[.ca] Health Care Data and the SAS System (ISBN 1580258654)

4 1/2* Excellent within its Limits:
This is a very good book which focuses a bit more on descriptive statistics and data set management and manipulation than on any of the inferential statistics common to health care research. There is little or no mention of cost analyses (a popular topic these days), or of epidemiology or case-mix adjustment. The book probably best for the intermediate user who wants SAS programming instruction than for the advanced user testing hypotheses. Still, the book includes several topics that are endemic to the health care field, including an explanation of diagnostic, procedure, and national drug codes, and brief overviews of some of the major health care databases (e.g., HCFA, UB92). There is a very good appendix listing database and health information organizations on the net, such as the National Health Information Resource Center, the Agency for Health Care Research and Quality, the Centers for Disease Control, and the Substance Abuse and Mental Health Administration (which includes some programs), although VA sites are not listed. In general, there are several gaps in the book, perhaps not unexpected given its relatively brief length (274 pages). There's an excellent glossary, although, annoyingly, some of the terms are not discussed in the text itself (e.g., RUG). Some complicated procedures are admirably described (Proc SQL); others are mentioned with very little explanation (ARRAY). The authors are aware that of the space limitations, and helpfully list other books in the SAS-published "Books By Users" series (including the excellent SAS books by Ron Cody). This book is best viewed as an introduction and overview of the subject.