Increased risk of pulmonary embolism and deep vein thrombosis with COVID-19 pneumonia in comparison to influenza pneumonia: insights from the National Inpatient Sample database

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), principally a respiratory virus causing coronavirus disease 2019 (COVID-19) pneumonia, results in a hypercoagulable state (1). Numerous observational studies have linked acute COVID-19 pneumonia with an increased risk of venous thromboembolism (VTE), including pulmonary embolism (PE) and lower extremity deep vein thrombosis (DVT) (2-7). However, it is unclear if this finding is due to intrinsic viral characteristics such as vascular tropism or a product of the inflammatory state associated with critical illness. Clinically, we sought to answer this question by comparing the inpatient prevalence of acute PE and acute lower extremity DVT among patients hospitalized with COVID-19 pneumonia versus patients hospitalized with influenza pneumonia using a large, nationally representative sample. Influenza pneumonia was selected as a comparator viral pneumonia because it has a similar clinical course and has been associated with an increased risk of myocardial infarction (MI), ischemic stroke, and VTE (3-5).

Previous studies comparing the prevalence of VTE in patients with COVID-19 pneumonia and patients with influenza pneumonia have shown that COVID-19 pneumonia is associated with a higher risk of PE and DVT; however, these studies do not adjust for the severity and acuity of the respective viral pneumonia (6,7). Therefore, we designed our study to compare the risk of acute PE and DVT after extensive adjustment for critical illness to investigate whether COVID-19 pneumonia continues to be associated with a higher risk of PE and DVT when compared with influenza pneumonia. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1674/rc).

Methods

Data source

This retrospective study used the National Inpatient Sample (NIS) database from the years 2019 and 2020; this comprises hospitalizations in the United States (U.S.) from January 1^st, 2019, to December 31^st, 2020 (8). The NIS is maintained by the Healthcare Cost and Utilization Project (HCUP) and is the largest publicly available all-payer inpatient healthcare database designed to produce U.S. regional and national estimates of inpatient utilization, access, cost, quality, and outcomes. Unweighted, it contains data from approximately seven million hospital stays each year. Weighted, it estimates around 35 million hospitalizations nationally. The NIS is drawn from all states participating in HCUP, covering more than 97% of the U.S. population and approximates 20% of discharges from U.S. community hospitals (8). The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was exempt from institutional review board approval as the NIS 2020 database is comprised of previously collected, de-identified data.

Study population and baseline demographics

Our study contained two populations: (I) patients 18 years of age and older who were hospitalized with a primary diagnosis of COVID-19 pneumonia between January 1^st, 2020, and December 31^st, 2020; and (II) patients 18 years of age and older who were hospitalized with a primary diagnosis of influenza pneumonia between January 1^st, 2019 and December 31^st, 2019. Exclusion criteria included patients who were transferred out of the index hospital to another acute care hospital and patients admitted with a diagnosis of acute trauma. This was necessary in order to minimize duplicates since the NIS database does not track patients after hospital discharge. The COVID-19 pneumonia population was identified using International Classification of Diseases, 10^th Revision (ICD-10) codes U07.1 and J12.89 (other viral pneumonia), while excluding influenza pneumonia, ICD-10 codes J09, J10, and J11. Similarly, the influenza pneumonia population was identified using the corresponding ICD-10 codes (Table S1) during the year 2019. We felt that the advantage of using 2019 influenza data and 2020 COVID-19 data is that it lowers the likelihood of including patients who may have been coinfected. The following variables were extracted from the NIS database:

• Demographic characteristics: age, sex, and race;
• Medical comorbidities, Elixhauser comorbidity index, and venous thromboembolic risk factors were based on the Padua prediction score for VTE and literature analysis. These were extracted using ICD-10 codes.

Primary outcomes and statistical analysis

The two primary outcomes of this study were inpatient acute PE and inpatient acute lower extremity deep venous thrombosis, which were also determined using ICD-10 codes (Tables S1,S2).

Descriptive statistics were used to report the mean for continuous variables and the proportion for categorical variables within the baseline populations. Univariate logistic regression was used to compare baseline characteristics of the two populations (Table 1). Univariate logistic regression was also used to determine the unadjusted odds ratios (ORs) for the two primary outcomes, comparing patients with COVID-19 pneumonia and patients with influenza pneumonia. Risk factor analysis was performed for all listed baseline characteristics (Table 1) with respect to both primary outcomes to investigate association. In our multivariate model, we created a variable with influenza pneumonia as the reference and COVID-19 pneumonia as the predictor. Using this variable and other risk factors for VTE, we created two multivariate models—one for PE and another one for DVT. Variables with significant P values on the univariate model or those with important clinical relevance were used to construct a multivariate logistic regression model to determine if COVID-19 pneumonia is independently associated with PE and DVT when compared to influenza pneumonia (Table 2).

All analysis was carried out using Stata statistical software: release 17 (2021; StataCorp LLC, College Station, TX, USA). Stata’s svy command and appropriate weights were used in all estimations. The overall fit was assessed using receiver operating characteristic (ROC) curves, and sensitivity analysis was performed using the evalue package. The study was exempt from institutional review board approval as the database uses previously collected de-identified data.

Results

Baseline characteristics

A total of 1,172,410 COVID-19 patients and 184,980 influenza patients were included in the study. At baseline, patients with COVID-19 pneumonia were older, had a higher proportion of males, and had more racial multiplicity, including larger populations of Blacks and Hispanics, than patients with influenza pneumonia. Patients with COVID-19 pneumonia also had lower Elixhauser comorbidity indices, lower prevalence of solid, hematologic, and metastatic malignancies, lower prevalence of previous VTE, and were less likely to be smokers. The only pre-existing risk factors for VTE that had an increased prevalence among patients with COVID-19 pneumonia were chronic estrogen use and known thrombophilia.

There are numerous clinical markers indicating a high acuity of illness among patients with COVID-19 pneumonia, including higher rates of vasopressor use, renal replacement therapy, and intubation with mechanical ventilation (in contrast to patients with influenza pneumonia who had more use of non-invasive positive-pressure ventilation). Similarly, patients with COVID-19 pneumonia had significantly longer hospital stays and higher inpatient mortality.

There was no difference in pre-admission antiplatelet use between the two groups. With respect to in-hospital complications, there were no differences in the prevalence of acute stroke or septic shock. Most complications were more prevalent in the COVID-19 group, except for a few, such as MI, severe sepsis, and the need for non-invasive ventilation.

Differential risk of acute PE and its risk factors

After multivariate adjustment, patients with COVID-19 pneumonia maintained significantly higher odds of developing acute inpatient PE when compared to patients with influenza pneumonia [adjusted OR (aOR): 2.48; 95% confidence interval (CI): 2.16–2.86; P<0.001]. On sensitivity analysis, this was associated with an expectation value (E-value) of 4.39 and an E-value (CI) of 3.74, indicative of a strong association.

Both male sex and black race (OR: 1.16; 95% CI: 1.14–1.18; P<0.001) were associated with a higher risk of PE among patients with COVID-19 pneumonia when compared to patients with influenza pneumonia. As expected, traditional risk factors for VTE, including estrogen-based hormonal therapy, known thrombophilia, malignancies, and paralysis, had a stronger risk association for PE among patients with COVID-19 pneumonia. With respect to chronic comorbidities, congestive heart failure, congenital heart disease, connective tissue disease, chronic malnutrition, and obesity were all associated with higher risk for PE among patients with COVID-19 pneumonia. The higher acuity of care characterized by dependence on vasopressors, intubation, or non-invasive positive pressure ventilation was a stronger risk factor for PE among patients with COVID-19 pneumonia when compared to patients with influenza pneumonia.

Differential risk of acute lower extremity DVT and its risk factors

After multivariate adjustment, patients with COVID-19 pneumonia had a significantly higher risk of DVT (aOR: 1.66; 95% CI: 1.41–1.96; P<0.001) when compared to patients with influenza pneumonia. On sensitivity analysis, this yielded an E-value of 2.70 and an E-value (CI) of 2.1.

Both male sex and black race (OR: 1.17; 95% CI: 1.07–1.27; P<0.001) were associated with a higher risk of DVT among patients with COVID-19 pneumonia when compared to patients with influenza pneumonia. As expected, traditional risk factors for VTE, including estrogen-based hormonal therapy, known thrombophilia, prior VTE, malignancies, and paralysis, had a stronger risk association for DVT among patients with COVID-19 pneumonia. Tobacco use was associated with slightly lower odds of PE and DVT. With respect to chronic comorbidities, chronic kidney disease, congestive heart failure, congenital heart disease, chronic neurologic disease, connective tissue disease, diabetes mellitus, chronic malnutrition, and obesity were all associated with higher risk for DVT among patients with COVID-19 pneumonia. The higher acuity of care characterized by dependence on vasopressors, intubation, non-invasive positive pressure ventilation, or renal replacement therapy was a stronger risk factor for DVT among patients with COVID-19 pneumonia when compared to patients with influenza pneumonia.

Discussion

The primary finding of our study is that patients hospitalized with COVID-19 pneumonia have a significantly increased risk of acute inpatient PE and lower extremity deep venous thrombosis when compared to patients hospitalized with influenza pneumonia. More importantly, even after adjusting for the higher acuity of illness observed among COVID-19 patients during the early pandemic, the risk of PE and lower extremity DVT remains significantly elevated. To our knowledge, this study utilizes the largest sample of patients with COVID-19 and influenza pneumonia and is the first study to comprehensively adjust for the higher acuity of illness observed during the early pandemic.

Secondly, we identified congenital heart disease and malnutrition as novel risk factors for acute PE in COVID-19 patients.

At baseline, the overall population of patients with COVID-19 pneumonia had a lower burden of medical comorbidities when compared to patients with influenza pneumonia, as signified by the mean Elixhauser Comorbidity Indices. There are two primary exceptions; patients with COVID-19 pneumonia had a higher prevalence of obesity and diabetes mellitus. Existing literature demonstrates that obesity is independently associated with a 6.2-fold increased risk of VTE outside of COVID-19 pneumonia (9). Similarly, diabetes mellitus has also been independently associated with a significantly increased risk of VTE (aOR: 1.74; 95% CI: 1.21–2.51; P<0.001) (10). With respect to traditional VTE risk factors, patients with COVID-19 pneumonia had a lower prevalence of solid, hematologic, and metastatic malignancies, previous VTE, and tobacco use when compared to patients with influenza pneumonia. One possible explanation for the lower odds of VTE in tobacco users is residual confounding, as tobacco users are generally younger, and PE and DVT were more common in older patients. Our analysis did not show any risk modification effect of tobacco use. We did note an increased prevalence of chronic estrogen use and known thrombophilia among patients with COVID-19 pneumonia. In summary, while the prevalence of obesity, diabetes mellitus, estrogen use, and known thrombophilia likely contributed to the burden of in-hospital PE and DVT in patients with COVID-19 pneumonia, we found an independent risk of PE and DVT with COVID-19 pneumonia after multivariate adjustment.

In our risk factor analysis (Table 3), most of our findings were as expected. Known VTE risk factors were associated with a higher risk of PE and DVT among patients with COVID-19 pneumonia when compared to patients with influenza pneumonia. This is congruent with the higher background risk of PE and DVT in patients with COVID-19 pneumonia, as demonstrated in our study—in other words, risk factors are likely additive to the higher intrinsic risk of PE and DVT in COVID-19 pneumonia. We observed two risk factors that were not reported in previous literature: congenital heart disease and chronic malnutrition. Possible mechanisms of increased PE risk (likely more so than DVT) in congenital heart disease may include alterations in pulmonary vasculature blood flow or epithelial characteristics; in any case, these are a heterogeneous group of diseases that require further analysis to elucidate the underlying mechanisms of thrombotic disease in COVID-19 pneumonia. The link between chronic malnutrition and thrombosis can likely be explained by alterations in the hepatic production of coagulant proteins (11).

While hospitalized, patients with COVID-19 pneumonia had markers of high acuity illness characterized by more complications, longer hospital stays, and increased mortality. We believe that this finding likely stems from decreased host immunity to SARS-CoV-2 in the absence of widespread messenger RNA (mRNA) vaccine use during 2020, as well as intrinsic differences in virus characteristics (12). The acuity of illness is an important factor with respect to PE and DVT, as a higher acuity of illness and need for intensive care is independently associated with increased VTE risk regardless of the underlying disease process (13). Similarly, inflammatory and infectious diseases generally increase the risk of VTE (14,15). The strength of our analysis is that we adjusted for the acuity of illness and medical comorbidities and found that patients with COVID-19 pneumonia were at an independently high risk of VTE, and particularly acute PE, more so than lower extremity DVT. The higher risk of acute inpatient PE over lower extremity DVT in patients with COVID-19 pneumonia in comparison to patients with influenza pneumonia is a noteworthy observation.

These findings may support the “thrombus in situ” theory of SARS-CoV-2 affecting the pulmonary arteries and inducing pulmonary thrombosis rather than the traditional mechanism of peripheral deep venous clots embolizing to the pulmonary vasculature. Thrombotic disease in COVID-19 is likely due to two distinct mechanisms: (I) a hypercoagulable state leading to large-vessel thrombosis and thromboembolism; and (II) endothelial injury responsible for in situ microvascular thrombosis (16). Additionally, it has been proposed that SARS-CoV-2 directly infects endothelial cells, including pulmonary vasculature, causing inflammation and impairing the coagulation cascade (17,18). This mechanism of direct endothelial infection may explain the specific increase in PE risk among patients with COVID-19 pneumonia when compared to patients with influenza pneumonia.

Limitations

Our study has several limitations. First, the NIS is an administrative database that relies on ICD-10 codes for diagnosis. This may have underestimated the true prevalence of many risk factors, such as anticoagulant and antiplatelet use, medical comorbidities, and acute in-hospital complications due to under-coding. Moreover, there is a possibility of detection bias, as the ICD-10 code for COVID-19 pneumonia was created later in 2020, likely resulting in the under-reporting of COVID-19 pneumonia cases. Secondly, there remains a possibility of confounding bias in this retrospective study despite our thorough multivariate adjustment and sensitivity analysis. Thus, as with any observational study, association does not imply causation, and conclusions should be drawn cautiously. Thirdly, it is complicated to extrapolate data based on initial SARS-CoV-2 strains to currently evolving strains, including Omicron BA.5, as these may be accompanied by the possibility of modified thrombogenic properties. Lastly, our study utilizes data before widespread mRNA vaccine availability for COVID-19; we anticipate that host immune responses modulate thrombotic complications of COVID-19, including DVT and PE.

Conclusions

The primary finding of our study is that patients hospitalized with COVID-19 pneumonia have a significantly increased risk of acute inpatient PE and lower extremity deep venous thrombosis when compared to patients hospitalized with influenza pneumonia. More importantly, even after adjusting for the higher acuity of illness observed among COVID-19 patients during the early pandemic and for traditional risk factors of VTE, the risk of PE and DVT remains significantly elevated in this population. To our knowledge, this study utilizes the largest sample of patients with COVID-19 and influenza pneumonia and is the first study to comprehensively adjust for the higher acuity of illness observed during the early COVID-19 pandemic. In this study, we also discovered that congenital heart disease and malnutrition are emerging risk factors for acute PE in individuals with COVID-19. The higher occurrence of inpatient PE compared to lower extremity DVT in COVID-19 pneumonia patients suggests a possible “thrombus in situ” mechanism for SARS-CoV-2-induced pulmonary thrombosis. This highlights the need for clinicians to remain vigilant about the possibility of PE, even in the absence of DVT, in COVID-19 pneumonia patients. It is imperative to adhere to current evidence-based guidelines for the diagnosis and management of this condition (19).

Acknowledgments

Funding: None.

Footnote

Provenance and Peer Review: This article was a standard submission to the series “Thoracic Diseases and Big Data” published in Journal of Thoracic Disease. The article has undergone external peer review.

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1674/rc

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1674/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-23-1674/coif). The series “Thoracic Diseases and Big Data” was commissioned by the editorial office without any funding or sponsorship. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was exempt from institutional review board approval as the NIS 2020 database is comprised of previously collected, de-identified data.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Spiezia L, Boscolo A, Poletto F, et al. COVID-19-Related Severe Hypercoagulability in Patients Admitted to Intensive Care Unit for Acute Respiratory Failure. Thromb Haemost 2020;120:998-1000. [Crossref] [PubMed]
2. Poissy J, Goutay J, Caplan M, et al. Pulmonary Embolism in Patients With COVID-19: Awareness of an Increased Prevalence. Circulation 2020;142:184-6. [Crossref] [PubMed]
3. Kwong JC, Schwartz KL, Campitelli MA, et al. Acute Myocardial Infarction after Laboratory-Confirmed Influenza Infection. N Engl J Med 2018;378:345-53. [Crossref] [PubMed]
4. Boehme AK, Luna J, Kulick ER, et al. Influenza-like illness as a trigger for ischemic stroke. Ann Clin Transl Neurol 2018;5:456-63. [Crossref] [PubMed]
5. Bunce PE, High SM, Nadjafi M, et al. Pandemic H1N1 influenza infection and vascular thrombosis. Clin Infect Dis 2011;52:e14-7. [Crossref] [PubMed]
6. Stals MAM, Grootenboers MJJH, van Guldener C, et al. Risk of thrombotic complications in influenza versus COVID-19 hospitalized patients. Res Pract Thromb Haemost 2021;5:412-20. [Crossref] [PubMed]
7. Lo Re V 3rd, Dutcher SK, Connolly JG, et al. Association of COVID-19 vs Influenza With Risk of Arterial and Venous Thrombotic Events Among Hospitalized Patients. JAMA 2022;328:637-51. [Crossref] [PubMed]
8. AHRQ. National (Nationwide) Inpatient Sample (NIS) Database Documentation. 2022. Accessed October 17, 2022. Available online: https://hcup-us.ahrq.gov/db/nation/nis/nisdbdocumentation.jsp
9. Hotoleanu C. Association between obesity and venous thromboembolism. Med Pharm Rep 2020;93:162-8. [Crossref] [PubMed]
10. Piazza G, Goldhaber SZ, Kroll A, et al. Venous thromboembolism in patients with diabetes mellitus. Am J Med 2012;125:709-16. [Crossref] [PubMed]
11. Mureșan AV, Hălmaciu I, Arbănași EM, et al. Prognostic Nutritional Index, Controlling Nutritional Status (CONUT) Score, and Inflammatory Biomarkers as Predictors of Deep Vein Thrombosis, Acute Pulmonary Embolism, and Mortality in COVID-19 Patients. Diagnostics (Basel) 2022;12:2757. [Crossref] [PubMed]
12. Raj K, Yeruva K, Jyotheeswara Pillai K, et al. Population Risk Factors for Severe Disease and Mortality in COVID-19 in the United States during the Pre-Vaccine Era: A Retrospective Cohort Study of National Inpatient Sample. Med Sci (Basel) 2022;10:67. [Crossref] [PubMed]
13. Minet C, Potton L, Bonadona A, et al. Venous thromboembolism in the ICU: main characteristics, diagnosis and thromboprophylaxis. Crit Care 2015;19:287. [Crossref] [PubMed]
14. Smeeth L, Cook C, Thomas S, et al. Risk of deep vein thrombosis and pulmonary embolism after acute infection in a community setting. Lancet 2006;367:1075-9. [Crossref] [PubMed]
15. Tichelaar YI, Kluin-Nelemans HJ, Meijer K. Infections and inflammatory diseases as risk factors for venous thrombosis. A systematic review. Thromb Haemost 2012;107:827-37. [Crossref] [PubMed]
16. Iba T, Levy JH, Levi M, et al. Coagulopathy of Coronavirus Disease 2019. Crit Care Med 2020;48:1358-64. [Crossref] [PubMed]
17. Varga Z, Flammer AJ, Steiger P, et al. Endothelial cell infection and endotheliitis in COVID-19. Lancet 2020;395:1417-8. [Crossref] [PubMed]
18. Ackermann M, Verleden SE, Kuehnel M, et al. Pulmonary Vascular Endothelialitis, Thrombosis, and Angiogenesis in Covid-19. N Engl J Med 2020;383:120-8. [Crossref] [PubMed]
19. Raj K, Chandna S, Doukas SG, et al. Combined Use of Wells Scores and D-dimer Levels for the Diagnosis of Deep Vein Thrombosis and Pulmonary Embolism in COVID-19: A Retrospective Cohort Study. Cureus 2021;13:e17687. [Crossref] [PubMed]