Characteristics and Outcomes of COVID-19-Associated Pulmonary Embolism

Abstract

Dear Editor, Pulmonary embolism (PE) in the setting of COVID-19 is related to the procoagulant state as evidenced by the increased D-dimer levels and in situ thrombosis [1,2] and it usually involves the segmental/subsegmental arteries. [2] Some data have shown that mortality is higher among patients with COVID-19 and PE compared with non-COVID-19 PE. [3] The COVID-19 pandemic has led to marked reductions in cardiovascular testing in the United States (US). [4] However, the impact of the COVID-19 pandemic on the diagnosis and management of PE is not well studied. To better describe the characteristics and the impact of COVID-19 infection on the management strategies and outcomes of patients with acute PE, we performed a comprehensive analysis of a nationally representative database in the US. We used the Nationwide Readmissions Database (NRD) for the years 2019 and 2020 to extract the study cohort. All patients aged 18 years or older with any discharge diagnosis of PE in the years 2019 and 2020 were included. Pregnancy-related PE (n=2,871) and admissions with missing data on mortality (n=169) were excluded. We examined the differences in all-cause in-hospital mortality, length of stay (LOS), cost, intracranial hemorrhage (ICH), non-ICH bleeding, and 30-day urgent readmission rates between admissions with and without COVID-19 in the year 2020. We also examined the differences in outcomes among admissions with non-COVID-19 PE in the year 2019 versus 2020. All analyses were performed according to the Healthcare Cost and Utilization Project (HCUP) regulations for statistical analysis. Multivariable logistic regression analyses were used to adjust for the differences in baseline and hospital characteristics, PE severity and different management strategies. Additionally, we performed a subgroup analysis for patients with high-risk PE (i.e., cardiogenic shock). Statistical analyses were performed using IBM SPSS Statistics for Windows (version 28.0. Armonk, NY: IBM Corp). A total of 786,963 weighted hospitalizations with acute PE were included in the analysis, 374,122 (47.5%) were admitted in 2019 and 412,842 (52.5%) in 2020. Among those admitted in 2020, 46,825 (11.3%) weighted hospitalizations had concomitant COVID-19 infection. The use of surgical embolectomy, catheter-directed interventions (CDI) and mechanical circulatory support (MCS) devices was less common among admissions with COVID-19 infection. The use of vasopressors and mechanical ventilation was higher among admissions with underlying COVID-19 infection. (Table 1 ) Table 1 Characteristics and outcomes of patients with acute PE admitted in the year 2020. Table 1 PE without COVID-19 (n=366,017) PE with COVID-19 (n=46,825) P value Age, median (IQR) 66 (54-76) 66 (55-76) 0.04 Female 183,709 (50.2%) 19,462 (41.6%) <0.001 Smoking 85,873 (23.5%) 10,022 (21.4%) <0.001 Comorbidities Morbid obesity 46,862 (12.8%) 6,343 (13.5%) 0.007 Hypertension 233,817 (63.9%) 29,628 (63.3%) 0.118 Diabetes mellitus 96,420 (26.3%) 16,634 (35.5%) <0.001 Anemia 98,656 (27.0%) 10,384 (22.2%) <0.001 Coagulopathy 49,278 (13.5%) 8,127 (17.4%) <0.001 Pulmonary hypertension 41,094 (11.2%) 2,460 (5.3%) <0.001 Chronic pulmonary disease 96,833 (26.5%) 10,084 (21.5%) <0.001 Atrial fibrillation/flutter 62,518 (17.1%) 6,857 (14.6%) <0.001 Heart failure 87,636 (23.9%) 7,595 (16.2%) <0.001 Chronic kidney disease 61,036 (16.7%) 7,431 (15.9%) 0.008 Chronic liver disease 4,443 (1.2%) 248 (0.5%) <0.001 Connective tissue diseases 13,867 (3.8%) 1,482 (3.2%) <0.001 CAD 69,080 (18.9%) 7,197 (15.4%) <0.001 PAD 14,340 (3.9%) 1,163 (2.5%) <0.001 Carotid disease 3,012 (0.8%) 253 (0.5%) <0.001 Prior stroke 30,798 (8.4%) 3,217 (6.9%) <0.001 Malignancy 80,199 (21.9%) 3,026 (6.5%) <0.001 Metastatic 43,930 (12.0%) 1,005 (2.1%) <0.001 Presentation and severity Saddle PE 26,195 (7.2%) 1,950 (4.2%) <0.001 Acute cor pulmonale 27,267 (7.4%) 1,803 (3.9%) <0.001 High-risk PE (Cardiogenic shock) 8,647 (2.4%) 846 (1.8%) <0.001 Concomitant DVT 122,654 (33.5%) 8,811 (18.8%) <0.001 Hospital characteristics Large hospital 202,418 (55.3%) 24,998 (53.4%) 0.018 Teaching hospital 273,508 (74.7%) 34,137 (72.9%) 0.003 Medicare 203,423 (55.6%) 24,366 (52.0%) <0.001 Transferred from other hospital 13,297 (3.6%) 2,203 (4.7%) <0.001 Treatment modalities Systemic thrombolysis 8,687 (2.4%) 1,088 (2.3%) 0.70 Surgical embolectomy 596 (0.2%) 17 (0.04%) <0.001 CDT 9,549 (2.6%) 468 (1%) <0.001 CDE 410 (1.8%) 41 (0.8%) <0.001 IVC filter 22,557 (6.2%) 1,122 (2.4%) <0.001 Circulatory and ventilatory support Vasopressors 5,628 (1.5%) 1,443 (3.1%) <0.001 Mechanical ventilation 31,681 (8.7%) 9,739 (20.8%) <0.001 Mechanical circulatory support 1,459 (0.4%) 84 (0.2%) <0.001 Impella 392 (0.1%) 22 (0.0%) 0.03 ECMO 702 (0.2%) 55 (0.1%) 0.03 IABP 537 (0.1%) 10 (0.0%) <0.001 Outcomes In-hospital mortality 27,240 (7.4%) 9,285 (19.8%) <0.001 Intracranial hemorrhage (ICH) 5,847 (1.6%) 604 (1.3%) <0.001 Non-ICH 39,744 (10.9%) 5,261 (11.2%) <0.001 Length of stay, days (IQR) 4 (2-8) 7 (4-14) <0.001 Cost of stay, US Dollars (IQR) 12,462 (7,056-25,364) 17,212 (9,079-37,833) <0.001 30-day unplanned readmissions* 46,890/308,437 (15.2%) 2,596/25,967 (10.0%) <0.001 ⁎ After excluding those who died during the index admissions and those who were admitted in December of each calendar year. CDI: catheter-directed intervention, PE: pulmonary embolism, IQR: interquartile range, MI: myocardial infarction, PCI: percutaneous coronary intervention, CABG: Coronary artery bypass grafting, DVT: deep venous thrombosis, CDT: catheter-directed thrombolysis, CDE: catheter-directed embolectomy, US: ultrasound, IVC: inferior vena cava, ECMO: extracorporeal membrane oxygenation, IABP: intra-aortic balloon pump, IQR: interquartile range. The rate of all-cause in-hospital mortality was higher among admissions with COVID-19 infection (7.4% vs. 19.8%, P<0.001). The rate of ICH was slightly lower (1.6% vs. 1.3%, P<0.001) and non-ICH was slightly higher (10.9%. vs. 11.2%, P<0.001) in admissions with COVID-19 infection. Admissions with COVID-19 infection had longer LOS and higher cost. (Table 1) On multivariable analysis, COVID-19 infection was independently associated with higher mortality (adjusted odds ratio [aOR] 2.71, 95% confidence interval [CI] 2.56, 2.87, P<0.001), higher risk of non-ICH (aOR 1.10, 95% CI 1.05, 1.15, P<.001), and lower risk ICH (aOR 0.59, 95% CI 0.51, 0.68, P<0.001). Among admissions with high-risk PE, in-hospital mortality was higher among those with COVID-19 (36.9% vs. 69.7%, aOR 2.64, 95% CI 2.10, 3.33, P<0.001). In the analysis restricted to non-COVID-19 PE admissions, the prevalence of saddle PE (6.2% vs. 7.2%, P<0.001), cor pulmonale (6.8% vs. 7.4%, P=0.008), and cardiogenic shock (2.1% vs. 2.4%, P=0.007) was higher in 2020. There was no difference in the rate of utilization of systemic thrombolysis (2.4% vs. 2.4%, P=0.76), surgical embolectomy (0.2% vs. 0.2%, P=0.85), catheter-directed thrombolysis (2.5% vs. 2.6%, P=0.52), vasopressors (1.3% vs. 1.5%, P=0.20), and MCS devices (0.4% vs. 0.4%, P=0.92) between 2019 and 2020. All-cause in-hospital mortality (7.0% vs. 7.4%, P<0.001) was slightly higher in 2020. However, after adjustment, there was no difference in mortality between both years (2020 vs. 2019: aOR 1.04, 95% CI 0.99, 1.09, P=0.14). In this nationwide analysis, we examined the association between COVID-19 infection and the management and outcomes of acute PE during the early wave of the pandemic. The main findings are as follows: 1) Patients with COVID-19 and PE were less likely to receive CDI, surgical embolectomy, and MCS devices, compared with those without COVID-19 infection. 2) COVID-19 infection was independently associated with a higher incidence of all-cause in-hospital mortality, higher costs, and longer LOS compared to patients without COVID-19. 3) In analyses restricted to non-COVID-19 PE, patients admitted in 2020 versus 2019 were sicker but there was no difference in the utilization of advanced therapies and in in-hospital mortality. Similar to prior studies, [3] we found that patients with COVID-19 and PE had fewer comorbidities and risk factors for PE, indicating that COVID-19 itself is the predisposing factor for PE. Also, fewer patients with COVID-19 PE had DVT or saddle PE, suggesting that in situ thrombosis plays a role in the pathogenies of COVID-19 PE. [5] In the current analysis, COVID-19 infection was associated with a 2.7-fold increase in in-hospital mortality among patients with PE. This can be related to severe COVID-19 infection itself leading to systemic inflammation, shock, multiorgan failure and respiratory failure with involvement of the lung parenchyma. [3,6] Additionally, PE with COVID-19 infection is associated with an increased risk of mechanical ventilation and ICU admission. [7] In our analysis, patients with PE and COVID-19 infection during the early wave of the pandemic were less likely to receive surgical embolectomy, CDI, or MCS. This may be attributed to patients' isolation, instability, and difficulties in transferring patients safely to the catheterization laboratory or operating room without exposing the healthcare team. Additionally, patients may have not been offered invasive procedures due to the risk of transmission or poor prognosis. We found that the care of patients with acute PE was generally not affected during the early wave of the COVID-19 pandemic. We noticed that in 2020, compared with 2019, patients with non-COVID PE were sicker suggesting that only patients with more severe symptoms presented to the emergency departments (ED) or that patients with less severe presentations were discharged directly from the ED. Despite that, there was no decline in the utilization of advanced therapies and no difference in in-hospital mortality between 2019 and 2020 for patients without COVID-19, which suggests that care for essential services such as PE was maintained in 2020 despite significant system constraints. There are some limitations to this study. Being a retrospective observational study, it is prone to selection bias. Given the administrative nature of the NRD, the study is subject to coding errors and data quality at the site of collection, without the ability to adjudicate accuracy. Clinical, laboratory, and imaging data as well as data on prescribed medications including the type and dose of thrombolytics are lacking from the NRD, which may have impacted the clinical outcomes. Long-term outcomes could not be assessed, and we could not also ascertain PE-specific mortality from NRD. [8] Finally, our findings are restricted to the early wave of the pandemic when the therapies for COVID-19 were limited, and widespread vaccines were unavailable. In this nationwide observational cohort of patients admitted with PE during the early wave of the pandemic, COVID-19 infection was independently associated with a higher risk of all-cause in-hospital mortality among PE admissions. There was no decline in utilization of advanced therapies and no difference in-hospital mortality among patients with acute PE without COVID-19 infection in the year 2019 vs. 2020. Sources of funding None Declaration of Competing Interests None