Stratified breast cancer follow-up using a partially observable Markov decision process

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

1 Citation (Scopus)
10 Downloads (Pure)

Abstract

Frequency and duration of follow-up for patients with breast cancer is still under discussion. Current follow-up consists of annual mammography for the first five years after treatment and does not depend on the personal risk of developing a locoregional recurrence (LRR) or second primary tumor. Aim of this study is to gain insight in how to allocate resources for optimal and personal follow-up. We formulate a discrete-time Partially Observable Markov Decision Process (POMDP) with a finite horzion in which we aim to maximize the total expected number of quality-adjusted life years (QALYs). Transition probabilities were obtained from data from the Netherlands Cancer Registry (NCR). Twice a year the decision is made whether or not a mammography will be performed. Recurrent disease can be detected by both mammography or women themselves (self-detection). The optimal policies were determined for three risk categories based on differentiation of the primary tumor. Our results suggest a slightly more intensive follow-up for patients with a high risk and poorly differentiated tumor, and a less intensive schedule for the other risk groups.
Original languageEnglish
Title of host publicationMarkov Decision Processes in Practice
EditorsRichardus J. Boucherie, Nico van Dijk
Place of PublicationCham
PublisherSpringer
Pages223-244
Number of pages22
ISBN (Electronic)978-3-319-47766-4
ISBN (Print)978-3-319-47764-0
DOIs
Publication statusPublished - Mar 2017

Publication series

NameInternational Series in Operations Research & Management Science
PublisherSpringer International Publishing
Volume248
ISSN (Print)0884-8289

Fingerprint

Markov Chains
Mammography
Breast Neoplasms
Neoplasms
Quality-Adjusted Life Years
Netherlands
Registries
Appointments and Schedules
Recurrence
Therapeutics

Keywords

  • EWI-27869
  • Partially observable markov decision process
  • Breast cancer
  • Optimal policies
  • IR-104507
  • Stratified follow-up

Cite this

Otten, J. W. M., Witteveen, A., Vliegen, I., Siesling, S., Timmer, J. B., & IJzerman, M. J. (2017). Stratified breast cancer follow-up using a partially observable Markov decision process. In R. J. Boucherie, & N. van Dijk (Eds.), Markov Decision Processes in Practice (pp. 223-244). (International Series in Operations Research & Management Science; Vol. 248). Cham: Springer. https://doi.org/10.1007/978-3-319-47766-4_7
Otten, Jan Willem Maarten ; Witteveen, Annemieke ; Vliegen, Ingrid ; Siesling, Sabine ; Timmer, Judith B. ; IJzerman, Maarten Joost. / Stratified breast cancer follow-up using a partially observable Markov decision process. Markov Decision Processes in Practice. editor / Richardus J. Boucherie ; Nico van Dijk. Cham : Springer, 2017. pp. 223-244 (International Series in Operations Research & Management Science).
@inbook{b7649942c90c4603b7aa2b36da07ae98,
title = "Stratified breast cancer follow-up using a partially observable Markov decision process",
abstract = "Frequency and duration of follow-up for patients with breast cancer is still under discussion. Current follow-up consists of annual mammography for the first five years after treatment and does not depend on the personal risk of developing a locoregional recurrence (LRR) or second primary tumor. Aim of this study is to gain insight in how to allocate resources for optimal and personal follow-up. We formulate a discrete-time Partially Observable Markov Decision Process (POMDP) with a finite horzion in which we aim to maximize the total expected number of quality-adjusted life years (QALYs). Transition probabilities were obtained from data from the Netherlands Cancer Registry (NCR). Twice a year the decision is made whether or not a mammography will be performed. Recurrent disease can be detected by both mammography or women themselves (self-detection). The optimal policies were determined for three risk categories based on differentiation of the primary tumor. Our results suggest a slightly more intensive follow-up for patients with a high risk and poorly differentiated tumor, and a less intensive schedule for the other risk groups.",
keywords = "EWI-27869, Partially observable markov decision process, Breast cancer, Optimal policies, IR-104507, Stratified follow-up",
author = "Otten, {Jan Willem Maarten} and Annemieke Witteveen and Ingrid Vliegen and Sabine Siesling and Timmer, {Judith B.} and IJzerman, {Maarten Joost}",
year = "2017",
month = "3",
doi = "10.1007/978-3-319-47766-4_7",
language = "English",
isbn = "978-3-319-47764-0",
series = "International Series in Operations Research & Management Science",
publisher = "Springer",
pages = "223--244",
editor = "Boucherie, {Richardus J.} and {van Dijk}, Nico",
booktitle = "Markov Decision Processes in Practice",

}

Otten, JWM, Witteveen, A, Vliegen, I, Siesling, S, Timmer, JB & IJzerman, MJ 2017, Stratified breast cancer follow-up using a partially observable Markov decision process. in RJ Boucherie & N van Dijk (eds), Markov Decision Processes in Practice. International Series in Operations Research & Management Science, vol. 248, Springer, Cham, pp. 223-244. https://doi.org/10.1007/978-3-319-47766-4_7

Stratified breast cancer follow-up using a partially observable Markov decision process. / Otten, Jan Willem Maarten; Witteveen, Annemieke; Vliegen, Ingrid; Siesling, Sabine; Timmer, Judith B.; IJzerman, Maarten Joost.

Markov Decision Processes in Practice. ed. / Richardus J. Boucherie; Nico van Dijk. Cham : Springer, 2017. p. 223-244 (International Series in Operations Research & Management Science; Vol. 248).

Research output: Chapter in Book/Report/Conference proceedingChapterAcademicpeer-review

TY - CHAP

T1 - Stratified breast cancer follow-up using a partially observable Markov decision process

AU - Otten, Jan Willem Maarten

AU - Witteveen, Annemieke

AU - Vliegen, Ingrid

AU - Siesling, Sabine

AU - Timmer, Judith B.

AU - IJzerman, Maarten Joost

PY - 2017/3

Y1 - 2017/3

N2 - Frequency and duration of follow-up for patients with breast cancer is still under discussion. Current follow-up consists of annual mammography for the first five years after treatment and does not depend on the personal risk of developing a locoregional recurrence (LRR) or second primary tumor. Aim of this study is to gain insight in how to allocate resources for optimal and personal follow-up. We formulate a discrete-time Partially Observable Markov Decision Process (POMDP) with a finite horzion in which we aim to maximize the total expected number of quality-adjusted life years (QALYs). Transition probabilities were obtained from data from the Netherlands Cancer Registry (NCR). Twice a year the decision is made whether or not a mammography will be performed. Recurrent disease can be detected by both mammography or women themselves (self-detection). The optimal policies were determined for three risk categories based on differentiation of the primary tumor. Our results suggest a slightly more intensive follow-up for patients with a high risk and poorly differentiated tumor, and a less intensive schedule for the other risk groups.

AB - Frequency and duration of follow-up for patients with breast cancer is still under discussion. Current follow-up consists of annual mammography for the first five years after treatment and does not depend on the personal risk of developing a locoregional recurrence (LRR) or second primary tumor. Aim of this study is to gain insight in how to allocate resources for optimal and personal follow-up. We formulate a discrete-time Partially Observable Markov Decision Process (POMDP) with a finite horzion in which we aim to maximize the total expected number of quality-adjusted life years (QALYs). Transition probabilities were obtained from data from the Netherlands Cancer Registry (NCR). Twice a year the decision is made whether or not a mammography will be performed. Recurrent disease can be detected by both mammography or women themselves (self-detection). The optimal policies were determined for three risk categories based on differentiation of the primary tumor. Our results suggest a slightly more intensive follow-up for patients with a high risk and poorly differentiated tumor, and a less intensive schedule for the other risk groups.

KW - EWI-27869

KW - Partially observable markov decision process

KW - Breast cancer

KW - Optimal policies

KW - IR-104507

KW - Stratified follow-up

U2 - 10.1007/978-3-319-47766-4_7

DO - 10.1007/978-3-319-47766-4_7

M3 - Chapter

SN - 978-3-319-47764-0

T3 - International Series in Operations Research & Management Science

SP - 223

EP - 244

BT - Markov Decision Processes in Practice

A2 - Boucherie, Richardus J.

A2 - van Dijk, Nico

PB - Springer

CY - Cham

ER -

Otten JWM, Witteveen A, Vliegen I, Siesling S, Timmer JB, IJzerman MJ. Stratified breast cancer follow-up using a partially observable Markov decision process. In Boucherie RJ, van Dijk N, editors, Markov Decision Processes in Practice. Cham: Springer. 2017. p. 223-244. (International Series in Operations Research & Management Science). https://doi.org/10.1007/978-3-319-47766-4_7