Systematic review and meta-analysis of the prevalence of venous thromboembolic events in novel coronavirus disease-2019 patients

Abstract

Background: Emerging clinical evidence has shown that patients with the novel coronavirus disease-2019 (COVID-19) have complications that include venous thromboembolism (VTE), consisting of deep vein thrombosis (DVT) and pulmonary embolism (PE). The prevalence of VTE in patients hospitalized with COVID-19 is unclear.

Methods: Eligible studies on COVID-19 were collected from PubMed, Web of Science, and Embase. Patient characteristics and information were extracted for three categories of patients: consecutive, ICU, and non-ICU group. All PEs and DVTs were diagnosed by computed tomographic pulmonary arteriography and duplex ultrasound examination, respectively. A subgroup analysis of testing strategies in ICU and non-ICU patients for PE and DVT was also performed.

Results: Forty clinical studies involving 7966 patients hospitalized with COVID-19 were included. Pooled VTE prevalence was 13% in consecutive patients (95% confidence interval [CI], 0.05-0.24; I² = 97%), 7% in non-ICU patients (95% CI, 0.01-0.18; I² = 93%), and 31% in ICU patients (95% CI, 0.22-0.42; I² = 91%). ICU patients had the highest prevalence of PE among the three groups (17% [95% CI, 0.12-0.23] vs 8% in consecutive patients [95% CI, 0.04-0.13], 4% in non-ICU patients [95% CI, 0.01-0.08]). ICU patients also had the highest DVT prevalence (25% [95% CI, 0.14-0.37] vs 7% in consecutive patients [95% CI, 0.03-0.14], and 7% in non-ICU [95% CI, 0.02-0.14]). The subgroup analysis showed a three-fold improvement in the PE and DVT detection rates in both ICU and non-ICU patients with COVID-19 when the screening test for VTE was applied. In the settings of screening tests for VTE, ICU patients have a significantly higher prevalence of PE (37% vs 10%; P < .0001) and DVT (40% vs 12%; P = .0065) compared with non-ICU patients.

Conclusions: VTE is common in patients hospitalized with COVID-19, especially among ICU patients. Screening tests for PE and DVT may significantly improve detection rates in both ICU and non-ICU patients with COVID-19 than tests based on clinical suspicion.

Keywords: COVID-19; Deep vein thrombosis; Meta-analysis; Pulmonary embolism; Venous thromboembolism.

Fig 1

The PRISMA flowchart. CTPA, Computed tomographic pulmonary arteriography; DUS, duplex ultrasound.

Fig 2

Pooled results of VTE, PE, and DVT prevalence in patients with COVID-19. DVT, Deep vein thrombosis; ICU, intensive care unit; PE, pulmonary embolism; VTE, venous thromboembolic event.

Fig 3

Subgroup analysis of prevalence of DVT and PE in patients with COVID-19 by different test strategies. A, Subgroup analysis of PE. B, Subgroup analysis of DVT. DVT, Deep vein thrombosis; ICU, intensive care unit; PE, pulmonary embolism.

Supplementary Fig 1 (online only)

Publication bias analysis of VTE, DVT, and PE in consecutive patients. A, C and E, Potential asymmetry of trials' distribution, which indicates a lack of relevant trials with high quality included in the comparisons. The results of Egger's test presented in B, D and F suggested potential publication bias in the results of PE, DVT, and VTE prevelance of consecutive patients. DVT, Deep vein thrombosis; PE, pulmonary embolism; VTE, venous thromboembolism.

Supplementary Fig 2 (online only)

Publication bias analysis of VTE, DVT, and PE in ICU patients. In (A, C and E) Potential asymmetry of trials' distribution, which indicates a lack of relevant trials with high quality included in the comparisons. The results of the Egger's test in B, D and F suggested potential publication bias in the results of DVT prevalence in ICU patients, and no publication bias of PE and VTE prevalence in ICU patients. DVT, Deep vein thrombosis; ICU, intensive care unit; PE, pulmonary embolism; VTE, venous thromboembolism.

Supplementary Fig 3 (online only)

Publication bias analysis of VTE, DVT, and PE in non-ICU patients. In (A, C and E) Potential asymmetry of trials' distribution, which indicates a lack of relevant trials with high quality included in the comparisons. The results of the Egger's test in B, D and F suggested potential publication bias in the results of PE, DVT, and VTE prevalence in non-ICU patients. DVT, Deep vein thrombosis; ICU, intensive care unit; PE, pulmonary embolism; VTE, venous thromboembolism.

Supplementary Fig 4 (online only)

Publication bias analysis of DVT and PE diagnosed by different test strategy. DVT, Deep vein thrombosis; ICU, intensive care unit; PE, pulmonary embolism.