Saskia Comess

All Publications

• Modelling spatial heterogeneity in exposure buffers and risk: a hierarchical Bayesian approach JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES C-APPLIED STATISTICS Comess, S., Ho, D. E., Warren, J. L. 2026

  View details for DOI 10.1093/jrsssc/qlag020

  View details for Web of Science ID 001729261400001

• A Comprehensive Dataset of Factory Farms in California Compiled Using Computer Vision and Human Validation. Scientific data Magesh, V., Rothbacher, N., Comess, S., Maneri, E., Rodolfa, K., Tartof, S., Casey, J., Nachman, K., Ho, D. E. 2025; 12 (1): 1826

  Abstract

  Concentrated Animal Feeding Operations (CAFOs) house livestock at high densities for prolonged periods of time, posing substantial risks to environmental and human health. However, limited public information on CAFOs has constrained efforts to quantify their impacts on proximate communities. Gaps in permitting and reporting have severely limited studies that rely primarily on administrative records. We introduce Cal-FF, a near-complete census of CAFOs in California, a large and agriculturally significant state in the United States, with richer facility data than existing administrative data. Cal-FF was constructed using computer vision on satellite imagery, along with extensive human validation. We focus on California, which accounts for about 20% of US livestock production and has been at the forefront of CAFO regulatory innovation. We estimate that our 2,121 facility dataset captures 98% (95% CI [82%, 98%]) of all California CAFOs as of 2017, identifying 222 locations not present in state regulatory records. In addition to improved accuracy, Cal-FF adds a wealth of information about each facility, including validated permit information, land parcel data, satellite imagery, and annotated facility features. These data provide numerous opportunities for research, analysis, and monitoring.

  View details for DOI 10.1038/s41597-025-06082-6

  View details for PubMedID 41258138

  View details for PubMedCentralID PMC12630652

• A Bayesian framework for incorporating exposure uncertainty into health analyses with application to air pollution and stillbirth. Biostatistics (Oxford, England) Comess, S., Chang, H. H., Warren, J. L. 2022

  Abstract

  Studies of the relationships between environmental exposures and adverse health outcomes often rely on a two-stage statistical modeling approach, where exposure is modeled/predicted in the first stage and used as input to a separately fit health outcome analysis in the second stage. Uncertainty in these predictions is frequently ignored, or accounted for in an overly simplistic manner when estimating the associations of interest. Working in the Bayesian setting, we propose a flexible kernel density estimation (KDE) approach for fully utilizing posterior output from the first stage modeling/prediction to make accurate inference on the association between exposure and health in the second stage, derive the full conditional distributions needed for efficient model fitting, detail its connections with existing approaches, and compare its performance through simulation. Our KDE approach is shown to generally have improved performance across several settings and model comparison metrics. Using competing approaches, we investigate the association between lagged daily ambient fine particulate matter levels and stillbirth counts in New Jersey (2011-2015), observing an increase in risk with elevated exposure 3 days prior to delivery. The newly developed methods are available in the R package KDExp.

  View details for DOI 10.1093/biostatistics/kxac034

  View details for PubMedID 35984351

• Statistical Modeling for Practical Pooled Testing During the COVID-19 Pandemic STATISTICAL SCIENCE Comess, S., Wang, H., Holmes, S., Donnat, C. 2022; 37 (2): 229-250

  View details for DOI 10.1214/22-STS857

  View details for Web of Science ID 000798149000006

• Exposure to atmospheric metals using moss bioindicators and neonatal health outcomes in Portland, Oregon ENVIRONMENTAL POLLUTION Comess, S., Donovan, G., Gatziolis, D., Deziel, N. C. 2021; 284: 117343

  Abstract

  Studying the impacts of prenatal atmospheric heavy-metal exposure is challenging, because biological exposure monitoring does not distinguish between specific sources, and high-resolution air monitoring data is lacking for heavy metals. Bioindicators - animal or plant species that can capture environmental quality - are a low-cost tool for evaluating exposure to atmospheric heavy-metal pollution that have received little attention in the public-health literature. We obtained birth records for Portland, Oregon live births (2008-2014) and modeled metal concentrations derived from 346 samples of moss bioindicators collected in 2013. Exposure estimates were assigned using mother's residential address at birth for six metals with known toxic and estrogenic effects (arsenic, cadmium, chromium, cobalt, nickel, lead). Associations were evaluated for continuous (cts) and quartile-based (Q) metal estimates and three birth outcomes (preterm birth (PTB; <37 weeks)), very PTB (vPTB; <32 weeks), small for gestational age (SGA; 10th percentile of weight by age and sex)) using logistic regression models with adjustment for demographic characteristics, and stratified by maternal race. Chromium and cobalt were associated with increased odds of vPTB (chromium - odds ratio (OR)cts = 1.09, 95% CI: 1.00, 1.17; cobalt - ORQ4vsQ1 = 1.33, 95% CI: 1.03, 1.71). Cobalt, chromium and cadmium were significantly associated with odds of SGA, although the direction of association differed by metal (cobalt - ORcts = 1.04, 95% CI: 1.01, 1.07; chromium - ORQ3vsQ1 = 0.91, 95% CI: 0.83, 0.99; cadmium - ORcts = 0.96, 95% CI: 0.93, 1.00). In stratified analyses, odds of SGA were significantly different among non-white mothers compared to white mothers with exposure to chromium, cobalt, lead and nickel. This novel application of a moss-based exposure metric found that exposure to some atmospheric metals is associated with adverse birth outcomes. These findings are consistent with previous literature and suggest that moss bioindicators are a useful complement to traditional exposure-assessment methods.

  View details for DOI 10.1016/j.envpol.2021.117343

  View details for Web of Science ID 000672534900011

  View details for PubMedID 34030082