Tests for cured proportion for recurrent event count data – Uncensored case with covariates

Sumathi, K and Rao, Aruna K (2014) Tests for cured proportion for recurrent event count data – Uncensored case with covariates. IOSR Journal of Mathematics, 10 (2). pp. 47-59. ISSN 2278-3008

[img] PDF
IOSR_paper.pdf - Published Version
Restricted to Registered users only

Download (587kB) | Request a copy
Official URL: http://www.iosrjournals.org

Abstract

Sumathi and Rao (2008) proposed a cure model for the recurrent event count data. The proposed model was based on the zero inflated Poisson (ZIP) distribution. Several tests were proposed for testing the cured proportion for recurrent event count data (Sumathi and Rao (2010)). In any health related studies, whether epidemiological or long term follow up studies, it is very essential for an investigator to record the information of patients regarding their demographic and socio-economic status as well as their medical history, at the time of clinical examination. These factors, commonly known as covariates, do play a vital role in influencing the health status of the individuals. The present paper is an extension of the work of Sumathi and Rao (2010). It proposes tests for testing the cured proportion in the presence of covariates when the data are uncensored. The covariates are related to the mean parameter of the proposed model, using the log link function. Although testing for p  0 has been done in the past (Broek (1995)), testing for 0 p  p has not been studied for the ZIP model. The small sample performances of the proposed tests are studied using simulations.

Item Type: Article
Uncontrolled Keywords: cure model, recurrent event count data, inflated Poisson distribution, covariates, size and power of the test.
Subjects: Engineering > MIT Manipal > Mathematics
Depositing User: MIT Library
Date Deposited: 14 Jul 2016 09:50
Last Modified: 14 Jul 2016 09:50
URI: http://eprints.manipal.edu/id/eprint/146574

Actions (login required)

View Item View Item