Alison Mondul
Associate Professor
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (30)
COVID-19 ICU admission among cancer (n = 12,995) and non-cancer patients (n = 34,693) who tested positive and had non-missing covariates. The results were from the model logit P(YICU = 1|X, Cancer, Covariate) = β0 + βXX + βCancerCancerStatus + βIntX × CancerStatus + βCovCovariate, where Covariate = Age + Race/Ethnicity + Sex + Disadvantage Index (quartile) + Comorbidity Score. Reference groups were age: 18 to 35, sex: female, BMI: 18.5 to 25, race: white/non-Hispanic, smoking status: never. A unit change in the continuous age variable is 10 years. The comorbidity score ranges from 0 to 6 and is the sum of indicators for presence of six categories of preexisting conditions in the electronic health record.
Year:
2025
COVID-19 hospitalization among recent cancer (within three years of positive COVID-19 test; n = 6,143) and non-cancer patients (n = 41,545) who tested positive and have non-missing covariates. The results were from the model logit P(YHospitalization = 1|X, Cancer, Covariate) = β0 + βXX + βCancerCancer + βIntX × Cancer + βCovCovariate, where Covariate = Age + Race/Ethnicity + Sex + Disadvantage Index (quartile) + Comorbidity Score. Reference groups were age: 18 to 35, sex: female, BMI: 18.5 to 25, race: white/non-Hispanic, smoking status: never. A unit change in the continuous age variable is 10 years. The comorbidity score ranges from 0 to 6 and is the sum of indicators for presence of six categories of preexisting conditions in the electronic health record.
Year:
2025
Severe COVID-19 among cancer (n = 12,995) and non-cancer patients (n = 34,693) who tested positive and had non-missing covariates. The results were from the model logit P(YSevereCOVID = 1|X, Cancer, Covariate) = β0 + βXX + βCancerCancer + βIntX × Cancer + βCovCovariate, where Covariate = Age + Race/Ethnicity + Sex + Disadvantage Index (quartile) + Comorbidity Score. Reference groups were age: 18 to 35, sex: female, BMI: 18.5 to 25, race: white/non-Hispanic, smoking status: never. A unit change in the continuous age variable is 10 years. The comorbidity score ranges from 0 to 6 and is the sum of indicators for presence of six categories of preexisting conditions in the electronic health record.
Year:
2025
A schematic representing electronic health record scenarios and their resulting (panel A) classification of individuals with cancer for the primary and cancer recency analyses and (panel B) vaccination status for vaccine-related analyses.
Year:
2025
An UpSet diagram showing the distribution of cancer treatment patterns among COVID-19 positive individuals with cancer (n = 13,752). Treatment-based analyses were performed using three mutually exclusive categories: chemotherapy (n = 2,910), radiation only (n = 690), and surgery only (n = 172). Treatment was defined according to Table S2.
Year:
2025
COVID-19 ICU admission among recent cancer (within three years of positive COVID-19 test; n = 6,143) and non-cancer patients (n = 41,545) who tested positive and have non-missing covariates. The results were from the model logit P(YICU = 1|X, Cancer, Covariate) = β0 + βXX + βCancerCancer + βIntX × Cancer + βCovCovariate, where Covariate = Age + Race/Ethnicity + Sex + Disadvantage Index (quartile) + Comorbidity Score. Reference groups were age: 18 to 35, sex: female, BMI: 18.5 to 25, race: white/non-Hispanic, smoking status: never. A unit change in the continuous age variable is 10 years. The comorbidity score ranges from 0 to 6 and is the sum of indicators for presence of six categories of preexisting conditions in the electronic health record.
Year:
2025
Additional description of methods including derivation of analytic cohort, description of the neighborhood disadvantage index, and additional details of statistical analyses for transparency, replicability, and reproducibility.
Year:
2025
Severe COVID-19 among recent cancer (within three years of positive COVID-19 test; n = 6,143) and non-cancer patients (n = 41,545) who tested positive and have non-missing covariates. The results were from the model logit P(YSevereCOVID = 1|X, Cancer, Covariate) = β0 + βXX + βCancerCancer + βIntX × Cancer + βCovCovariate, where Covariate = Age + Race/Ethnicity + Sex + Disadvantage Index (quartile) + Comorbidity Score. Reference groups were age: 18 to 35, sex: female, BMI: 18.5 to 25, race: white/non-Hispanic, smoking status: never. A unit change in the continuous age variable is 10 years. The comorbidity score ranges from 0 to 6 and is the sum of indicators for presence of six categories of preexisting conditions in the electronic health record.
Year:
2025
COVID-19 hospitalization among cancer (n = 12,995) and non-cancer patients (n = 34,693) who tested positive and had non-missing covariates. The results were from the model logit P(YHospitalization = 1|X, Cancer, Covariate) = β0 + βXX + βCancerCancerStatus + βIntX × CancerStatus + βCovCovariate, where Covariate = Age + Race/Ethnicity + Sex + Disadvantage Index (quartile) + Comorbidity Score. Reference groups were age: 18 to 35, sex: female, BMI: 18.5 to 25, race: white/non-Hispanic, smoking status: never. A unit change in the continuous age variable is 10 years. The comorbidity score ranges from 0 to 6 and is the sum of indicators for presence of six categories of preexisting conditions in the electronic health record.
Year:
2025
Electronic health records (EHRs) are valuable for public health and clinical research but are prone to many sources of bias, including missing data and non-probability selection. Missing data in EHRs is complex due to potential non-recording, fragmentation, or clinically informative absences. This study explores whether polygenic risk score (PRS)-informed multiple imputation for missing traits, combined with sample weighting, can mitigate missing data and selection biases in estimating disease-exposure associations. Simulations were conducted for missing completely at random (MCAR), missing at random (MAR), and missing not at random (MNAR) conditions under different sampling mechanisms. PRS-informed multiple imputation showed generally lower bias, particularly when combined with sample weighting. For example, in biased samples of 10,000 with exposure and outcome MAR data, PRS-informed imputation had lower percent bias (3.8%) and better coverage rate (0.883) compared to PRS-uninformed (4.5%; 0.877) and complete case analyses (10.3%; 0.784) in covariate-adjusted, weighted, multiple imputation scenarios. In a case study using Michigan Genomics Initiative (n=50,026) data, PRS-informed imputation aligned more closely with a sample-weighted All of Us-derived benchmark than analyses ignoring missing data and selection bias. Researchers should consider leveraging genetic data and sample weighting to address biases from missing data and non-probability sampling in biobanks.
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Collaborators (5)
Bhramar Mukherjee
-
Seunggeun Lee
Professor
Seoul National University
Lars Fritsche
Research Associate Professor
University of Michigan–Ann Arbor
Xu Shi
University of Michigan
Christopher Friese
Professor
University of Michigan School of Nursing
Make your next move.
Prepare your application
Review drafts before you send them.