Multivariate left‐censored Bayesian modeling for predicting exposure using multiple chemical predictors

Groth, Caroline P; Banerjee, Sudipto; Ramachandran, Gurumurthy; Stenzel, Mark; Stewart, Patricia A.

doi:10.1002/env.2505

Cited by 11 publications

(2 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our design represents merely one possible way in which we might seek to model such data, and we shall pursue alternatives in future work. In situations where the emphasis lies firmly on prediction, for example, there is value in exploring a design similar to that of Groth et al (2018), in which each component of the multivariate response is regressed on the others. This in turn opens up new options for capturing and describing between‐depth relationships.…”

Section: Conclusion and Future Researchmentioning

confidence: 99%

A Nearest-Neighbour Gaussian Process Spatial Factor Model for Censored, Multi-Depth Geochemical Data

Davies¹,

Banerjee

Martin

et al. 2022

Journal of the Royal Statistical Society Series C: Applied Statistics

Self Cite

View full text Add to dashboard Cite

We investigate the relationships between local environmental variables and the geochemical composition of the Earth in a region spanning over 26,000 km 2 in the lower South Island of New Zealand. Part of the Southland-South Otago geochemical baseline survey-a pilot study pre-empting roll-out across the country-the data comprise the measurements of 59 chemical trace elements, each at two depth prescriptions, at several hundred spatial sites. We demonstrate construction of a hierarchical spatial factor model that captures inter-depth dependency; handles imputation of left-censored readings in a statistically principled manner; and exploits sparse approximations to Gaussian processes to deliver inference. The voluminous results provide a novel impression of the underlying processes and are presented graphically via simple web-based applications. These both confirm existing knowledge and provide a basis from which new research hypotheses in geochemistry might be formed.

show abstract

Section: Conclusion and Future Researchmentioning

confidence: 99%

A Nearest-Neighbour Gaussian Process Spatial Factor Model for Censored, Multi-Depth Geochemical Data

Davies¹,

Banerjee

Martin

et al. 2022

Journal of the Royal Statistical Society Series C: Applied Statistics

Self Cite

View full text Add to dashboard Cite

show abstract

“…While most priors were left vague to allow the data to drive the inference, we did restrict the GSDs in the repeated measures ANOVA in order to restrict possible GSDs to ranges typically seen in personal timeweighted averages. We also assume that measurements below the limit of detection follow similar trends as the observed measurements (52,53). Additional assumptions associated with ANOVA include normality of errors, independence of individuals (or observations within a non-repeated measures ANOVA), and constant variances within-and between-workers.…”

Section: Limitations and Further Researchmentioning

confidence: 99%

Exposures and Emissions in Coffee Roasting Facilities and Cafés: Diacetyl, 2,3-Pentanedione, and Other Volatile Organic Compounds

LeBouf

Hawley

Fortner

et al. 2020

Front. Public Health

Self Cite

View full text Add to dashboard Cite

Roasted coffee and many coffee flavorings emit volatile organic compounds (VOCs) including diacetyl and 2,3-pentanedione. Exposures to VOCs during roasting, packaging, grinding, and flavoring coffee can negatively impact the respiratory health of workers. Inhalational exposures to diacetyl and 2,3-pentanedione can cause obliterative bronchiolitis. This study summarizes exposures to and emissions of VOCs in 17 coffee roasting and packaging facilities that included 10 cafés. We collected 415 personal and 760 area full-shift, and 606 personal task-based air samples for diacetyl, 2,3-pentanedione, 2,3-hexanedione, and acetoin using silica gel tubes. We also collected 296 instantaneous activity and 312 instantaneous source air measurements for 18 VOCs using evacuated canisters. The highest personal full-shift exposure in part per billion (ppb) to diacetyl [geometric mean (GM) 21 ppb; 95th percentile (P95) 79 ppb] and 2,3-pentanedione (GM 15 ppb; P95 52 ppb) were measured for production workers in flavored coffee production areas. These workers also had the highest percentage of measurements above the NIOSH Recommended Exposure Limit (REL) for diacetyl (95%) and 2,3-pentanedione (77%). Personal exposures to diacetyl (GM 0.9 ppb; P95 6.0 ppb) and 2,3-pentanedione (GM 0.7 ppb; P95 4.4 ppb) were the lowest for non-production workers of facilities that did not flavor coffee. Job groups with the highest personal full-shift exposures to diacetyl and 2,3-pentanedione were flavoring workers (GM 34 and 38 ppb), packaging workers (GM 27 and 19 ppb) and grinder operator (GM 26 and 22 ppb), respectively, in flavored coffee facilities, and packaging workers (GM 8.0 and 4.4 ppb) and production workers (GM 6.3 and 4.6 ppb) in non-flavored coffee facilities. Baristas in cafés had mean full-shift exposures below the RELs (GM 4.1 ppb diacetyl; GM 4.6 ppb 2,3pentanedione). The tasks, activities, and sources associated with flavoring in flavored LeBouf et al. Exposures Among Coffee Workers coffee facilities and grinding in non-flavored coffee facilities, had some of the highest GM and P95 estimates for both diacetyl and 2,3-pentanedione. Controlling emissions at grinding machines and flavoring areas and isolating higher exposure areas (e.g., flavoring, grinding, and packaging areas) from the main production space and from administrative or non-production spaces is essential for maintaining exposure control.

show abstract

Quantifying Worker Exposures Using Bayesian Statistical Methods in Industrial Hygiene

Groth¹,

Banerjee²,

Ramachandran³

et al. 2022

Wiley StatsRef: Statistics Reference Online

View full text Add to dashboard Cite

Workplaces are dynamic and complex locations in which workers can be exposed to a variety of harmful chemicals. One of the industrial hygienists' primary responsibilities is the evaluation of workers' exposures to such chemicals. To do so, appropriate statistical modeling strategies are needed to accurately describe these exposures. Bayesian methods provide a useful framework to quantify and characterize workers' exposures in the complex work environment. This article presents several Bayesian modeling strategies to improve the understanding of workers' exposures while accounting for measurements below the analytic limit of detection. These methods allow industrial hygienists to characterize exposures by summarizing measured exposures, thereby investigating relationships between exposures to multiple chemicals, statistically identifying and describing exposure determinants, and estimating exposures when measurements are unavailable.

show abstract

Multivariate left‐censored Bayesian modeling for predicting exposure using multiple chemical predictors

Cited by 11 publications

References 14 publications

A Nearest-Neighbour Gaussian Process Spatial Factor Model for Censored, Multi-Depth Geochemical Data

A Nearest-Neighbour Gaussian Process Spatial Factor Model for Censored, Multi-Depth Geochemical Data

Exposures and Emissions in Coffee Roasting Facilities and Cafés: Diacetyl, 2,3-Pentanedione, and Other Volatile Organic Compounds

Quantifying Worker Exposures Using Bayesian Statistical Methods in Industrial Hygiene

Contact Info

Product

Resources

About