Rank tests for clustered survival data.

In a clinical trial, we may randomize subjects (called clusters) to different treatments (called groups), and make observations from multiple sites (called units) of each subject. In this case, the observations within each subject could be dependent, whereas those from different subjects are independent. If the outcome of interest is the time to an event, we may use the standard rank tests proposed for independent survival data, such as the logrank and Wilcoxon tests, to test the equality of marginal survival distributions, but their standard error should be modified to accommodate the possible intracluster correlation.

A parametric model for long-term follow-up data from phase III breast cancer clinical trials.

We propose a parametric version of a univariate gamma frailty model. The proposed model is shown to be flexible enough to model long-term follow-up survival data from breast cancer clinical trials when the treatment effect diminishes as time progresses, a case for which neither the proportional hazards nor proportional odds assumptions are satisfied. The observed information matrix is computed to evaluate the variances of parameter estimates.

Update on late relapse of germ cell tumor: a clinical and molecular analysis.

Purpose: Analysis of patients with late relapse (LR) of germ cell tumor (GCT) with reports on clinical characteristics, outcomes, and molecular and cytogenetic features.

Patients And Methods: Eighty-three patients evaluated at Indiana University from 1993 through 2000 for relapse of GCT more than 2 years from initial therapy were reviewed. Available specimens were investigated for expression of the transcription regulator FoxD3 and apurinic/apyrimidinic endonuclease and the presence of chromosome 12 abnormalities.

Sample size calculation for rank tests comparing K survival distributions.

Rank tests, such as logrank or Wilcoxon rank sum tests, have been popularly used to compare survival distributions of two or more groups in the presence of right censoring. However, there has been little research on sample size calculation methods for rank tests to compare more than two groups. An existing method is based on a crude approximation, which tends to underestimate sample size, i.

Application of an adjusted chi2 statistic to site-specific data in observational dental studies.

Background: When a binary response is observed on teeth from each subject belonging to 2 or more exposure groups, application of the usual Pearson chi2 tests is invalid, since such responses within the same subject are not independent. Consequently, special statistical methods are needed to control for the correlation among teeth (sites) within the same subject. A simple adjustment to the Pearson chi2 statistic has been proposed for comparing proportions in site-specific data.

Etoposide, ifosfamide and cisplatin (VIP) plus concurrent radiation therapy for previously untreated limited small cell lung cancer (SCLC): a Hoosier Oncology Group (HOG) phase II study.

Results of a previous Hoosier Oncology Group (HOG) study revealed a small survival advantage for VIP versus etoposide and cisplatin (EP) for patients with extensive stage small cell lung cancer (SCLC). This phase II study evaluated VIP with concurrent thoracic radiotherapy in patients with limited stage SCLC. Eligible patients had a Karnofsky Performance Score > or = 50, no prior chemotherapy or radiotherapy, and adequate end organ function.