Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 27 min ago
Peer-reviewedCC BYTrials

Computing assumption-lean bounds for per-protocol effects

  • Paul N. Zivich1
  • Jessie K. Edwards1
  • Valerie A. Lucas1
  • Bonnie E. Shook-Sa2
  • M. Alan Brookhart3
  • Stephen R. Cole1
  1. 1Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina at Chapel Hill
  2. 2Department of Biostatistics, Gillings School of Global Public Health, University of North Carolina at Chapel Hill
  3. 3Department of Population Health Science, Duke University
Read full textDOICite

Abstract

Background A common approach used to estimate per-protocol effects, or the effect of perfect adherence to a defined protocol on a specified outcome, is to artificially censor individuals when they deviate from the protocol. As adherence is not under control of the investigator, selection bias is a threat to the validity of per-protocol analyses. Methods Here, we describe nonparametric bounds for the per-protocol effect. Importantly, these bounds describe the range of all possible per-protocol effects concordant with the observed data under the sole additional assumption—beyond those of a standard intent-to-treat analysis—that adherence was accurately measured. Further, these bounds naturally follow from an intuitive understanding of the “best” and “worst” cases among trial participants. Finally, the per-protocol bounds are straightforward to compute with standard statistical analysis software. Results We describe and illustrate these bounds with two open-source data sets, one in the context of a binary outcome and the other with a survival outcome. These examples are accompanied by open-source code in several different software programs to aid implementation. Conclusions Per-protocol bounds offer a simple and intuitive assessment of the per-protocol effect under relatively weak assumptions and offer insight into the strength of the assumptions like no selection bias.

Related

  1. Assumption-Lean Inference for Spectral Differential Network Analysis of High-Dimensional Time Series
    Michael Hellstern, Byol Kim, Ali Shojaie
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  2. Covariate Adjustment for Time-Restricted Marginal Win Statistics in Randomized Trials: Assumption-Lean Inference and Guaranteed Efficiency Gain
    Baoshan Zhang, Laine E. Thomas, Yuan Wu +2
    arXivPreprintarXiv licenceFull textStatisticsarXiv
  3. Assumption-lean logistic regression with missing covariates
    Jyotishka Ray Choudhury, Kabir Aladin Verchand, Richard J. Samworth +1
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  4. Sparse departures from independence in two-way tables: a heteroscedasticity profile and detection boundary, an adaptive higher-criticism gate, and an assumption-lean exact anchor
    William J. Dwyer
    arXivPreprintCC BY-NC-ND 4.0Full textStatisticsarXiv
  5. Toward a Lean Formalization of Analog Computing with Microwaves
    Matteo Nerini, Xuekang Liu, Bruno Clerckx
    arXivPreprintCC BY 4.0Full textGeneral mathematicsApplied physics & instrumentationarXiv

Cite

BibTeXRIS
BibTeX
@article{zivich2026computing,
  title = {{Computing assumption-lean bounds for per-protocol effects}},
  author = {Paul N. Zivich and Jessie K. Edwards and Valerie A. Lucas and Bonnie E. Shook-Sa and M. Alan Brookhart and Stephen R. Cole},
  journal = {Trials},
  year = {2026},
  month = jul,
  doi = {10.1186/s13063-026-09927-6},
  url = {https://doi.org/10.1186/s13063-026-09927-6},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Computing assumption-lean bounds for per-protocol effects
AU  - Paul N. Zivich
AU  - Jessie K. Edwards
AU  - Valerie A. Lucas
AU  - Bonnie E. Shook-Sa
AU  - M. Alan Brookhart
AU  - Stephen R. Cole
PY  - 2026
DA  - 2026/07/01
JO  - Trials
DO  - 10.1186/s13063-026-09927-6
UR  - https://doi.org/10.1186/s13063-026-09927-6
AB  - Background A common approach used to estimate per-protocol effects, or the effect of perfect adherence to a defined protocol on a specified outcome, is to artificially censor individuals when they deviate from the protocol. As adherence is not under control of the investigator, selection bias is a threat to the validity of per-protocol analyses. Methods Here, we describe nonparametric bounds for the per-protocol effect. Importantly, these bounds describe the range of all possible per-protocol effects concordant with the observed data under the sole additional assumption—beyond those of a standard intent-to-treat analysis—that adherence was accurately measured. Further, these bounds naturally follow from an intuitive understanding of the “best” and “worst” cases among trial participants. Finally, the per-protocol bounds are straightforward to compute with standard statistical analysis software. Results We describe and illustrate these bounds with two open-source data sets, one in the context of a binary outcome and the other with a survival outcome. These examples are accompanied by open-source code in several different software programs to aid implementation. Conclusions Per-protocol bounds offer a simple and intuitive assessment of the per-protocol effect under relatively weak assumptions and offer insight into the strength of the assumptions like no selection bias.
SN  - 1745-6215
ER  -