Detecting multiple change points in piecewise constant hazard functions

Melody Goodman, Yi Li, Ram C. Tiwari

Research output: Contribution to journalArticle

Abstract

The National Cancer Institute (NCI) suggests a sudden reduction in prostate cancer mortality rates, likely due to highly successful treatments and screening methods for early diagnosis. We are interested in understanding the impact of medical breakthroughs, treatments, or interventions, on the survival experience for a population. For this purpose, estimating the underlying hazard function, with possible time change points, would be of substantial interest, as it will provide a general picture of the survival trend and when this trend is disrupted. Increasing attention has been given to testing the assumption of a constant failure rate against a failure rate that changes at a single point in time. We expand the set of alternatives to allow for the consideration of multiple change-points, and propose a model selection algorithm using sequential testing for the piecewise constant hazard model. These methods are data driven and allow us to estimate not only the number of change points in the hazard function but where those changes occur. Such an analysis allows for better understanding of how changing medical practice affects the survival experience for a patient population. We test for change points in prostate cancer mortality rates using the NCI Surveillance, Epidemiology, and End Results dataset.

Original languageEnglish (US)
Pages (from-to)2523-2532
Number of pages10
JournalJournal of Applied Statistics
Volume38
Issue number11
DOIs
StatePublished - Nov 2011

Fingerprint

Hazard Function
Constant function
Change Point
Prostate Cancer
Mortality Rate
Failure Rate
Cancer
Sequential Testing
Hazard Models
Time Change
Epidemiology
Data-driven
Model Selection
Surveillance
Expand
Screening
Likely
Testing
Change point
Hazard function

Keywords

  • Cancer
  • Change points
  • Hazard function
  • Piecewise constant
  • Survival analysis

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Detecting multiple change points in piecewise constant hazard functions. / Goodman, Melody; Li, Yi; Tiwari, Ram C.

In: Journal of Applied Statistics, Vol. 38, No. 11, 11.2011, p. 2523-2532.

Research output: Contribution to journalArticle

@article{6bb25193f69c497980b11a7a63161150,
title = "Detecting multiple change points in piecewise constant hazard functions",
abstract = "The National Cancer Institute (NCI) suggests a sudden reduction in prostate cancer mortality rates, likely due to highly successful treatments and screening methods for early diagnosis. We are interested in understanding the impact of medical breakthroughs, treatments, or interventions, on the survival experience for a population. For this purpose, estimating the underlying hazard function, with possible time change points, would be of substantial interest, as it will provide a general picture of the survival trend and when this trend is disrupted. Increasing attention has been given to testing the assumption of a constant failure rate against a failure rate that changes at a single point in time. We expand the set of alternatives to allow for the consideration of multiple change-points, and propose a model selection algorithm using sequential testing for the piecewise constant hazard model. These methods are data driven and allow us to estimate not only the number of change points in the hazard function but where those changes occur. Such an analysis allows for better understanding of how changing medical practice affects the survival experience for a patient population. We test for change points in prostate cancer mortality rates using the NCI Surveillance, Epidemiology, and End Results dataset.",
keywords = "Cancer, Change points, Hazard function, Piecewise constant, Survival analysis",
author = "Melody Goodman and Yi Li and Tiwari, {Ram C.}",
year = "2011",
month = "11",
doi = "10.1080/02664763.2011.559209",
language = "English (US)",
volume = "38",
pages = "2523--2532",
journal = "Journal of Applied Statistics",
issn = "0266-4763",
publisher = "Routledge",
number = "11",

}

TY - JOUR

T1 - Detecting multiple change points in piecewise constant hazard functions

AU - Goodman, Melody

AU - Li, Yi

AU - Tiwari, Ram C.

PY - 2011/11

Y1 - 2011/11

N2 - The National Cancer Institute (NCI) suggests a sudden reduction in prostate cancer mortality rates, likely due to highly successful treatments and screening methods for early diagnosis. We are interested in understanding the impact of medical breakthroughs, treatments, or interventions, on the survival experience for a population. For this purpose, estimating the underlying hazard function, with possible time change points, would be of substantial interest, as it will provide a general picture of the survival trend and when this trend is disrupted. Increasing attention has been given to testing the assumption of a constant failure rate against a failure rate that changes at a single point in time. We expand the set of alternatives to allow for the consideration of multiple change-points, and propose a model selection algorithm using sequential testing for the piecewise constant hazard model. These methods are data driven and allow us to estimate not only the number of change points in the hazard function but where those changes occur. Such an analysis allows for better understanding of how changing medical practice affects the survival experience for a patient population. We test for change points in prostate cancer mortality rates using the NCI Surveillance, Epidemiology, and End Results dataset.

AB - The National Cancer Institute (NCI) suggests a sudden reduction in prostate cancer mortality rates, likely due to highly successful treatments and screening methods for early diagnosis. We are interested in understanding the impact of medical breakthroughs, treatments, or interventions, on the survival experience for a population. For this purpose, estimating the underlying hazard function, with possible time change points, would be of substantial interest, as it will provide a general picture of the survival trend and when this trend is disrupted. Increasing attention has been given to testing the assumption of a constant failure rate against a failure rate that changes at a single point in time. We expand the set of alternatives to allow for the consideration of multiple change-points, and propose a model selection algorithm using sequential testing for the piecewise constant hazard model. These methods are data driven and allow us to estimate not only the number of change points in the hazard function but where those changes occur. Such an analysis allows for better understanding of how changing medical practice affects the survival experience for a patient population. We test for change points in prostate cancer mortality rates using the NCI Surveillance, Epidemiology, and End Results dataset.

KW - Cancer

KW - Change points

KW - Hazard function

KW - Piecewise constant

KW - Survival analysis

UR - http://www.scopus.com/inward/record.url?scp=79960892854&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79960892854&partnerID=8YFLogxK

U2 - 10.1080/02664763.2011.559209

DO - 10.1080/02664763.2011.559209

M3 - Article

VL - 38

SP - 2523

EP - 2532

JO - Journal of Applied Statistics

JF - Journal of Applied Statistics

SN - 0266-4763

IS - 11

ER -