2016-04-15

Ying Zhang, Ph.D., Professor, Department of Biostatistics, IUPUI

"Nonparametric Inference with Misclassified Competing Risks Data"

Competing risks data play an important role in medicine, epidemiology and public health. However, a frequent complication in biomedical research is that ascertainment of the causes of event is subject to error and this can lead to seriously biased inference. To deal with this issue when misclassification probabilities are not a-priory known, a double-sampling design can be adopted to randomly select a small sample of subjects with terminal event and then using a gold standard, which could be a very expensive outcome ascertainment procedure, to determine the true cause. Based on this additional information and a parametric model for the probability of the misclassification on the causes, we develop a closed form nonparametric pseudolikelihood estimator (NPMPLE) of the cause-specific cumulative incidences and we show that the estimator is uniformly consistent and converges weakly to a tight zero-mean Gaussian random field. We conduct simulation studies to justify the validity of the proposed method. Finally, the method is applied to a motivating example from a large HIV/AIDS study in Sub-Saharan Africa to evaluate the PEPFAR-funded HIV healthcare programs, where serious death under-reporting results in classifying deceased patients as being disengagers from HIV care.

Stay connected TwitterFacebookLinkedInYouTubeInstagram