key: cord-0973726-zug8gmwu
authors: Guan, Wei-jie; Liang, Wen-hua; Shi, Ying; Gan, Lan-xia; Wang, Hai-bo; He, Jian-xing; Zhong, Nan-shan
title: Chronic respiratory diseases and the outcomes of COVID-19: A nationwide retrospective cohort study of 39,420 cases
date: 2021-03-06
journal: J Allergy Clin Immunol Pract
DOI: 10.1016/j.jaip.2021.02.041
sha: 293d67502f671929448f0de618ed433d38e66158
doc_id: 973726
cord_uid: zug8gmwu

Background Chronic respiratory diseases (CRD) are common among patients with coronavirus disease 2019 (COVID-19). Objectives We sought to determine the association between CRD (including disease overlap) and the clinical outcomes of COVID-19. Methods Data of diagnoses, comorbidities, medications, laboratory results and clinical outcomes were extracted from the national COVID-19 reporting system. CRD was diagnosed based on ICD-10 codes. The primary endpoint was the composite outcome of needing invasive ventilation, admission to intensive care unit, or death within 30 days after hospitalization. The secondary endpoint was death within 30 days after hospitalization. Results We included 39,420 laboratory-confirmed patients from the electronic medical records as of May 6th, 2020. Any CRD and CRD overlap was present in 2.8% and 0.2% of patients, respectively. COPD was most common (56.6%), followed by bronchiectasis (27.9%) and asthma (21.7%). COPD-bronchiectasis overlap was the most common combination (50.7%), followed by COPD-asthma (36.2%) and asthma-bronchiectasis overlap (15.9%). After adjustment for age, sex and other systemic comorbidities, patients with COPD (OR: 1.71, 95% CI: 1.44-2.03) and asthma (OR: 1.45, 95%CI: 1.05-1.98), but not bronchiectasis, were more likely to reach to the composite endpoint compared with those without at day 30 after hospitalization. Patients with CRD were not associated with a greater likelihood of dying from COVID-19 compared to those without. Patients with CRD overlap did not have a greater risk of reaching the composite endpoint compared to those without. Conclusion CRD was associated with the risk of reaching the composite endpoint, but not death, of COVID-19.

Study patients 153 In this retrospective cohort study, data were derived from the national COVID-19 reporting system, a 154 platform of in-patient Electronic Medical Records (EMR) authorized by National Health 155 Commission. Since the initial outbreak, submission of the EMR from individual hospitals designated 156 for admitting patients with COVID-19 was requested by the National Health Commission. We 157 extracted the data of the clinical diagnoses, comorbidities, medications, laboratory results and 158 clinical outcomes from the EMR. As of May 6 th , 2020 (the data cut-off date for our study), the 159 database consisted of 42,218 in-patient EMR records, covering 558 designated hospitals. To be 160 eligible for data inclusion in our analysis, all hospitalized patients had to have a diagnosis of 161 laboratory-confirmed COVID-19. All patients had established chronic respiratory diseases prior to 162 admission. We excluded patients without any information on the comorbidities and the clinical 163 outcomes (dead or alive, receipt of mechanical ventilation, and admission to intensive care unit). 164 This study was approved by the ethics committee of the First Affiliated Hospital of Guangzhou 165 Medical University (Institutional Review Board: 202092) . Written informed consent form was 166 waived due to the anonymize data extraction of the EMR. 167 168 This was a retrospective cohort study which was conducted between December 2019 and May 6 th , 169 2020. All hospitalized patients were prospectively followed-up until 30 days after hospitalization. 170 Within the EMR, each standardized in-patient discharge summary consisted of the following items: 1) 171 demographics (i.e., gender, date of birth, occupation, and geographic location); 2) the primary and 172 secondary discharge description, coded based on the International Classification of Diseases-10; 3) 173 the main treatment description and discharge records; 4) in-hospital outcomes (i.e., death, length of 174 hospital stay); 5) discharge or death summary (i.e., medications, discharge outcomes). 182 At the request of the National Health Commission, all medical records were stored centrally in 183 the Tianhe-2 supercomputer, the data processing center in Guangzhou. A team of experienced 184 computing scientists and bioinformatics data managers formulated the clinical data and electronically 185 extracted the data with a customized operating system from the clinical charts and the portable 186 document format files. Data were exported into a computerized database for further cross-check. 187 Study definitions 188 Chronic respiratory disease overlap denoted at least two co-existing chronic respiratory diseases. At 189 hospital admission, patients were stratified into having non-severe (common type), or severe 190 (respiratory rate ≥30/min, dyspnea, oxygenation index <300) or critical illness of (needing intensive care), based on the criteria established by The Diagnosis and Treatment Protocol 192 for [13] . The primary endpoint, the composite outcome, was defined as 193 needing invasive ventilation, admission to intensive care unit, or death within 30 days after 194 hospitalization [5] . The secondary endpoint was death within 30 days after hospitalization.

Statistical analysis 196 In this study, we took a stepwise approach for examining the completeness of the core data sets. 197 Specifically, we initially verified the completeness of data pertaining to the age and sex, followed by 9 displayed as the counts and percentages. 213 We included 39,420 (93.4%) laboratory-confirmed patients out of 42,218 patients after excluding 214 patients with missing data [age or sex (n=456), discharge records (n=1647), and admission date 215 (n=695) (Figure 1) . 2,053 (5.21%) deaths were recorded. Patients who were included in our analysis 216 had comparable demographic characteristics compared with those who were not (Table 1) .

Baseline characteristics 218 Any chronic respiratory diseases and chronic respiratory disease overlap was present in 2.8% 219 (n=1123) and 0.2% (n=69) of all patients, respectively. COPD was the most common chronic 220 respiratory disease (n=636, 56.6%), followed by bronchiectasis (n=313, 27.9%) and asthma (n=244, 221 21.7%). For chronic respiratory disease overlap, COPD-bronchiectasis overlap was the most 222 common combination (n=35, 50.7%), followed by COPD-asthma overlap ( Table 2) . 271 At day 30 after hospitalization, patients with chronic respiratory diseases had an increased risk 272 of dying from COVID-19 than those without chronic respiratory diseases in the unadjusted analysis 273 (OR: 2.05, 95%CI: 1.68-2.51). As shown in Table 3 Table E1 and E2, respectively. 293 The risk of being admitted to the intensive care unit was higher in patients with chronic should be interpreted with caution. 344 Age, sex and the presence of other systemic comorbidities have also been associated with the 345 clinical outcomes of COVID-19, which was consistent with the findings reported previously [5, 7, 9] . 346 Considering that these factors might have confounded our analysis, we have performed the 347 regression analysis that mutually adjusted for these variables in our study. The models have 348 reaffirmed the significant association of these variables with the clinical outcomes of COVID-19. b Adjusted with the presence of any chronic respiratory disease/COPD/asthma/bronchiectasis, female sex and age. c Adjusted with any chronic respiratory disease/COPD/asthma/bronchiectasis, any other systemic comorbidities, and age. d Adjusted with any chronic respiratory disease/COPD/asthma/bronchiectasis, any other systemic comorbidities and female sex Abbreviation: OR: odds ratio; 95%CI: 95% confidence interval. 

Mortality Hazard ratio 0.79(95%CI 0.63-1) log-rank p=0

Any CRD (n=1123,100.0%) Unknown (n=182,16.2%)

Bronchiectasis (n=313,27.9%)

Asthma (n=244,21.7%)

COPD (n=636,56.6%)

Death (n=111,9.9%)

Non-severe Covid-19 (n=377,33.6%)

COPD (n=636,56.6%)

Asthma (n=244,21.7%)

Bronchiectasis (n=313,27.9%) ICU (n=187,16.7%)

Unknown (n=182,16.2%)

Non-severe Covid-19 (n=377,33.6%)

Any CRD (n=1123,100.0%)

COPD (n=636,56.6%)

Asthma (n=244,21.7%)

Bronchiectasis (n=313,27.9%) IPPV (n=113,10.1%)

Unknown (n=182,16.2%)

Non-severe Covid-19 (n=377,33.6%)

COPD (n=636,56.6%)

Asthma (n=244,21.7%)

Bronchiectasis (n=313,27.9%) IPPV (n=113,10.1%)

Unknown (n=182,16.2%)