Agreement Between Methods of Measurement with Multiple Observations Per Individual

Department of Health Sciences, University of York, York, UK.
Journal of Biopharmaceutical Statistics (Impact Factor: 0.59). 02/2007; 17(4):571-82. DOI: 10.1080/10543400701329422
Source: PubMed


Limits of agreement provide a straightforward and intuitive approach to agreement between different methods for measuring the same quantity. When pairs of observations using the two methods are independent, i.e., on different subjects, the calculations are very simple and straightforward. Some authors collect repeated data, either as repeated pairs of measurements on the same subject, whose true value of the measured quantity may be changing, or more than one measurement by one or both methods of an unchanging underlying quantity. In this paper we describe methods for analysing such clustered observations, both when the underlying quantity is assumed to be changing and when it is not.

    • "Comparison of logger temperatures recorded at the different body sites before and after herding and immobilization was tested using paired t-tests. The agreement between abdominal temperature and the temperatures measured with microchips was estimated using the limit of agreement methods for repeated paired measurements (Bland and Altman, 2007). For each microchip location, limits of agreement were calculated using 42 paired measurements from 14 individuals. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Hyperthermia is described as the major cause of morbidity and mortality associated with capture, immobilization and restraint of wild animals. Therefore, accurately determining the core body temperature of wild animals during capture is crucial for monitoring hyperthermia and the efficacy of cooling procedures. We investigated if microchip thermometry can accurately reflect core body temperature changes during capture and cooling interventions in the springbok (Antidorcas marsupialis), a medium-sized antelope. Subcutaneous temperature measured with a temperature-sensitive microchip was a weak predictor of core body temperature measured by temperature-sensitive data loggers in the abdominal cavity (R2=0.32, bias >2 °C). Temperature-sensitive microchips in the gluteus muscle, however, provided an accurate estimate of core body temperature (R2=0.76, bias=0.012 °C). Microchips inserted into muscle therefore provide a convenient and accurate method to measure body temperature continuously in captured antelope, allowing detection of hyperthermia and the efficacy of cooling procedures.
    No preview · Article · Dec 2015 · Journal of Thermal Biology
  • Source
    • "Correlations between each method and the reference RR, and the correlations between the two methods, were calculated using the Pearson's product-moment correlations. Measurements from both methods were compared to the reference RR and to each other using the Bland–Altman analysis for repeated measurements[16,17]. Bias, limits of agreement (LoA), percentage error (PE), and within-subject and between-subject variances were reported. "
    [Show abstract] [Hide abstract]
    ABSTRACT: OBJECTIVE: To compare the accuracy and efficiency of the respiratory rate (RR) RRate mobile application to the WHO ARI Timer. METHODS: Volunteers used both devices to measure RR from reference videos of infants and children. Measurements were compared using correlation, Bland-Altman analysis, error metrics and time taken. RESULTS: Measurements with either device were highly correlated to the reference (r = 0.991 and r = 0.982), and to each other (r = 0.973). RRate had a larger bias than the ARI Timer (0.6 vs. 0.04 br/min), but tighter limits of agreement (-4.5 to 3.3 br/min vs. -5.5 to 5.5 br/min). RRate was more accurate than the ARI Timer (percentage error 10.6% vs. 14.8%, root mean square error 2.1 vs. 2.8 br/min and normalized root mean square error 5.6% vs. 7.5%). RRate measurements were 52.7 seconds (95% CI 50.4 s to 54.9 s) faster. CONCLUSION: During video observations, RRate measured RR quicker with a similar accuracy compared to the ARI Timer.
    Preview · Article · Dec 2015 · Journal of Healthcare Engineering
  • Source
    • "Statistical significance was assumed at values of P < 0.05. Agreement between measured and predicted _ Vo 2 values with multiple observations per participant was assessed graphically with Bland–Altman plots (Bland and Altman 2007). Variables are presented as means AE SD, and 95% Confidence Intervals (CI), unless otherwise stated. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Estimation of human oxygen uptake (_ Vo 2) during exercise is often used as an alternative when its direct measurement is not feasible. The American College of Sports Medicine (ACSM) suggests estimating human _ Vo 2 during exercise on a cycle ergometer through an equation that considers individual's body mass and external work rate, but not pedaling rate (PR). We hypothesized that including PR in the ACSM equation would improve its _ Vo 2 prediction accuracy. Ten healthy male participants' (age 19–48 years) were recruited and their steady-state _ Vo 2 was recorded on a cycle ergometer for 16 combinations of external work rates (0, 50, 100, and 150 W) and PR (50, 70, 90, and 110 revolutions per minute). _ Vo 2 was calculated by means of a new equation, and by the ACSM equation for comparison. Kinematic data were collected by means of an infrared 3-D motion analysis system in order to explore the mechanical determinants of _ Vo 2. Including PR in the ACSM equation improved the accuracy for prediction of sub-maximal _ Vo 2 during exercise (mean bias 1.9 vs. 3.3 mL O 2 kg À1 min À1) but it did not affect the accuracy for prediction of maximal _ Vo 2 (P > 0.05). Confirming the validity of this new equation, the results were replicated for data reported in the literature in 51 participants. We conclude that PR is an important determinant of human _ Vo 2 during cycling exercise, and it should be considered when predicting oxygen consumption.
    Full-text · Article · Sep 2015
Show more