CMStatistics 2022: Start Registration
View Submission - CMStatistics
Title: An augmented likelihood approach incorporating error-prone auxiliary data into a survival analysis Authors:  Noorie Hyun - Medical College of Wisconsin (United States) [presenting]
Pamela Shaw - Kaiser Permanente Washington Health Research Institute (United States)
Abstract: Substantial clinical data collection is available from large healthcare community studies or electronic health records (EHR) in health systems. However, data accuracy can vary according to measurement methods. For example, self-reported medical history can include bias, such as recall bias or response bias. In contrast, biomarkers from a laboratory test are less likely to be biased. We are motivated to study what benefit we can gain by augmenting error-prone self-reported and biomarker-based disease diagnoses in regression for time-to-disease onset. The proposed model addresses left-truncation and interval-censoring in time-to-disease onset outcomes. Also, self-reported disease diagnosis errors are corrected using sensitivity and specificity parameters in the joint likelihood. Compared to other models using biomarker or self-reported data, comprehensive simulation studies found appealing finite sample properties of the proposed augmenting model, including the smallest mean square error. The proposed model is applied to the Hispanic Community Health Study/ Study of Latino data to quantify risk factors associated with diabetes onset.