Abstract and Introduction
Coronavirus disease 2019 (COVID-19) is primarily a respiratory illness, although increasing evidence indicates that infection with SARS-CoV-2, the virus that causes COVID-19, can affect multiple organ systems. Data that examine all in-hospital complications of COVID-19 and that compare these complications with those associated with other viral respiratory pathogens, such as influenza, are lacking. To assess complications of COVID-19 and influenza, electronic health records (EHRs) from 3,948 hospitalized patients with COVID-19 (March 1–May 31, 2020) and 5,453 hospitalized patients with influenza (October 1, 2018–February 1, 2020) from the national Veterans Health Administration (VHA), the largest integrated health care system in the United States,* were analyzed. Using International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes, complications in patients with laboratory-confirmed COVID-19 were compared with those in patients with influenza. Risk ratios were calculated and adjusted for age, sex, race/ethnicity, and underlying medical conditions; proportions of complications were stratified among patients with COVID-19 by race/ethnicity. Patients with COVID-19 had almost 19 times the risk for acute respiratory distress syndrome (ARDS) than did patients with influenza, (adjusted risk ratio [aRR] = 18.60; 95% confidence interval [CI] = 12.40–28.00), and more than twice the risk for myocarditis (2.56; 1.17–5.59), deep vein thrombosis (2.81; 2.04–3.87), pulmonary embolism (2.10; 1.53–2.89), intracranial hemorrhage (2.85; 1.35–6.03), acute hepatitis/liver failure (3.13; 1.92–5.10), bacteremia (2.46; 1.91–3.18), and pressure ulcers (2.65; 2.14–3.27). The risks for exacerbations of asthma (0.27; 0.16–0.44) and chronic obstructive pulmonary disease (COPD) (0.37; 0.32–0.42) were lower among patients with COVID-19 than among those with influenza. The percentage of COVID-19 patients who died while hospitalized (21.0%) was more than five times that of influenza patients (3.8%), and the duration of hospitalization was almost three times longer for COVID-19 patients. Among patients with COVID-19, the risk for respiratory, neurologic, and renal complications, and sepsis was higher among non-Hispanic Black or African American (Black) patients, patients of other races, and Hispanic or Latino (Hispanic) patients compared with those in non-Hispanic White (White) patients, even after adjusting for age and underlying medical conditions. These findings highlight the higher risk for most complications associated with COVID-19 compared with influenza and might aid clinicians and researchers in recognizing, monitoring, and managing the spectrum of COVID-19 manifestations. The higher risk for certain complications among racial and ethnic minority patients provides further evidence that certain racial and ethnic minority groups are disproportionally affected by COVID-19 and that this disparity is not solely accounted for by age and underlying medical conditions.
The study population comprised two cohorts of hospitalized adult (aged ≥18 years) VHA patients: 1) those with nasopharyngeal (90%) or other specimens that had tested positive for SARS-CoV-2 by real-time reverse transcription–polymerase chain reaction (RT-PCR) during March 1–May 31, 2020, and 2) those with laboratory-confirmed influenza A or B by rapid antigen assay, real-time RT-PCR, direct or indirect fluorescent staining, or viral culture, during October 1, 2018–February 1, 2020. Patients who received an influenza diagnosis after February 1, 2020, were excluded to minimize the possible inclusion of patients co-infected with SARS-CoV-2. Patients were restricted to those with a COVID-19 or influenza test during hospitalization or in the 30 days preceding hospitalization (including inpatient care at a nursing home). Patients who were still hospitalized as of July 31, 2020, or who were admitted >14 days before receiving testing were excluded from the analysis.
Data from EHRs were extracted from VHA Praedico Surveillance System, a biosurveillance application used for early detection, monitoring, and forecasting of infectious disease outbreaks† and Corporate Data Warehouse. Data included age, sex, race/ethnicity, ICD-10-CM diagnosis codes, hospital admission and discharge date, and, if applicable, date of intensive care unit (ICU) admission and date of death. Thirty-three acute complications (not mutually exclusive) were identified using ICD-10-CM codes from the hospitalization HER. Underlying medical conditions were identified using ICD-10-CM codes from inpatient, outpatient, and problem list records from at least 14 days before the specimen collection date.
Categorical variables were compared using Chi-squared or Fisher’s exact test and continuous variables with Wilcoxon rank sum test. Two-sided p-values <0.05 were considered statistically significant. Among patients with COVID-19, the risk for complications was compared among racial/ethnic groups using log-binomial models, adjusting for age and underlying medical conditions, with White patients as the reference group. Relative risk for complications in patients with COVID-19 compared with those with influenza were estimated using log-binomial models, adjusting for age, sex, race/ethnicity, and underlying medical conditions. To assess bias from seasonality in complications unrelated to influenza or COVID-19, a sensitivity analysis restricted to cases diagnosed during March–May of 2019 (influenza) and March–May of 2020 (COVID-19) was conducted. All analyses were performed using SAS (version 9.4; SAS Institute). The data used in this analysis were obtained for the purpose of public health operations in VHA.§ Because no additional analyses were performed outside public health operational activities, the activity was determined to meet the requirements of public health surveillance as defined in 45 CFR 46.102(l)(2), and Institutional Review Board review was not required.
During October 1, 2018–February 1, 2020, 5,746 hospitalized patients received a positive influenza test result and during March 1–May 31, 2020, 4,305 hospitalized patients received a positive SARS-CoV-2 test result. For both groups, testing occurred during the 30 days preceding hospitalization or while hospitalized. A total of 132 patients admitted >14 days before testing were excluded, as were 518 patients who were still hospitalized as of July 31, 2020, leaving 5,453 influenza patients and 3,948 COVID-19 patients for analysis.
Patients with COVID-19 were slightly older than were those with influenza (median = 70 years; interquartile range [IQR] = 61–77 years versus 69 years; IQR = 61–75 years) (p = 0.001), but patients with influenza had higher prevalences of most underlying medical conditions than did those with COVID-19 (Table 1). Black patients accounted for 48.3% of COVID-19 patients and 24.7% of influenza patients; the proportion of Hispanic patients was similar in both groups. The percentage of COVID-19 patients admitted to an ICU (36.5%) was more than twice that of influenza patients (17.6%); the percentage of COVID-19 patients who died while hospitalized (21.0%) was more than five times that of influenza patients (3.8%); and the duration of hospitalization was almost three times longer for COVID-19 patients (median 8.6 days; IQR = 3.9–18.6 days) than that for influenza patients (3.0 days; 1.8–6.5 days) (p<0.001 for all).
Among patients with COVID-19, 76.8% had respiratory complications, including pneumonia (70.1%), respiratory failure (46.5%), and ARDS (9.3%). Nonrespiratory complications were frequent, including renal (39.6%), cardiovascular (13.1%), hematologic (6.2%), and neurologic complications (4.1%), as well as sepsis (24.9%) and bacteremia (4.7%); 24.1% of COVID-19 patients had complications involving three or more organ systems. Among COVID-19 patients, nine complications were more prevalent among racial and ethnic minority patients, including respiratory, neurologic, and renal complications, even after adjustment for age and underlying medical conditions (Table 2).
Compared with patients with influenza, patients with COVID-19 had two times the risk for pneumonia; 1.7 times the risk for respiratory failure; 19 times the risk for ARDS; 3.5 times the risk for pneumothorax; and statistically significantly increased risks for cardiogenic shock, myocarditis, deep vein thrombosis, pulmonary embolism, disseminated intravascular coagulation, cerebral ischemia or infarction, intracranial hemorrhage, acute kidney failure, dialysis initiation, acute hepatitis or liver failure, sepsis, bacteremia, and pressure ulcers (Figure). Patients with COVID-19 had a lower risk for five complications (asthma exacerbation, COPD exacerbation, acute myocardial infarction (MI) or unstable angina, acute congestive heart failure (CHF), and hypertensive crisis), although acute MI or unstable angina, acute CHF, and hypertensive crisis were not statistically significant when restricting to patients diagnosed during the same seasonal months.
Adjusted relative risk* for selected acute respiratory and nonrespiratory complications in hospitalized patients with COVID-19 (March 1– May 31, 2020), compared with historically hospitalized patients with influenza (October 1, 2018–February 1, 2020) — Veterans Health Administration, United States†,§,¶
Abbreviations: ARDS = acute respiratory distress syndrome; CHF = congestive heart failure; COPD = chronic obstructive pulmonary disease; DIC = disseminated intravascular coagulation; MI = myocardial infarction.
*95% confidence intervals (CIs) indicated with error bars.
†When restricted to patients with influenza during the same seasonal months (March–May), aRRs and 95% CIs for acute MI or unstable angina, acute CHF, and hypertensive crisis were 0.90 (0.74–1.11), 1.03 (0.82–1.28), and 0.75 (0.44–1.29), respectively.
§Dialysis during hospitalization was identified using International Classification of Diseases, Tenth Revision, Clinical Modification and current procedural terminology codes, and new initiation of dialysis was determined by excluding patients with indication of dialysis within the past year.
¶Separate crude and adjusted log-binomial models were run for each complication (which were not mutually exclusive). All adjusted models adjusted for age, sex, race/ethnicity, and outcome-specific underlying conditions. Specifically, respiratory complication models controlled for COPD, asthma, and other lung diseases; neurologic complication models controlled for underlying cerebrovascular diseases, neurological/musculoskeletal conditions, heart disease, and heart failure; cardiovascular and hematologic condition models controlled for heart disease, heart failure, renal conditions, diabetes mellitus, and extreme obesity; the acute kidney failure model controlled for underlying renal disease, diabetes mellitus, and hypertension. Complications related to the worsening of a chronic medical condition were restricted to those patients with that underlying medical condition.