Interlaboratory Comparison of University Hospital Referee Laboratories and Community Hospital Laboratories, Using Results of the 1968 College of American Pathologists Clinical Chemistry Survey

Publisher Website
Google Scholar
Abstract
The performances of 58 to 112 university hospital laboratories and 1,300 to 1,700 clinical laboratories participating in the 1968 College of American Pathologists Comprehensive Clinical Chemistry Survey were compared. Each laboratory received eight identical lyophilized samples which were analyzed for eight to 12 chemical constituents. Although the university hospitals performed better, as would be expected, the difference from other laboratories was small. The ± 2 SD “state of the art” criterion used to evaluate performance in national surveys was found to be valid because it provides equally narrow or narrower ranges than those provided by the 99.73% range (3 SD) of university hospitals. Certain systematic differences were observed: these included (1) participant values too high for low sodium levels and too low for high sodium levels; (2) participant values too high for cholesterol and too low for uric acid. The findings indicate a need for particular attention to causes of inter laboratory variability, which are shown to contribute a significantly greater amount than intra laboratory variability to the total variability of clinical chemical analyses of identical survey samples.

This publication has 0 references indexed in Scilit: