Air pollution and COVID-19 mortality in the United States: Strengths and limitations of an ecological regression analysis
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ScreenIT)
- Evaluated articles (NCRC)
- High interest articles (NCRC)
Abstract
We describe the challenges and opportunities of analyzing links between exposure to air pollution and vulnerability to COVID-19.
Article activity feed
-
SciScore for 10.1101/2020.04.05.20054502: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
NIH rigor criteria are not applicable to paper type.Table 2: Resources
No key resources detected.
Results from OddPub: Thank you for sharing your code and data.
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:It is important to acknowledge that this study has limitations, mainly due to the fact that this is an ecological study with data available at the county level and that this is a cross-sectional study. High quality, nationwide individual-level COVID-19 outcome data are unavailable at this time and for the foreseeable future, thus necessitating the use of an ecologic study design for these analyses. Due to the potential for ecologic …
SciScore for 10.1101/2020.04.05.20054502: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
NIH rigor criteria are not applicable to paper type.Table 2: Resources
No key resources detected.
Results from OddPub: Thank you for sharing your code and data.
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:It is important to acknowledge that this study has limitations, mainly due to the fact that this is an ecological study with data available at the county level and that this is a cross-sectional study. High quality, nationwide individual-level COVID-19 outcome data are unavailable at this time and for the foreseeable future, thus necessitating the use of an ecologic study design for these analyses. Due to the potential for ecologic bias, our results should be interpreted in the context of this design and should not be used to make individual-level inferential statements. Also, unmeasured confounding bias is a threat to the validity of our conclusions. Unfortunately, in the midst of a pandemic it is not feasible to design a study and collect the data at the ideal level of spatial and temporal resolution to minimize all sources of bias. Yet, conditional on the data available, we have endeavored to adjust for confounding bias by all of the most important factors, including population density, time since the beginning of the outbreak, social isolation measures, behavior, weather, age structure, ethnicity, access to health care, and socio-economic factors. We also conducted 68 additional analyses to assess the robustness of the results to many modelling choices. Furthermore, we computed the E-value to demonstrate that the confounding effect of any unmeasured confounder would need to be much stronger than that of any of our observed confounders in order to explain away the relation...
Results from TrialIdentifier: No clinical trial numbers were referenced.
Results from Barzooka: We did not find any issues relating to the usage of bar graphs.
Results from JetFighter: We did not find any issues relating to colormaps.
Results from rtransparent:- Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
- Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
- No protocol registration statement was detected.
-
-
Our take
This ecological study of 3,087 counties in the United States (covering ~98% of the population), published as a preprint and thus not yet peer reviewed, found a statistically significant, positive association between average levels of fine particulate matter (PM-2.5) for the period 2000-2016 and COVID-19 death rates of counties up to April 22, 2020. Findings remained robust even after several sensitivity analyses. However, given the ecological study design and the possibility that confounding factors may be unaccounted for, particularly as COVID-19 is still a new and evolving disease, caution should be taken in interpreting the study findings. These ecological results should be interpreted with caution and followed up with individual-level designs and exploration of potential mechanisms for the observed associations.
Our take
This ecological study of 3,087 counties in the United States (covering ~98% of the population), published as a preprint and thus not yet peer reviewed, found a statistically significant, positive association between average levels of fine particulate matter (PM-2.5) for the period 2000-2016 and COVID-19 death rates of counties up to April 22, 2020. Findings remained robust even after several sensitivity analyses. However, given the ecological study design and the possibility that confounding factors may be unaccounted for, particularly as COVID-19 is still a new and evolving disease, caution should be taken in interpreting the study findings. These ecological results should be interpreted with caution and followed up with individual-level designs and exploration of potential mechanisms for the observed associations.
Study design
ecological
Study population and setting
The aim of this study was to examine the association between long-term air pollution exposure (2000-2016) and COVID-19 mortality during the first wave of the COVID-19 pandemic in the United States. The pollutant of interest was particulate matter <2.5 micrometers in diameter (PM-2.5). Particulate matter, which refers to the mix of solid particles and liquid droplets found in the air, come in various shapes and sizes, and can contain hundreds of different chemicals. PM-2.5 are of particular interest because their small size allows them to penetrate deep into the lungs, and even into the bloodstream. In order to estimate average PM-2.5 levels per county, the authors first estimated monthly PM-2.5 exposure levels at 0.01 degree X 0.01 degree grid resolution across the US—combining satellite, modeled and monitored PM-2.5 data in a geographically weighted regression. Values for all grid points within a zip code were then averaged. Then, the PM-2.5 levels were averaged across all zip codes in each county. The authors obtained COVID-19 death counts for 3,087 counties in the United States (covering ~98% of the population) up to April 22, 2020. Death counts were obtained from the Johns Hopkins Center for Systems Science and Engineering Coronavirus Resource Center which summarizes data reported by CDC and state health departments. Mortality rate per county was calculated as the total COVID-19 deaths in the county divided by the county’s population size. A negative binomial mixed model incorporating a random intercept by state to account for correlation in counties within the same state, and several potential confounding factors (e.g., time since stay at home orders, population density, median household income, percent black, percent Hispanic, percent current smokers) was fit to assess the relationship between the PM-2.5 level and COVID-19 deaths outcome.
Summary of main findings
Higher COVID-19 death rates were observed in the Mid-Atlantic, upper Midwest, and Gulf Coast regions of the US, similar to patterns for high population density and high PM-2.5 levels. After adjustment for covariates, the mortality rate ratio was 1.08 (95% CI: 1.02, 1.15), indicating that for each 1 ug/m^3 increase in average PM-2.5 levels, there was an associated statistically significant 8% increase in the COVID-19 death rate.
Study strengths
Use of an extensively cross-validated PM-2.5 prediction model to estimate each county’s long-term average PM-2.5 level, and several sensitivity analyses to examine robustness of the findings.
Limitations
This was an ecological study, with data available at the county level only. Thus, these findings cannot be used to make individual-level inferences. In addition, limited testing capacity in the country at the time of the analyses made it difficult to accurately quantify number of COVID-19 cases, a limitation somewhat diminished by utilization of death versus diagnosis. It is unclear whether the period examined for PM-2.5 (2000-2016) may be most relevant for the COVID-19 death outcome (assessed in 2020).
Value added
This was the first nationwide study to examine associations between long-term PM-2.5 exposure and COVID-19 death rates in the United States. This study provides a base for further investigations into how air pollution might impact infection and mortality from COVID-19.
-
SciScore for 10.1101/2020.04.05.20054502: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Institutional Review Board Statement not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Sex as a biological variable not detected. Table 2: Resources
Software and Algorithms Sentences Resources Acknowledgments: The computations in this paper were run on ( 1 ) the Odyssey cluster supported by the FAS Division of Science , Research Computing Group at Harvard University; and ( 2 ) the Research Computing Environment supported by the Institute for Quantitative Social Science in the Faculty of Arts and Sciences at Harvard University . Artssuggested: NoneResults from OddPub: We did not detect open …
SciScore for 10.1101/2020.04.05.20054502: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Institutional Review Board Statement not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Sex as a biological variable not detected. Table 2: Resources
Software and Algorithms Sentences Resources Acknowledgments: The computations in this paper were run on ( 1 ) the Odyssey cluster supported by the FAS Division of Science , Research Computing Group at Harvard University; and ( 2 ) the Research Computing Environment supported by the Institute for Quantitative Social Science in the Faculty of Arts and Sciences at Harvard University . Artssuggested: NoneResults from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
About SciScore
SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore is not a substitute for expert review. SciScore checks for the presence and correctness of RRIDs (research resource identifiers) in the manuscript, and detects sentences that appear to be missing RRIDs. SciScore also checks to make sure that rigor criteria are addressed by authors. It does this by detecting sentences that discuss criteria such as blinding or power analysis. SciScore does not guarantee that the rigor criteria that it detects are appropriate for the particular study. Instead it assists authors, editors, and reviewers by drawing attention to sections of the manuscript that contain or should contain various rigor criteria and key resources. For details on the results shown here, including references cited, please follow this link.
-
-