Dataset Information

Time-dependent classification accuracy curve under marker-dependent sampling.

ABSTRACT: Evaluating the classification accuracy of a candidate biomarker signaling the onset of disease or disease status is essential for medical decision making. A good biomarker would accurately identify the patients who are likely to progress or die at a particular time in the future or who are in urgent need for active treatments. To assess the performance of a candidate biomarker, the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) are commonly used. In many cases, the standard simple random sampling (SRS) design used for biomarker validation studies is costly and inefficient. In order to improve the efficiency and reduce the cost of biomarker validation, marker-dependent sampling (MDS) may be used. In a MDS design, the selection of patients to assess true survival time is dependent on the result of a biomarker assay. In this article, we introduce a nonparametric estimator for time-dependent AUC under a MDS design. The consistency and the asymptotic normality of the proposed estimator is established. Simulation shows the unbiasedness of the proposed estimator and a significant efficiency gain of the MDS design over the SRS design.

SUBMITTER: Zhu Z

PROVIDER: S-EPMC4930889 | biostudies-other | 2016 Jul

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Time-dependent classification accuracy curve under marker-dependent sampling.

Zhu Zhaoyin Z Wang Xiaofei X Saha-Chaudhuri Paramita P Kosinski Andrzej S AS George Stephen L SL

Biometrical journal. Biometrische Zeitschrift 20160427 4

Evaluating the classification accuracy of a candidate biomarker signaling the onset of disease or disease status is essential for medical decision making. A good biomarker would accurately identify the patients who are likely to progress or die at a particular time in the future or who are in urgent need for active treatments. To assess the performance of a candidate biomarker, the receiver operating characteristic (ROC) curve and the area under the ROC curve (AUC) are commonly used. In many cas ...[more]

PMID: 27119599

Similar Datasets

Project description:BACKGROUND:An extended-release, once-daily, oral formulation of tacrolimus is currently used after kidney transplantation as a substitute for the conventional twice-daily formulation. The purpose of this study was to provide a limited sampling strategy with minimum and optimum sampling points to predict the tacrolimus area under the concentration-time curve (AUC) after administration of once-daily tacrolimus in de novo adult kidney transplant patients. METHODS:A total of 36 adult Japanese kidney transplant patients receiving once-daily tacrolimus were included: 31 were allocated to a study group to develop limited sampling strategy (LSS) model equations based on multiple stepwise linear regression analysis, and 5 were allocated to a validation group to estimate the precision of the LSS equations developed by the study group. Twelve-hour AUC (AUC0-12) was calculated by the trapezoidal rule, and the relationship between individual concentration points and AUC0-12 were determined by multiple linear regression analysis. The coefficient of determination (R2) was used to assess the goodness-of-fit of the regression models. Three error indices (mean error, mean absolute error, and root mean squared prediction error) were calculated to evaluate predictive bias, accuracy, and precision, respectively. Quality of the statistical models was compared with Akaike's information criterion (AIC). RESULTS:A four-point model using C0, C2, C4 and C6 gave the best fit to predict AUC0-12 (R2 = 0.978). In the three- and two-point models, the best fits were at time points C2, C4, and C6 (R2 = 0.973), and C2 and C6 (R2 = 0.962), respectively. All three models reliably estimated tacrolimus AUC0-12, consistent with evaluations by the three error indices and Akaike's information criterion. Practically, the two-point model with C2 and C6 was considered to be the best combination, providing a highly accurate prediction and the lowest blood sampling frequency. CONCLUSIONS:The two-point model with C2 and C6 may be valuable in reducing the burden on patients, as well as medical costs, for once-daily tacrolimus monitoring.

Dataset Information

Time-dependent classification accuracy curve under marker-dependent sampling.

Publications

Time-dependent classification accuracy curve under marker-dependent sampling.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets