key: cord-1056556-a5oz6joh
authors: Chhiba, Krishan D.; Patel, Gayatri B.; Vu, Thanh Huyen T.; Chen, Michael M.; Guo, Amina; Kudlaty, Elizabeth; Mai, Quan; Yeh, Chen; Muhammad, Lutfiyya N.; Harris, Kathleen E.; Bochner, Bruce S.; Grammer, Leslie C.; Greenberger, Paul A.; Kalhan, Ravi; Kuang, Fei Li; Saltoun, Carol A.; Schleimer, Robert P.; Stevens, Whitney W.; Peters, Anju T.
title: Prevalence and characterization of asthma in hospitalized and non-hospitalized patients with COVID-19
date: 2020-06-15
journal: J Allergy Clin Immunol
DOI: 10.1016/j.jaci.2020.06.010
sha: 5f2f10594f0967be91306a6b6f711e1190a30131
doc_id: 1056556
cord_uid: a5oz6joh

Abstract: Background The Centers for Disease Control and Prevention advises that patients with moderate-to-severe asthma belong to a high-risk group that is susceptible to severe COVID-19. However, the association between asthma and COVID-19 has not been well-established. Objective The primary objective was to determine the prevalence of asthma among COVID-19 patients in a major U.S. health system. We assessed the clinical characteristics and comorbidities in asthmatic and non-asthmatic COVID-19 patients. We also determined the risk of hospitalization associated with asthma and/or inhaled corticosteroid use. Methods Medical records of patients with COVID-19 were searched by a computer algorithm (March 1–April 15, 2020), and chart review was used to validate the diagnosis of asthma and medications prescribed for asthma. All patients were PCR-confirmed COVID-19. Demographics and clinical features were characterized. Regression models were used to assess the associations between asthma and corticosteroid use and the risk of COVID-19-related hospitalization. Results Of 1,526 patients identified with COVID-19, 220 (14%) were classified as having asthma. Asthma was not associated with an increased risk of hospitalization (RR of 0.96 [95%CI: 0.77-1.19]) after adjusting for age, sex, gender, and comorbidities. The ongoing use of ICS did not increase the risk of hospitalization in a similar adjusted model (RR of 1.39 [95%CI: 0.90-2.15]). Conclusions Despite a substantial prevalence of asthma in our COVID-19 cohort, asthma was not associated with an increased risk of hospitalization. Similarly, the use of ICS with or without systemic corticosteroids was not associated with COVID-19-related hospitalization.

severe COVID-19. However, the association between asthma and COVID-19 has not 48 been well-established. 49 50 Objective: The primary objective was to determine the prevalence of asthma among 51 COVID-19 patients in a major U.S. health system. We assessed the clinical 52 characteristics and comorbidities in asthmatic and non-asthmatic COVID-19 patients. 53

We also determined the risk of hospitalization associated with asthma and/or inhaled 54 corticosteroid use. 55 56 Methods: Medical records of patients with COVID-19 were searched by a computer 57 algorithm (March 1-April 15, 2020), and chart review was used to validate the diagnosis 58 of asthma and medications prescribed for asthma. All patients were PCR-confirmed 59 COVID-19. Demographics and clinical features were characterized. Regression models 60 were used to assess the associations between asthma and corticosteroid use and the 61 

Severe acute respiratory syndrome coronavirus 2 113

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV 2) is a novel 115 betacoronavirus that was first detected in December 2019. The coronavirus disease 116 2019 (COVID-19) has rapidly spread globally causing severe pneumonia along with 117 additional complications including death in the most severely affected individuals. 118

Community spread likely has occurred rapidly because the virus transmits easily, even 119 in asymptomatic patients, and remains viable in respiratory droplets and fomites. 1 Three 120 months after first emerging, fueled by community transmission, there were 121 approximately 2.6 million cases reported globally -including 900,000 cases in the 122 United States and 40,000 cases in Illinois according to the Centers for Disease Control 123

and Prevention (CDC). The outcomes of COVID-19 are worsened by several 124 comorbidities, including hypertension, chronic obstructive pulmonary disease, diabetes 125 mellitus, cardiovascular disease, and obesity. 2, 3 Whether asthma stands among these 126 exacerbating factors requires further study. suggest that the prevalence of asthma in the COVID-19 population in China was <1%. 10, 11 The reported prevalence of asthma in patients with 

Data collected from RT-PCR-confirmed COVID-19 patients (N=1,542) were 177 subsequently stratified based on the presence (N=236) or absence (N=1,306) of 1). Manual chart review of all asthmatics was then performed to confirm a diagnosis of 180 asthma. The criteria used to classify asthma included either a physician diagnosis of 181 asthma or self-reported history of asthma. Patients with a diagnosis of childhood 182 asthma (N=16) but no diagnosis of asthma as an adult were excluded. 183 184

Automated chart review was performed to identify clinical characteristics including age, 186 gender, race/ethnicity, smoking status, and obesity (body mass index (BMI) ≥30). ICD-9 187 and ICD-10 codes were used to identify clinical comorbidities including hypertension 188 Table 1 ). 193

For each patient with asthma, a manual chart review was performed to document a 196 When available, laboratory measurements including white blood cell counts, absolute 202 eosinophil counts, absolute lymphocyte counts, platelet counts as well as ferritin, lactate 203 dehydrogenase (LDH), D-dimer, creatinine, and C-reactive protein (CRP) levels were 204 evaluated in each study patient at the time of COVID-19 diagnosis. If more than one lab 205 value was available, the first value obtained up to 4 weeks after the diagnosis of 206 COVID-19 was used for this study. 207 208

Demographic data and clinical characteristics were computed for all included 210 participants and compared using Chi-square tests. Differences in laboratory values 211

were compared using non-parametric Mann-Whitney tests or Kruskal-Wallis, where 212 appropriate. Poisson regression models were used to calculate the relative risk (RR) of 213 hospital admission (inpatient with or without intensive care unit (ICU) versus outpatient). 214

For the analysis samples with all COVID-19 patients (N=1,526), the association 215 between asthma and COVID-19 hospitalization was determined. Model covariables 216 included: (1) age, gender, and race/ethnicity (Model 1), and (2) age, gender, 217 race/ethnicity, smoking status, and comorbidities (Model 2). Comorbidities included 218 obesity, HTN, DM, OSA, CAD and COPD, AR, rhinosinusitis, and immunodeficiency. 219

Similar models were used for the analysis sample of only COVID-19 patients with 220 asthma (N=220) in which the association between ICS use and hospitalization was 221 tested. There were only a small number of patients (N=15) among 220 asthmatics 222 receiving systemic corticosteroids. In a sensitivity analysis, we repeated the analysis 223 after excluding these 15 patients to examine if systemic corticosteroids may have any impact on the association of using ICS with the risk of hospitalization. Data were 225 displayed and statistics were performed using SAS statistical software version 9.4 (SAS 226

Institute Inc., Cary, NC) and GraphPad Prism 8 (GraphPad Software, La Jolla, CA). 227

An automated electronic review of patient medical records identified 1,837 patients with 231 an ICD-10 diagnosis code of COVID-19 in our system between March 1, 2020 and April 232 15, 2020. Of these, 1,542 (84%) had confirmed disease by RT-PCR and were included 233 in subsequent analyses ( Table 1) 

We assessed and compared various demographic and clinical characteristics in COVID-242 19 patients with and without comorbid asthma (Table 1 ). The majority (55.3%) of 243 COVID-19 patients were between 40-69 years of age regardless of asthma status. 244

Slightly more than half (53%) of all COVID-19 patients were female with a significant 245 female predominance in the asthma cohort (70.9%). The primary race/ethnicities of the 246 total COVID-19 cohort were non-Hispanic White (42.1%), non-Hispanic African 247

American (23.5%), and Hispanic or Latino (21.2%). Within those with asthma, the 248 percentage of patients identifying as non-Hispanic African American was 35.5% which 249 was significantly higher compared to 21.4% in the non-asthma cohort. Although representation was even higher in the non-asthma group (22.7%). Hospitalization rate 252 and mortality did not significantly differ between COVID-19 patients with asthma or 253 without asthma. severity, 15 were all significantly lower in COVID-19 patients with asthma compared to Chhiba and Patel et al. 15 COVID-19 patients without asthma (P < 0.0001, 0.048, 0.0004, respectively). D-dimer 275 was also lower in asthmatics compared to non-asthmatics although this was not 276 statistically significant (P = 0.052). Absolute lymphocyte counts (x1000/µL) (median 277 

We used two different models to evaluate if asthma was associated with an increased 283 risk of hospitalization for COVID-19. After adjusting for baseline age, gender, and 284 race/ethnicity (Model 1), asthmatics did not have a higher risk of COVID-19-related 285 hospitalization compared to non-asthmatics (RR 1.01; 95%CI: 0.83-1.24) ( Table 2 ). These demographic risks for hospitalization were present 298 irrespective of asthma status. Even when adjusting for comorbidities using Model 2, 299

Hispanics continued to be at increased risk of hospitalization due to COVID-19 (RR 300 1.35; 95%CI: 1.12-1.63; Supplemental Table 2 ). However, in this model, non-Hispanic 301 African Americans no longer had a significantly elevated relative risk of hospitalization 302 compared to non-Hispanic White patients (Supplemental Table 2 ). Age (≥70 years), 303 male gender, and comorbid diagnoses of diabetes (RR 1.16; 95% CI: 1.00-1.36), and 304 OSA (RR 1.23; 95% CI: 1.01-1.49) also elevated the relative risk of COVID-19 305 hospitalization regardless of asthma status (Supplemental Table 2 , Model 2). 306

Rhinosinusitis was associated with a significantly lower risk of hospitalization compared 307 to the absence of rhinosinusitis (RR 0.78; 95% CI: 0.61-0.99) (Supplemental Table 2 , 308

Model 2). Patients with allergic rhinitis also showed a trend towards lower 309 hospitalization although not statistically significant (RR 0.83; 95% CI: 0.64-1.07). These 310 associations with rhinosinusitis and allergic rhinitis were observed in COVID-19 patients 311 with or without asthma. 312 313

We also explored the relationship between inhaled corticosteroids and the risk of 315 hospitalization in COVID-19 patients with asthma using two different statistical models. 316

Over half (52%, N=114) of COVID-19 patients with asthma were not prescribed either 317 ICS or ICS/LABA at the time of diagnosis (Supplemental Table 3 ). Whereas, among 318 those with asthma, 11.8% and 36.4% had documentation of ICS (N=26) or ICS/LABA among COVID-19 patients by the level of medical care is shown in Figure 4 . Although, 321 the percentage of COVID-19 patients with asthma stratified by ICS use and level of 322 medical care was not statistically different (P=0.10), the proportion of patients not using 323 ICS or ICS/LABA was highest (57.1%) in the outpatient group, and lowest (31.6%) in 324 the ICU group. The proportion of patients using ICS/LABA was lowest in the outpatient 325 group (28.6%), and highest in the ICU group (57.9%). (Table 3 ). The individual baseline risk factors used to adjust for relative risk assessing 332 ICS use and COVID-19-related hospital admission are listed in Supplemental Table 4 . and the use of ICS did not increase or decrease the risk of COVID-19 hospitalization in was receiving an asthma-related biologic (omalizumab). This patient required an ICU 344 stay and was intubated for COVID-19 but was successfully discharged after 16 days of 345 hospitalization. 346

To the best of our knowledge, this is the first comprehensive cohort study of patients 349 with COVID-19 and comorbid asthma. In this study, asthma was present in 14.4% of 350 COVID-19 patients which included both hospitalized and non-hospitalized patients. 351

Compared to the general U.S. and metropolitan Chicago population which is estimated 352

to have an asthma prevalence of 8-9% and 9.5%, respectively, asthma is enriched in 353 our COVID-19 population. 4, 5, 16 Among only hospitalized patients with COVID-19 in this 354 cohort, the prevalence of asthma was 13.5%, which supports recent published U.S. 355 data observing asthma prevalence between 7.4%-17% in COVID-19 hospitalized 356 patients. 2, 12, 13, 17 This is in stark contrast to the low prevalence of asthma (<1%) noted 357 in China. 10, 11 Geographic differences in the frequency of asthma or methods of 358 ascertainment may be contributing to these heterogenous findings. 359 360 Importantly, despite the high prevalence of asthma in our study, we observed no 361 significant difference in risk of hospitalization or mortality due to COVID-19 in asthmatic 362 compared to non-asthmatic patients. The overall mortality rate (4.7%) in our COVID-19 363 population aligned closely with the national mortality rate of 6.0% during this time period 364 as published on the Johns Hopkins Coronavirus Resource Center (May 6, 2020). In this 365 cohort, the mortality rate (3.6%) in the COVID-19 population with asthma at the time of 366 this study was not different than the mortality rate in the COVID-19 population without 367 asthma (4.9%). 368

Well-established comorbidities that are associated with COVID-19 were present in this 370 cohort of asthma ( Figure 2) . Interestingly, patients with asthma and COVID-19, 371 compared to COVID-19 patients without asthma, had an increased prevalence of 372 multiple comorbidities. Previous studies have shown that obesity, OSA, and GERD are 373 associated with asthma. [18] [19] [20] In the general COVID-19 cohort, DM and OSA were 374 associated with a higher risk of hospitalization; however, this was no longer true when 375 evaluating the asthma subgroup alone. Further investigation is needed to determine 376 why these comorbidities, despite being more prevalent in asthmatics, do not appear to 377 worsen COVID-19-related outcomes. 378 379 Dramatic racial disparities have been reported during the COVID-19 pandemic and this 380 was true in our study. Non-Hispanic African Americans made up almost one-quarter of 381 our overall COVID-19 cohort despite the 6.1% prevalence of African Americans in our 382 healthcare system. Moreover, African Americans were disproportionately higher in the 383 asthma group (36%) compared to the non-asthma group (21%). Of the COVID-19 384 patients with asthma in this study, 12.7% were Hispanic or Latino. This data is in 385 contrast to the national findings. According to the CDC, African Americans and 386

Hispanics comprise 9.6% and 6.0% of the adult asthma population, respectively. 5 After 387 controlling for age, sex, and race, African Americans had a higher risk of COVID-19-388 related hospitalization in the general COVID-19 cohort. Depending on the model used, 389 the risk of COVID-19-related hospitalization was even higher in an adjusted analysis for 390 the Hispanic or Latino population (35-44%).

The assessment of laboratory values demonstrates that patients with asthma had 393 significantly lower levels of ferritin, CRP, and LDH, compared to non-asthma patients. 394

These are markers of disease severity in COVID-19. This is the first report to our 395 knowledge to describe a potential decreased inflammatory burden in COVID-19 patients 396 with comorbid asthma, despite these patients having higher levels of other comorbid 397 diseases compared to non-asthmatics. These findings suggest that underlying immune 398 modulation either due to asthma or asthma treatment may have a mitigating effect on 399 COVID-19, but more studies are needed to understand this. 400 401 Interestingly, asthma did not increase the risk of hospitalization after adjusting for 402

covariates. This is notable as it has been anticipated that underlying chronic lung 403 disease such as asthma, which are typically triggered by a viral illness, would place 404 these patients at increased risk of severe exacerbations. 21 The role of ICS in asthma 405 patients and COVID-19 is not established and has brought concern to many patients. 22, 406 23 Almost half (48%) of the patients with asthma were using ICS before COVID-19 in our 407 study. After controlling for baseline risk factors, the use of ICS did not increase the risk 408 of COVID-19-related hospitalization. In this study, only fifteen patients were prescribed 409 systemic corticosteroids before diagnosis, so this limits our ability to make any 410 conclusion specifically regarding oral corticosteroid use in COVID-19. However, it is 411 reassuring that in the model assessing the risk of ICS, oral corticosteroids did not 412 change the risk of hospitalization. 413

It has been postulated that type 2 immune modulation decreases expression of ACE2, 415 the known receptor for COVID-19 cellular entry. [24] [25] [26] There are several limitations to our study. Data were obtained retrospectively so we are 433 limited to drawing associations rather than causal inferences. Our study population and 434 some of the variables used for analyses were based on ICD codes which may have 435 mis-captured data. To minimize this, we performed chart reviews for the asthma cohort the level of care required for COVID-19. Also, because of the study design, we cannot 438 assume adherence with the prescribed medications. An additional limitation of our study 439 is that we did not assess the contribution of asthma severity or control to COVID-19-440 related hospitalization as we were limited by our study design. Although we cannot 441 make inferences based on asthma severity, COVID-19-associated level of care (ICU vs. 442 non-ICU) was not significantly different between patients prescribed ICS or ICS/LABA 443 and those not on ICS or ICS/LABA. Our findings are based on data collected between 444

March 1-April 15 (with the exception of mortality assessed until April 30, 2020) and 445 might change as additional data is collected after the study period. While it may be 446 possible that patients with asthma were more likely to be tested as asthma is a chronic 447 lung disease, our asthma prevalence data was similar to the prevalence reported by the 448 Morbidity and Mortality Weekly Report from the CDC during this study period. 13 Lastly, 449 widespread COVID-19 testing was not available during our data collection period so 450 selected patients may represent a bias towards more severe COVID-19 disease. 451

In summary, we found that asthma prevalence was 14% in our cohort of COVID-19 453 patients. Despite a high prevalence of comorbid diseases that are associated with 454 COVID-19 severity, it is reassuring that neither asthma nor the use of ICS was 455 associated with an increased risk of COVID-19 hospitalization. With this in mind, 456 physicians need to be vigilant of older patients, those with comorbidities (especially DM 457

and OSA based on this study), African Americans, and Hispanics who present with 458 COVID-19 symptoms since they are at increased risk of hospitalization. This is true in 459 the general population as well as in asthmatics, according to this study. Immunodeficiency includes patients with a diagnosis of immunodeficiency, antibody 555 deficiency, or IgA deficiency. Obesity was determined based on reported BMI (≥ 30). 556

For two patients who were younger than 20 years old, the weight-for-age percentile was 557 (1) not taking ICS, (2) using ICS alone or (3) using ICS/LABA at the time of COVID-19 577 diagnosis. Oral steroids were used by 15 out of 220 asthma patients: outpatient (N=7), 578 inpatient -no ICU (N=8) and inpatient -ICU (N=0). Bars represent mean ± SEM. 579

Statistics were performed using Chi-square test (P = 0.10). Percentage of COVID-19 asthmatics by level of care 

exacerbation of asthma and chronic obstructive pulmonary disease

Regional, age and respiratory-secretion-488 specific prevalence of respiratory viruses associated with asthma exacerbation: a 489 literature review

Epidemiological and 491 clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan

China: a descriptive study

Risk factors for severity and 494 mortality in adult COVID-19 inpatients in Wuhan

Covid-19 in Critically Ill Patients in the Seattle Region -Case Series

Hospitalization Rates and Characteristics of Patients Hospitalized with 499

Laboratory-Confirmed Coronavirus Disease 2019-COVID-NET, 14 States

Characteristics of Hospitalized Adults 502

With COVID-19 in an Integrated Health Care System in California

Mild versus severe COVID-19: laboratory markers

Chicago Department of Public Health HCS

Chicago Health

Characteristics of Hospitalized Adults 510

With COVID-19 in an Integrated Health Care System in California

 Chhiba and Patel et al. 24 investigation is necessary to understand the possible protective role of type 2 461 inflammation in asthma and