The LZIP: A Bayesian Latent Factor Model for Correlated Zero-Inflated Counts

Neelon, Brian; Chung, Dongjun

doi:10.1111/biom.12558

Cited by 16 publications

(16 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fourth, it was challenging to model deaths because most counties reported no deaths on any given day. Future studies could employ zero-inflated models to better account for this aspect of the data [43][44][45]. Future work could also examine temporal trends in locations of correctional facilities, long-term care facilities, nursing homes, Indian reservations and Tribal lands, and other places with high rates of infection [46][47][48][49].…”

Section: Discussionmentioning

confidence: 99%

Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States

Neelon

Mutiso

Mueller

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Background: Emerging evidence suggests that socially vulnerable communities are at higher risk for coronavirus disease 2019 (COVID-19) outbreaks in the United States. However, no prior studies have examined temporal trends and differential effects of social vulnerability on COVID-19 incidence and death rates. The purpose of this study was to examine temporal trends among counties with high and low social vulnerability and to quantify disparities in these trends over time. We hypothesized that highly vulnerable counties would have higher incidence and death rates compared to less vulnerable counties and that this disparity would widen as the pandemic progressed. Methods: We conducted a retrospective longitudinal analysis examining COVID-19 incidence and death rates from March 1 to August 31, 2020 for each county in the US. We obtained daily COVID-19 incident case and death data from USAFacts and the Johns Hopkins Center for Systems Science and Engineering. We classified counties using the Social Vulnerability Index (SVI), a percentile-based measure from the Centers for Disease Control and Prevention in which higher scores represent more vulnerability. Using a Bayesian hierarchical negative binomial model, we estimated daily risk ratios (RRs) comparing counties in the first (lower) and fourth (upper) SVI quartiles. We adjusted for percentage of the county designated as rural, percentage in poor or fair health, percentage of adult smokers, county average daily fine particulate matter (PM2.5), percentage of primary care physicians per 100,000 residents, and the proportion tested for COVID-19 in the state. Results: In unadjusted analyses, we found that for most of March 2020, counties in the upper SVI quartile had significantly fewer cases per 100,000 than lower SVI quartile counties. However, on March 30, we observed a crossover effect in which the RR became significantly greater than 1.00 (RR = 1.10, 95% PI: 1.03, 1.18), indicating that the most vulnerable counties had, on average, higher COVID-19 incidence rates compared to least vulnerable counties. Upper SVI quartile counties had higher death rates on average starting on March 30 (RR = 1.17, 95% PI: 1.01,1.36). The death rate RR achieved a maximum value on July 29 (RR = 3.22, 95% PI: 2.91, 3.58), indicating that most vulnerable counties had, on average, 3.22 times more deaths per million than the least vulnerable counties. However, by late August, the lower quartile started to catch up to the upper quartile. In adjusted models, the RRs were attenuated for both incidence cases and deaths, indicating that the adjustment variables partially explained the associations. We also found positive associations between COVID-19 cases and deaths and percentage of the county designated as rural, percentage of resident in fair or poor health, and average daily PM2.5. Conclusions: Results indicate that the impact of COVID-19 is not static but can migrate from less vulnerable counties to more vulnerable counties over time. This highlights the importance of protecting vulnerable populations as the pandemic unfolds.

show abstract

Section: Discussionmentioning

confidence: 99%

Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States

Neelon

Mutiso

Mueller

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Second, it was challenging to model deaths because many counties reported no deaths on any given day. Future studies could employ zero-inflated models to better account for this aspect of the data [47][48][49]. Future work could also examine temporal trends in locations of correctional facilities, long-term care PLOS ONE…”

Section: Plos Onementioning

confidence: 99%

Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States

et al. 2021

Self Cite

View full text Add to dashboard Cite

Background Socially vulnerable communities may be at higher risk for COVID-19 outbreaks in the US. However, no prior studies examined temporal trends and differential effects of social vulnerability on COVID-19 incidence and death rates. Therefore, we examined temporal trends among counties with high and low social vulnerability to quantify disparities in trends over time. Methods We conducted a longitudinal analysis examining COVID-19 incidence and death rates from March 15 to December 31, 2020, for each US county using data from USAFacts. We classified counties using the Social Vulnerability Index (SVI), a percentile-based measure from the Centers for Disease Control and Prevention, with higher values indicating more vulnerability. Using a Bayesian hierarchical negative binomial model, we estimated daily risk ratios (RRs) comparing counties in the first (lower) and fourth (upper) SVI quartiles, adjusting for rurality, percentage in poor or fair health, percentage female, percentage of smokers, county average daily fine particulate matter (PM2.5), percentage of primary care physicians per 100,000 residents, daily temperature and precipitation, and proportion tested for COVID-19. Results At the outset of the pandemic, the most vulnerable counties had, on average, fewer cases per 100,000 than least vulnerable SVI quartile. However, on March 28, we observed a crossover effect in which the most vulnerable counties experienced higher COVID-19 incidence rates compared to the least vulnerable counties (RR = 1.05, 95% PI: 0.98, 1.12). Vulnerable counties had higher death rates starting on May 21 (RR = 1.08, 95% PI: 1.00,1.16). However, by October, this trend reversed and the most vulnerable counties had lower death rates compared to least vulnerable counties. Conclusions The impact of COVID-19 is not static but can migrate from less vulnerable counties to more vulnerable counties and back again over time.

show abstract

“…This situation often occurs with ZI models. Thus, the WAIC has been increasingly used in model selection of Bayesian ZI models; see, for example, Neelon and Chung (2017), Song et al (2018), andNeelon (2019). The LPML statistic requires calculation of the conditional predictive ordinate (CPO) statistic of Gelfand, Dey, and Chang (1992).…”

Section: Inference Considerationsmentioning

confidence: 99%

Zero‐inflatedmodeling part I: Traditionalzero‐inflatedcount regression models, their applications, and computational tools

Young

Roemmele

Yeh

2020

WIREs Computational Stats

View full text Add to dashboard Cite

Count regression models maintain a steadfast presence in modern applied statistics as highlighted by their usage in diverse areas like biometry, ecology, and insurance. However, a common practical problem with observed count data is the presence of excess zeros relative to the assumed count distribution. The seminal work of Lambert (1992) was one of the first articles to thoroughly treat the problem of zero-inflated count data in the presence of covariates. Since then, a vast literature has emerged regarding zero-inflated count regression models. In this first of two review articles, we survey some of the classic and contemporary literature on parametric zero-inflated count regression models, with emphasis on the utility of different univariate discrete distributions. We highlight some of the primary computational tools available for estimating and assessing the adequacy of these models. We concurrently emphasize the diverse data problems to which these models have been applied.

show abstract

The LZIP: A Bayesian Latent Factor Model for Correlated Zero-Inflated Counts

Cited by 16 publications

References 16 publications

Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States

Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States

Spatial and temporal trends in social vulnerability and COVID-19 incidence and death rates in the United States

Zero‐inflatedmodeling part I: Traditionalzero‐inflatedcount regression models, their applications, and computational tools

Contact Info

Product

Resources

About