A comparison of cholesterol measurements via various blood sample types

No Thumbnail Available
Macklin, Diana C.
Kaminsky, Leonard A., 1955-
Issue Date
Thesis (M.S.)
School of Physical Education
Other Identifiers

There has been inconsistencies in the performance of dry-chemistry analyzers using different blood sample types. Therefore, the purpose of this study was to determine precision and accuracy of both capillary and venous whole blood analyzed by the Reflotron (Boehringer Mannheim Diagnostics, Indianapolis, Indiana) and capillary plasma analyzed by the Ektachem DT60 (Eastman Kodak Company, Rochester, New York). Fifty subjects were recruited to provide a representative sample of cholesterol concentrations. One technician performed two fingerstick punctures and one venipuncture on each subject and analyzed the blood sample types in duplicate using each of the dry-chemistry analyzers. The methods and sample types utilized for comparison of total cholesterol are summarized below.ReflotronEktachem DT60Sigma2-Fingerstick whole2-Fingerstick plasma2-Venipuncture plasmablood2-Venipuncture whole2-Venipuncture plasmablood2-Veni uncture lasmaThe mean percent variation of the duplicate samples analyzed revealed all sample types, with the exception of fingerstick whole blood analyzed by theReflotron, met the LSP ideal goal for precision of 5 3% CV. Fingerstick wholeblood CV was 3.1%, meeting the current LSP standard of _5 5% CV for precision. The Sigma wet-chemistry assay for determination of total cholesterol was used as the reference for assessment of bias of each of the sample types. Fingerstick whole blood, via the Reflotron method, produced a positive 5.5% bias when compared to the reference, failing to meet the current LSP goal for acceptable accuracy (±5% bias). Venous whole blood analyzed using the Reflotron met this goal with a bias of +3.3%. Fingerstick plasma, via the Ektachem DT60 method, produced a bias of +2.1%, meeting the ideal LSP goal of ±3% bias. Venous plasma as measured by both the Reflotron and Ektachem DT60 also met this ideal goal (+2.0% and +1.8% bias, respectively). Overall, precision and accuracy of all sample types, with the exception of fingerstick whole blood, when analyzed by their respective dry-chemistry analyzer was acceptable.