Next Article in Journal
Investigating Dual Character of Atmospheric Ammonia on Particulate NH4NO3: Reducing Evaporation Versus Promoting Formation
Previous Article in Journal
Temporal Trends and Meteorological Associations of Particulate Matter and Gaseous Air Pollutants in Tehran, Iran (2017–2021)
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 16
Issue 6
10.3390/atmos16060684
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

COVID-19 Mortality Among Hospitalized Medicaid Patients in Kentucky (2020–2021): A Geospatial Study of Social, Medical, and Environmental Risk Factors

by
Shaminul H. Shakib
1,*,
Bert B. Little
1,2,
Seyed M. Karimi
1,2 and
Michael Goldsby
1
1
Department of Health Management and Systems Sciences, School of Public Health and Information Sciences, University of Louisville, Louisville, KY 40202, USA
2
Department of Bioinformatics and Biostatistics, School of Public Health and Information Sciences, University of Louisville, Louisville, KY 40202, USA
*
Author to whom correspondence should be addressed.
Atmosphere 2025, 16(6), 684; https://doi.org/10.3390/atmos16060684
Submission received: 1 May 2025 / Revised: 28 May 2025 / Accepted: 2 June 2025 / Published: 5 June 2025
(This article belongs to the Section Air Quality and Health)
Browse Figures
Review Reports Versions Notes

Abstract

(1) Background: Geospatial associations for COVID-19 mortality were estimated using a cohort of 28,128 hospitalized Medicaid patients identified from the 2020–2021 Kentucky Health Facility and Services administrative claims data. (2) Methods: County-level patient information (age, sex, chronic obstructive pulmonary disease [COPD], and mechanical ventilation use [96 hrs. plus]); social deprivation index (SDI) scores; physician and nurse rates per 100,000; and annual average particulate matter 2.5 (PM2.5) were used as the predictors. Ordinary least-squares (OLS) regression and multiscale geographically weighted regression (MGWR) with the dependent variable, COVID-19 mortality per 100,000, were performed to compute global and local effects, respectively. (3) Results: MGWR (adjusted R2: 0.52; corrected Akaike information criterion [AICc]: 292.51) performed better at explaining the association between the dependent variable and predictors than the OLS regression (adjusted R2: 0.36; AICc: 301.20). The percentages of patients with COPD and who were mechanically ventilated (96 hrs. plus) were significantly associated with COVID-19 mortality, respectively (OLS standardized βCOPD: 0.22; βventilation: 0.53; MGWR mean βCOPD: 0.38; βventilation: 0.57). Other predictors were not statistically significant in both models. (4) Conclusions: A risk of COVID-19 mortality was observed among patients with COPD and prolonged mechanical ventilation use, after controlling for social determinants, the healthcare workforce, and PM2.5 in rural and Appalachian counties of Kentucky. These counties are characterized by persistent poverty, healthcare workforce shortages, economic distress, and poor population health outcomes. Improving population health protection through multisector collaborations in rural and Appalachian counties may help reduce future health burdens.

1. Introduction

As of 12 April 2025, 1,226,565 COVID-19 deaths have been reported in the United States (U.S.) [1]. Hospitalized COVID-19 patients with lung disease, i.e., chronic obstructive pulmonary disease (COPD), are at higher risk of in-hospital mortality [2]. Social determinants of health (SDOH) risk factors among COVID-19 inpatients include older age, sex, race, poverty, unemployment, education, overcrowded housing units, and air quality [3]. Acute respiratory distress syndrome (ARDS) is another common risk factor among COVID-19 patients that increases the risk of in-hospital mortality and often requires mechanical ventilation [4].
Positive associations between air quality indicators—particulate matter 2.5 (PM2.5) and other respiratory viruses, including the influenza virus—have been reported, indicating the potential role of particulate matter as a transporter for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) [5,6]. Microorganisms with a diameter less than 5 micrometers (μm) can pair with suspended microscopic dust in the air. Although SARS-CoV-2 particles are much smaller—approximately 70 to 90 nanometers (nm)—they may become embedded within or adsorbed onto larger particles such as PM2.5. This could facilitate indirect airborne transmission by allowing viral particles to remain suspended in the atmosphere for extended periods or travel longer distances [7,8]. Multiple studies have reported positive associations between PM2.5 and COVID-19-related mortality [9].
Current mortality risk estimates among hospitalized COVID-19 patients, adjusting for COPD status, SDOH factors, and PM2.5, are from non-spatial studies that use patient-level data [9,10,11,12,13,14,15,16]. Furthermore, limited studies have focused on the COVID-19-associated risk of in-hospital mortality among COVID-19-positive Medicaid beneficiaries [17,18].
Traditional regression methods, i.e., ordinary least-squares (OLS) and binary logistic regression, can identify a global association between an outcome and one or more independent variables. However, these regression models assume spatial stationarity—meaning the relationship between the independent variables and outcome is constant across all locations in the study area. This assumption can limit the model’s ability to account for local variations and lead to misleading or biased estimates.
For instance, the effect of chronic conditions like COPD on COVID-19 mortality may be stronger in rural counties with limited healthcare access than in urban counties with more resources. Failing to account for such heterogeneity can mask high-risk areas and produce under-informed interventions.
The spatial regression method, i.e., geographically weighted regression (GWR), overcomes this limitation by computing local associations between a dependent variable and independent variables at each spatial location, with smooth changes in regression coefficients across the locations in the study area.
GWR does this by computing an individual weighted regression model for each location in the study area using data from neighboring locations. Weights are calculated based on the geographical distance between a location and its spatial neighbors. Neighbors in close proximity to the location receive higher weights and have a greater effect on the local regression model [19,20].
While limited in numbers, there are COVID-19-related studies that have used the GWR model and identified the issue of spatial heterogeneity [21,22,23]. However, these studies do not account for the variance in spatial heterogeneity at different scales for COVID-19 outcomes across locations in the study area, a limitation of GWR.
Multiscale geographically weighted regression (MGWR), an extension of GWR, can address this limitation as it analyzes data across multiple scales, therefore allowing the associations between dependent and independent variables to vary both across locations and at different scales [24].
Medicaid patients generally have a high prevalence of COPD, limited access to care, and greater exposure to adverse social determinants of health, making them more susceptible to COVID-19-related hospitalization and mortality [25]. Accordingly, Kentucky is an ideal location for examining COVID-19-related mortality, as it has a high prevalence of COPD and a large Medicaid-enrolled population [26,27]. The state’s mix of rural, urban, and Appalachian counties introduces considerable geographic and socioeconomic diversity. Moreover, access to healthcare services varies widely across counties, with rural and Appalachian areas experiencing greater provider shortages [28].
Approximately half of the Kentucky physicians (49.1%) work in Fayette (19.6%; physician ratio [PR]: 1:148) or Jefferson (29.5%; PR: 1:244) counties. University of Louisville Health and University of Kentucky Healthcare, two of Kentucky’s largest healthcare organizations, are located in Jefferson and Fayette counties, respectively [28]. Therefore, patients from neighboring counties to Fayette or Jefferson have better access to care compared to far-distant counties.
The spatial heterogeneity observed in county-level mortality among hospitalized COVID-19 Medicaid patients provides a unique opportunity to measure non-stationary geographic variation across counties in Kentucky (Figure 1). To the best of our knowledge, no prior study has conducted a spatial analysis focused on COVID-19-related mortality among Medicaid patients.
This study investigated spatial associations of county-level mortality, controlling for COPD, SDOH factors, the healthcare workforce, and PM2.5 among hospitalized COVID-19 Medicaid patients using the MGWR model in Kentucky.
Atmosphere 16 00684 g001
Figure 1. Mortality per 100,000 among hospitalized COVID-19 Medicaid patients across counties in Kentucky: 2020–2021.
Figure 1. Mortality per 100,000 among hospitalized COVID-19 Medicaid patients across counties in Kentucky: 2020–2021.
Atmosphere 16 00684 g001

2. Materials and Methods

2.1. Study Sample

A total of 28,128 hospitalized non-pregnant adult Medicaid beneficiaries diagnosed with COVID-19 were identified from the Kentucky Health Facility and Services (KHFS) inpatient administrative claims data for the years 2020–2021 [29]. Inclusion criteria were (1) enrollment in Kentucky Medicaid during 2020–2021; (2) an age between 18 and 64 years; (3) a confirmed COVID-19 diagnosis; and (4) hospitalization for COVID-19, excluding pregnancy-, childbirth-, and puerperium-related cases. The International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) diagnostic codes B97.29 and/or U07.1 were used to record COVID-19-positive hospitalized patients [30].
KFHS data provides the following patient demographic information: (1) adult age groups: 18–20, 21–24, 25–29, 30–34, 35–39, 40–44, 45–49, 50–54, 55–59, and 60–64; (2) sex: male and female. The CDC’s 2021 national COVID-19 death counts were used to identify the age-specific burden of mortality [31]. This identification served as guidance to recode the original age groups into the following categories: (1) 18–54 and (2) 55–64.
Patients with COPD were identified using the latest set of ICD-10-CM diagnosis codes from the Centers for Medicare & Medicaid Services (CMS) Chronic Conditions Data Warehouse (CCW) [32]. A binary variable for mechanical ventilation use by duration of usage (hrs.) was created using the following ICD-10 Procedure Coding System (PCS) codes: ventilation greater than 96 consecutive hrs. (5A095 and 5A195) [33]. A discharge status code of 20 was used to identify patients who died by the end of their hospital stay.
The unit of analysis for this geospatial ecological cross-sectional study was the county (n = 120). All the patient-level information captured from the KHFS inpatient data was aggregated at the county level. This study abided by the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines [34].

2.2. Outcome

The outcome of interest was county-level COVID-19-related in-hospital mortality per 100,000 ([count of COVID-19 in-hospital mortality/count of COVID-19-positive patients] × 100,000).

2.3. Independent Variables

Independent variables included the following county-level percents ([count of respective variables/count of COVID-19-positive patients] × 100): age groups: 18–54 (reference [ref]) and 55–64; sex: female (ref) and male; COPD; and mechanical ventilation use: 96 hrs. plus. In addition to this, the county-level social deprivation index (SDI) score, physician and nurse rates per 100,000, and annual average PM2.5 were also included as independent variables.
Robert Graham Center’s SDI is a composite measure of area-level deprivation based on the following seven demographic parameters obtained from the American Community Survey: (1) the percentage living in poverty; (2) the percentage with less than 12 years of education; (3) the percentage of single-parent households; (4) the percentage living in rented housing units; (5) the percentage living in an overcrowded housing unit; (6) the percentage of households without a car; and (7) the percentage of non-employed adults under 65 years of age.
The SDI can help evaluate socioeconomic heterogeneity in health outcomes. SDI scores range from 0 to 100. Members of communities with higher SDI scores are at increased risk of disease and have limited access to care, which contributes to unmet healthcare needs and poor patient outcomes [35].
The 2021 county-level physician and nurse rates per 100,000 were obtained from the Area Health Resources Files (AHRFs) [36]. The 2020 county-level highest annual average concentration of PM2.5 was obtained from the Centers for Disease Control and Prevention (CDC) National Environmental Public Health Tracking Network [37]. Corresponding data for 2021 were not available.
County-level SDI scores, physician and nurse rates per 100,000, and PM2.5 concentrations were merged with aggregated patient data at the county level using Federal Information Processing Standards (FIPS) codes. Descriptive statistics for the dependent and independent variables are reported (Table 1).

2.4. Statistical Analysis

In an OLS model, the relationships between the dependent variable and the independent variables are stationary and constant [38]. An OLS model was performed to understand the statewide global effect of independent variables on the dependent variables using the county-level data.
In the OLS model, the following county-level independent variables (unit: percent) were regressed against COVID-19-related in-hospital mortality per 100,000: age group: 18–54 (ref) and 55–64; sex: female (ref) and male; COPD; and mechanical ventilation use: 96 hrs. plus (Table 2). The county-level SDI scores, physician and nurse rates per 100,000, and annual average PM2.5 were also included as independent variables.
The variance inflation factor (VIF) is a measure of collinearity between independent variables in an OLS regression. Multicollinearity is present if the VIF is higher than 5 [39]. None of the independent variables had a VIF greater than 2.5 in the OLS model (Table 2).
An MGWR model was performed using the same set of variables as the OLS model to determine if the effects of independent variables on the dependent variable differed by counties at different scales in Kentucky (Table 3). MGWR, unlike linear models, has effects that vary across locations by strength and direction. This assists in capturing county-specific variation in trends. The MGWR method computes a local regression model for every county (i) by using data from enclosed counties (j) that are within county i’s neighborhood. The number of nearest neighbors (that forms the j) from which data will be used is the bandwidth. The size of the bandwidth differs for each variable, as not all relationships take place at the same scale in MGWRs. Below is the MGWR model expression:
y i = ∑ j = 1 M β b w j X i j + ε i
βbwj is the estimated coefficient for county i, bwj is the optimal bandwidth size, and εi is the random error term. Bandwidths for each of the independent variables reveal the scale at which spatial non-stationarity occurs in the MGWR model. Smaller bandwidths suggest greater local variance, while bigger bandwidths indicate a more global response, similar to OLS. A maximum of 120 bandwidths were possible, which corresponds to the number of counties (N = 120). A bi-square kernel with a bandwidth (the number of nearest neighbors j) was used to form a neighborhood from which data was used in the MGWR model. The golden-section search algorithm was used to optimize the bandwidth that minimized the corrected Akaike information criterion (AICc).
To understand the local county-level effects of the independent variables on the dependent variable, the following scales of effects were defined based on the percentage of each variable’s bandwidth relative to the total number of counties ([bandwidth/120] × 100): small regional effect (0–20%), medium regional effect (21–40%), large regional effect (41–60%), and global statewide effect (>60%).
For each county, MGWR produced a local coefficient (N = 120) for the independent variables. Independent variables with significant local coefficients for one or more counties were mapped (base layer: OpenStreetMap; classification method: natural breaks; color scheme: viridis) (Figure 2, Figure 3 and Figure 4). Local coefficients for each county that were not significant for these maps were recoded to 0 for better visual illustration (Figure 2, Figure 3 and Figure 4). Mean coefficients and COVID-19-related in-hospital mortality per 100,000, as well as bandwidth and scale of effect for counties with significant local coefficients by independent variables, are reported (Table 4).
The Akaike information criterion (AIC) estimates the model’s prediction error, allowing one to compare and determine the relative quality of statistical models. The model with the lowest AIC value is considered to be the best compared to other models [40]. The adjusted R2 and AICc were used to evaluate the OLS regression and MGWR model performance. The critical value for statistical significance (p-value) was set at ≤0.05. Analyses were conducted using KNIME v5.1, STATA 18, and MGWR 2.0 [41].

3. Results

3.1. Descriptive

The county-level mean and standard deviation (SD) of COVID-19-related in-hospital mortality per 100,000 among the patients were 9737.43 and 4426.83, respectively. Following were the means and SDs for independent variables, respectively: (1) age group: % of 18–54 (25.62, 5.19) and % of 55–64 (53.47, 5.90); (2) sex: % of females (46.94, 6.21) and % of males (53.06, 6.21); (3) % of COPD (18.46, 5.23); (4) mechanical ventilation use: % of 96 hrs. plus (11.27, 4.00); (5) SDI score (66.74, 20.91); (6) physicians per 100,000 (109.67, 118.60); (7) nurses per 100,000 (131.44, 87.25); and (8) PM2.5 (7.95, 0.63) (Table 1).
Table 1. Descriptive statistics of county-level study sample: hospitalized COVID-19 patients in Kentucky: 2020–2021.
Table 1. Descriptive statistics of county-level study sample: hospitalized COVID-19 patients in Kentucky: 2020–2021.
CategoryDescriptionDependent VariableMean2 SDMinMaxObs.
COVID-19-Related Inpatient Mortality(county count of COVID-19-related inpatient mortality)/(county count of patient) × 100,000Inpatient Mortality per 100,0009737.434426.832380.9533,333.34120
Independent Variables
Age Group(County count of patients by respective age group)/(county count of patients) × 100% of 18–5425.625.190.0036.75120
% of 55–6453.475.9030.0068.00120
Sex(County count of patients by respective sex)/(county count of patients) × 100% of Female46.946.2132.5677.78120
% of Male53.066.2122.2267.44120
Lung Disease(County count of COPD patients)/(county count of patients) × 1001 % of COPD18.465.230.0033.33120
Mechanical Ventilation Use(County count of patients mechanically ventilated 96 hrs. plus)/(county count of patients) × 100% of 96 hrs. Plus11.274.002.8622.22120
Social Deprivation Index (SDI)County-level SDI score ranging from 0 to 100SDI Score (0:100) 66.7420.913.0096.00120
Healthcare AccessCounty-level physician rate per 100,000 populationPhysician per 100k109.67118.600.00848.99120
County-level nurse rate per 100,000 populationNurse per 100k131.4487.2520.54483.37120
Air QualityAnnual average particulate matter 2.5PM2.57.950.635.609.80120
1 Chronic obstructive pulmonary disease (COPD). 2 SD = standard deviation. Note: Patient refers to COVID-19-positive Medicaid beneficiaries hospitalized in Kentucky during 2020–2021.
Counties with the highest COVID-19-related in-hospital mortality per 100,000 were the following: Robertson (33,333.33), McLean (28,571.43), Nicholas (20,454.55), Trimble (19,444.44), Mercer (17,532.47), and Russell (17,142.86) (Figure 1).

3.2. OLS Regression and MGWR Model

Results from OLS regression and the MGWR model for COVID-19-related in-hospital mortality per 100,000 were compared to better understand these spatial patterns. The MGWR model (adjusted R2: 0.52; AICc: 292.51) performed better at explaining the association between the predictors (COPD, SDOH factors, the healthcare workforce, and PM2.5) and COVID-19-related in-hospital mortality per 100,000 than the OLS regression (adjusted R2: 0.36; AICc: 301.20). The difference of 16% in adjusted R2 and −8.69 in AICc between MGWR and OLS regression attests to MGWR’s ability to capture spatial heterogeneity across counties at different scales in Kentucky.
The following independent variables were not statistically significant in both OLS regression (unstandardized β [β], 95% confidence intervals [95% CIs], and p-value [p]) and MGWR (mean β, SD, bandwidth [BW]): sex (ref: % of female): % of male (β: −65.55, 95% CI: −176.43−45.33, p = 0.24, mean β: 0.08, SD: 0.01, BW: 119); SDI score (β: −10.11, 95% CI: −52.73–32.50, p = 0.64, mean β: 0.06, SD: 0.04, BW: 117); physicians per 100,000 (β: −1.46, 95% CI: −9.74–6.82, p = 0.73, mean β: −0.02, SD: 0.02, BW: 119); nurses per 100,000 (β: −3.99, 95% CI: −15.06–7.07, p = 0.48, mean β: −0.06, SD: 0.01, BW: 119); and PM2.5 (β: −237.60, 95% CI: −1602.45–1127.25, p = 0.73, mean β: 0.05, SD: 0.01, BW: 119) (Table 2 and Table 3). In the MGWR model, all βs across counties for each of these independent variables were not significant. Furthermore, the small changes in βs across counties and large bandwidths yield a global statewide effect.
Table 2. Ordinary least-squares (OLS) regression of county-level mortality per 100,000 among hospitalized COVID-19 patients.
Table 2. Ordinary least-squares (OLS) regression of county-level mortality per 100,000 among hospitalized COVID-19 patients.
Unstandardized βStandardized β1 SE95% CIp-Value2 VIF
LowerUpper
Intercept4629.87 8487.99−12,189.6421,449.380.59
Age Group (ref: % of 18–54)
% of 55–6438.310.0558.58−77.78154.400.521.13
Sex (ref: % Female)
% of Male−65.55−0.0955.96−176.4345.330.241.14
Lung Disease
% of COPD189.680.2267.5455.83323.520.011.18
Mechanical Ventilation Use
% of 96 hrs. Plus581.450.5384.85413.30749.59<0.0011.09
Social Deprivation Index (SDI)
SDI Score (0:100)−10.11−0.0521.51−52.7332.500.641.91
Healthcare Workforce
Physicians per 100k−1.46−0.044.18−9.746.820.732.32
Nurses per 100k−3.99−0.085.58−15.067.070.482.24
Air Quality
PM2.5−237.60−0.03688.77−1602.451127.250.731.78
1 SE = Standard Error of Coefficients. 2 VIF = variance inflation factor.
Table 3. Multiscale geographically weighted regression (MGWR) of county-level mortality per 100,000 among hospitalized COVID-19 patients.
Table 3. Multiscale geographically weighted regression (MGWR) of county-level mortality per 100,000 among hospitalized COVID-19 patients.
1 Mean1 SD1 Min1 Median1 Max2 Counties (%)3 Bandwidth4 Scale of Effect (%)
Intercept0.090.020.040.090.120.00117Statewide Global97.50
Age Group (ref: % of 18–54)
% of 55–640.140.15−0.080.180.4521.6765Large Regional54.17
Sex (ref: % Female)
% of Male0.080.010.070.070.100.00119Statewide Global99.17
Lung Disease
% of COPD0.210.25−0.430.240.5548.3345Medium Regional37.50
Mechanical Ventilation Use
% of 96 hrs. Plus0.450.240.030.460.9369.1744Medium Regional36.67
Social Deprivation Index (SDI)
SDI Score (0:100)0.060.04−0.040.080.080.00117Statewide Global97.50
Healthcare Access
Physicians per 100 k−0.020.02−0.03−0.020.030.00119Statewide Global99.17
Nurses per 100 k−0.060.01−0.08−0.07−0.030.00119Statewide Global99.17
Air Quality
PM2.50.050.010.020.050.070.00119Statewide Global99.17
1 Mean, median, standard deviation, minimum, and maximum of county-specific local coefficients for respective independent variables from MGWR. 2 Percentage of counties with a significant result (p ≤ 0.05) for 120 local coefficients from MGWR. 3 Optimal bandwidth for the respective independent variables using the golden-section search algorithm from MGWR. 4 Scale of effect (bandwidth/120) × 100: small regional scale (0–20%), medium regional scale (20–40%), large regional scale (40–60%), and statewide global scale (>60%).
The OLS regression results showed that the percentage of patients aged 55–64 was not significantly associated with COVID-19-related in-hospital mortality per 100,000 (β: 38.31, 95% CI: −77.78–154.40, p = 0.52) (Table 2). However, the percentage of patients aged 55–64 was significantly associated with COVID-19-related in-hospital mortality per 100,000 among 21.67% (n = 26) of the total counties located in the south-central region in the MGWR model (mean β: 0.41, SD: 0.01) (Table 4). The block of counties stretched from Mercer (west to Fayette) to Monroe and McCreary, located on the state borderline in south central Kentucky (Figure 2). The scale of this effect was a large regional effect (BW: 65).
Table 4. Significant local coefficients from multiscale geographically weighted regression (MGWR) of county-level mortality per 100,000 among hospitalized COVID-19 patients.
Table 4. Significant local coefficients from multiscale geographically weighted regression (MGWR) of county-level mortality per 100,000 among hospitalized COVID-19 patients.
1 Mean1 SD1 Min1 Median1 Max2 Counties (%)3 Bandwidth4 Scale of Effect (%)
Age Group (ref: % of 18–54)
% of 55–640.350.060.240.340.4521.6765Large Regional54.17
Lung Disease
% of COPD0.380.23−0.430.430.5548.3345Medium Regional37.50
Mechanical Ventilation Use
% of 96 hrs. Plus0.570.180.260.560.9369.1744Medium Regional36.67
1 Mean, median, standard deviation, minimum, and maximum of county-specific significant local coefficients for respective independent variables from MGWR. 2 Percentage of counties with a significant result (p ≤ 0.05) for 120 local coefficients from MGWR. 3 Optimal bandwidth for the respective independent variables using the golden-section search algorithm from MGWR. 4 Scale of effect (bandwidth/120) × 100: small regional scale (0–20%), medium regional scale (20–40%), large regional scale (40–60%), and statewide global scale (>60%).
Atmosphere 16 00684 g002
Figure 2. Significant local coefficients of age group: % of 55−64 from the MGWR model.
Figure 2. Significant local coefficients of age group: % of 55−64 from the MGWR model.
Atmosphere 16 00684 g002
Results from both OLS regression and the MGWR model indicated a significant association between the percentage of patients with COPD and COVID-19-related in-hospital mortality per 100,000 (β: 189.68, 95% CI: 55.83–323.52, p = 0.01, mean β: 0.38, SD: 0.23) (Table 2 and Table 4). While the OLS regression implied a positive significant association, the MGWR model provided county-specific significant βs ranging from negative to positive. Among 58 counties with significant βs, 5 of those were in the northwest stretching from Hancock to Muhlenberg, indicating a negative association. However, the majority of the rest of the remaining 53 counties were located between north-central and south-central Kentucky (Figure 3). The scale of this effect was a medium regional effect (BW: 45).
Atmosphere 16 00684 g003
Figure 3. Significant local coefficients of lung disease: % of COPD patients from the MGWR model.
Figure 3. Significant local coefficients of lung disease: % of COPD patients from the MGWR model.
Atmosphere 16 00684 g003
Both OLS regression and the MGWR model showed a strong positive significant association between the percentage of patients mechanically ventilated for more than 96 hrs. and COVID-19-related in-hospital mortality per 100,000 (β: 581.45, 95% CI: 413.30–749.59, p < 0.001, mean β: 0.57, SD: 0.18) (Table 2 and Table 4). A bit more than two-thirds of counties across Kentucky (N = 83, 69.17%) were found to have significant local coefficients in the MGWR model. The highest set of local coefficients was found among counties stretching from Kenton (north-central) to Powell (mid-central) and Owen (north to Fayette) to Lewis (northeast) (Figure 4). Relatively lower significant local coefficients were found among most of the counties located from the west to the northeast and southwest, followed by the south-central to the northeast.
Atmosphere 16 00684 g004
Figure 4. Significant local coefficients of mechanical ventilation use: % of 96 hrs. plus from the MGWR model.
Figure 4. Significant local coefficients of mechanical ventilation use: % of 96 hrs. plus from the MGWR model.
Atmosphere 16 00684 g004

4. Discussion

To the best of our knowledge, no study has attempted to identify non-stationary spatial associations between mortality among hospitalized COVID-19 Medicaid patients and county-level social determinants, COPD, the healthcare workforce, and PM2.5 across different spatial scales. The county-level COVID-19-related mortality among hospitalized Medicaid patients during 2020–2021 substantially differed across counties in Kentucky, indicating the presence of spatial heterogeneity (Figure 1).
Although results from this study did not reveal any independent effect of social determinants (e.g., sex or SDI score), the healthcare workforce, or air quality (PM2.5) on county-level COVID-19-related mortality among Medicaid patients, numerous studies have demonstrated statistically significant associations between COVID-19 outcomes—particularly mortality—and social determinants of health among marginalized communities [3].
There are limited studies that have explored the relationship between healthcare workforce availability and COVID-19 mortality in the U.S. Findings from these studies indicate a significant association between provider shortages and increased mortality [42,43]. However, the current study found no significant relationship between county-level physician and nurse rates and COVID-19-related mortality among Medicaid patients in either the global OLS regression or the local MGWR model.
This absence of significance may reflect broader structural mechanisms within hospitals. The relationship between healthcare workforce availability and patient outcomes may be influenced by external factors such as infection control policies and practices, ICU surge capacity, staffing ratios, the availability of ventilators, and the operational readiness of hospitals to manage surges [44,45]. These nuances are not captured by our model.
Several ecological studies have associated long-term exposure to PM2.5 with COVID-19 mortality [46]. But these studies do not account for the individual-level confounding effects of social determinants or the availability of healthcare resources and medical capacity that vary across geographical locations. Sheppard et al. (2023) conducted a systematic review focused on the risk associated with PM2.5 exposure to COVID-19 outcomes and only included studies that used individual-level data [47]. While associations were observed between PM2.5 and COVID-19-related mortality in several studies, the effect was found not to be significant, indicating a weak relationship.
Non-stationary spatial effects with regional-scale variation were observed in some counties located in south-central Kentucky among hospitalized older adults aged 55–64 for COVID-19-related mortality (Figure 2). While it is well studied that the risk of COVID-19-related mortality significantly increases with age, with older adults aged 50 plus experiencing a much higher mortality rate compared to younger individuals [48], there is a lack of associated geosocial factors that may further explain this risk.
Most of the counties with a significant association between the percentage of hospitalized patients aged 55–64 and COVID-19-related mortality are in the Kentucky Appalachian Region (Figure 2). The Kentucky Appalachian Region has a long history of persistent poverty and subsistence living. The Appalachian Regional Commission’s (ARC) county economic status classifies counties in Appalachia as either “distressed” or “at risk.” Distressed counties are the most economically disadvantaged counties. They are in the bottom 10% of the nation’s counties. At-risk counties are those that are on the verge of becoming economically distressed. They rank among the lowest 10 to 25% of the nation’s counties [26]. While analyses from this study did not reveal an independent effect of SDI scores on COVID-19-related inpatient mortality, the finding on age accounting for spatial effects alludes to poverty and subsistence living—a major component of the SDI score—as a potential geosocial factor.
COPD weakens the innate and adaptive immune responses and delays the clearance of respiratory infections, thereby contributing to a higher mortality risk observed among COVID-19 patients [49]. A strong positive association between COPD and COVID-19-related mortality among hospitalized Medicaid patients in counties between north- and south-central Kentucky was observed in this study (Figure 3). The majority of these counties are rural/Appalachian areas faced with higher rates of poverty, limited employment opportunities, poor health status, and lower rates of educational attainment. On the contrary, a negative association between COPD and COVID-19-related mortality was identified in some selected counties in northwestern Kentucky, namely, Hancock, Daviess, and McLean. These counties are urbanized, populous areas contributing to Kentucky’s economic growth [26].
Mechanical ventilation use among COVID-19 patients with severe respiratory distress has been associated with an increased risk of mortality [50,51]. In the current study, widespread non-stationary spatial effects with regional-scale variation were observed in two-thirds of the counties across Kentucky among patients who received prolonged mechanical ventilation (96 hrs. plus) for COVID-19-related mortality (Figure 4). Highly significant associations were identified among counties extending from north-central and northeastern Kentucky midway through to the south-central region (Figure 4). While some of these counties are urban areas, the surrounding counties are rural areas. This association spreads outward with a relatively lower effect across the region. Other significant associations were mostly observed among rural counties located in western Kentucky.
The differences in associations observed among patients with COPD or those who were mechanically ventilated (96 hrs. plus) across regions may be due to variations in population health protections available in urban and rural areas of Kentucky. Population health activities—monitoring adverse health events and outbreaks, assessing and implementing community health needs, conducting health inspections, and developing health education and policies to encourage prevention and behavior change—greatly vary in rural and urban areas of Kentucky. Further, collaboration among a diverse set of multisector partners for population health protection is far more prevalent in urban areas in the state [52].
Population health interventions, particularly those focused on promoting physical activity and addressing risk factors such as smoking, can significantly impact the health of individuals with COPD [53]. Urban areas generally have better access to early population health interventions for COPD compared to rural areas [54]. These interventions may help achieve better lung functioning and avoid critical care, i.e., mechanical ventilation, thereby improving overall health outcomes.
Notably, differences in public health funding between rural and urban areas are key drivers of variations in population health activities and interventions. Urban centers often receive a larger share of resources, while rural communities may experience limitations in access to services and funding [55,56]. Future research should directly measure and evaluate these contextual factors to better understand their contribution to geographic disparities in COVID-19-related outcomes.

5. Strengths and Limitations

Other studies have evaluated the risk of COVID-19 mortality. However, analyses from those studies assumed that the associations between mortality and other relevant factors are spatially stationary [9,10,11,12,13,14,15,16]. Further, limited studies have focused on Medicaid patients. Use of the MGWR technique helped capture the spatial and scale-dependent associations between COVID-19-related mortality and relevant determinants. The inclusion of the healthcare workforce and air quality (PM2.5) strengthened the study design [17,18].
KHFS—deidentified administrative claims data—was used to construct county-level data for analysis; therefore, it was not possible to capture hospital re-admission, a key limitation. The current study used county-level counts of hospitalized COVID-19-positive Medicaid patients from Kentucky, which limits its generalizability. Ecological fallacy is another major limitation, as county-level data was used in this study. Temporality is another concern as air quality (PM2.5) is subject to seasonal change and may have varying effects on COVID-19-related mortality. Future studies should use spatiotemporal analytical models to further investigate the effects of PM2.5 on COVID-19 outcomes with respect to time and location.

6. Conclusions

In this geospatial ecological cross-sectional study, a disparity in COVID-19-related mortality among Kentucky Medicaid patients struggling with COPD or receiving prolonged mechanical ventilation was identified, adjusting for social determinants, the healthcare workforce, and PM2.5. The MGWR model revealed that these patients who reside in rural and Appalachian counties were significantly at a higher risk of mortality. Care begins in the community. An increase in population health protection through multisector collaborations across rural and Appalachian counties in Kentucky may help reduce the health disparity.

Author Contributions

S.H.S.: conceptualization, methodology, software, validation, formal analysis, investigation, data curation, writing—original draft, writing—review and editing, and visualization. B.B.L.: conceptualization, methodology, validation, formal analysis, investigation, writing—review and editing, and visualization. S.M.K.: conceptualization, methodology, formal analysis, investigation, writing—review and editing, and visualization. M.G.: conceptualization, methodology, software, validation, formal analysis, investigation, data curation, writing—review and editing, and visualization. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

The study did not require ethical approval.

Informed Consent Statement

Not applicable.

Data Availability Statement

Kentucky Health Facility and Services (KHFS) inpatient data was used to identify Medicaid patients. This data can be requested from the Commonwealth Institute of Kentucky, University of Louisville. Other datasets used in this study are publicly available.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Centers for Disease Control and Prevention. COVID Data Tracker. 2025. Available online: https://covid.cdc.gov/covid-data-tracker/#maps_percent-covid-deaths (accessed on 21 April 2025).
  2. Puebla Neira, D.A.; Watts, A.; Seashore, J.; Duarte, A.; Nishi, S.P.; Polychronopoulou, E.; Kuo, Y.F.; Baillargeon, J.; Sharma, G. Outcomes of Patients with COPD Hospitalized for Coronavirus Disease 2019. Chronic Obstr. Pulm. Dis. 2021, 8, 517–527. [Google Scholar] [CrossRef]
  3. Brakefield, W.S.; Olusanya, O.A.; White, B.; Shaban-Nejad, A. Social Determinants and Indicators of COVID-19 Among Marginalized Communities: A Scientific Review and Call to Action for Pandemic Response and Recovery. Disaster Med. Public Health Prep. 2022, 17, e193. [Google Scholar] [CrossRef] [PubMed]
  4. Potere, N.; Valeriani, E.; Candeloro, M.; Tana, M.; Porreca, E.; Abbate, A.; Spoto, S.; Rutjes, A.W.S.; Di Nisio, M. Acute complications and mortality in hospitalized patients with coronavirus disease 2019: A systematic review and meta-analysis. Crit. Care 2020, 24, 389. [Google Scholar] [CrossRef]
  5. Su, W.; Wu, X.; Geng, X.; Zhao, X.; Liu, Q.; Liu, T. The short-term effects of air pollutants on influenza-like illness in Jinan, China. BMC Public Health 2019, 19, 1319. [Google Scholar] [CrossRef] [PubMed]
  6. Chen, F.; Liu, Z.; Huang, T.; Wang, B.; Sun, Z.; Gao, X.; Wang, W. Short-Term Effects of Air Pollution on the Risk of Influenza in Jinan, China during 2020–2021: A Time-Series Analysis. Atmosphere 2023, 14, 53. [Google Scholar] [CrossRef]
  7. Kim, J.M.; Chung, Y.S.; Jo, H.J.; Lee, N.J.; Kim, M.S.; Woo, S.H.; Park, S.; Kim, J.W.; Kim, H.M.; Han, M.G. Identification of Coronavirus Isolated from a Patient in Korea with COVID-19. Osong Public Health Res. Perspect. 2020, 11, 3–7. [Google Scholar] [CrossRef]
  8. Nor, N.S.M.; Yip, C.W.; Ibrahim, N.; Jaafar, M.H.; Rashid, Z.Z.; Mustafa, N.; Hamid, H.H.A.; Chandru, K.; Latif, M.T.; Saw, P.E.; et al. Particulate matter (PM(2.5)) as a potential SARS-CoV-2 carrier. Sci. Rep. 2021, 11, 2508. [Google Scholar] [CrossRef]
  9. Shao, L.; Cao, Y.; Jones, T.; Santosh, M.; Silva, L.F.O.; Ge, S.; da Boit, K.; Feng, X.; Zhang, M.; BeruBe, K. COVID-19 mortality and exposure to airborne PM(2.5): A lag time correlation. Sci. Total Environ. 2022, 806, 151286. [Google Scholar] [CrossRef]
  10. Cummings, M.J.; Baldwin, M.R.; Abrams, D.; Jacobson, S.D.; Meyer, B.J.; Balough, E.M.; Aaron, J.G.; Claassen, J.; Rabbani, L.E.; Hastie, J.; et al. Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: A prospective cohort study. Lancet 2020, 395, 1763–1770. [Google Scholar] [CrossRef]
  11. Denslow, S.; Wingert, J.R.; Hanchate, A.D.; Rote, A.; Westreich, D.; Sexton, L.; Cheng, K.; Curtis, J.; Jones, W.S.; Lanou, A.J.; et al. Rural-urban outcome differences associated with COVID-19 hospitalizations in North Carolina. PLoS ONE 2022, 17, e0271755. [Google Scholar] [CrossRef]
  12. Garg, S.; Kim, L.; Whitaker, M.; O’Halloran, A.; Cummings, C.; Holstein, R.; Prill, M.; Chai, S.J.; Kirley, P.D.; Alden, N.B.; et al. Hospitalization Rates and Characteristics of Patients Hospitalized with Laboratory-Confirmed Coronavirus Disease 2019—COVID-NET, 14 States, March 1–30, 2020. MMWR Morb. Mortal. Wkly. Rep. 2020, 69, 458–464. [Google Scholar] [CrossRef]
  13. Gold, J.A.W.; Wong, K.K.; Szablewski, C.M.; Patel, P.R.; Rossow, J.; da Silva, J.; Natarajan, P.; Morris, S.B.; Fanfair, R.N.; Rogers-Brown, J.; et al. Characteristics and Clinical Outcomes of Adult Patients Hospitalized with COVID-19—Georgia, March 2020. MMWR Morb. Mortal. Wkly. Rep. 2020, 69, 545–550. [Google Scholar] [CrossRef] [PubMed]
  14. Imam, Z.; Odish, F.; Gill, I.; O’Connor, D.; Armstrong, J.; Vanood, A.; Ibironke, O.; Hanna, A.; Ranski, A.; Halalau, A. Older age and comorbidity are independent mortality predictors in a large cohort of 1305 COVID-19 patients in Michigan, United States. J. Intern. Med. 2020, 288, 469–476. [Google Scholar] [CrossRef]
  15. Roth, G.A.; Emmons-Bell, S.; Alger, H.M.; Bradley, S.M.; Das, S.R.; de Lemos, J.A.; Gakidou, E.; Elkind, M.S.V.; Hay, S.; Hall, J.L.; et al. Trends in Patient Characteristics and COVID-19 In-Hospital Mortality in the United States During the COVID-19 Pandemic. JAMA Netw. Open 2021, 4, e218828. [Google Scholar] [CrossRef] [PubMed]
  16. Mendy, A.; Wu, X.; Keller, J.L.; Fassler, C.S.; Apewokin, S.; Mersha, T.B.; Xie, C.; Pinney, S.M. Air pollution and the pandemic: Long-term PM(2.5) exposure and disease severity in COVID-19 patients. Respirology 2021, 26, 1181–1187. [Google Scholar] [CrossRef]
  17. Howland, R.E.; Wang, S.; Ellen, I.G.; Glied, S. Not a New Story: Place- and Race-Based Disparities in COVID-19 and Influenza Hospitalizations among Medicaid-Insured Adults in New York City. J. Urban Health 2022, 99, 345–358. [Google Scholar] [CrossRef] [PubMed]
  18. Jacobson, M.; Chang, T.Y.; Shah, M.; Pramanik, R.; Shah, S.B. Racial and Ethnic Disparities in SARS-CoV-2 Testing and COVID-19 Outcomes in a Medicaid Managed Care Cohort. Am. J. Prev. Med. 2021, 61, 644–651. [Google Scholar] [CrossRef]
  19. Brunsdon, C.; Fotheringham, S.; Charlton, M. Geographically weighted regression. J. R. Stat. Soc. Ser. D (Stat.) 1998, 47, 431–443. [Google Scholar] [CrossRef]
  20. Leung, Y.; Mei, C.-L.; Zhang, W.-X. Statistical tests for spatial nonstationarity based on the geographically weighted regression model. Environ. Plan. A Econ. Space 2000, 32, 9–32. [Google Scholar] [CrossRef]
  21. Liu, F.; Wang, J.; Liu, J.; Li, Y.; Liu, D.; Tong, J.; Li, Z.; Yu, D.; Fan, Y.; Bi, X.; et al. Predicting and analyzing the COVID-19 epidemic in China: Based on SEIRD, LSTM and GWR models. PLoS ONE 2020, 15, e0238280. [Google Scholar] [CrossRef]
  22. Jiao, J.; Chen, Y.; Azimian, A. Exploring temporal varying demographic and economic disparities in COVID-19 infections in four U.S. areas: Based on OLS, GWR, and random forest models. Comput. Urban Sci. 2021, 1, 27. [Google Scholar] [CrossRef] [PubMed]
  23. Wu, X.; Zhang, J. Exploration of spatial-temporal varying impacts on COVID-19 cumulative case in Texas using geographically weighted regression (GWR). Environ. Sci. Pollut. Res. 2021, 28, 43732–43746. [Google Scholar] [CrossRef] [PubMed]
  24. Fotheringham, A.S.; Yang, W.; Kang, W. Multiscale geographically weighted regression (MGWR). Ann. Am. Assoc. Geogr. 2017, 107, 1247–1265. [Google Scholar] [CrossRef]
  25. Bazell, C.; Alston, M.; Feigler, N.; Germack, H.D.; Leary, S.; Fopalan, W.; Mannino, D. Variation in Prevalence and Burden of Chronic Obstructive Pulmonary Disease by State and Insurance Type in the United States. Chronic Obstr. Pulm. Dis. 2025, 12, 158–174. [Google Scholar] [CrossRef]
  26. UK Center of Excellence in Rural Health. Demographics, Analytics, Trends and Access (K-DATA); UK Center of Excellence in Rural Health: Hazard, KY, USA, 2022. [Google Scholar]
  27. Kamour, A.; David, M.; Kanotra, S. Prevalence and Comorbidities of Chronic Obstructive Pulmonary Disease Among Adults in Kentucky Across Gender and Area Development Districts, 2011. Chronic Obstr. Pulm. Dis. 2015, 2, 296–312. [Google Scholar] [CrossRef]
  28. UK Center of Excellence in Rural Health. Kentucky Physician Report; University of Kentucky (UK): Lexington, KY, USA, 2022. [Google Scholar]
  29. CHFS. Health Facility and Services Data. 2024. Available online: https://www.chfs.ky.gov/agencies/ohda/Pages/hfsd.aspx (accessed on 15 March 2024).
  30. NCHS. COVID-19 Hospital Encounters by Week from Selected Hospitals. 2020. Available online: https://www.cdc.gov/nchs/covid19/nhcs/hospital-encounters-by-week.htm (accessed on 28 June 2024).
  31. Ahmad, F.B.; Cisewski, J.A.; Xu, J.; Anderson, R.N. Provisional Mortality Data—United States, 2022. MMWR Morb. Mortal. Wkly. Rep. 2023, 72, 488–492. [Google Scholar] [CrossRef]
  32. Chronic Conditions Data Warehouse. Condition Categories. 2024. Available online: https://www2.ccwdata.org/web/guest/condition-categories (accessed on 28 June 2024).
  33. Bruce, S.S.; Navi, B.B.; Zhang, C.; Kim, J.; Devereux, R.B.; Schenck, E.J.; Sedrakyan, A.; Díaz, I.; Kamel, H. Transesophageal echocardiography and risk of respiratory failure in patients who had ischemic stroke or transient ischemic attack: An IDEAL phase 4 study. BMJ Surg. Interv. Health Technol. 2022, 4, e000116. [Google Scholar] [CrossRef]
  34. von Elm, E.; Altman, D.G.; Egger, M.; Pocock, S.J.; Gotzsche, P.C.; Vandenbroucke, J.P.; Initiative, S. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. Lancet 2007, 370, 1453–1457. [Google Scholar] [CrossRef]
  35. Butler, D.C.; Petterson, S.; Phillips, R.L.; Bazemore, A.W. Measures of social deprivation that predict health care access and need within a rational area of primary care service delivery. Health Serv. Res. 2013, 48, 539–559. [Google Scholar] [CrossRef]
  36. US Department of Health and Human Services. Area Health Resources Files (AHRF). 2021. Available online: https://data.hrsa.gov/topics/health-workforce/ahrf (accessed on 25 April 2024).
  37. Centers for Disease Control and Prevention. National Environmental Public Health Tracking Network. 2025. Available online: https://ephtracking.cdc.gov/ (accessed on 25 April 2024).
  38. Mahanty, C.; Kumar, R.; Mishra, B.K. Analyses the effects of COVID-19 outbreak on human sexual behaviour using ordinary least-squares based multivariate logistic regression. Qual. Quant. 2021, 55, 1239–1259. [Google Scholar] [CrossRef]
  39. Kim, J.H. Multicollinearity and misleading statistical results. Korean J. Anesth. 2019, 72, 558–569. [Google Scholar] [CrossRef] [PubMed]
  40. Klee, E.W. Data mining for biomarker development: A review of tissue specificity analysis. Clin. Lab. Med. 2008, 28, 127–143. [Google Scholar] [CrossRef]
  41. Oshan, T.M.; Li, Z.; Kang, W.; Wolf, L.J.; Fotheringham, A.S. mgwr: A Python Implementation of Multiscale Geographically Weighted Regression for Investigating Process Spatial Heterogeneity and Scale. ISPRS Int. J. Geo-Inf. 2019, 8, 269. [Google Scholar] [CrossRef]
  42. Ku, B.S.; Druss, B.G. Associations Between Primary Care Provider Shortage Areas and County-Level COVID-19 Infection and Mortality Rates in the USA. J. Gen. Intern. Med. 2020, 35, 3404–3405. [Google Scholar] [CrossRef] [PubMed]
  43. Janke, A.T.; Mei, H.; Rothenberg, C.; Becher, R.D.; Lin, Z.; Venkatesh, A.K. Analysis of Hospital Resource Availability and COVID-19 Mortality Across the United States. J. Hosp. Med. 2021, 16, 211–214. [Google Scholar] [CrossRef]
  44. Kadri, S.S.; Sun, J.; Lawandi, A.; Strich, J.R.; Busch, L.M.; Keller, M.; Babiker, A.; Yek, C.; Malik, S.; Krack, J.; et al. Association Between Caseload Surge and COVID-19 Survival in 558 U.S. Hospitals, March to August 2020. Ann. Intern. Med. 2021, 174, 1240–1251. [Google Scholar] [CrossRef]
  45. Ranney, M.L.; Griffeth, V.; Jha, A.K. Critical Supply Shortages—The Need for Ventilators and Personal Protective Equipment during the COVID-19 Pandemic. N. Engl. J. Med. 2020, 382, e41. [Google Scholar] [CrossRef]
  46. Yu, K.; Zhang, Q.; Wei, Y.; Chen, R.; Kan, H. Global association between air pollution and COVID-19 mortality: A systematic review and meta-analysis. Sci. Total Environ. 2024, 906, 167542. [Google Scholar] [CrossRef]
  47. Sheppard, N.; Carroll, M.; Gao, C.; Lane, T. Particulate matter air pollution and COVID-19 infection, severity, and mortality: A systematic review and meta-analysis. Sci. Total Environ. 2023, 880, 163272. [Google Scholar] [CrossRef]
  48. Biswas, M.; Rahaman, S.; Biswas, T.K.; Haque, Z.; Ibrahim, B. Association of Sex, Age, and Comorbidities with Mortality in COVID-19 Patients: A Systematic Review and Meta-Analysis. Intervirology 2020, 64, 36–47. [Google Scholar] [CrossRef]
  49. Graziani, D.; Soriano, J.B.; Del Rio-Bermudez, C.; Morena, D.; Diaz, T.; Castillo, M.; Alonso, M.; Ancochea, J.; Lumbreras, S.; Izquierdo, J.L. Characteristics and Prognosis of COVID-19 in Patients with COPD. J. Clin. Med. 2020, 9, 3259. [Google Scholar] [CrossRef] [PubMed]
  50. Richardson, S.; Hirsch, J.S.; Narasimhan, M.; Crawford, J.M.; McGinn, T.; Davidson, K.W.; the Northwell COVID-19 Research Consortium; Barnaby, D.P.; Becker, L.B.; Chelico, J.D.; et al. Presenting Characteristics, Comorbidities, and Outcomes Among 5700 Patients Hospitalized With COVID-19 in the New York City Area. JAMA 2020, 323, 2052–2059. [Google Scholar] [CrossRef]
  51. Auld, S.C.; Caridi-Scheible, M.; Blum, J.M.; Robichaux, C.; Kraft, C.; Jacob, J.T.; Jabaley, C.S.; Carpenter, D.; Kaplow, R.; Hernandez-Romieu, A.C.; et al. ICU and ventilator mortality among critically ill adults with COVID-19. medRxiv 2020. [Google Scholar] [CrossRef]
  52. Hogg-Graham, R.; Carman, A.; Mays, G.P.; Zephyr, P.M.D. Geographic Variation in the Structure of Kentucky’s Population Health Systems: An Urban, Rural, and Appalachian Comparison. J. Appalach. Health 2020, 2, 14–25. [Google Scholar] [CrossRef] [PubMed]
  53. Ambrosino, N.; Bertella, E. Lifestyle interventions in prevention and comprehensive management of COPD. Breathe 2018, 14, 186–194. [Google Scholar] [CrossRef] [PubMed]
  54. Croft, J.B.; Wheaton, A.G.; Liu, Y.; Xu, F.; Lu, H.; Matthews, K.A.; Cunningham, T.J.; Wang, Y.; Holt, J.B. Urban-Rural County and State Differences in Chronic Obstructive Pulmonary Disease—United States, 2015. MMWR Morb. Mortal. Wkly. Rep. 2018, 67, 205–211. [Google Scholar] [CrossRef]
  55. Cunningham, M.; Patel, K.; McCall, T.; Hall, K.; Garofalini, C.; Lee, J.; Alford, A. 2022 National Profile of Local Health Departments; National Association of County and City Health Officials: Washington, DC, USA, 2024. [Google Scholar]
  56. Leider, J.P.; Meit, M.; McCullough, J.M.; Resnick, B.; Dekker, D.; Alfonso, Y.N.; Bishai, D. The State of Rural Public Health: Enduring Needs in a New Decade. Am. J. Public Health 2020, 110, 1283–1290. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Shakib, S.H.; Little, B.B.; Karimi, S.M.; Goldsby, M. COVID-19 Mortality Among Hospitalized Medicaid Patients in Kentucky (2020–2021): A Geospatial Study of Social, Medical, and Environmental Risk Factors. Atmosphere 2025, 16, 684. https://doi.org/10.3390/atmos16060684

AMA Style

Shakib SH, Little BB, Karimi SM, Goldsby M. COVID-19 Mortality Among Hospitalized Medicaid Patients in Kentucky (2020–2021): A Geospatial Study of Social, Medical, and Environmental Risk Factors. Atmosphere. 2025; 16(6):684. https://doi.org/10.3390/atmos16060684

Chicago/Turabian Style

Shakib, Shaminul H., Bert B. Little, Seyed M. Karimi, and Michael Goldsby. 2025. "COVID-19 Mortality Among Hospitalized Medicaid Patients in Kentucky (2020–2021): A Geospatial Study of Social, Medical, and Environmental Risk Factors" Atmosphere 16, no. 6: 684. https://doi.org/10.3390/atmos16060684

APA Style

Shakib, S. H., Little, B. B., Karimi, S. M., & Goldsby, M. (2025). COVID-19 Mortality Among Hospitalized Medicaid Patients in Kentucky (2020–2021): A Geospatial Study of Social, Medical, and Environmental Risk Factors. Atmosphere, 16(6), 684. https://doi.org/10.3390/atmos16060684

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop