key: cord-0848212-9nqk1gi5
authors: Karlsson Sundbaum, Johanna; Konradsen, Jon R.; Vanfleteren, Lowie E.G.W.; Axelsson Fisk, Sten; Pedroletti, Christophe; Sjöö, Yvonne; Syk, Jörgen; Sterner, Therese; Lindberg, Anne; Tunsäter, Alf; Nyberg, Fredrik; Ekberg-Jansson, Ann; Stridsman, Caroline
title: Uncontrolled asthma predicts severe COVID-19: a report from the Swedish National Airway Register
date: 2022-04-18
journal: Ther Adv Respir Dis
DOI: 10.1177/17534666221091183
sha: a3c6d5fb60c76efe832bc747e4bb8286eec2dccc
doc_id: 848212
cord_uid: 9nqk1gi5

BACKGROUND: Severe asthma increases the risk of severe COVID-19 outcomes such as hospitalization and death. However, more studies are needed to understand the association between asthma and severe COVID-19. METHODS: A cohort of 150,430 adult asthma patients were identified in the Swedish National Airway Register (SNAR) from 2013 to December 2020. Data on body mass index, smoking habits, lung function, and asthma control test (ACT) were obtained from SNAR, and uncontrolled asthma was defined as ACT ⩽19. Patients with severe COVID-19 were identified following hospitalization or in death certificates based on ICD-10 codes U07.1 and U07.2. The Swedish Prescribed Drug register was used to identify comorbidities and data from Statistics Sweden for educational level. Multivariate logistic regression analyses were used to estimate associations with severe COVID-19. RESULTS: Severe COVID-19 was identified in 1067 patients (0.7%). Older age (OR = 1.04, 95% CI = 1.03–1.04), male sex (1.42, 1.25–1.61), overweight (1.56, 1.27–1.91), obesity (2.12, 1.73–2.60), high-dose inhaled corticosteroids in combination with long-acting β-agonists (1.40, 1.22–1.60), dispensed oral corticosteroids ⩾2 (1.48, 1.25–1.75), uncontrolled asthma (1.64, 1.35–2.00), cardiovascular disease (1.20, 1.03–1.40), depression (1.47, 1.28–1.68), and diabetes (1.52, 1.29–1.78) were associated with severe COVID-19, while current smoking was inversely associated (0.63, 0.47–0.85). When comparing patients who died from COVID-19 with those discharged alive from hospital until 31 December 2020, older age, male sex, and current smoking were associated with COVID-19 death. CONCLUSION: Patients with uncontrolled asthma and high disease burden, including increased asthma medication intensity, should be identified as risk patients for severe COVID-19. Furthermore, current smoking is strongly associated with COVID-19 death in asthma.

relatively low, 6 and several meta-analyses show that patients with asthma are not at risk of severe COVID-19-related outcomes such as hospitalization, intensive care, or mortality. 7, 8 It has even been speculated that inhaled corticosteroid treatment might have a protective effect against COVID- 19. 9 Despite the low risk of patients with asthma of acquiring the infection, 5, 6 studies are indicating that severe asthma and non-allergic asthma can increase the risk of severe COVID-19-related outcomes. [10] [11] [12] Yet, there is need for further studies in larger asthma cohorts on factors associated with severe COVID-19 in asthma. These factors could be important when identifying patients at risk of COVID-19 and developing guidelines in asthma-related COVID-19 care. The current study is based on a large and well-characterized cohort of patients with physician-diagnosed asthma within the Swedish National Airway Register (SNAR). 6, 13, 14 In this cohort, patients hospitalized or who died from COVID-19 have been identified, giving a unique opportunity to study the burden of asthma and its association with severe COVID-19.

The Swedish National Airway Register. SNAR is a national quality register launched in 2013 and includes data on patients from primary and secondary care with a physician diagnosis of asthma (children and adults) or chronic obstructive pulmonary disease (COPD). 6, 13, 14 In this study, we included adult patients (⩾18 years) with asthma diagnoses (International Classification of Diseases, version 10; ICD-10 J45) identified in extracted data from SNAR on 1 February 2020. Patients with COPD diagnosis (ICD-10 J44) were excluded from the study population.

Severe COVID-19 detected by national registers. As cases, we identified asthma patients with severe COVID-19 by linking data extracted from SNAR with data obtained up to 31 December 2020 from the National Patient Register (NPR), the Swedish Intensive Care Registry (SIR), and the Swedish Cause of Death Register (SCDR). 15 Those registers provide data about COVID-19 based on the ICD-10 codes U07.1 (COVID-19 confirmed by laboratory testing) and U07.2 (COVID-19 clinically or epidemiologically diagnosed, but laboratory testing is inconclusive or not available). Severe COVID-19 was defined as inpatient care identified as primary discharge diagnosis in NPR or care at the intensive care unit in SIR, or COVID-19 death registered as an underlying cause of death in SCDR. The asthma patients without severe COVID-19 were defined as controls.

Clinical data. Clinical data were obtained from SNAR including the latest uploaded information about body mass index (BMI), smoking habits, Forced Expiratory Volume in 1 s as percent of predicted value (FEV 1 %), and Asthma Control Test (ACT). BMI was used as a continuous variable but also categorized as underweight (BMI <19), normal weight (BMI [19] [20] [21] [22] [23] [24] [25] , overweight (BMI [26] [27] [28] [29] [30] , and obese (BMI > 30). Smoking was categorized as current smokers, ex-smokers (including those who quit smoking during the last 6 months), and non-smokers. The Swedish reference values by Hedenström were used to calculate FEV 1 % predicted, 16, 17 and if post-values were missing, pre-values were used. ACT scores were used as a continuous variable (range, 5-25; poor control to complete control), and ACT ⩽19 was used to define uncontrolled asthma. 18 Data on asthma medications. Data on dispensed asthma medications between 1 January 2019 and 28 February 2020 were retrieved from the Swedish Prescribed Drug Register (SPDR) 14 using Anatomical Therapeutic Chemical (ATC) codes. Asthma medication was classified as short-acting β-agonists (SABA) (R03AC02-03), short-acting muscarinic antagonists (SAMA) (R03BB01), long-acting β-agonists (LABA) (R03AC12, 13, 18, 19) , long-acting muscarinic antagonists (LAMA) (R03BB04-07), inhaled corticosteroids (ICS) (R03BA), LABA-LAMA combinations (R03AL03-07), LABA-ICS combinations (R03AK), LABA-LAMA-ICS combinations (R03AL08,09, 11,12), leukotriene antagonists (LTRA) (R03DC03), acetylcysteine (R05CB01), and oral corticosteroids (OCS) (H02AB01,02,06). OCS were classified into ⩾1 and ⩾2 dispensations. Based on the different ATC codes, and a 4-month dispensed coverage of single or combination medication, dispensed inhaled medications were further divided into groups: (1) SABA and/or SAMA only, ( Data on level of education, asthma healthcare utilization, and comorbidities. Data on level of education were obtained from Statistics Sweden 19 and classified into primary school (usually 9 years), secondary school (usually 12 years), and tertiary education (longer than 12 years). Prior inpatient or secondary care due to asthma was identified from the NPR, defined as patients who had been hospitalized or had a secondary care visit in 2019 with a primary ICD-10 code for asthma (J45 or J46). Medication-treated comorbidities were identified through the SPDR with ATC codes for dispensed medication between 1 January 2019 and 28 February 2020 and classified into cardiovascular disease (C01-03, C08, C09), depression (N06), diabetes (A10A, A10B), and rhinitis (R01A).

Data analyses were conducted using SAS 9.4 for Windows (SAS Institute Inc, Cary, NC, USA). Baseline data of cases and controls were expressed as means, standard deviations (SD), and percentage of relative frequencies. For comparative analyses between the cases and controls, t-tests were used for continuous data and Chi-square tests for categorical data. All p-values <0.05 were considered statistically significant. Odds ratios (OR) and 95% confidence intervals (CI) were generated using multivariable logistic models, with severe COVID-19 as a dependent variable. Independent variables were age, sex, level of education, BMI, smoking habits, FEV 1 % predicted, ACT scores ⩽19, asthma medication (high-dose ICS in combination with LABA and OCS), prior asthma inpatient/secondary care, and concomitant comorbidities (cardiovascular disease, depression, diabetes, and rhinitis). Missing values were handled in two ways: missing as a separate category (Model 1) and a complete case analysis (Model 2). 20 We further explored factors associated with COVID-19-related death by comparing patients dying from COVID-19 with those discharged from hospital alive. This analysis was conducted with COVID-19 death as the dependent variable and the same covariates as described above.

The study complies with the Declaration of Helsinki and was approved by the Swedish Ethical Review Authority (2020-02777).

Up to 31 December 2020, 150,430 living adult asthma patients in SNAR were identified, of whom 1067 (0.7%) fulfilled the definition of severe COVID-19. In total, 875 patients were identified following hospitalization due to COVID-19 (n = 777 NPR and n = 98 NPR and SIR) and an additional 192 patients were identified in death certificates (SCDR) (Figure 1 ). A majority of cases, 1059 (99%), were confirmed by laboratory testing (ICD-10 U07.1). 

In a multivariable logistic regression model, comparing those who died from COVID-19 (n = 192) with those discharged alive from hospital (n = 875), older age (OR = 1.12, 95% CI = 1.09-1.13), male sex (OR = 1.89 95% CI = 1.24-2.78), and current smoking (OR = 5.35 95% CI = 2. 29-12.48) were associated with COVID-19 death when adjusted for covariates ( Figure 2 ).

In this large register-based Swedish asthma cohort, asthma-related factors such as uncontrolled asthma increased asthma medication intensity, and prior asthma healthcare utilization was associated with severe COVID-19. In addition, known risk factors in the general population such as male sex, older age, obesity, cardiovascular disease, and diabetes were associated with severe COVID-19. Surprisingly, current smoking was inversely associated with severe COVID-19 but was identified as a predictor for COVID-19 death together with male sex and older age.

While severe COVID-19 was more common among patients with decreased lung function and lower level of education, these factors were not associated with severe COVID-19 in the adjusted models. However, it has been reported that patients with uncontrolled asthma have an increased risk of more severe viral exacerbations. 21 In the light of our results, it is now evident that uncontrolled asthma, defined as an ACT score of ⩽19, and a high disease burden are associated with severe COVID-19. We used dispensed asthma medication, prior asthma inpatient or secondary care, and the use of OCS to determine a high disease burden. Our result is consistent with previously reported associations between COVID-19 and a severe asthma. 10,11,22-24 Zein et al. 11 showed that severe asthma, based on GINA treatment steps 3 and 4, influenced the association with COVID-19-related outcomes such as hospitalization and admission to the intensive care unit (ICU). Also reported from the United States is an increased risk of severe COVID-19-related outcomes in asthma patients who required clinical asthma care during the last year before the COVID-19 diagnosis. 22 In a study from the United Kingdom, an increased risk of COVID-19 related hospital death in asthma patients who recently had been prescribed OCS was observed, 23 and worse COVID-19 outcome (death, ICU care, or need for ventilation) in patients with severe asthma was found in Italy. 24 All these factors are relevant aspects to consider when identifying patients at risk of severe COVID-19.

During the pandemic, the GINA guidance has recommended that patients should maintain their controller medication, both ICS alone and ICS in combination with LABA. If necessary, OCS should also be prescribed to achieve good asthma control. 5 The issue whether ICS can modify the risk of severe COVID-19 in asthma patients has been of major concern ever since the first wave of the pandemic, 9 and an association with ICS use and increased risk of upper respiratory infections has been previously reported. 25 Yet, early pandemic reports indicated a low prevalence of asthma among severely affected COVID-19 patients, arising the question whether ICS indeed could have a protective effect. 9 However, over time, several studies have reported a lack of association between ICS use and the risk of severe COVID-19-related outcomes. 26, 27 Similar to our study, others have shown an increased risk of death in asthma patients on high-dose ICS compared with SABA only. 28 Importantly, these results might reflect patients with severe asthma in need of higher doses of ICS and need for increased asthma medication intensity, rather than an ICS-mediated effect.

In the general population as well in asthma patients, obesity is a known risk factor for severe 29 In our study, it was clear that overweight patients with asthma had an increased risk of contracting severe COVID-19. Importantly, BMI is a factor that affects the level of asthma control, 30 and several possible mechanisms link a higher BMI to uncontrolled asthma. First, it may be due to systemic inflammation in obese, the overweight influencing lung mechanics and metabolic factors. 31 Second, it has been reported that obese patients with asthma are less responsive to corticosteroid therapy and are exposed to a more disease-related severe outcome. 32 Higher BMI in asthma is often linked to the non-allergic phenotype, 32 and studies from the UK biobank and South Korea have investigated asthma phenotypes and COVID-19 outcomes. 12, 33 In both studies, patients with non-allergic asthma had worse clinical outcomes, whereas there was no association between severe COVID-19 and allergic asthma. This supports the lack of association between dispensed medication for rhinitis and severe COVID-19 in our study. However, there is still limited research about different asthma phenotypes and COVID-19, which needs to be further studied.

Smoking is a well-known risk factor of respiratory complications, and meta-analyses have shown that current smokers have an increased risk of worse COVID-19 prognosis than never smokers. 34 On the contrary, in a large Swedish population-based study, no association was found between current smoking and COVID-19 hospitalization or death. 35 Surprisingly in our study, current smoking was an inverse predictor of severe COVID-19, and the same trend has been shown in a cohort of patients with COPD. 36 We can speculate that our results could be related to protective strategies, that is, current smokers with asthma might have followed basic infection-control measures such as social distancing, especially during the first months of the pandemic, thereby reducing the risk of the infection. However, in our analysis of fatal outcomes, current smoking was a strong predictor for death, also shown by others. 37 This highlights the importance of smoking cessation support among current smokers with asthma.

In our study, we identified cardiovascular disease and diabetes as comorbidities associated with severe COVID-19. These are the same risk factors of severe COVID-19 previously reported in the general population 1,2 and in asthma patients. 29, 38 An interesting finding was the association between depression and severe COVID-19, which might be related to the close relationship between depression and poor asthma control. 39, 40 Strengths and limitations A major strength of our study is the large database of SNAR with well-characterized physiciandiagnosed asthma patients and that patients with a concomitant COPD diagnosis were excluded. Furthermore, the possibility to link SNAR with other national registers allows studying severe COVID-19 among asthma patients in Sweden.

In a registry study, we can reduce the risk of collider bias compared with hospital-based observational studies on COVID-19. 41 An important limitation, in relation to the use of register-based studies with clinical data from healthcare, is the variable amount of missing data. The associations between severe COVID-19 and missing data on BMI, lung function, and ACT scores indicate selection bias. A further limitation was lack of data on pack-years, since others have shown a dose-response association between pack-years and adverse COVID-19 outcomes. 42 To handle missing data, we conducted different multivariable models to validate our result. Furthermore, data on asthma medication are based on dispensed medications from Swedish Prescribed Drug Register, which have some limitations: (1) Patients with severe asthma cannot be identified when linking data. Thus, we used the term 'a high symptom burden' as a proxy for severe or difficult-to-treat asthma. (2) We also lack data on adherence, and therefore the ICS dose can be under-or overestimated. (3) Dispensed OCS and dispensed medication for rhinitis could be related to other conditions than asthma exacerbations or allergic rhinitis. (4) Data do not cover all comorbidities and therapies that could affect the risk of COVID-19.

Apart from known risk factors in the general population such as older age, male sex, obesity, cardiovascular disease, and diabetes, we identified uncontrolled asthma and a high disease burden including increased asthma medication intensity to be associated with severe COVID-19. Current smoking was an inverse predictor for severe COVID-19, but an analysis on fatal outcomes showed that current smoking was associated with COVID-19 death. Our results highlight the importance of identifying and optimizing treatment for patients with uncontrolled asthma. Equally important, for the asthma patient in general, is to maintain good symptom control and thereby reduce the risk of severe COVID-19. Furthermore, smoking cession support is essential also in relation to the risk of COVID-19 death.

Writing -review & editing

Vanfleteren: Writing -review & editing

Sten Axelsson Fisk: Writing -review & editing

Writing -review & editing

Writing -review & editing

Writing -review & editing

Therese Sterner: Writing -review & editing

Writing -review & editing

Writing -review & editing

Writing -review & editing

Ann Ekberg-Jansson: Writing -review & editing

Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing -original draft

Risk factors for adverse clinical outcomes with COVID-19 in China: a multicenter, retrospective, observational study

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

Features of 20 133 UK patients in hospital with covid-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study

Risk factors for coronavirus disease 2019 (COVID-19)-associated hospitalization: COVID-19-Associated Hospitalization Surveillance Network and Behavioral Risk Factor Surveillance System

Global Initiative for Asthma. Global strategy for asthma management and prevention

Severe COVID-19 among patients with asthma and COPD: a report from the Swedish National Airway Register

Asthma does not influence the severity of COVID-19: a metaanalysis

Asthma and risk of infection, hospitalization, ICU admission and mortality from COVID-19: systematic review and meta-analysis

Reduced risk of COVID-19 hospitalization in asthmatic and COPD patients: a benefit of inhaled corticosteroids?

Risk of adverse outcomes in patients with underlying respiratory conditions admitted to hospital with COVID-19: a national, multicentre prospective cohort study using the ISARIC WHO Clinical Characterisation Protocol UK

The relationship of asthma severity to COVID-19 outcomes

Association of asthma and its genetic predisposition with the risk of severe COVID-19

The Swedish National Airway Register (SNAR): development, design and utility to date

The Swedish National Board of Health and Welfare. Statistical databases, www. socialstyrelsen.se/en/statistics-and-data/registers

Reference values for lung function tests in men: regression equations with smoking variables

Reference values for lung function tests in females. regression equations with smoking variables

Asthma control test: reliability, validity, and responsiveness in patients not previously followed by asthma specialists

Missing covariate data in clinical research: when and when not to use the missing-indicator method for analysis

The influence of asthma control on the severity of virus-induced asthma exacerbations

Asthma disease status, COPD, and COVID-19 severity in a large multiethnic population

Factors associated with COVID-19-related death using OpenSAFELY

Asthma in a large COVID-19 cohort: prevalence, features, and determinants of COVID-19 disease severity

Inhaled corticosteroids and risk of upper respiratory tract infection in patients with asthma: a meta-analysis

Prevalence and characterization of asthma in hospitalized and nonhospitalized patients with COVID-19

Inhaled corticosteroids and COVID-19 risk and mortality: a nationwide cohort study

Risk of COVID-19-related death among patients with chronic obstructive pulmonary disease or asthma prescribed inhaled corticosteroids: an observational cohort study using the OpenSAFELY platform

Characteristics and outcomes of asthmatic patients with COVID-19 pneumonia who require hospitalisation

Obesity and poor asthma control in patients with severe asthma

Obesity and asthma

Immunologic mechanisms in asthma

Allergic disorders and susceptibility to and severity of COVID-19: a nationwide cohort study

The effect of smoking on COVID-19 severity: a systematic review and meta-analysis

Association of cardiometabolic risk journals.sagepub.com/home/tar factors with hospitalisation or death due to COVID-19: population-based cohort study in Sweden (SCAPIS)

Predictors of severe COVID-19 in a registry-based Swedish cohort of patients with chronic obstructive pulmonary disease (COPD)

Impact of smoking on COVID-19 outcomes: a HOPE Registry subanalysis

Risk factors for hospitalization, intensive care, and mortality among patients with asthma and COVID-19

Anxiety, depression, and asthma control: changes after standardized treatment

Depression is associated with poor control of symptoms in asthma and rhinitis: a population-based study

Collider bias undermines our understanding of COVID-19 disease risk and severity

Association of smoking and cumulative pack-year exposure with COVID-19 outcomes in the Cleveland Clinic COVID-19 Registry

Special thanks to Caddie Zhou for statistical analyses. Further acknowledgements are given to all the patients and healthcare professionals who contributed to registrations in SNAR and to the steering committee and register coordinators. Thanks also to the Centre of Registers Västra Götaland for infrastructure and data management support. 

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: CS has served in an advisory board and/or participated in education arranged by AstraZeneca, Boehringer-Ingelheim, and Novartis. LV has received grants and personal fees from AstraZeneca and personal fees from GSK, Novartis, Boehringer-Ingelheim, Menarini, Resmed, Chiesi, AGA Linde, and Pulmonx. TS has received personal fees from ALK Abello for lectures at sponsored meetings. AL has received personal fees from AstraZeneca, Novartis, Boehringer-Ingelheim, and GlaxoSmithKline for advisory boards and/or lectures at sponsored meetings. AT has received personal fees from AstraZeneca, Novartis, Boehringer-Ingelheim, and GlaxoSmithKline for advisory boards and/or lectures at sponsored meetings. FN was an employee of AstraZeneca until 2019 and holds some AstraZeneca shares. JRK, SAF, CP, YS, JS, AEJ, and JKS have no conflicts of interest.

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Swedish Heart-Lung Foundation under Grant 20200308 and The Asthma and Allergy Association.