Pith. sign in

REVIEW 3 major objections 7 minor 71 references

Improving the efficiency of infectious disease prevention trials using negative control outcome event times

T0 review · 3 major / 7 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Adjusting for the time to a treatment-unaffected 'negative control' infection can improve the precision of prevention efficacy estimates, even when both the outcome and the control infection are right-censored.

desk verdict A genuinely new and mostly rigorous method for adjusting for a right-censored negative control event time in prevention trials, but the headline 27% variance reduction rests on an underpowered check of a strong causal assumption. read the letter →

arxiv 2608.05261 v1 pith:W6EWOUK7 submitted 2026-08-05 stat.ME

Ethan Ashby , Zhewei Zhang , Tanya P. Garcia , Ting Ye , Holly Janes , Bo Zhang This is my paper · ORCID
classification stat.ME MSC 62N0262P1062D20
keywords negativecontroloutcomeefficientinfluencefunctionsurvivalanalysismultiplerobustnessrandomizedpreventiontrialsrightcensoringHIVprecisiongain
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that the precision of treatment-effect estimates in infectious disease prevention trials can be improved by adjusting for a negative control outcome (NCO) event time: the time to an infection that is unaffected by the intervention but shares exposure mechanisms with the primary infection. The authors show that naively adjusting for the observed, right-censored NCO fails because censoring and post-treatment selection distort the target, and instead derive an efficient influence function for the treatment-arm-specific survivor function when both the primary outcome and the NCO are right-censored. They build a cross-fitted, one-step estimator that is multiply robust and asymptotically efficient, and an EM algorithm for the difficult nuisance estimation. In the HVTN 704/HPTN 085 trial of VRC01 against HIV-1, adjusting for time to rectal gonorrhea or syphilis reduced the estimated variance of prevention efficacy by about 27%, versus about 2.5% for baseline covariate adjustment. If the assumptions hold, the method offers a way to gain power without changing trial design, enrollment, or budget.

What carries the argument

The key object is the observed-data efficient influence function of the treatment-arm-specific survivor function S_a(t0) in a model where both the primary event time Y and the negative control event time N are right-censored. The paper shows identification of S_a(t0) fails when one adjusts for the observed censored N directly, and succeeds via inverse probability weighting of complete-N cases with the NCO-censoring survival function G_N. The EIF is the projection of the full-data EIF onto the observed data tangent space, and it combines an inverse-probability-weighted complete-case term with a martingale integral over the NCO censoring process, using the conditional mean Q of the full-data EIF on the at-risk set. The cross-fitted one-step estimator built on this EIF gains efficiency from the NCO through the conditional dependence of Y on N, and an EM algorithm based on redistribute-to-the-right data augmentation estimates the required conditional survival function S_Y(y|a,x,n).

What would settle it

A falsifying observation would be a prevention trial in which the treatment significantly changes the incidence of the candidate negative control outcome while the estimator is applied; the paper's claim that adjustment is valid requires detecting no such effect, as its own empirical check does for rectal gonorrhea in HVTN 704/HPTN 085.

Watch

Extended reading notes

Core claim

The central claim is that the full time-to-event information in a causally null infection outcome can be leveraged to estimate the treatment-arm-specific survivor function more efficiently, even when both that NCO event time and the primary outcome are right-censored. The paper proves that the observed-data efficient influence function of S_a(t0) is an inverse-probability-weighted complete-case term plus a martingale augmentation that uses partial information from NCO-censored participants, and that the resulting one-step estimator is multiply robust (consistent if any of four nuisance subsets is correct) and asymptotically normal when nuisances converge at product-of-rates conditions. This theoretical result is what licenses the practical claim that adjusting for time-to-STI in HVTN 704/HPTN 085 cuts the estimated variance of prevention efficacy by roughly 27%.

Load-bearing premise

The entire adjustment is valid only if the negative control outcome is completely unaffected by the treatment (N(0)=N(1)); if the intervention changes exposure-related behavior or otherwise affects the control infection, the adjustment can introduce selection bias.

Editorial extensions

If this is right

  • In prevention trials where a valid NCO event time is measured, the estimator can yield substantial precision gains that scale with how strongly the NCO is prognostic for the primary outcome; in simulations the gain reached about 33% efficiency when the association was strong.
  • Because the estimator is multiply robust, it remains consistent under partly incorrect nuisance models, making it suitable for use with flexible machine-learned nuisances under cross-fitting.
  • The method costs little when the NCO turns out uninformative: simulation shows performance essentially matching the unadjusted estimator when the NCO and primary outcome are independent.
  • The case study demonstrates that routinely collected STI surveillance data can serve as negative controls for HIV-1 acquisition, with estimated variance reductions of about 27% from rectal gonorrhea or syphilis adjustment, compared to 2.5% from baseline covariates.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same identification and efficiency logic could be ported to other disease settings where a treatment-invariant infection acts as an exposure marker, such as non-target respiratory pathogens in influenza or SARS-CoV-2 vaccine trials; the paper notes this possibility but does not develop it.
  • The authors suggest that multivariate and recurrent NCO endpoints, or quantitative STI loads, may unlock further efficiency gains; if the per-outcome gain compounds, future trials could see even larger variance reductions than the 27% reported.
  • A practical test before use in a new trial is to check the NCO null empirically, as done in Section 4.1; the method's acceptance in regulatory settings may hinge on demonstrating such checks with adequate power.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 7 minor

Summary. The paper proposes a cross-fitted, one-step estimator of the treatment-arm-specific survivor function of a primary time-to-event outcome in a randomized prevention trial, using a right-censored negative control outcome (NCO) event time as an adjustment variable. The central methodological claims are (i) an efficient influence function for the survivor function when both the primary outcome and the NCO event time are right-censored, (ii) a multiply robust cross-fitted estimator based on that influence function, and (iii) an EM-type nuisance estimation procedure built on a one-way factorization of the joint distribution of the primary and NCO event times. The method is evaluated in simulations and applied to HVTN 704/HPTN 085, where adjusting for time-to-first rectal gonorrhea is reported to reduce the estimated variance of the prevention efficacy estimate by approximately 27%, compared with roughly 2.5% for baseline covariate adjustment.

Significance. If the causal assumptions hold, the paper fills a genuine gap: prevention trials routinely collect time-to-infection data for other pathogens, and these data have not previously been used as right-censored adjustment variables for a right-censored primary outcome. The semiparametric theory is largely standard but carefully assembled, and the proof strategy in Appendix A is coherent; the simulation results are supportive, with low bias and coverage near nominal. The paper is also commendably explicit in Section 5 about the two main limitations of the approach. At the same time, the headline practical claim is conditional on an untestable NCO sharp null, and the implemented estimator does not achieve all of the multiple robustness regimes claimed in the abstract and Theorem 2. These issues are load-bearing for the paper's central claims and need to be addressed before publication.

major comments (3)
  1. [Section 4.1 and Assumption 5] The multiple robustness claim is not realized by the recommended implementation. Under the one-way factorization, the nuisance f_N(n|a,x,ey,delta_y) is derived from f_N(n|a,x) and S_Y(y|a,x,n) by Bayes rule, so f_N and S_Y are variation dependent. Consequently, the four consistency regimes in Condition C3 cannot be chosen independently; for example, regime (pi,G_Y,f_N) in Theorem 2 would require the derived f_N to be correct even when S_Y is misspecified, which is generally impossible under the factorization. The paper acknowledges this in Section 2.5 and states that the estimator remains doubly robust, but the abstract, the introduction, and Theorem 2 still claim multiple robustness. Please restate the property as applying to the influence-function structure with arbitrary nuisance estimators, and characterize precisely which of the four regimes survive under the one-way factorization used in the implemented estimator.
  2. [Section 2.2, Assumption 4 and Proposition 1] The practical claim of a 27% variance reduction depends on the sharp NCO null, Assumption 5, which cannot be established by the reported empirical check. The comparison of pooled VRC01 versus placebo RGC incidence is a failure-to-reject with limited power; it cannot rule out modest risk-compensation effects or indirect effects on STI acquisition. If Assumption 5 fails, the adjusted estimator targets E[P(Y(a)>t | X, N(a))] rather than S_a(t), and the shift in the point estimate seen in Table 2 (logRR moving from -0.309 unadjusted to -0.248 RGC-adjusted) would be bias rather than a pure precision gain. The manuscript should include a sensitivity analysis or quantitative bounds on the bias as a function of a plausible NCO treatment effect, so that the variance reduction reported in the abstract can be interpreted alongside the bias risk.
  3. [Section 5, Discussion] The independent censoring assumption is stated as (C_Y, C_N) perpendicular (Y, N) given (A, X). This is stronger than what the identification argument actually requires, and it excludes plausible mechanisms in the application, such as NCO censoring driven by the primary event or dropout driven by evolving risk. The derivation in Appendix A.2 uses Assumption 4 to establish C_N perpendicular N given (A, X, e_Y, Delta_Y), which is the condition actually needed for the IPW identification, together with conditions ensuring that G_Y does not depend on N. Please state the weaker coarsening-at-random conditions directly and discuss their plausibility in HVTN 704/HPTN 085. As written, the identification result in Proposition 1 may appear inapplicable to the case study if NCO follow-up is terminated at HIV infection or if dropout is related to risk behavior.
minor comments (7)
  1. [Section 2.3, Eq. (4)] The text says sample sizes are n = {2400, 4000}, but Table 1 reports rows for n = 2000 and n = 4000; please align the text and the table.
  2. [Section 2.2, Proposition 1] In Eq. (4), the first term uses the symbol delta_n, which appears to be a typo for Delta_N; the notation should be consistent with the surrounding text.
  3. [Section 2.5, Assumption 6] The statement of Proposition 1 uses G_0N(e_N | A, X, e_Y, Delta_Y), while the proof in Appendix A.1.2 writes G_0(e_N | A, X); please align the conditioning sets in the proposition and its proof.
  4. [Section 2.4, Theorem 3] The phrase 'U_N and U_Y may not be unequal' should be corrected to 'may not be equal' or 'may be unequal'.
  5. [References] In the paragraph before Theorem 3, 'Condition 3' should be 'Condition C4'.
  6. [Section 4.1] References [25] and [66] appear to cite the same Etievant et al. paper with different years and page ranges; please consolidate to a single reference.
  7. The text states that results were similar for syphilis, but no corresponding figure or table is presented for syphilis in that section; please add the supporting display or refer to the relevant results in Table 2.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the estimator and its efficiency claim follow from standard semiparametric projection theory, and the reported 27% variance reduction is an empirical estimate, not a fitted prediction.

full rationale

The derivation chain is self-contained. The target parameter S_a(t) is defined causally via potential outcomes, and identification (Proposition 1) follows from standard IPW/G-computation arguments under Assumptions 1-5, not from the estimating procedure itself. The full-data EIF is imported from Westling et al. (2023), an external source, and the observed-data EIF is obtained by the standard coarsening/projection argument from Tsiatis (2006); no step defines the estimand in terms of the fitted EIF. Multiple robustness and asymptotic linearity are proven from the EIF structure with explicit rate conditions (C1-C4) and do not rely on the application data. In the simulation, the oracle estimator uses nuisance values computed from an independent external dataset, while the cross-fitted estimator uses standard machine learning; neither is tuned to produce a target efficiency gain. In the application, the variance reduction is reported as an estimated quantity, not as a prediction derived from any fitted parameter. The only self-citation ([24]) is motivational background about respiratory negative control outcomes and is not load-bearing. The NCO sharp null (Assumption 5) is a causal assumption with a direct empirical check; although the check and the precision-gain estimate use the same trial, that is a limitation of the application rather than a circular step in the derivation.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The method introduces no new physical entities. Its validity rests on causal assumptions about treatment, censoring, and the negative control outcome, plus algorithmic hyperparameters. The absence of fitted physical parameters is a strength: the reported variance reduction is not the result of tuning a constant to make the data come out a certain way.

free parameters (3)
  • NCO event type choice = rectal gonorrhea; syphilis
    In the case study the NCO events were selected based on prior literature and confirmed to be prognostic in the same trial, a hand-chosen modeling decision that affects the reported variance reduction.
  • EM iteration cap = 5 iterations
    The EM algorithm is stopped after 5 iterations or when relative weight change is below 1e-4; chosen by hand, results may depend on the cap.
  • Cross-fitting folds K = 6 (simulation), 8 (case study)
    The number of folds is chosen by the authors; standard practice, but a tuning choice.
assumptions (6)
  • standard math Consistency and SUTVA: Y = Y(A), N = N(A), no interference, no multiple versions of treatment (Assumption 1)
    Standard causal inference assumptions, usually plausible in individually randomized trials.
  • standard math Positivity: 0 < P(A=a|X=x) for a=0,1 (Assumption 2)
    Satisfied by design in a randomized trial.
  • domain assumption Strong ignorability: A independent of potential outcomes and censoring times given X (Assumption 3)
    Guaranteed by randomization.
  • domain assumption Independent censoring: (C^Y, C^N) independent of (Y,N) given (A,X) (Assumption 4)
    Requires dropout and administrative censoring to be unconfounded conditional only on baseline variables; this is stronger than typical survival assumptions and is load-bearing for IPW identification.
  • domain assumption NCO sharp null: N(0)=N(1) (Assumption 5)
    The NCO must be unaffected by treatment; falsifiable but only weakly testable in the case study.
  • ad hoc to paper Approximate U-comparability: NPSEM (Assumption 6)
    A structural model introduced to justify why N can be prognostic for Y; not needed for identification but needed for efficiency gains.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improving the efficiency of infectious disease prevention trials using negative control outcome event times." pith.science (2026). https://pith.science/paper/W6EWOUK7

@misc{pith2026260805261,
  author       = {Pith},
  title        = {Pith review of: Improving the efficiency of infectious disease prevention trials using negative control outcome event times},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W6EWOUK7}},
  note         = {Machine review of arXiv:2608.05261}
}
read the original abstract

Baseline covariate adjustment can enhance the efficiency of randomized trials by improving precision of treatment effect estimates. However, the precision gain depends on how strongly the baseline covariates are prognostic for the primary outcome. In randomized trials of infectious disease prevention interventions (e.g., vaccines or passively administered antibodies), an individual's exposure to the pathogen is a leading prognostic factor but is rarely measurable at baseline. Hence, conventional covariate adjustment offers limited precision gain in prevention trials. We propose adjusting for a negative control outcome (NCO) event time, which is causally unaffected by the intervention but shares overlapping exposure mechanisms with the primary outcome. We formalize assumptions under which adjustment for the NCO event time is valid, and show that right-censoring of the NCO event time further complicates adjustment. We derive the efficient influence function for the treatment-arm-specific survivor function of the primary outcome when both the primary outcome and the NCO event time are right-censored, and use it to construct a cross-fitted, one-step estimator that is multiply robust to nuisance misspecification and asymptotically efficient when the nuisances are estimated accurately. In numerical experiments, our estimator compares comparably to benchmarks when the NCO event time is uninformative, and gains precision as the NCO event time is more prognostic for the primary outcome. We apply our method to HVTN 704/HPTN 085, a randomized, double-blinded trial of VRC01, a broadly neutralizing antibody against HIV-1. Adjusting for the time to a bacterial sexually transmitted infection --- a negative control outcome for HIV-1 acquisition --- reduced the estimated variance of the prevention efficacy estimate by approximately 27%, compared to roughly 2.5% for baseline covariate adjustment.

Figures

Figures reproduced from arXiv: 2608.05261 by the authors.

Figure 1
Figure 1. Graphical causal model illustrating assumptions underlying approach. [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. A visual depiction of the EM algorithm used to estimate [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Top: cumulative incidence functions for Rectal Gonorrhea (RGC) over 72 weeks as a function [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

71 extracted references · 68 canonical work pages

  1. [1]

    Guideline on adjustment for baseline covariates in clinical tri- als, March 2015

    European Medicines Agency. Guideline on adjustment for baseline covariates in clinical tri- als, March 2015. URLhttps://www.ema.europa.eu/en/documents/scientific-guideline/ guideline-adjustment-baseline-covariates-clinical-trials_en.pdf

    work page 2015

  2. [2]

    Food and Drug Administration

    U.S. Food and Drug Administration. Adjusting for covariates in randomized clinical trials for drugs and biological products: Draft guidance for industry, May 2021. URLhttps://www.fda.gov/ media/148910/download. Draft guidance

    work page 2021

  3. [3]

    Li Yang and Anastasios A. Tsiatis. Efficiency study of estimators for a treatment effect in a pretest- posttest trial.The American Statistician, 55:314–321, 2001

    work page 2001

  4. [4]

    Tsiatis, Marie Davidian, Min Zhang, and Xiaomin Lu

    Anastasios A. Tsiatis, Marie Davidian, Min Zhang, and Xiaomin Lu. Covariate adjustment for two-sample treatment comparisons in randomized clinical trials: A principled yet flexible approach. Statistics in medicine, 27(23):4658–4677, October 2008. ISSN 0277-6715. doi: 10.1002/sim.3113. URLhttps://www.ncbi.nlm.nih.gov/pmc/articles/PMC2562926/

    doi:10.1002/sim.3113 2008

  5. [5]

    Tsiatis, and Marie Davidian

    Min Zhang, Anastasios A. Tsiatis, and Marie Davidian. Improving Efficiency of In- ferences in Randomized Clinical Trials Using Auxiliary Covariates.Biometrics, 64 (3):707–715, 2008. ISSN 1541-0420. doi: 10.1111/j.1541-0420.2007.00976.x. URL https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1541-0420.2007.00976.x. eprint: https://onlinelibrary.wiley.com/do...

    arXiv 2008

  6. [6]

    Agnostic notes on regression adjustments to experimental data: Reexamining freed- man’s critique.Annals of Applied Statistics, 7:295–318, 2013

    Winston Lin. Agnostic notes on regression adjustments to experimental data: Reexamining freed- man’s critique.Annals of Applied Statistics, 7:295–318, 2013

    work page 2013

  7. [7]

    Toward better practice of covariate adjustment in analyzing randomized clinical trials.Journal of the American Statistical Association, 2022

    Ting Ye, Jun Shao, Yanyao Yi, and Qingquan Zhao. Toward better practice of covariate adjustment in analyzing randomized clinical trials.Journal of the American Statistical Association, 2022. 15

    work page 2022

  8. [8]

    Statistical principles for clinical trials (e9)

    International Council for Harmonisation (ICH). Statistical principles for clinical trials (e9). ICH Harmonised Tripartite Guideline E9, International Council for Harmonisation of Technical Require- ments for Pharmaceuticals for Human Use, 1998

    work page 1998

Show all 71 references
  1. [9]

    Covariate adjustment in randomized controlled trials: General concepts and practical considerations.Clinical Trials, 2024

    Kelly Van Lancker, Frank Bretz, and Oliver Dukes. Covariate adjustment in randomized controlled trials: General concepts and practical considerations.Clinical Trials, 2024

    2024

  2. [10]

    Farrington, M.N

    C.P. Farrington, M.N. Kanaan, and N.J. Gay. Estimation of the basic reproduction number for infectious diseases from age-stratified serological survey data.Applied Statistics, 50:251–292, 2001

    2001

  3. [11]

    Correlated infections: quantifying individual heterogeneity in the spread of infectious diseases.Am

    C Paddy Farrington, Heather J Whitaker, Steffen Unkel, and Richard Pebody. Correlated infections: quantifying individual heterogeneity in the spread of infectious diseases.Am. J. Epidemiol., 177(5): 474–486, March 2013

    2013

  4. [12]

    Schroder, Michael P

    Kerstin E.E. Schroder, Michael P. Carey, and Peter A. Vanable. Methodological challenges in research on sexual risk behavior: Ii. accuracy of self-reports.Annals of Behavioral Medicine, 26: 104—-123, 2003

    2003

  5. [13]

    Paddy Farrington and Heather J Whitaker

    C. Paddy Farrington and Heather J Whitaker. Contact Surface Models for Infectious Diseases: Estimation From Serologic Survey Data.Journal of the American Statistical Association, 100(470): 370–379, June 2005. ISSN 0162-1459, 1537-274X. doi: 10.1198/016214504000001754. URLhttp:...

    2005 doi

  6. [14]

    Paddy Farrington, Heather J

    C. Paddy Farrington, Heather J. Whitaker, Steffen Unkel, and Richard Pebody. Correlated Infec- tions: Quantifying Individual Heterogeneity in the Spread of Infectious Diseases.American Journal of Epidemiology, 177(5):474–486, March 2013. ISSN 1476-6256, 0002-9262. doi: 10.1093...

    2013 doi

  7. [15]

    Effect of HSV-2 infection on subsequent HIV acquisition: an updated systematic review and meta-analysis.Lancet Infect

    Katharine J Looker, Jocelyn A R Elmes, Sami L Gottlieb, Joshua T Schiffer, Peter Vickerman, Katherine M E Turner, and Marie-Claude Boily. Effect of HSV-2 infection on subsequent HIV acquisition: an updated systematic review and meta-analysis.Lancet Infect. Dis., 17(12):1303– 1...

    2017

  8. [16]

    Strong associations between national prevalence of various STIs suggests sexual network connectivity is a common underpinning risk factor.BMC Infect

    Chris Kenyon. Strong associations between national prevalence of various STIs suggests sexual network connectivity is a common underpinning risk factor.BMC Infect. Dis., 17(1):682, October 2017

    2017

  9. [17]

    Ross A Baiers, Daniel T Ryan, Antonia Clifford, Erik Munson, Richard D’Aquila, Michael E New- comb, and Brian Mustanski. Asymptomatic rectal bacterial pathogens show large prospective re- lationships with HIV incidence in a cohort of young sexual and gender minorities: Implica...

    2024

  10. [18]

    Trends in HIV prevalence by self-report among MSM diagnosed and reported with gonorrhea in six united states jurisdictions from 2010 to 2019.AIDS, 35(15):2523–2530, December 2021

    Mark Richard Stenger, Preeti Pathela, Christina Schumacher, Nicole Burghardt, Rachel Amiya, Robbie Madera, Trang Q Nguyen, and Elizabeth Torrone. Trends in HIV prevalence by self-report among MSM diagnosed and reported with gonorrhea in six united states jurisdictions from 201...

    2010

  11. [19]

    Risk of HIV diagnosis following bacterial sexually transmitted infections in tennessee, 2013-2017.Sex

    Heather N Grome, Peter F Rebeiro, Meredith Brantley, Dyanne Herrera-Vasquez, Samantha A Mathieson, and April C Pettit. Risk of HIV diagnosis following bacterial sexually transmitted infections in tennessee, 2013-2017.Sex. Transm. Dis., 48(11):873–880, November 2021

    2013

  12. [20]

    Charu Mullick and Jeffrey Murray. Correlations between human immunodeficiency virus (HIV) infection and rectal gonorrhea incidence in men who have sex with men: Implications for future HIV preexposure prophylaxis trials.J. Infect. Dis., 221(2):214–217, January 2020

    2020

  13. [21]

    Rectal gonorrhoea as an independent risk factor for HIV infection in a cohort of homosexual men.Genitourin

    K J Craib, D R Meddings, S A Strathdee, R S Hogg, J S Montaner, M V O’Shaughnessy, and M T Schechter. Rectal gonorrhoea as an independent risk factor for HIV infection in a cohort of homosexual men.Genitourin. Med., 71(3):150–154, June 1995

    1995

  14. [22]

    Effect of syphilis infection on HIV acquisition: a systematic review and meta-analysis.Sex

    Meng Yin Wu, Hui Zi Gong, Kui Ru Hu, He-Yi Zheng, Xia Wan, and Jun Li. Effect of syphilis infection on HIV acquisition: a systematic review and meta-analysis.Sex. Transm. Infect., 97(7): 525–533, November 2021. 16

    2021

  15. [23]

    Rates of co-infection between SARS-CoV-2 and other respiratory pathogens.JAMA, 323(20):2085–2086, May 2020

    David Kim, James Quinn, Benjamin Pinsky, Nigam H Shah, and Ian Brown. Rates of co-infection between SARS-CoV-2 and other respiratory pathogens.JAMA, 323(20):2085–2086, May 2020

    2020

  16. [24]

    Gilbert, Honghong Zhou, Xiaowei Wang, Bethany Girard, James B

    Ethan Ashby, Holly Janes, Dean Follmann, Peter B. Gilbert, Honghong Zhou, Xiaowei Wang, Bethany Girard, James B. Kublin, Lawrence Corey, Kathleen M. Neuzil, Lindsey R. Baden, Hana M. El Sahly, and Bo Zhang. Validating and leveraging non-sars-cov-2 respiratory infection as a ne...

    2025 doi

  17. [25]

    Sampson, and Mitchell H

    Lola Etievant, Joshua N. Sampson, and Mitchell H. Gail. Increasing efficiency and reducing bias when assessing HPV vaccination efficacy by using nontargeted HPV strains.Biometrics, pages 1– 12, 2022. doi: 10.1111/biom.13663. URLhttps://onlinelibrary.wiley.com/doi/abs/10.1111/ ...

    2022 doi

  18. [26]

    Chapman and Hall/CRC, 2020

    Miguel A Hern´ an and James M Robins.Causal Inference: What If. Chapman and Hall/CRC, 2020

    2020

  19. [27]

    Rosenbaum

    Paul R. Rosenbaum. The consequences of adjustment for a concomitant variable that has been affected by the treatment.Journal of the Royal Statistical Society: Series B, 147, 1984

    1984

  20. [28]

    Negative Controls: A Tool for Detecting Confounding and Bias in Observational Studies.Epidemiology, 21(3):383–388, May 2010

    Marc Lipsitch, Eric Tchetgen Tchetgen, and Ted Cohen. Negative Controls: A Tool for Detecting Confounding and Bias in Observational Studies.Epidemiology, 21(3):383–388, May 2010. ISSN 1044-3983. doi: 10.1097/EDE.0b013e3181d61eeb. URLhttps://www.ncbi.nlm.nih.gov/pmc/ articles/P...

    2010 doi

  21. [29]

    A Selective Review of Negative Control Meth- ods in Epidemiology.Current epidemiology reports, 7(4):190–202, December 2020

    Xu Shi, Wang Miao, and Eric Tchetgen Tchetgen. A Selective Review of Negative Control Meth- ods in Epidemiology.Current epidemiology reports, 7(4):190–202, December 2020. ISSN 2196-

    2020

  22. [30]

    Sur les applications de la theorie des probabilites aux experiences agricoles: Essai des principes.Statistical Science, 5, 1921

    Jerzy Neyman. Sur les applications de la theorie des probabilites aux experiences agricoles: Essai des principes.Statistical Science, 5, 1921

    1921

  23. [31]

    Causal inference using potential outcomes.Journal of the American Statistical Association, 100, 2005

    Donald B Rubin. Causal inference using potential outcomes.Journal of the American Statistical Association, 100, 2005

    2005

  24. [32]

    Bayesian inference for causal effects: The role of randomization.The Annals of Statistics, 6, 1978

    Donald B Rubin. Bayesian inference for causal effects: The role of randomization.The Annals of Statistics, 6, 1978

    1978

  25. [33]

    Gilbert, and Marco Carone

    Ted Westling, Alex Luedtke, Peter B. Gilbert, and Marco Carone. Inference for treatment-specific survival curves using machine learning.Journal of the American Statistical Association, 2023

    2023

  26. [34]

    Springer Series in Statistics

    Anastasios A Tsiatis.Semiparametric Theory and Missing Data. Springer Series in Statistics. Springer, New York, NY, 2006 edition, June 2006

    2006

  27. [35]

    Bickel, Chris A.J

    Peter J. Bickel, Chris A.J. Klaassen, Ya’acov Ritov, and Jon A. Wellner.Efficient and adaptive estimation for semiparametric models. Johns Hopkins University Press Baltimore, 1993

    1993

  28. [36]

    On adaptive estimation.Ann

    P J Bickel. On adaptive estimation.Ann. Stat., 10(3):647–671, September 1982

    1982

  29. [37]

    Double/debiased/Neyman machine learning of treatment effects.Am

    Victor Chernozhukov, Denis Chetverikov, Mert Demirer, Esther Duflo, Christian Hansen, and Whit- ney Newey. Double/debiased/Neyman machine learning of treatment effects.Am. Econ. Rev., 107 (5):261–265, May 2017

    2017

  30. [38]

    Double/debiased machine learning for treatment and structural parameters.The Econometrics Journal, 21, 2018

    Victor Chernozhukov, Denis Chetverikov, Mert Demirer, Esther Duflo, Christian Hansen, Whit- ney Newey, and James Robins. Double/debiased machine learning for treatment and structural parameters.The Econometrics Journal, 21, 2018

    2018

  31. [39]

    Super learner.Statistical Applications in Genetics and Molecular Biology, 6:number 1, 2007

    Mark J van der Laan, Eric C Polley, and Alan E Hubbard. Super learner.Statistical Applications in Genetics and Molecular Biology, 6:number 1, 2007

    2007

  32. [40]

    James Press

    Barry C Arnold and S. James Press. Compatible conditional distributions.J. Am. Stat. Assoc., 84 (405):152, March 1989

    1989

  33. [41]

    Compatibility of conditionally specified models.Stat

    Hua Yun Chen. Compatibility of conditionally specified models.Stat. Probab. Lett., 80(7-8):670–677, April 2010. 17

    2010

  34. [42]

    he two sample problem with censored data

    Bradley Efron. he two sample problem with censored data. InProceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, volume 4, pages 831–853. University of California Press, Berkeley, CA, 1967

    1967

  35. [43]

    The calculation of posterior distributions by data aug- mentation.J

    Martin A Tanner and Wing Hung Wong. The calculation of posterior distributions by data aug- mentation.J. Am. Stat. Assoc., 82(398):528, June 1987

    1987

  36. [44]

    Survival estimation and testing via multiple imputation.Statistics and Probability Letters, 58:221–232, 2002

    Jeremy MG Taylor, Susan Murray, and Chiu-Hsieh Hsu. Survival estimation and testing via multiple imputation.Statistics and Probability Letters, 58:221–232, 2002

    2002

  37. [45]

    Gilbert, Michal Juraska, David C

    Lawrence Corey, Peter B. Gilbert, Michal Juraska, David C. Montefiori, Lynn Morris, Shelly T. Karuna, Srilatha Edupuganti, Nyaradzo M. Mgodi, Allan C. deCamp, Erika Rudnicki, Yunda Huang, Pedro Gonzales, Robinson Cabello, Catherine Orrell, Javier R. Lama, Fatima Laher, Er- ica...

    2021

  38. [46]

    van der Laan, E

    Mark J. van der Laan, E. Polley, and A. Hubbard. Super learner.Statistical Applications in Genetics and Molecular Biology, 6, 2007

    2007

  39. [47]

    Karuna, Philip Andrew, Erika Rudnicki, Nidhi Kochar, Allan deCamp, Robert De La Grecca, Maija Anderson, Carissa Karg, India Tindale, Eliz- abeth Greene, Gail B

    Srilatha Edupuganti, Nyaradzo Mgodi, Shelly T. Karuna, Philip Andrew, Erika Rudnicki, Nidhi Kochar, Allan deCamp, Robert De La Grecca, Maija Anderson, Carissa Karg, India Tindale, Eliz- abeth Greene, Gail B. Broder, Jonathan Lucas, John Hural, Jorge A. Gallardo-Cartagena, Pe- ...

    2021

  40. [48]

    An application of propensity score weighting to quantify the causal effect of rectal sexually transmitted infections on incident HIV among men who have sex with men.BMC Med

    Adam S Vaughan, Colleen F Kelley, Nicole Luisi, Carlos del Rio, Patrick S Sullivan, and Eli S Rosenberg. An application of propensity score weighting to quantify the causal effect of rectal sexually transmitted infections on incident HIV among men who have sex with men.BMC Med...

    2015

  41. [49]

    New hu- man immunodeficiency virus diagnosis independently associated with rectal gonorrhea and chlamy- dia in men who have sex with men.Sex

    Lindley A Barbee, Christine M Khosropour, Julia C Dombrowksi, and Matthew R Golden. New hu- man immunodeficiency virus diagnosis independently associated with rectal gonorrhea and chlamy- dia in men who have sex with men.Sex. Transm. Dis., 44(7):385–389, July 2017

    2017

  42. [50]

    Maximum likelihood estimation in generalized linear models with multiple covariates subject to detection limits.Stat

    Ryan C May, Joseph G Ibrahim, and Haitao Chu. Maximum likelihood estimation in generalized linear models with multiple covariates subject to detection limits.Stat. Med., 30(20):2551–2561, September 2011

    2011

  43. [51]

    Multiple imputation forM-regression with censored co- variates.J

    Huixia Judy Wang and Xingdong Feng. Multiple imputation forM-regression with censored co- variates.J. Am. Stat. Assoc., 107(497):194–204, March 2012

    2012

  44. [52]

    Statistical methods for generalized linear models with covariates subject to detection limits.Stat

    Paul W Bernhardt, Huixia J Wang, and Daowen Zhang. Statistical methods for generalized linear models with covariates subject to detection limits.Stat. Biosci., 7(1):68–89, May 2015

    2015

  45. [53]

    Multiple imputation of a randomly censored covariate improves logistic regression analysis.J

    Folefac D Atem, Jing Qian, Jacqueline E Maye, Keith A Johnson, and Rebecca A Betensky. Multiple imputation of a randomly censored covariate improves logistic regression analysis.J. Appl. Stat., 43(15):2886–2896, March 2016

    2016

  46. [54]

    Atem, Jing Qian, Jacqueline E

    Folefac D. Atem, Jing Qian, Jacqueline E. Maye, Keith A. Johnson, and Rebecca A. Betensky. Linear regression with a randomly censored covariate: application to an alzheimer’s study.Journal of the Royal Statistical Society. Series C (Applied Statistics), 66, 2017. 18

    2017

  47. [55]

    A semiparametric imputation approach for regression with censored covariate with application to an AMD progression study.Stat

    Ying Ding, Shengchun Kong, Shan Kang, and Wei Chen. A semiparametric imputation approach for regression with censored covariate with application to an AMD progression study.Stat. Med., 37(23):3293–3308, October 2018

    2018

  48. [56]

    Atem, Roland A

    Folefac D. Atem, Roland A. Matsouaka, and Vincent E. Zimmern. Cox regression model with randomly censored covariates.Biometrical Journal, 61, 2019

    2019

  49. [57]

    Matsouaka and Folefac D

    Roland A. Matsouaka and Folefac D. Atem. Regression with a right-censored predictor using inverse probability weighting methods.Statistics in Medicine, 39, 2020

    2020

  50. [58]

    Semiparametric analysis of a generalized linear model with multiple covariates subject to detection limits.Stat

    Ling-Wan Chen, Jason P Fine, Eric Bair, Victor S Ritter, Thomas F McElrath, David E Cantonwine, John D Meeker, Kelly K Ferguson, and Shanshan Zhao. Semiparametric analysis of a generalized linear model with multiple covariates subject to detection limits.Stat. Med., 41(24):479...

    2022

  51. [59]

    Understanding the implications of a complete case analysis for regression models with a right-censored covariate.Am

    Marissa C Ashner and Tanya P Garcia. Understanding the implications of a complete case analysis for regression models with a right-censored covariate.Am. Stat., 78(3):335–344, 2024

    2024

  52. [60]

    Making sense of censored covariates: Statistical methods for studies of huntington’s disease.Annu

    Sarah C Lotspeich, Marissa C Ashner, Jesus E Vazquez, Brian D Richardson, Kyle F Grosser, Benjamin E Bodek, and Tanya P Garcia. Making sense of censored covariates: Statistical methods for studies of huntington’s disease.Annu. Rev. Stat. Appl., 11(1):255–277, April 2024

    2024

  53. [61]

    Estimation of regression coefficients when some regressors are not always observed.J

    James M Robins, Andrea Rotnitzky, and Lue Ping Zhao. Estimation of regression coefficients when some regressors are not always observed.J. Am. Stat. Assoc., 89(427):846–866, September 1994

    1994

  54. [62]

    Robust and efficient estimation in the presence of a randomly censored covariate.arXiv, 2024

    Seong-Ho Lee, Brian D Richardson, Yanyuan Ma, Karen S Marder, and Tanya P Garcia. Robust and efficient estimation in the presence of a randomly censored covariate.arXiv, 2024

    2024

  55. [63]

    Robust estimation under outcome dependent right censoring in huntington disease: Estimators for low and high censoring rates.arXiv, November 2025

    Jesus E Vazquez, Yanyuan Ma, Karen Marder, and Tanya P Garcia. Robust estimation under outcome dependent right censoring in huntington disease: Estimators for low and high censoring rates.arXiv, November 2025

    2025

  56. [64]

    Super doubly robust and efficient estimator for informative covariate censoring.arXiv, November 2025

    Zhewei Zhang, Yanyuan Ma, Karen Marder, and Tanya P Garcia. Super doubly robust and efficient estimator for informative covariate censoring.arXiv, November 2025

    2025

  57. [65]

    How estimators change when adapted from the missing covariate problem to the right-censored covariate problem.International Statistical Review, 2026

    Jesus E Vazquez, Marissa C Ashner, Yanyuan Ma, Karen Marder, and Tanya P Garcia. How estimators change when adapted from the missing covariate problem to the right-censored covariate problem.International Statistical Review, 2026. doi: https://doi.org/10.1111/insr.70043

    2026 doi

  58. [66]

    Sampson, and Mitchell H

    Lola Etievant, Joshua M. Sampson, and Mitchell H. Gail. Increasing efficiency and reducing bias when assessing hpv vaccination efficacy by using nontargeted hpv strains.Biometrics, 79(2):1534— -1545, 2023

    2023

  59. [67]

    The role of sexually transmitted infections in HIV-1 progression: A comprehensive review of the literature.J

    Helen M Chun, Robert J Carpenter, Grace E Macalino, and Nancy F Crum-Cianflone. The role of sexually transmitted infections in HIV-1 progression: A comprehensive review of the literature.J. Sex. Transm. Dis., 2013:176459, June 2013

    2013

  60. [68]

    Sexually transmitted infections and HIV in the era of antiretroviral treatment and prevention: the biologic basis for epidemiologic synergy.J

    Myron S Cohen, Olivia D Council, and Jane S Chen. Sexually transmitted infections and HIV in the era of antiretroviral treatment and prevention: the biologic basis for epidemiologic synergy.J. Int. AIDS Soc., 22 Suppl 6(S6):e25355, August 2019

    2019

  61. [69]

    Principal stratification in causal inference.Biometrics, 58:21–29, 2002

    CE Frangakis and DB Rubin. Principal stratification in causal inference.Biometrics, 58:21–29, 2002

    2002

  62. [70]

    I(N≤C) G0( eN|A, X) I(A=a 0) π0(a0 |X) P0(Y > t|A=a 0, X,eN) # 20 WhereE 0[·] refers to marginalizing over the distribution of (A, X, N, C). E0

    Aad W Van Der Vaart and Jon A Wellner. Weak convergence. InWeak convergence and empirical processes: with applications to statistics, pages 16–28. Springer, 1996. A Key proofs A.1 Proposition 1 As a reminder, Proposition 1 consists of two parts (a) applying G-computation with ...

    1984

  63. [2995]

    URLhttps://www.ncbi.nlm.nih.gov/pmc/articles/ PMC8118596/

    doi: 10.1007/s40471-020-00243-4. URLhttps://www.ncbi.nlm.nih.gov/pmc/articles/ PMC8118596/

    doi

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.