Health care data, the epidemiologist's sand: comments on the quantity and quality of data.

Massive amounts of health care data are currently available for epidemiologic review through improvements in computerization and electronic communication. Multiple abstracts of patient care data are collected, stored, retrieved, and analyzed to study health care practice and outcome. The high level of variation in data from these sources is noted. Examples of these data collections are reviewed and the issues of the quality of these data for research and evaluation are discussed. Increased amounts of poor quality data will not be helpful. Collections from the National Center for Health Statistics and other sources are cited as models for improved standards for quality data banks and registries, including the Centers for Disease Control National Nosocomial Infections Surveillance collection. Throughout, a metaphor relating quality of sand for the production of lens instruments to view scientific change is used.