Search PubMed⌕ Search
Home/Archive/PubMed

PubMed · 7016436

Precision and accuracy: concepts and assessment by method evaluation testing.

J O Westgard
View original source ↗Export metadata ↓

Abstract

Achieving precision and accuracy in routine clinical analyses is a complex task, requiring the identification, estimation, and elimination of sources of analytical error. This review first considers concepts of precision and accuracy, including discussions of the meaning of measurement process, analytical method, state of statistical control, precision, imprecision, accuracy, inaccuracy, systematic error, overall or total error, true value, traceability, and compatability. These concepts provide the basis upon which the performance of analytical methods can be evaluated. The second part of the review considers how precision and accuracy are assessed by the use of method evaluation experiments. The approach emphasizes the development of an evaluation protocol based on the analytical characteristics which represent the performance of the method. This includes discussions of the familiarization period; testing analytic range and linearity; testing precision by a replication experiment; testing accuracy by recovery, interference, and comparison of methods experiments; the selection of a comparative analytical method; the statistical analysis of method comparison data, including the interpretation of that data; the collaborative testing.

Explore related subjects

Aspartate AminotransferasesChemistry, ClinicalClinical Laboratory TechniquesDiagnostic ErrorsEvaluation Studies as TopicHumansMethodsQuality ControlStatistics as TopicJournal ArticleReview

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

J O Westgard. 1981. Precision and accuracy: concepts and assessment by method evaluation testing.. https://doi.org/10.3109/10408368109106450

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2007

Confidence intervals for multinomial logistic regression in sparse data.

Logistic regression is one of the most widely used regression models in practice, but alternatives to conventional maximum likelihood estimation methods may be more appropriate for small or sparse samples. Modification of the logistic regression score function to remove first-order bias is equivalent to penalizing the likelihood by the Jeffreys prior, and yields penalized maximum likelihood estimates (PLEs) that always exist, even in samples in which maximum likelihood estimates (MLEs) are infinite. PLEs are an attractive alternative in small-to-moderate-sized samples, and are preferred to exact conditional MLEs when there are continuous covariates. We present methods to construct confidence intervals (CI) in the penalized multinomial logistic regression model, and compare CI coverage and length for the PLE-based methods to that of conventional MLE-based methods in trinomial logistic regressions with both binary and continuous covariates. Based on simulation studies in sparse data sets, we recommend profile CIs over asymptotic Wald-type intervals for the PLEs in all cases. Furthermore, when finite sample bias and data separation are likely to occur, we prefer PLE profile CIs over MLE methods.

Aspartate Aminotransferases↗
PubMed2006

Significance of low or normal serum gamma glutamyl transferase level in infants with idiopathic neonatal hepatitis.

INTRODUCTION: We evaluated the significance of low/normal serum gamma glutamyl transferase (GGT) level in infants with idiopathic neonatal hepatitis (INH). MATERIALS AND METHODS: A retrospective review of the hospital records of 103 infants less than 3 months of age who were diagnosed with INH between August 1991 and November 2000 was performed. Variables including age at which jaundice was noticed, age at presentation, perinatal risk factors, family history of liver disease, parental consanguinity, initial ultrasound scan, liver biopsy, laboratory values at the first visit, the peak levels of total bilirubin, aspartate aminotransferase (AST), GGT and alkaline phosphatase (ALP) in the first 3 months of follow-up and interval for normalisation of serum bilirubin and AST were compared between infants presenting with low/normal GGT ( 100 U/L). RESULTS AND DISCUSSION: Infants with low/normal GGT levels presented earlier (median 36.5 days versus 44 days; p=0.016) and had significantly higher bilirubin and AST levels at presentation (bilirubin 167.5 micromol/L versus 133 micromol/L; p<0.005 and AST 187.5 U/L versus 106 U/L; p<0.001) and at peak levels (bilirubin 170 micromol/L versus 146 micromol/L; p=0.024 and AST 210.5 U/L versus 129 U/L; p=0.001). A significant correlation was also found between GGT levels and serum albumin levels (p=0.004). Patients with low/normal GGT levels were more likely to have giant cell hepatitis on histology (p=0.015). There was no difference in time taken to recovery. CONCLUSION: Low/normal levels of GGT in INH infants may be a predictor of more severe but recoverable disease.

Aspartate Aminotransferases↗