Determination of the Accuracy of Quantitative Measurement Methods in the Clinical Chemistry Laboratory at Machakos Level V Hospital, Kenya

Abstract

Accuracy is the extent to which measurements are closely related to the actual values of the analyte. Accuracy is one of the key requirements in Laboratory method verification. Verification is a method of ensuring that tests are carried out per the manufacturer's specifications when tested by laboratory staff and patients at the facility. Other elements of verification include precision, linearity verified with reportable range, uncertainty of measurement, carry-over studies, reference ranges and limit of detection. This study aims to determine the accuracy of the quantitative measurement methods in the Clinical Chemistry laboratory in the month of January 2023. It also provides an alternative method for meeting accreditation requirements on accuracy by medical laboratories seeking accreditation in far-stretched laboratories where interlaboratory comparison may not be feasible. A systematic review and meta-analysis at Machakos County, Kenya. The procedure involved analysing commercially available internal quality control material five (5) times a day for eight (8) days bearing the same lot number. Roche Diagnostics, Mannheim, Germany supplied all the reagents, internal quality control materials and calibrators. Accuracy was determined using Roche Cobas® Intergra 400 Clinical Chemistry analyser to perform a comparative descriptive analysis of albumin, alanine aminotransferase (ALT), alkaline phosphatase (Alk Phos), aspartate aminotransferase (AST), chloride, creatinine, Direct Bilirubin (D. Bil), Gamma-glutamyl transferase (GGT), potassium, sodium, total bilirubin, total protein, and urea. The study followed guidelines issued by the Clinical Laboratory Improvements Amendment 1988 (CLIA) and Clinical Laboratory Standards Institute (2014). Data analysis was carried out using Excel Windows 10 MS Office 2021. Mean, standard deviation (SD), Z-score, bias %, coefficient of variation (CV) %, and total error allowable (Tea) were calculated from the results of measurement of the analytes. All the analytes achieved a mean which fell within the manufacturer verification interval and hence passed. The bias % score for the analytes was as follows 5 analytes level (19.2%) scored 0 - ±1, 10 analytes level (38.5%) scored ±1.1 - ±2, 7 analytes level (26.9%) ±2.1 - ±3 and 4 analytes level (15.4%) scored above ±3. All analytes were found to have an excellent Z-score performance between 0 to ± 1.96. The total error allowable was found to fall within CLIA and CLSI specifications limits except for chloride PCC1/ Normal control which failed at 5.82% (CLIA target being ± 5%). Analytes' mean was expected to fall within the given manufacturer's mean. The bias % for the analytes which was at zero or near ± zero was considered an excellent score the further away the score was, the poorer the performance. The Z-score was also calculated to establish how far the observed mean fell from the target mean. All analytes were found to have an excellent Z-score performance between 0 to ± 1.96. A Z-score of ± 1.96 to ± 2.57 indicates good performance while a Z-score of over ± 2.58 indicates failed performance. These good results are attributable to the fact that analysis was carried out within a very short duration by the same person, on the state-of-the-art instrument, in a well-controlled environment hence no room for analytical variation. Chloride being an electrolyte failed since it is physiologically controlled in a strict and narrow range.