We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: Semiparametric Methods for Exposure Misclassification in Propensity Score-Based Time-to-Event Data Analysis

Abstract: In epidemiology, identifying the effect of exposure variables in relation to a time-to-event outcome is a classical research area of practical importance. Incorporating propensity score in the Cox regression model, as a measure to control for confounding, has certain advantages when outcome is rare. However, in situations involving exposure measured with moderate to substantial error, identifying the exposure effect using propensity score in Cox models remains a challenging yet unresolved problem. In this paper, we propose an estimating equation method to correct for the exposure misclassification-caused bias in the estimation of exposure-outcome associations. We also discuss the asymptotic properties and derive the asymptotic variances of the proposed estimators. We conduct a simulation study to evaluate the performance of the proposed estimators in various settings. As an illustration, we apply our method to correct for the misclassification-caused bias in estimating the association of PM2.5 level with lung cancer mortality using a nationwide prospective cohort, the Nurses' Health Study (NHS). The proposed methodology can be applied using our user-friendly R function published online.
Comments: Withdrawn due to grant related requirements
Subjects: Methodology (stat.ME); Statistics Theory (math.ST); Applications (stat.AP); Machine Learning (stat.ML)
Cite as: arXiv:1903.07782 [stat.ME]
  (or arXiv:1903.07782v2 [stat.ME] for this version)

Submission history

From: Yingrui Yang [view email]
[v1] Tue, 19 Mar 2019 00:57:16 GMT (46kb)
[v2] Sat, 8 Jun 2019 07:50:43 GMT (0kb,I)

Link back to: arXiv, form interface, contact.