OpenLB
Open Library of Bioscience
Advanced Search
ESMO open
Jan, 2025;10 (1): 104094.

Informative censoring in externally controlled clinical trials: a potential source of bias.

T Rudra Gupta, D E Schwartz, R Saha, P Y Wen, R Rahman, L Trippa
Author Information
  1. T Rudra Gupta: Dana-Farber Cancer Institute, Boston, USA; Harvard School of Public Health, Boston, USA. Electronic address: tulika@ds.dfci.harvard.edu.
  2. D E Schwartz: Dana-Farber Cancer Institute, Boston, USA; Harvard School of Public Health, Boston, USA.
  3. R Saha: Dana-Farber Cancer Institute, Boston, USA; Harvard School of Public Health, Boston, USA.
  4. P Y Wen: Dana-Farber Cancer Institute, Boston, USA; Harvard Medical School, Boston, USA.
  5. R Rahman: Harvard Medical School, Boston, USA; Brigham and Women's Hospital, Boston, USA.
  6. L Trippa: Dana-Farber Cancer Institute, Boston, USA; Harvard School of Public Health, Boston, USA.
PMID: 39754980 DOI: 10.1016/j.esmoop.2024.104094

Abstract

BACKGROUND: Cancer researchers frequently consider the use of single-arm and randomized controlled clinical trial designs that leverage external data. The literature has reported extensively on how the use of external data can introduce bias through a variety of distortion mechanisms. In this article, we focus on a distortion mechanism that is often overlooked: informative censoring. Informative censoring arises when there is statistical dependence between patients' censoring times and survival times.
MATERIALS AND METHODS: We used simulations to investigate how informative censoring of external controls (ECs) can influence the results of cancer clinical trials. Our simulations included the following: (i) model-based replicates of clinical trials and in silico glioblastoma trials obtained by resampling patients from completed phase III trials; (ii) single-arm and randomized controlled cancer clinical trial designs; and (iii) different types of informative censoring, with positive or negative associations between censoring times and survival times.
RESULTS: Our simulations showed that informative censoring of EC data can bias cancer clinical trial results. The direction of the bias depends on the censoring mechanism. Similarly, informative censoring can inflate or reduce type I error and power.
CONCLUSIONS: Selection of EC data and the decision to leverage these data in the analysis of clinical trials should account for the risk of bias due to informative censoring. Analyses to detect informative censoring are recommended when the clinical trial design involves external data.

Keywords

clinical trials external control data informative censoring survival analysis

References

Biometrics. 1998 Sep;54(3):1176-82 [PMID: 9840972]
Biometrics. 2023 Mar;79(1):394-403 [PMID: 34694626]
Biom J. 2011 Sep;53(5):822-37 [PMID: 21770046]
Stat Med. 2014 Mar 30;33(7):1242-58 [PMID: 24122911]
Blood. 2010 Aug 5;116(5):679-86 [PMID: 20385792]
J Clin Oncol. 2014 Sep 20;32(27):3068-74 [PMID: 25113767]
Biometrics. 2001 Jun;57(2):404-13 [PMID: 11414563]
Pharm Stat. 2014 Jan-Feb;13(1):41-54 [PMID: 23913901]
Neuro Oncol. 2024 May 3;26(5):796-810 [PMID: 38254183]
J Biopharm Stat. 2022 Jan 2;32(1):158-169 [PMID: 34756158]
Clin Cancer Res. 2019 Aug 15;25(16):4993-5001 [PMID: 31175098]
AIDS Res Hum Retroviruses. 2014 Dec;30(12):1170-7 [PMID: 25183195]
Blood Adv. 2023 Oct 10;7(19):5680-5690 [PMID: 36534147]
J Med Internet Res. 2023 Apr 5;25:e43484 [PMID: 37018021]
Nat Rev Drug Discov. 2020 Dec;19(12):821-822 [PMID: 32811986]
Stat Med. 2007 Dec 30;26(30):5411-21 [PMID: 18058850]
Adv Ther. 2011 Apr;28(4):334-40 [PMID: 21432029]
Lancet Oncol. 2021 Oct;22(10):e456-e465 [PMID: 34592195]
Am Stat. 2024;78(1):76-87 [PMID: 38680760]
Pharm Stat. 2023 Jul-Aug;22(4):721-738 [PMID: 36794571]
Curr Control Trials Cardiovasc Med. 2002 Jan 8;3(1):4 [PMID: 11985778]
Annu Rev Public Health. 1997;18:83-104 [PMID: 9143713]
Health Serv Res. 2014 Oct;49(5):1701-20 [PMID: 24779867]
Lifetime Data Anal. 1995;1(3):241-54 [PMID: 9385104]
Biostatistics. 2012 Apr;13(2):341-54 [PMID: 22223746]
Neuro Oncol. 2022 Feb 1;24(2):247-256 [PMID: 34106270]
J Clin Epidemiol. 2009 Oct;62(10):1095-102 [PMID: 19251392]
Kidney Int. 2008 Jan;73(2):148-53 [PMID: 17978812]
Biostatistics. 2006 Jan;7(1):29-40 [PMID: 15923407]
Stat Biosci. 2022;14(1):79-89 [PMID: 34178164]
Ther Innov Regul Sci. 2018 Sep;52(5):546-559 [PMID: 29909645]
Pharm Stat. 2024 Sep-Oct;23(5):645-661 [PMID: 38442919]
Pharm Stat. 2021 Jul;20(4):783-792 [PMID: 33655598]
Pharmacoepidemiol Drug Saf. 2020 Nov;29(11):1382-1392 [PMID: 32964514]
Clin Pharmacol Ther. 2020 Apr;107(4):806-816 [PMID: 31725899]
J Clin Oncol. 2008 Dec 10;26(35):5775-82 [PMID: 19001321]
Nat Commun. 2022 Oct 2;13(1):5783 [PMID: 36184621]
Clin Cancer Res. 2023 Jun 13;29(12):2194-2198 [PMID: 36939557]
Eur J Cancer. 2024 Jan;196:113443 [PMID: 38008030]
Pharmacoepidemiol Drug Saf. 2020 Oct;29(10):1228-1235 [PMID: 32162381]
J Biopharm Stat. 2022 May 4;32(3):400-413 [PMID: 35675348]
Eur J Cancer. 2021 Aug;153:190-202 [PMID: 34186504]

Grants

  1. R01 LM013352/NLM NIH HHS
  2. T32 CA009337/NCI NIH HHS

MeSH Term

Humans
Bias
Computer Simulation
Randomized Controlled Trials as Topic
Research Design
Neoplasms
Clinical Trials, Phase III as Topic
Glioblastoma
Journal Article

Word Cloud

Similar Articles

Cited By