A likelihood ratio test for cancer screening trials

S. G. Self, Ruth Etzioni

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

In randomized cancer screening trials, mortality rates for the screened group relative to those of the control group are not likely to be constant as a function of years from randomization due to the inherent lag between initiation of screening and any putative effects of screening on mortality. In this situation, a log rank test for differences in mortality between the randomization groups will not be optimal. Although optimality could potentially be recovered by use of a weighted log rank statistic, the optimal weights are difficult to specify a priori and the potential loss of power by use of poorly specified weights is great. We describe a likelihood ratio test with two degrees of freedom for use in this situation which is based on a fit of a weakly structured full model. Computation of an approximate significance level for this test is described and a large sample justification for this approximation is given. Size and power properties of the proposed statistic are compared to that of several other statistics in a small simulation study and the statistic is applied to data from the HIP Breast Cancer Screening Trial.

Original languageEnglish (US)
Pages (from-to)44-50
Number of pages7
JournalBiometrics
Volume51
Issue number1
DOIs
StatePublished - May 30 1995
Externally publishedYes

Fingerprint

Likelihood Ratio Test
Early Detection of Cancer
Screening
Cancer
statistics
Statistics
screening
neoplasms
Random Allocation
Randomisation
Mortality
Statistic
Log-rank Statistic
testing
Log-rank Test
Weights and Measures
Hot isostatic pressing
Mortality Rate
Significance level
Breast Cancer

Keywords

  • Cancer
  • Likelihood ratio
  • Screening Trial

ASJC Scopus subject areas

  • Agricultural and Biological Sciences(all)
  • Public Health, Environmental and Occupational Health
  • Agricultural and Biological Sciences (miscellaneous)
  • Applied Mathematics
  • Statistics and Probability

Cite this

A likelihood ratio test for cancer screening trials. / Self, S. G.; Etzioni, Ruth.

In: Biometrics, Vol. 51, No. 1, 30.05.1995, p. 44-50.

Research output: Contribution to journalArticle

Self, S. G. ; Etzioni, Ruth. / A likelihood ratio test for cancer screening trials. In: Biometrics. 1995 ; Vol. 51, No. 1. pp. 44-50.
@article{bfded1024a484c0c86fd9403a58300c6,
title = "A likelihood ratio test for cancer screening trials",
abstract = "In randomized cancer screening trials, mortality rates for the screened group relative to those of the control group are not likely to be constant as a function of years from randomization due to the inherent lag between initiation of screening and any putative effects of screening on mortality. In this situation, a log rank test for differences in mortality between the randomization groups will not be optimal. Although optimality could potentially be recovered by use of a weighted log rank statistic, the optimal weights are difficult to specify a priori and the potential loss of power by use of poorly specified weights is great. We describe a likelihood ratio test with two degrees of freedom for use in this situation which is based on a fit of a weakly structured full model. Computation of an approximate significance level for this test is described and a large sample justification for this approximation is given. Size and power properties of the proposed statistic are compared to that of several other statistics in a small simulation study and the statistic is applied to data from the HIP Breast Cancer Screening Trial.",
keywords = "Cancer, Likelihood ratio, Screening Trial",
author = "Self, {S. G.} and Ruth Etzioni",
year = "1995",
month = "5",
day = "30",
doi = "10.2307/2533313",
language = "English (US)",
volume = "51",
pages = "44--50",
journal = "Biometrics",
issn = "0006-341X",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - A likelihood ratio test for cancer screening trials

AU - Self, S. G.

AU - Etzioni, Ruth

PY - 1995/5/30

Y1 - 1995/5/30

N2 - In randomized cancer screening trials, mortality rates for the screened group relative to those of the control group are not likely to be constant as a function of years from randomization due to the inherent lag between initiation of screening and any putative effects of screening on mortality. In this situation, a log rank test for differences in mortality between the randomization groups will not be optimal. Although optimality could potentially be recovered by use of a weighted log rank statistic, the optimal weights are difficult to specify a priori and the potential loss of power by use of poorly specified weights is great. We describe a likelihood ratio test with two degrees of freedom for use in this situation which is based on a fit of a weakly structured full model. Computation of an approximate significance level for this test is described and a large sample justification for this approximation is given. Size and power properties of the proposed statistic are compared to that of several other statistics in a small simulation study and the statistic is applied to data from the HIP Breast Cancer Screening Trial.

AB - In randomized cancer screening trials, mortality rates for the screened group relative to those of the control group are not likely to be constant as a function of years from randomization due to the inherent lag between initiation of screening and any putative effects of screening on mortality. In this situation, a log rank test for differences in mortality between the randomization groups will not be optimal. Although optimality could potentially be recovered by use of a weighted log rank statistic, the optimal weights are difficult to specify a priori and the potential loss of power by use of poorly specified weights is great. We describe a likelihood ratio test with two degrees of freedom for use in this situation which is based on a fit of a weakly structured full model. Computation of an approximate significance level for this test is described and a large sample justification for this approximation is given. Size and power properties of the proposed statistic are compared to that of several other statistics in a small simulation study and the statistic is applied to data from the HIP Breast Cancer Screening Trial.

KW - Cancer

KW - Likelihood ratio

KW - Screening Trial

UR - http://www.scopus.com/inward/record.url?scp=0029005031&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0029005031&partnerID=8YFLogxK

U2 - 10.2307/2533313

DO - 10.2307/2533313

M3 - Article

C2 - 7766795

AN - SCOPUS:0029005031

VL - 51

SP - 44

EP - 50

JO - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 1

ER -