Statistical methods for incomplete and distributed data

In many cohort studies, a covariate is a time-to-event variable that is only partly observed. In Huntington disease studies, for example, age at clinical diagnosis is unknown for participants who have not yet been diagnosed at their last visit. I develop robust and efficient estimators for regression models with right-censored covariates, study the links between right-censored and missing covariates, and apply the estimators to Huntington disease progression data.

• How Estimators Change When Adapted from the Missing Covariate Problem to the Right-Censored Covariate Problem
  Vazquez, J.E.; Ashner, M.C.; Ma, Y.; Marder, K.; Garcia, T.P.
  Published International Statistical Review, 2026 DOI
  Abstract

  While right-censored time-to-event outcomes have been studied for decades, handling time-to-event covariates, also known as right-censored covariates, is now of growing interest. So far, the literature has treated right-censored covariates as distinct from missing covariates, overlooking the potential applicability of estimators to both scenarios. We bridge this gap by establishing connections between right-censored and missing covariates under various assumptions about censoring and missingness, allowing us to identify parallels and differences to determine when estimators can be used in both contexts. These connections reveal adaptations to five estimators for right-censored covariates in the unexplored area of informative covariate right-censoring and to formulate a new estimator for this setting, where the event time depends on the censoring time. We establish the asymptotic properties of the six estimators, evaluate their robustness under incorrect distributional assumptions, and establish their comparative efficiency. We conducted a simulation study to confirm our theoretical results, and then applied all estimators to a Huntington disease observational study to analyze cognitive impairments as a function of time to clinical diagnosis.

• Making sense of censored covariates: statistical methods for studies of Huntington's disease
  Lotspeich, S.C.; Ashner, M.C.; Vazquez, J.E.; Richardson, B.D.; Grosser, K.F.; Bodek, B.E.; Garcia, T.P.
  Published Annual Review of Statistics and Its Application, 2024 PMID
  Abstract

  The landscape of survival analysis is constantly being revolutionized to answer biomedical challenges, most recently the statistical challenge of censored covariates rather than outcomes. There are many promising strategies to tackle censored covariates, including weighting, imputation, maximum likelihood, and Bayesian methods. Still, this is a relatively fresh area of research, different from the areas of censored outcomes (i.e., survival analysis) or missing covariates. In this review, we discuss the unique statistical challenges encountered when handling censored covariates and provide an in-depth review of existing methods designed to address those challenges. We emphasize each method's relative strengths and weaknesses, providing recommendations to help investigators pinpoint the best approach to handling censored covariates in their data.

• Robust Estimation under Outcome Dependent Right Censoring in Huntington Disease: Estimators for Low and High Censoring Rates
  Vazquez, J.E.; Ma, Y.; Marder, K.; Garcia, T.P.
  Under review Statistics in Medicine, expected 2026 arXiv
• Multiply-robust inverse probability weighting estimation for right-censored covariates
  Raman, M.; Ma, Y.; Marder, K.; Garcia, T.P.; Vazquez, J.E.†
  In preparation
• Identifiability of Regression Models with a Right-Censored Covariate: Extending Canonical Causal Diagrams
  Lin, D.; Ma, Y.; Garcia, T.P.; Vazquez, J.E.†
  In preparation
• Nonparametric Conditional Mean Imputation in Linear Models of Huntington Disease
  Grosser, K.; Ma, Y.; Vazquez, J.E.; Rahman, M.; Lotspeich, S.C.; Garcia, T.P.
  In preparation
• Longitudinal regression models with a right-censored covariate
  Vazquez, J.E.; Ma, Y.; Marder, K.; Garcia, T.P.
  In preparation
• Robustness of parametric maximum likelihood estimation for handling right-censored covariates in generalized linear models
  Vazquez, J.E.; Lotspeich, S.C.; Ma, Y.; Garcia, T.P.
  In preparation

Observational studies of treatment effects are limited when an important confounder, such as disease aggressiveness, is not measured, and in slowly progressive diseases the available proxies of the confounder are often right-censored. I formalize identification conditions for this setting and develop weighted estimators, with applications to antidepressant use in Huntington disease.

• Causal inference with a right-censored marker of disease progression: Adjusting for latent severity in Huntington disease
  Vazquez, J.E.; Ringlein, G.; Bang, J.; Stuart, E.A.; Nguyen, T.
  In preparation
• Target Trial Emulation of Antidepressant Initiation and Depression Outcomes in Huntington Disease: A Clone-Censor-Weight Approach with Grace Periods
  Vazquez, J.E.; Nguyen, T.; Ogilvie, A.; Bang, J.; Garcia, T.P.; Stuart, E.A.
  In preparation

Clinical sites often cannot share individual-level data, and each site can have its own missingness patterns. I develop federated inverse probability weighting estimators that combine site-level summaries into valid joint inference, together with a variance correction that accounts for the estimated weights.

• Federated Learning with Missing Data: A Weighted Approach with Variance Correction
  Vazquez, J.E.; Sheng, Y.; Akulian, J.; Hochberg, C.; Iwashyna, T.J.; Stuart, E.A.; Tong, J.
  Under review Statistics in Medicine, expected 2026 arXiv

The methods above are motivated by studies of Huntington disease. Applied projects with clinical collaborators cover disease progression, family-level outcomes, and clinical characteristics of patients across cohorts.

• Higher educational attainment in Huntington disease families: evidence from the Enroll-HD study
  Vazquez, J.E.; Lin, D.; Mendizabal, A.; Ogilvie, A.C.; Stuart, E.A.; Garcia, T.P.
  Published Orphanet Journal of Rare Diseases, 2026 DOI
• Maternal Predominance Among Affected Parents in Huntington Disease: Evidence from the Enroll-HD Study
  Vazquez, J.E.; Mendizabal, A.; Ogilvie, A.C.; Nguyen, T.; Garcia, T.P.; Stuart, E.A.
  In preparation
• Multi-state time-to-event modeling of Huntington Disease stage progression
  Rahman, M.; Vazquez, J.E.; Lin, D.; Cross, S.; Krishnan, A.; Lotspeich, S.C.; Garcia, T.P.
  In preparation
• Clinical, Molecular, and Survival Characteristics of Huntington's Disease in Mexico: A 30-Year Cohort from the National Referral Center
  Ochoa-Morales, A.; Vazquez, J.E.; Jara-Prado, A.; Yescas-Gómez, P.; Guerrero-Camacho, J.; Cruz-López, F.; Orozco-Pla, M.; Gómez-Pérez, C.A.; Fernandez-Ruiz, J.; Ramírez-García, M.A.
  In preparation

I collaborate with clinical and public health researchers on statistical design and analysis. Recent projects cover cardiovascular and kidney health after preterm birth, pulmonary infection, dermatology, physical activity measurement, neuroimaging, and environmental health.

• Perinatal predictors of abnormal adolescent cardiovascular-kidney health after extremely preterm birth: A machine learning analysis of the Extremely Low Gestational Age Newborn–Environmental Child Health Outcomes (ELGAN-ECHO) cohort
  Sanderson, K.; Vazquez, J.E.; Flythe, J.E.; Askenazi, D.; Kosorok, M.R.; Lu, M.; Makker, K.; Laughon, M.; Kistler, C.; Fry, R.C.; O'Shea, T.M.
  Published Kidney International, 2026 DOI
• Resting State fMRI seizure onset localization meta-analysis: Comparing rs-fMRI to other seizure localizing modalities and surgical outcomes
  Boerwinkle, V.; Nowlene, M.; Vazquez, J.E.; et al.
  Published Frontiers in Neuroimaging, 2024 DOI
• Concurrent validity of the global physical activity questionnaire to accelerometry in Hispanic/Latino adults: The Hispanic Community Health Study/Study of Latinos (HCHS/SOL)
  Vazquez, J.E.; Sotres-Alvarez, D.; Carlson, J.; Talavera, G.; Gallo, L.; Castañeda, S.; Evenson, K.
  Published Journal of Science and Medicine in Sport, 2024 DOI
  Abstract

  Objectives
  The validity of the Global Physical Activity Questionnaire has not been thoroughly evaluated among Hispanics/Latinos. In this cross-sectional study, we assessed the concurrent validity and correlates of discordance of the Global Physical Activity Questionnaire by comparing it to accelerometry in estimating sedentary behavior, moderate-to-vigorous physical activity, and meeting United States physical activity guidelines by sociodemographic, behavioral, and health characteristics.
  
  Design
  The Hispanic Community Health Study/Study of Latinos is a 4-site cohort study of United States adults aged 18–74 years enrolled from 2008 to 2011.
  
  Methods
  Participants (n=11,873) completed the Global Physical Activity Questionnaire and wore an accelerometer for 1 week. Lin's concordance and Pearson correlations assessed concurrent validity between self-reported and accelerometry-assessed measures of sedentary behavior and moderate-to-vigorous physical activity. Kappa coefficients assessed agreement of meeting physical activity guidelines. Linear and logistic regression models identified correlates of discordance.
  
  Results
  The overall Lin's concordance and Pearson correlations between the Global Physical Activity Questionnaire and accelerometry estimates were 0.10 (95% confidence interval 0.09, 0.12) and 0.24 (0.21, 0.27) for sedentary behavior, and 0.04 (0.03, 0.05) and 0.18 (0.15, 0.22) for moderate-to-vigorous physical activity, respectively. Agreement was poor for meeting the physical activity guideline classifications (Kappa coefficients: 0.12 to 0.26). Over a 16-hour day, sedentary behavior was under-reported by 3.8h and moderate-to-vigorous physical activity was over-reported by 1.9h.
  
  Conclusions
  The concurrent validity of the Global Physical Activity Questionnaire in measuring moderate-to-vigorous physical activity and sedentary behavior when compared to accelerometry was poor among Hispanic/Latino adults.

• Geospatial analysis of health risks and solid waste management behaviour
  Rahman, M.M.; Bohara, A.; Vazquez, J.E.†
  Published Journal of Environmental Economics and Policy, 2021 DOI
  Abstract

  The lack of proper waste management behaviour creates environmental debasement and induces health risks. This study employs primary survey data collected in 2016 to investigate the association between health risks and improper waste management behaviour by households in three locations of Rupandehi District, Nepal. The health risk is measured by a series of water-related symptoms such as Diarrhea, Jaundice, Typhoid Fever, Worms, and Cholera. This paper’s novel contribution is that we identify the spatial nature of the prevalence of waterborne diseases and related factors such as solid waste management behaviour, hygiene infrastructure and personal cleanliness, and socio-economic status of the households. We use a spatial autoregressive model under the negative binomial family, and the result indicates a significant spatial autocorrelation of waterborne diseases. Moreover, we find the significant effect of improper waste management practices on waterborne diseases. The result is consistent even after various robustness and falsification tests. The findings from this study indicate the acute need to raise awareness concerning the malicious effect of improper waste management and the urge to provide wider access to waste management services.

• CO₂ Laser Assisted Photodynamic Therapy as a Treatment for Basal Cell Carcinoma
  Park, M.; Vazquez, J.E.; Qaqish, B.; Jolly, P.S.
  Under review Journal of Skin Cancer, expected 2026
• Intrapleural Enzyme Therapy Failure in Pleural Infection: Distinct Factors Associated with Surgery and Mortality in an International Multicentre Cohort
  Choi, H.J.; Vazquez, J.E.; Qaqish, B.; Bedawi, E.O.; Rahman, N.M.; Yarmus, L.; Akulian, J.; Retrolysis and IPOG Study Groups.
  Under review CHEST, expected 2026
• Predicting Treatment Failure with Combination Intrapleural Enzyme Therapy for Complicated Pleural Space Infections: The SPARTAN Score
  Diddams, M.J.; Vazquez, J.E.; Bedawi, E.; Rahman, N.M.; Yarmus, L.; Qaqish, B.; Akulian, J.A.; Retrolysis and IPOG Study Groups.
  Under review expected 2026 SSRN
• A retrospective chart review on social determinants of health in pemphigus patients at a tertiary care referral center from 2014–2022 in the United States
  Alexander-Savino, C.V.; Vazquez, J.E.; Qaqish, B.; Culton, D.A.
  In preparation
• The IET Score: A 30- and 90-day mortality prediction model for patients presenting with complicated pleural infection and treated with intrapleural enzyme therapy
  Diddams, M.D.; Vazquez, J.E.; et al.
  In preparation
• Detection of multiple respiratory viruses associated with mortality and severity of illness among children with COVID-19
  Pizzuto, M.; Menozzi, S.E.; Vazquez, J.E.; Qaqish, B.
  In preparation