Assessing Lung Cancer Screening Programs under Uncertainty in a Heterogeneous Population

Henk Broekhuizen (Corresponding Author), Catharina G.M. Groothuis-Oudshoorn, Rozemarijn Vliegenthart, Harry J.M. Groen, Maarten J. IJzerman

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Background: Lung cancer screening can reduce cancer mortality. Most implementation studies focus only on low-dose computed tomography (LDCT) and clinical attributes of screening and do not include preferences of potential participants. In this study we evaluated the perceived value of screening programs based on LDCT, breath analysis (BA), or blood biomarkers (BB) according to the perspective of the target population. Methods: A multi-criteria decision analysis framework was adopted. The weights of seven attributes of screening (sensitivity, specificity, radiation burden, duration of screening process, waiting time until results are communicated, location of screening, and mode of screening) were obtained from an earlier study that included a broad sample from the Netherlands. Performance data for the screening modalities was obtained from clinical trials and expert opinion. Parameter uncertainty about clinical performances was incorporated probabilistically, while heterogeneity in preferences was analyzed through subgroup analyses. Results: The mean overall values were 0.58 (CI: 0.57 to 0.59), 0.57 (CI: 0.56 to 0.59), and 0.44 (CI: 0.43 to 0.45) for BB, BA, and LDCT, respectively. Seventy-seven per cent of respondents preferred BB or BA. For most subgroups, the overall values were similar to those of the entire sample. BA had the highest value for respondents who would have been eligible for earlier screening trials. Discussion: BB and BA seem valuable to participants because they can be applied in a primary care setting. Although LDCT still seems preferable given its strong and positive evidence base, it is important to take non-clinical attributes into account to maximize attendance.

Original languageEnglish
Pages (from-to)1269-1277
Number of pages9
JournalValue in health
Volume21
Issue number11
DOIs
Publication statusPublished - 1 Nov 2018

Fingerprint

Early Detection of Cancer
Uncertainty
Lung Neoplasms
Biomarkers
Tomography
Population
Decision Support Techniques
Health Services Needs and Demand
Expert Testimony
Netherlands
Primary Health Care
Clinical Trials
Radiation
Weights and Measures
Sensitivity and Specificity
Mortality
Neoplasms
Surveys and Questionnaires

Keywords

  • multicriteria decision analysis
  • public preferences
  • subgroup analysis
  • uncertainty
  • lung cancer screening

Cite this

Broekhuizen, Henk ; Groothuis-Oudshoorn, Catharina G.M. ; Vliegenthart, Rozemarijn ; Groen, Harry J.M. ; IJzerman, Maarten J. / Assessing Lung Cancer Screening Programs under Uncertainty in a Heterogeneous Population. In: Value in health. 2018 ; Vol. 21, No. 11. pp. 1269-1277.
@article{3acfa9ad50f741bf9a9abeceda779680,
title = "Assessing Lung Cancer Screening Programs under Uncertainty in a Heterogeneous Population",
abstract = "Background: Lung cancer screening can reduce cancer mortality. Most implementation studies focus only on low-dose computed tomography (LDCT) and clinical attributes of screening and do not include preferences of potential participants. In this study we evaluated the perceived value of screening programs based on LDCT, breath analysis (BA), or blood biomarkers (BB) according to the perspective of the target population. Methods: A multi-criteria decision analysis framework was adopted. The weights of seven attributes of screening (sensitivity, specificity, radiation burden, duration of screening process, waiting time until results are communicated, location of screening, and mode of screening) were obtained from an earlier study that included a broad sample from the Netherlands. Performance data for the screening modalities was obtained from clinical trials and expert opinion. Parameter uncertainty about clinical performances was incorporated probabilistically, while heterogeneity in preferences was analyzed through subgroup analyses. Results: The mean overall values were 0.58 (CI: 0.57 to 0.59), 0.57 (CI: 0.56 to 0.59), and 0.44 (CI: 0.43 to 0.45) for BB, BA, and LDCT, respectively. Seventy-seven per cent of respondents preferred BB or BA. For most subgroups, the overall values were similar to those of the entire sample. BA had the highest value for respondents who would have been eligible for earlier screening trials. Discussion: BB and BA seem valuable to participants because they can be applied in a primary care setting. Although LDCT still seems preferable given its strong and positive evidence base, it is important to take non-clinical attributes into account to maximize attendance.",
keywords = "multicriteria decision analysis, public preferences, subgroup analysis, uncertainty, lung cancer screening",
author = "Henk Broekhuizen and Groothuis-Oudshoorn, {Catharina G.M.} and Rozemarijn Vliegenthart and Groen, {Harry J.M.} and IJzerman, {Maarten J.}",
year = "2018",
month = "11",
day = "1",
doi = "10.1016/j.jval.2018.01.021",
language = "English",
volume = "21",
pages = "1269--1277",
journal = "Value in health",
issn = "1098-3015",
publisher = "Elsevier",
number = "11",

}

Assessing Lung Cancer Screening Programs under Uncertainty in a Heterogeneous Population. / Broekhuizen, Henk (Corresponding Author); Groothuis-Oudshoorn, Catharina G.M.; Vliegenthart, Rozemarijn; Groen, Harry J.M.; IJzerman, Maarten J.

In: Value in health, Vol. 21, No. 11, 01.11.2018, p. 1269-1277.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Assessing Lung Cancer Screening Programs under Uncertainty in a Heterogeneous Population

AU - Broekhuizen, Henk

AU - Groothuis-Oudshoorn, Catharina G.M.

AU - Vliegenthart, Rozemarijn

AU - Groen, Harry J.M.

AU - IJzerman, Maarten J.

PY - 2018/11/1

Y1 - 2018/11/1

N2 - Background: Lung cancer screening can reduce cancer mortality. Most implementation studies focus only on low-dose computed tomography (LDCT) and clinical attributes of screening and do not include preferences of potential participants. In this study we evaluated the perceived value of screening programs based on LDCT, breath analysis (BA), or blood biomarkers (BB) according to the perspective of the target population. Methods: A multi-criteria decision analysis framework was adopted. The weights of seven attributes of screening (sensitivity, specificity, radiation burden, duration of screening process, waiting time until results are communicated, location of screening, and mode of screening) were obtained from an earlier study that included a broad sample from the Netherlands. Performance data for the screening modalities was obtained from clinical trials and expert opinion. Parameter uncertainty about clinical performances was incorporated probabilistically, while heterogeneity in preferences was analyzed through subgroup analyses. Results: The mean overall values were 0.58 (CI: 0.57 to 0.59), 0.57 (CI: 0.56 to 0.59), and 0.44 (CI: 0.43 to 0.45) for BB, BA, and LDCT, respectively. Seventy-seven per cent of respondents preferred BB or BA. For most subgroups, the overall values were similar to those of the entire sample. BA had the highest value for respondents who would have been eligible for earlier screening trials. Discussion: BB and BA seem valuable to participants because they can be applied in a primary care setting. Although LDCT still seems preferable given its strong and positive evidence base, it is important to take non-clinical attributes into account to maximize attendance.

AB - Background: Lung cancer screening can reduce cancer mortality. Most implementation studies focus only on low-dose computed tomography (LDCT) and clinical attributes of screening and do not include preferences of potential participants. In this study we evaluated the perceived value of screening programs based on LDCT, breath analysis (BA), or blood biomarkers (BB) according to the perspective of the target population. Methods: A multi-criteria decision analysis framework was adopted. The weights of seven attributes of screening (sensitivity, specificity, radiation burden, duration of screening process, waiting time until results are communicated, location of screening, and mode of screening) were obtained from an earlier study that included a broad sample from the Netherlands. Performance data for the screening modalities was obtained from clinical trials and expert opinion. Parameter uncertainty about clinical performances was incorporated probabilistically, while heterogeneity in preferences was analyzed through subgroup analyses. Results: The mean overall values were 0.58 (CI: 0.57 to 0.59), 0.57 (CI: 0.56 to 0.59), and 0.44 (CI: 0.43 to 0.45) for BB, BA, and LDCT, respectively. Seventy-seven per cent of respondents preferred BB or BA. For most subgroups, the overall values were similar to those of the entire sample. BA had the highest value for respondents who would have been eligible for earlier screening trials. Discussion: BB and BA seem valuable to participants because they can be applied in a primary care setting. Although LDCT still seems preferable given its strong and positive evidence base, it is important to take non-clinical attributes into account to maximize attendance.

KW - multicriteria decision analysis

KW - public preferences

KW - subgroup analysis

KW - uncertainty

KW - lung cancer screening

UR - http://www.scopus.com/inward/record.url?scp=85056518605&partnerID=8YFLogxK

U2 - 10.1016/j.jval.2018.01.021

DO - 10.1016/j.jval.2018.01.021

M3 - Article

VL - 21

SP - 1269

EP - 1277

JO - Value in health

JF - Value in health

SN - 1098-3015

IS - 11

ER -