Quantitative Bias Analysis for Single-Arm Trials With External Control Arms

JAMA Network Open, vol. 8, pp. e252152

Abstract

Importance: Unmeasured confounding is a key concern for decision-makers when observational datasets are used to assemble external control arms (ECAs) for single-arm trials. Objective: To investigate the utility of quantitative bias analysis (QBA) for exploring the sensitivity to unmeasured confounding of nonrandomized analyses using ECAs. Design, Setting, and Participants: This study emulated 15 treatment comparisons using experimental arms from existing randomized trials in advanced non-small cell lung cancer (aNSCLC) conducted after 2011 and ECAs derived from observational data. Participants were eligible individuals diagnosed with aNSCLC between January 1, 2011, and March 1, 2020. After adjustment for measured baseline confounders, a prespecified QBA was conducted to address potential bias by known unmeasured and mismeasured confounders. The QBA relied on a synthesis of external evidence from a targeted literature search, randomized trial data, and clinician input. Hazard ratios from the original randomized trials were compared with those from their emulation based on ECA analyses. Analyses were completed from February 2022 to October 2023. Exposure: Initiation of systemic therapies for aNSCLC. Main outcomes and measures: Hazard ratios for all-cause death. Results: Sample sizes varied from 52 to 830 depending on the treatment group. The mean difference in the log hazard ratio estimates when using the original control arm vs the ECA for each trial was 0.247 in unadjusted analyses (ratio of hazard ratios, 1.36), 0.139 when adjusted for measured confounders (ratio of hazard ratios, 1.22), and 0.098 when adding external adjustment for unmeasured and mismeasured confounders (ratio of hazard ratios, 1.17). Conclusions and Relevance: QBA was feasible and informative in ECA analyses in which residual confounding was expected to be the most important source of bias. These findings encourage further exploration of how QBA can help quantify the impact of bias in other settings and when using other data sources.

Authors 14

1. Alind Gupta corresponding

   University of Toronto · Cytel (United States)

   Affiliation as printed

   Cytel, Waltham, Massachusetts

   Now with Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada

   Now with Inka Health, Toronto, Ontario, Canada

2. Grace G. Hsu

   Cytel (United States)

   Affiliation as printed

   Cytel, Waltham, Massachusetts

3. Seamus Kent

   Erasmus University Rotterdam

   Affiliation as printed

   Erasmus School of Health Policy and Management, Erasmus University Rotterdam, Rotterdam, the Netherlands

4. Stephen J. Duffield

   National Institute for Health and Care Excellence

   Affiliation as printed

   National Institute for Health and Care Excellence, London, United Kingdom

5. Evie Merinopoulou

   Cytel (United States)

   Affiliation as printed

   Cytel, Waltham, Massachusetts

6. Alexandre Lockhart

   Cytel (United States)

   Affiliation as printed

   Cytel, Waltham, Massachusetts

7. Paul Arora

   University of Toronto

   Affiliation as printed

   Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada

   Inka Health, Toronto, Ontario, Canada

8. Joshua A. Ray

   Roche (Switzerland)

   Affiliation as printed

   Global Access, F. Hoffmann–La Roche Ltd, Basel, Switzerland

9. Samantha Wilkinson

   Roche (Switzerland) · Merck Serono Ltd. (United Kingdom)

   Affiliation as printed

   Global Access, F. Hoffmann–La Roche Ltd, Basel, Switzerland

   Now with Merck Serono Ltd, Feltham, United Kingdom

10. Nicolas Scheuer

    Roche (United Kingdom)

    Affiliation as printed

    Roche Products Ltd, Welwyn Garden City, United Kingdom

11. Sreeram V Ramagopalan

    Roche (Switzerland)

    Affiliation as printed

    Global Access, F. Hoffmann–La Roche Ltd, Basel, Switzerland

12. Rolf H. H. Groenwold Aachen Department of Clinical Epidemiology Department of Biomedical Sciences

    Leiden University · Leiden University Medical Center

    Affiliation as printed

    Leiden University Medical Centre, Department of Clinical Epidemiology and Department of Biomedical Sciences, Leiden, the Netherlands

13. Sanjay Popat

    Institute of Cancer Research · Royal Marsden Hospital

    Affiliation as printed

    Royal Marsden Hospital and Institute of Cancer Research, London, United Kingdom

14. Miguel Ángel Hernán

    Harvard University

    Affiliation as printed

    CAUSALab, Departments of Epidemiology and Biostatistics, Harvard T.H. Chan School of Public Health, Boston, Massachusetts