Mutsvari, T.

Hierarchical modeling of agreement (Article)

2012-12-10T00:00:00Z

Kappa-like agreement indexes are often used to assess the agreement among examiners on a categorical scale. They have the particularity of correcting the level of agreement for the effect of chance. In the present paper, we first define two agreement indexes belonging to this family in a hierarchical context. In particular, we consider the cases of a random and fixed set of examiners. Then, we develop a method to evaluate the influence of factors on these indexes. Agreement indexes are directly related to a set of covariates through a hierarchical model. We obtain the posterior distribution of the model parameters in a Bayesian framework. We apply the proposed approach on dental data and compare it with the generalized estimating equations approach.

Examiner performance in calibration exercises compared with field conditions when scoring caries experience (Article)

2011-02-25T00:00:00Z

The objective of this study was to verify how valid misclassification measurements obtained from a 'pre-survey' calibration exercise are by comparing them to validation scores obtained in 'field' conditions. Validation data were collected from the 'Smile for Life' project, an oral health intervention study in Flemish children. A calibration exercise was organized under 'pre-survey' conditions (32 age-matched children examined by eight examiners and the benchmark scorer). In addition, using a pre-determined sampling scheme blinded to the examiners, the benchmark scorer re-examined between six and 11 children screened by each of the dentists during the survey. Factors influencing sensitivity and specificity for scoring caries experience (CE) were investigated, including examiner, tooth type, surface type, tooth position (upper/lower jaw, right/left side) and validation setting (pre-survey versus field). In order to account for the clustering effect in the data, a generalized estimating equations approach was applied. Sensitivity scores were influenced not only by the calibration setting (lower sensitivity in field conditions, p < 0.01), but also by examiner, tooth type (lower sensitivity in molar teeth, p < 0.01) and tooth position (lower sensitivity in the lower jaw, p < 0.01). Factors influencing specificity were examiner, tooth type (lower specificity in molar teeth, p < 0.01) and surface type (the occlusal surface with a lower specificity than other surfaces) but not the validation setting. Misclassification measurements for scoring CE are influenced by several factors. In this study, the validation setting influenced sensitivity, with lower scores obtained when measuring data validity in 'field' conditions. Results obtained in a pre-survey calibration setting need to be interpreted with caution and do not (always) reflect the actual performance of examiners during the field work.

Correcting for misclassification for a monotone disease process with an application in dental research (Article)

2010-12-30T00:00:00Z

Motivated by a longitudinal oral health study, we evaluate the performance of binary Markov models in which the response variable is subject to an unconstrained misclassification process and follows a monotone or progressive behavior. Theoretical and empirical arguments show that the simple version of the model can be used to estimate the prevalence, incidences, and misclassification parameters without the need of external information and that the incidence estimators associated with the model outperformed approaches previously proposed in the literature. We propose an extension of the simple version of the binary Markov model to describe the relationship between the covariates and the prevalence and incidence allowing for different classifiers. We implemented a Bayesian version of the extended model and show that, under the settings of our motivating example, the parameters can be estimated without any external information. Finally, the analyses of the motivating problem are presented. Copyright

Factors that influence data quality in caries experience detection: A multilevel modeling approach (Article)

2010-11-01T00:00:00Z

Caries experience detection is prone to misclassification. For this reason, calibration exercises which aim at assessing and improving the scoring behavior of dental raters are organized. During a calibration exercise, a sample of children is examined by the benchmark scorer and the dental examiners. This produces a 2 × 2 contingency table with the true and possibly misclassified responses. The entries in this misclassification table allow to estimate the sensitivity and the specificity of the raters. However, in many dental studies, the uncertainty with which sensitivity and specificity are estimated is not expressed. Further, caries experience data have a hierarchical structure since the data are recorded for the surfaces nested in the teeth within the mouth. Therefore, it is important to report the uncertainty using confidence intervals and to take the clustering into account. Here we apply a Bayesian logistic multilevel model for estimating the sensitivity and specificity. The main goal of this research is to find the factors that influence the true scoring of caries experience accounting for the hierarchical structure in the data. In our analysis, we show that the dentition type and tooth or surface type affect the quality of caries experience detection. Copyright

Measurement, analysis and interpretation of examiner reliability in caries experience surveys: some methodological thoughts (Article)

2010-10-13T00:00:00Z

Data obtained from calibration exercises are used to assess the level of agreement between examiners (and the benchmark examiner) and/or between repeated examinations by the same examiner in epidemiological surveys or large-scale clinical studies. Agreement can be measured using different techniques: kappa statistic, percentage agreement, dice coefficient, sensitivity and specificity. Each of these methods shows specific characteristics and has its own shortcomings. The aim of this contribution is to critically review techniques for the measurement and analysis of examiner agreement and to illustrate this using data from a recent survey in young children, the Smile for Life project. The above-mentioned agreement measures are influenced (in differing ways and extents) by the unit of analysis (subject, tooth, surface level) and the disease level in the validation sample. These effects are more pronounced for percentage agreement and kappa than for sensitivity and specificity. It is, therefore, important to include information on unit of analysis and disease level (in validation sample) when reporting agreement measures. Also, confidence intervals need to be included since they indicate the reliability of the estimate. When dependency among observations is present [as is the case in caries experience data sets with typical hierarchical structure (surface-tooth-subject)], this will influence the width of the confidence interval and should therefore not be ignored. In this situation, the use of multilevel modelling is necessary. This review clearly shows that there is a need for the development of guidelines for the measurement, interpretation and reporting of examiner reliability in caries experience surveys.