since this article does not have an abstract, we have provided the first 150 words of the full text PDF and any section headings.

In Reply.

—We agree with Dr Sadler on the importance of measuring the quality of Medicare data when using this database for scientific analyses. Our estimates published in this article are similar to our findings from a large data validation study.¹ However, objectivity requires us to be constantly critical, especially for untested data items such as rHuEPO.

Dr Sadler takes our minor discussion point out of context and thereby generalizes that we have "challenged the validity of the PMMIS." This is not our position. We basically agree with his four technical points (listed herein); however, our analyses show that our data quality estimates remain virtually unchanged.

First, the point that we "used incomplete PMMIS files": PMMIS files are always incomplete. We used a file that was over 15 months old since US Renal Data System analyses suggest that by 15 months these files can be considered complete for most . . . [Full Text PDF of this Article]

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?