Definition and application of performance specifications for measurement uncertainty of 23 common laboratory tests: linking theory to daily practice

Abstract

Laboratories should estimate and validate [using analytical performance specifications (APS)] the measurement uncertainty (MU) of performed tests. It is therefore essential to appropriately define APS for MU, but also to provide a perspective on suitability of the practical application of these APS. In this study, 23 commonly ordered measurands were allocated to the models defined during the 2014 EFLM Strategic Conference to derive APS for MU. Then, we checked if the performance of commercial measuring systems used in our laboratory may achieve them. Most measurands (serum alkaline phosphatase, aspartate aminotransferase, creatine kinase, γ-glutamyltransferase, lactate dehydrogenase, pancreatic amylase, total proteins, immunoglobulin G, A, M, magnesium, urate, and prostate-specific antigen, plasma homocysteine, and blood red and white cells) were allocated to the biological variation (BV) model and desirable APS were defined accordingly (2.65%, 4.75%, 7.25%, 4.45%, 2.60%, 3.15%, 1.30%, 2.20%, 2.50%, 2.95%, 1.44%, 4.16%, 3.40%, 3.52%, 1.55%, and 5.65%, respectively). Desirable APS for serum total cholesterol (3.00%) and urine albumin (9.00%) were derived using outcome-based model. Lacking outcome-based information, serum albumin, high-density lipoprotein cholesterol, triglycerides, and blood platelets were temporarily reallocated to BV model, the corresponding desirable APS being 1.25%, 2.84%, 9.90%, and 4.85%, respectively. A mix between the two previous models was employed for serum digoxin, with a 6.00% desirable APS. In daily practice by using our laboratory systems, 16 tests fulfilled desirable and five minimum APS, while two (serum albumin and plasma homocysteine) exceeded goals, needing improvements.

Application of the Analytical Performance Specification for Quality Management of Clinical Chemistry Tests

“…문헌 고찰을 통하여 대표적인 성과 연구 기반 분석수행사양 측 정 불확도(measurement uncertainty) 목표를 Table1에정리하였 다[9,10].2. 생물학적 변이 관련 용어 정의각 검사별 분석수행사양 목표 설정을 위한 근거 자료로 흔히 활 용되고 있는 생물학적 변이에 대한 용어와 정의는 다음과 같다 (Table2).CVI: 개체 내 생물학적 변이(within-subject BV), 개체로 이루어 진 집단의 합동 변이(pooled variation)로 추정되는 단일 개체 내 변이[35].CVG: 개체 간 생물학적 변이(between-subject BV), 여러 개체들 로 구성된 집단에서 각 개체의 평균 추정값 간의 변이[35].CVA: 분석변이(analytical variation, analytical imprecision), 특정 조건에서 검사방법의 비정밀도 추정치[35].www.labmedonline.org 257 RCV: 변화기준값(reference change value, RCV), 같은 대상자에 서 다른 시점에 2회 시행한 결과에 유의미한 차이가 있는지 판정 하는 기준값[29, 35].…”