Comparing two clinical measurements: a linear mixed model approach
Authors:
- 1 October 2005
- journal article
- Published by Taylor & Francis in Journal of Applied Statistics
- Vol. 32 (8) , 855-860
Abstract
In this article, we extended the widely used Bland-Altman graphical technique of comparing two measurements in clinical studies to include an analytical approach using a linear mixed model. The proposed statistical inferences can be conducted easily by commercially available statistical software such as SAS. The linear mixed model approach was illustrated using a real example in a clinical nursing study of oxygen saturation measurements, when functional oxygen saturation was compared against fractional oxy-hemoglobin.Accuracy, Bland-Altman method, linear mixed model, oxygen saturation,Keywords
This publication has 6 references indexed in Scilit:
- Statistical Methods in Assessing AgreementJournal of the American Statistical Association, 2002
- Modelling method comparison dataStatistical Methods in Medical Research, 1999
- Measuring agreement in method comparison studiesStatistical Methods in Medical Research, 1999
- Assay Validation Using the Concordance Correlation Coefficient1992
- STATISTICAL METHODS FOR ASSESSING AGREEMENT BETWEEN TWO METHODS OF CLINICAL MEASUREMENTThe Lancet, 1986
- Measurement in Medicine: The Analysis of Method Comparison StudiesJournal of the Royal Statistical Society: Series D (The Statistician), 1983