EUVIMEDEUVIMED entdecken
Crossref

A Simplified and Flexible Counterfactual Approach to Treatment Switching at a Secondary Baseline in Randomized Trials

Patrick M Schnell iD, Michael Johnson iD, Vladimir Svetnik iD, Shahrul Mt-Isa iD

Pharmaceutical Statistics · 2026-11 · 25 (6)

DOI: 10.1002/pst.70144 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Pharmaceutical Statistics, online erschienen am 18.11.2026. Verfasst von Patrick M Schnell und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Immunology, Genetics and Pathology, Uppsala Universitet, Uppsala, Sweden. Hauptthemen laut MEDLINE: Randomized Controlled Trials as Topic, Models, Statistical, Drug Substitution, Research Design. Die Arbeit nennt 15 Literaturangaben. Gefördert von: Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Treatment switching, common in randomized oncology trials, complicates outcome interpretation and causal attribution. One way to address these complications is to estimate causal effects that would have been realized if treatment switching were prevented or restricted. Although general tools such as g‐estimation and inverse probability of censoring weighting can be used to identify such estimands, they require controlling time‐dependent confounding over the entire follow‐up. The presence of a “secondary baseline” such as discontinuation of initial treatment can simplify matters if all of the confounding between treatment switching and outcome can be controlled at that timepoint. Prior methods taking advantage of such a secondary baseline were derived as simplifications from rank‐preserving structural failure time models and inherited some of the assumptions. We present identification and estimation methods by beginning with g‐estimation or inverse probability of censoring weighting and deriving the simplifications allowed by a secondary baseline. This approach results in more robust and flexible estimation than existing methods incorporating the secondary baseline while remaining simpler to implement than the more general methods.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Randomized Controlled Trials as TopicModels, StatisticalDrug SubstitutionResearch DesignHumansData Interpretation, StatisticalTreatment OutcomeProbabilityTime Factors

Stichwörter der Autor:innen

counterfactual estimandg‐computationinverse probability of censoring weightingrandomized trialtreatment switching

Autor:innen und Institutionen

  • Department of Immunology, Genetics and Pathology, Uppsala Universitet, Uppsala, Sweden.
  • Division of Biostatistics, The Ohio State University College of Public Health, Columbus, Ohio, USA.
  • Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden.
  • Merck & Co., Inc., Rahway, New Jersey, USA.
  • MSD, Zurich, Switzerland.

ORCID-Profile: Patrick M Schnell, Michael Johnson, Vladimir Svetnik, Shahrul Mt-Isa

Förderung

  • Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA

Zitieren

APA

Schnell, P. M., Johnson, M., Svetnik, V., & Mt-Isa, S. (2026). A Simplified and Flexible Counterfactual Approach to Treatment Switching at a Secondary Baseline in Randomized Trials. Pharmaceutical Statistics, 25(6). https://doi.org/10.1002/pst.70144

Vancouver

Schnell PM, Johnson M, Svetnik V, Mt-Isa S. A Simplified and Flexible Counterfactual Approach to Treatment Switching at a Secondary Baseline in Randomized Trials. Pharmaceutical Statistics. 2026;25(6). doi:10.1002/pst.70144

BibTeX
@article{Schnell2026,
  author = {Schnell, Patrick M and Johnson, Michael and Svetnik, Vladimir and Mt-Isa, Shahrul},
  title = {A Simplified and Flexible Counterfactual Approach to Treatment Switching at a Secondary Baseline in Randomized Trials},
  journal = {Pharmaceutical Statistics},
  year = {2026},
  volume = {25},
  number = {6},
  doi = {10.1002/pst.70144}
}
Kostenlos starten →