COVID-19 waves in an urban setting 2020–2022: an electronic medical record analysis

Background Global and national surveillance efforts have tracked COVID-19 incidence and clinical outcomes, but few studies have compared comorbid conditions and clinical outcomes across each wave of the pandemic. We analyzed data from the COVID-19 registry of a large urban healthcare system to determine the associations between presenting comorbidities and clinical outcomes during the pandemic. Methods We analyzed registry data for all inpatients and outpatients with COVID-19 from March 2020 through September 2022 (N = 44,499). Clinical outcomes were death, hospitalization, and intensive care unit (ICU) admission. Demographic and clinical outcomes data were analyzed overall and for each wave. Unadjusted and multivariable logistic regressions were performed to explore the associations between age, sex, race, ethnicity, comorbidities, and mortality. Results Waves 2 and 3 (Alpha and Delta variants) were associated with greater hospitalizations, ICU admissions, and mortality than other variants. Chronic pulmonary disease was the most common comorbid condition across all age groups and waves. Mortality rates were higher in older patients but decreased across all age groups in later waves. In every wave, mortality was associated with renal disease, congestive heart failure, cerebrovascular disease, diabetes, and chronic pulmonary disease. Multivariable analysis found that liver disease and renal disease were significantly associated with mortality, hospitalization, and ICU admission, and diabetes was significantly associated with hospitalization and ICU admission. Conclusion The COVID-19 registry is a valuable resource to identify risk factors for clinical outcomes. Our findings may inform risk stratification and care planning for patients with COVID-19 based on age and comorbid conditions.


Introduction
The COVID-19 pandemic continues, with more than 750 million individuals infected and more than 7 million deaths reported worldwide as of June 22, 2023 (1).The pandemic has been marked by waves of infection, driven by the evolution of the SARS-CoV-2 virus over time.Global and national surveillance efforts have documented the incidence of COVID-19 and its clinical outcomes, as well as the changes in dominant circulating SARS-CoV-2 variants in each wave (2).Throughout the pandemic, several studies have compared outcomes between early or late phases or between variant waves, but few studies have examined outcomes longitudinally across waves of SARS-Cov-2 variants.Four large studies have reported differences in inpatient outcomes (3), perinatal outcomes (4), ECMO-related outcomes (5), and outcomes in children diagnosed with multisystem inflammatory syndrome (MIS-C) (6) across multiple SARS-CoV-2 waves.A review by Lin et al. (7) found wild-type SARS-CoV-2 infection has less severe outcomes (e.g., intensive care unit [ICU] admission, hospitalization, and death) compared to other variants including Alpha, Delta, Gamma, and Beta.Several studies have found Omicron infection to have less severe outcomes including hospital admissions, hospital admissions for symptoms, ICU admission, ventilation, and death, compared to Delta (8)(9)(10)(11)(12)(13)(14).Studies are needed analyze outcomes by variant and risk factors.The primary objective of this study was to determine the associations between comorbidities in the Charlson Comorbidity Index (15) and mortality among patients with COVID-19 seen in an urban healthcare system in the United States, across all waves and for each wave of the pandemic from March 2020 to September 2022.As a secondary objective, we aimed to evaluate these same factors and their association with hospital admission and ICU admission.

Study participants and registry development
In partnership with Abbott Pandemic Defense Coalition (APDC), Rush University Medical Center (RUMC, Chicago, IL) established a COVID-19 registry to collect RUMC electronic health record (EHR) data on infections, patient demographics, and clinical outcomes.The Abbott Pandemic Defense Coalition (APDC) is a global multisector scientific and public health partnership whose primary objective is the early detection and mitigation of infectious disease threats of pandemic potential (16).
Each patient in the registry is assigned an identification number to protect privacy and deidentified data from the COVID-19 registry was analyzed for this study.The COVID-19 registry was developed utilizing Epic electronic health records data at RUMC.Specific data elements were first determined to develop the structure of the registry.SQL was used to conduct data extraction and organization of the data.The data was mapped to the Observational Medical Outcomes Partnership Common Data Model (OMOP) Common Data Model1 release 5.3.1 (17).The registry includes variables such as demographic characteristics, conditions, and COVID-19 testing information and related outcomes.This study was approved by the Institutional Review Board of Rush University Medical Center.
We analyzed data in the registry for all inpatients and outpatients who were under investigation for COVID-19 from March 2020 through September 2022 (N = 44,499).Patients were considered under investigation if they were tested for COVID-19 when they presented for care at RUMC.COVID-19 positivity was defined as any documented infection ("Infection of upper respiratory tract caused by 2019 novel coronavirus": OMOP condition concept ID 37310286) (17).COVID-19 positivity was defined as any documented infection ("Infection of upper respiratory tract caused by 2019 novel coronavirus": OMOP condition concept ID 37310286) (17).History of comorbidities using the Charlson Comorbidity Index (CCI) were identified if the patient ever had a corresponding International Classification of Diseases (ICD) code for the condition.The clinical outcomes of death, hospitalization, and intensive care unit (ICU) admission were recorded.Inpatient deaths were defined as deaths associated with an inpatient or emergency room visit.Outpatient deaths were defined as deaths associated with an outpatient or patient vehicle encounter, with the date of death identified through the patient EHR.Other demographic and clinical information were also captured for each patient.

Statistical analysis
Descriptive statistics were used to summarize COVID-19 registry data on patient demographics (age, sex, race, and ethnicity), comorbidities, hospitalization, ICU admissions, and deaths across all waves (overall) and by wave.Categorical variables were represented by proportion and continuous variables by medians and interquartile ranges (IQRs).Unadjusted and forward stepwise multivariable logistic regressions were performed to explore the associations between demographic variables (age, sex, race, ethnicity), comorbidities (as defined by the Charlson Comorbidity Index: congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatic disease, peptic ulcer disease, mild liver disease, diabetes without chronic complications, diabetes with chronic complications, hemiplegia or paraplegia, renal disease, cancer (any malignancy), moderate/severe liver disease, metastatic solid tumor, AIDS/HIV), and the three outcomes of hospitalization, ICU admissions, and deaths.Models reported odds ratios (OR) and 95% confidence intervals (CI) and a p < 0.05 was used to identify statistically significant variables for forward selection.Age-stratified analyses were conducted to identify differences in demographics and comorbidities for all COVID-19 patients and those reported as a death across age groups (18-49 years, 50-69 years, and 70-90 years) overall and by wave.No imputations were made for missing data and all analyses were conducted using R4.1.2.

Hospitalization, ICU admission, and
death in each wave of the pandemic

Mortality by wave
Most deaths occurred in Wave 1 (n = 627, 59.9%), followed by Wave 4 (Omicron BA.1 variant; n = 204, 19.5%) (Table 4).The mortality rate ranged from 3.2% in Wave 3 to 0.8% in Wave 5 and 0.7% in Wave 6, with similar rates in Waves 1, 2, and 4 (2.7, 2.6, and 2.0%, respectively).In each wave, the majority of deaths occurred among inpatients (Figure 2A) and those 50-69 and 70-90 years of age (Figure 2B).The mortality rate was consistently highest among those 70-90 years of age in each wave (Figure 3A).Deaths were slightly more frequent in males than females in Waves 1, 3, and 5 (Figure 2C), with a consistently higher mortality rate among males in each of the 6 waves (Figure 3B).The proportion of deaths in Waves 1 and 2 were similar for White and Black or African American patients; in Waves 3-6, White patients represented the highest proportion of deaths (Figure 2D).Mortality rates were slightly higher for White patients compared to other racial groups in most waves, with the exception of Wave 1, where the mortality rate was highest among Native Hawaiian or Other Pacific Islanders and Waves 3 and 4, where the mortality rate was highest among American Indians or Alaska Natives (Figure 3C).Note however, that these 2 groups made up a very small proportion of patients in the registry.In every wave of the pandemic, the most common comorbid conditions among patients reported as a death were renal disease, congestive heart failure, cerebrovascular diseases, diabetes, and chronic pulmonary disease, with similar frequencies in each wave (Figure 4).

Age-stratified analyses for deaths across all waves
Because the mortality rate was higher among older patients and those with certain comorbidities, we examined the association between age and comorbid conditions in all COVID-19 positive patients in the registry and those reported as a death.

Comorbidities by age
The frequency of comorbid conditions increased with increasing age (Figure 5A).Chronic pulmonary disease was the most common comorbidity in all three age groups across all waves of the pandemic.Chronic pulmonary disease was the most common comorbidity in patients 18-49 years of age, while diabetes without chronic    3.5 Age-stratified analyses in each wave

Comorbidities by age and by wave
In each wave of the pandemic, chronic pulmonary disease was consistently the most common comorbid condition in COVID-19positive patients 18-49 years of age, and diabetes without chronic complications was the second most common comorbidity (Figure 6).In Wave 1, patients 50-69 years of age were most likely to have diabetes without chronic complications or chronic pulmonary disease; in subsequent waves, several other comorbid conditions were seen with increased prevalence, notably diabetes with chronic complications, renal disease, and any cancer.Patients in the oldest group (70-90 years) had several comorbidities with similarly high frequencies in each wave; congestive heart failure, chronic pulmonary disease, renal disease, and cerebrovascular disease were consistently high in each wave.

Mortality for comorbidities by age and by wave
Among patients reported as a death, age-related comorbid conditions were slightly more frequent in the oldest age group in Waves 1-4, with consistent patterns of comorbid conditions among patients 50-69 and 70-90 years of age (Figure 7).Comorbid conditions within the youngest age group (18-49 years) varied widely with each pandemic wave.By Waves 5 and 6, the most common comorbid conditions were significantly different in each age group.In Wave 5, the youngest group of patients reported as a death were most likely to have renal disease, followed by diabetes with chronic complications, chronic pulmonary disease, cerebrovascular disease, and congestive heart failure.In contrast, patients 50-69 years of age were most likely to have peripheral vascular disease or congestive heart failure in Wave 5, whereas patients 70-90 years of age were most likely to have renal disease, diabetes with chronic complications, and congestive heart failure.Patterns of comorbid conditions in Wave 6 were similar for patients reported as a death in the 2 older age groups, though more patients 50-69 years of age presented with metastatic solid tumors and those 70-90 years of age most often presented with renal disease and congestive heart failure.

Multivariable logistic regression models
Multivariable logistic regression models were used to estimate the association of key factors with each of the outcomes adjusted for comorbidities, sex, ethnicity, and age (Figure 8).Liver disease was significantly associated with all outcomes: increased hospitalization (OR = 3.52, 95% CI 2.59-4.83),ICU admission (OR = 3.34, 95% CI 2.31-4.76),and mortality (OR = 5.68, 95% CI 3.84-8.27).Patients with metastatic solid tumors were more likely to be reported as a death (OR = 5.75, 95% CI 4.48-7.34).Dementia, being Black or African American, hemiplegia or paraplegia, and renal disease were found to be significantly associated with hospitalizations.Hemiplegia or paraplegia, renal disease, and diabetes with and without chronic complications were significantly associated with ICU admissions.Renal disease, myocardial infarction, and diabetes with and without chronic complications were significantly associated with mortality.

Discussion
This study leveraged the rich dataset from the RUMC COVID-19 registry to analyze the association between age-related comorbid conditions and death across and within each wave of the COVID-19 pandemic since March 2020.While several studies have examined risk factors associated with hospitalizations, ICU admissions, and death, few have evaluated differences in these factors across distinct waves of the pandemic (18)(19)(20)(21)(22).The Alpha and Delta variants were associated with greater hospitalizations, ICU admissions, and mortality compared to the other variants.

Overall mortality rates
Our finding of higher mortality rates in older patients, and a decrease in mortality across all age groups in later waves, is consistent with national trends (1).The association of mortality with renal disease, congestive heart failure, cerebrovascular disease, diabetes, and chronic pulmonary disease is also consistent with various reports in the US and elsewhere (23)(24)(25)(26).These findings underscore the value of COVID-19 registries.Several studies have illustrated the advantage of adding COVID-19 data points to existing registries in specific populations-such as cancer (27), inflammatory diseases (28), neurological disorders, transplantation (29), and cardiovascular disease (30)-in order to examine predictors of risk for COVID-19 outcomes.Further, evaluation of COVID-19 and comorbidities has been reviewed by several authors and have found significant associations.Hospitalizations and mortality for other respiratory infections were significantly decreased during the pandemic (31).The effect of COVID-19 on STEMI (32) and stoke (33) outcomes have also been demonstrated, resulting in higher in-patient mortality, length of stay and cost of hospitalization.Kapuria et al. found that patients with COVID-19 and cirrhosis had 3 times higher mortality than those without COVID-19 (34) and Pal

Mortality rates by wave
In our age-stratified analyses of patients recorded as a death, we found notable differences in the prevalence of specific comorbid conditions in each age group that changed with each wave of the COVID-19 pandemic.For example, younger patients reported as a death presented with various comorbid conditions in early waves of the pandemic but were most likely to have renal disease in later waves.Older patients recorded as a death also presented with various comorbid conditions in consistent patterns across early waves of the pandemic, but by Wave 5, those 50-69 years of age more often presented with peripheral vascular disease and congestive heart failure and those 70-90 years of age presented with renal disease, diabetes, and congestive heart failure.The mechanisms driving these differences in age-related comorbid conditions in patients reported as a death is unknown but may be related to differences in the predominant variant, previous infection history, vaccination status, or other patient-related factors, and may be important factors to consider when developing interventions (25).

Limitations
Although the registry contained pediatric patients, the modeling of risk factors and mortality focused on an adult population.Several Odds ratio (OR) and 95% CI of the association of key factors with hospitalizations, ICU admissions, and mortality with each of the outcomes adjusted for comorbidities, sex, ethnicity, and age.limitations should be considered and there are several bias sources to consider when using EHR data, particularly related to how data is collected as well as environmental aspects which can influence the quality of data (36).First, the retrospective design of the study has the potential for missing data and reporting bias, especially toward more severe symptoms or outcomes being reported more frequently.
Comorbidities were potentially underreported in the registry, and one-third of patients reported an undisclosed racial group in the registry, which limited our ability to draw strong conclusions about race as a risk factor for mortality across and within individual pandemic waves.The registry did not collect information about vaccination status or measure changes to clinical practice in treating COVID-19 as the pandemic unfolded, although the analysis by wave is a surrogate measure for both important factors.The registry source lacked sequencing data to confirm variants; instead, the predominant variant circulating in Chicago and epidemiological assumptions was used to define waves.There also exists the potential that patients attended other healthcare institutions in the Chicagoland area, thus underreporting death that occurred outside of RUMC.Data in the EHR are susceptible to data coding errors and the subsequent mapping to OMOP common data model could compound this.Comorbid conditions were classified based on any history of the disease, but it could not be ascertained from the data whether the patients were still suffering from these conditions.Given that most of these comorbid conditions are chronic and lifelong, this is a potential but small bias.Information regarding co-infections with other respiratory viruses was also not known.Lastly, no single institution can be generalizable to the general patient population.

Conclusion
The findings of this specific COVID-19 registry within a single localized population were representative of the trends seen in the US; namely, higher mortality rates in older patients and a decrease in mortality across all age groups with later variant waves.The association of mortality with the most at risk populations due to specific comorbidities of renal disease, congestive heart failure, cerebrovascular disease, diabetes, and chronic pulmonary disease was also generalizable to the US.
Future research should evaluate and include the impact of vaccination status and co-infection with other circulating respiratory viruses, as well as the effect of long-COVID on hospitalization, ICU admittance and death.
Overall, the RUMC COVID-19 registry provided a valuable resource for population-based analyses to identify risk factors for death and other outcomes and to understand the evolution with each wave of the pandemic.Our findings may have implications for risk stratification and care planning for patients with COVID-19 based on age and presenting comorbid conditions for subsequent variant waves.
disease as the most common comorbidity in all three age groups (Figure5B).Cerebrovascular disease was the second most common comorbidity for patients 18-49 years of age, whereas congestive heart failure was the second most common comorbidity for patients 50-69 and 70-90 years of age.
Of the 44,499 COVID-19-positive patients in the registry, 1,046 (2.4%) were reported as a death across all waves of the pandemic (

TABLE 2
Characteristics overall and by outcome (N = 44,499).

TABLE 3
Frequency of comorbid conditions among patients with hospitalization, ICU admission, or death (N = 44,499).

TABLE 4
Outcome by wave.