Agreed Statistics Measurement Method Comparison

Department of Health Sciences, University of York, Heslington, York, United Kingdom.
Anesthesiology (Impact Factor: 5.88). 11/2011; 116(1):182-5. DOI: 10.1097/ALN.0b013e31823d7784
Source: PubMed


Statistical Methods for Assessing Agreement between Two Methods of Clinical Measurement. By J. Martin Bland, Douglas G. Altman. Lancet 1986; 1(8476):307-10. Abstract reprinted with permission of Elsevier, copyright 1986. In clinical measurement comparison of a new measurement technique with an established one is often needed to see whether they agree sufficiently for the new to replace the old. Such investigations are often analyzed inappropriately, notably by using correlation coefficients. The use of correlation is misleading. An alternative approach, based on graphical techniques and simple calculations, is described, together with the relation between this analysis and the assessment of repeatability.

1 Follower
15 Reads
  • Source
    • "stature, R 2 ¼ 19–60%; body mass, R 2 ¼ 70–43%; lean lower limbs mass, R 2 ¼ 60–38%). The best model was obtained with an anthropometric estimate of lean body mass (Foster et al., 2012), that was noted as a non-biased assessment using DXA protocol as a reference (Bland & Altman, 2012; Holiday et al., 1995; Ludbrook, 2002; Ricker, 1973). By inference, lean body mass derived by anthropometry may be a reasonable and acceptable descriptor of the maximal metabolic capacity of muscles. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Peak oxygen uptake (VO2peak) is routinely scaled as mL O2 per kilogram body mass despite theoretical and statistical limitations of using ratios. Aim: To examine the contribution of maturity status and body size descriptors to age-associated inter-individual variability in VO2peak and to present static allometric models to normalize VO2peak in male youth soccer players. Subjects and methods: Total body and estimates of total and regional lean mass were measured with dual energy X-ray absorptiometry in a cross-sectional sample of Portuguese male soccer players. The sample was divided into three age groups for analysis: 8–12 years, 13–15 years and 16–18 years. VO2peak was estimated using an incremental maximal exercise test on a motorized treadmill. Static allometric models were used to normalize VO2peak. Results: The independent variables with the best statistical fit explained 72% in the younger group (lean body mass: k = 1.07), 52% in mid-adolescent players (lean body mass: k = 0.93) and 31% in the older group (body mass: k = 0.51) of variance in VO2peak. The inclusion of the exponential term pubertal status marginally increased the explained variance in VO2peak (adjusted R2 = 36–75%) and provided statistical adjustments to the size descriptors coefficients. Conclusion: The allometric coefficients and exponents evidenced the varying inter-relationship among size descriptors and maturity status with aerobic fitness from early to late-adolescence. Lean body mass, lean lower limbs mass and body mass combined with pubertal status explain most of the inter-individual variability in VO2peak among youth soccer players.
    Annals of Human Biology 10/2015; 42(2):125-33. DOI:10.3109/03014460.2014.932007 · 1.27 Impact Factor
  • Source
    • "The concentrations of sibutramine obtained with the three methods were statistically compared with a Bland– Altman test to determine the agreement between two separate techniques (Bland & Altman 2012). To assess the bias between methods, this test plots the difference between the measurements of two methods versus their average on the x-axis. "
    Dataset: mathon sibu
  • Source
    • "Microsoft™ Excel 2000 (Microsoft Corporation, USA), Graphpad Prism Version 5.00 for Mac (Graphpad Software, San Diego, USA), SPSS (SPSS Statistics for Windows, Version 17.0, USA) and StatsDirect (StatsDirect Ltd, UK) were used for statistical analyses described in this study. Bland-Altman plots were used to determine the agreement of the in house HIV assay with the commercial assay [17], [18]. Probit regression analysis was used to determine the theoretical lower limit of quantitation. "
    [Show abstract] [Hide abstract]
    ABSTRACT: HIV-1 viral quantitation is essential for treatment monitoring. An in-house assay would decrease financial barriers to access. A real-time competitive RT-PCR in house assay (Sing-IH) was developed in Singapore. Using HXB2 as reference, the assay's primers and probes were designed to generate a 183-bp product that overlaps a portion of the LTR region and gag region. A competitive internal control (IC) was included in each assay to monitor false negative results due to inhibition or human error. Clinical evaluation was performed on 249 HIV-1 positive patient samples in comparison with the commercially available Generic HIV Viral Load assay. Correlation and agreement of results were assessed for plasma HIV-1 quantification with both assays. The assay has a lower limit of detection equivalent to 126 copies/mL of HIV-1 RNA and a linear range of detection from 100-1000000 copies/mL. Comparative analysis with reference to the Generic assay demonstrated good agreement between both assays with a mean difference of 0.22 log10 copies/mL and 98.8% of values within 1 log10 copies/mL range. Furthermore, the Sing-IH assay can quantify HIV-1 group M subtypes A-H and group N isolates adequately, making it highly suitable for our region, where subtype B and CRF01_AE predominate. With a significantly lower running cost compared to commercially available assays, the broadly sensitive Sing-IH assay could help to overcome the cost barriers and serve as a useful addition to the currently limited HIV viral load assay options for resource-limited settings.
    PLoS ONE 03/2014; 9(3):e89826. DOI:10.1371/journal.pone.0089826 · 3.23 Impact Factor
Show more