Risk of posthospital venous thromboembolism in patients with COVID-19 varies by SARS-CoV-2 period and vaccination status

TO THE EDITOR:

Ambulatory patients with coronavirus disease 2019 (COVID-19) have a low risk of developing venous thromboembolism (VTE), whereas hospitalization is associated with a higher risk, necessitating thromboprophylaxis.^1,2 The use of extended thromboprophylaxis following COVID-19 hospitalization is controversial, and risk stratification tools are needed to better predict who may benefit. We report the 90-day incidence of VTE following hospitalization, stratified by COVID-19 surge period and vaccination status among adults tested for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).

We performed a retrospective cohort study of 63 920 adult Kaiser Permanente Northern California members who were hospitalized within 30 days of SARS-CoV-2 polymerase chain reaction testing from 1 December 2020 to 28 February 2022. If multiple SARS-CoV-2 tests were performed, the index date was the first date with a positive result or the first date with a negative result if all tests were negative. Based on the predominant SARS-CoV-2 variant circulating in California, we defined the pre-Delta period from 15 December 2020 to 31 May 2021, the Delta (B.1.617.2) period from 1 June 2021 to 30 November 2021, and the Omicron (B.1.1529) period from 1 December 2021 to 28 February 2022. We excluded subjects who were asymptomatic at the time of SARS-CoV-2 testing, had a history of VTE, or received anticoagulation in the prior year.

We assessed the incidence and timing of VTE within 90 days of hospital discharge using a combination of diagnosis codes, new anticoagulant prescriptions, and encounters for VTE with a centralized anticoagulation service.^3,4 We stratified patients by the International Medical Prevention Registry on Venous Thromboembolism (IMPROVE VTE) score to assess VTE risk in hospitalized patients for extended thromboprophylaxis (low/moderate [0-3]; high [4+] risk).⁵ COVID-19 vaccination was defined as the receipt of 2 doses of a mRNA vaccine (BNT162b2 [Pfizer-BioNTech]; or m-RNA-1973 [Moderna]) or 1 dose of the Ad.26.COV2.S [Janssen] vaccine 14 or more days before viral testing. Multivariable Cox regression was used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for 90-day posthospital VTE. The Kaiser Permanente Northern California institutional review board approved the study and waived informed consent.

Of the 63 920 hospitalized patients tested for SARS-CoV-2 (mean age [SD], 59 [20] years; 35 791 [56%] women), there were 10 084 (15.8%) unvaccinated patients with COVID-19 and 3443 (5.4%) subjects with COVID-19 vaccine breakthrough infections (Table 1). The 90-day posthospital VTE incidence among SARS-CoV-2 negative patients (Table 2) was 0.7% (343/50 393) and did not vary across the 3 variant periods (P = .86). The incidence of posthospital VTE was higher in both unvaccinated (1.6%; 160/10 084) and vaccinated (1.2%; 40/3443) patients with COVID-19 than in SARS-CoV-2 negative patients (P < .001). Time from hospital discharge to VTE diagnosis was shorter in patients with COVID-19 (12 days [IQR, 5-23]) than in SARS-CoV-2 negative patients (22 days [IQR, 9-40]; P < .001) and did not differ by surge period in patients with COVID-19 (P = .245).

During the pre-Delta period, there was a higher incidence of VTE events in unvaccinated cases (1.3% [60/4738]) than in SARS-CoV-2 negative cases (0.7% [144/20 644]; P = .001), and vaccine breakthrough hospitalizations were rare with only one 90-day post hospital VTE event (1/33 [3.0%]; P = .11). During the Delta surge period, 90-day post hospital VTE was higher among both unvaccinated (2.1% [79/3736) and vaccinated patients with COVID-19 (1.7% [17/1029]) than in SARS-CoV-2 negative patients (0.7% [100/14 531]; P = .001 for both). However, during the Omicron period, VTE incidence in vaccine breakthrough (0.9% [22/2381]) and SARS-CoV-2 negative patients (0.7% [99/15 218]) were similar but higher in unvaccinated patients compared with SARS-CoV-2 negative patients (1.3% [21/1610]; P < .001).

With multivariable regression, the adjusted HR for 90-day post-hospital VTE during the Delta surge period was 2.2 (95% CI, 1.3-3.6; P = .002) for vaccinated patients with COVID-19 and 4.1 (95% CI, 3.1-5.3; P < .001) for unvaccinated patients with COVID-19 compared with SARS-CoV-2 negative patients (Table 2). The HRs during the Omicron surge period were 1.5 (95% CI, 1.0-2.3; P = .083) and 2.2 (95% CI, 1.4-3.4; P < .001) for vaccinated and unvaccinated patients with COVID-19, respectively.

When examining patients with low/moderate IMPROVE VTE risk score during the Delta period, we again found higher 90-day post-hospital VTE incidence in both vaccinated (1.6% [16/1000]) and unvaccinated patients with COVID-19 (2.1% [76/3699]) than the SARS-CoV-2 negative patients (0.6% [86/14 117]; P < .001). However, during the Omicron period, post hospital VTE incidence was similar in vaccinated COVID-19 and SARS-CoV-2 negative patients with low/moderate IMPROVE VTE risk (0.7% [17/2327] vs 0.6% [83/14 835]; P = .313) but higher in unvaccinated patients with COVID-19 (1.3% [20/1579]; P < .05 for both).

We found that the incidence of 90-day post hospital VTE was higher in both unvaccinated and vaccinated patients with COVID-19 compared with SARS-CoV-2 negative patients, and, notably, that unvaccinated patients with COVID-19 had a higher risk of post hospital VTE than patients with COVID-19 vaccine breakthrough infections. When examined by study period, patients with COVID-19 had a higher incidence of VTE in the Delta period than either the pre-Delta or Omicron period. Moreover, patients with vaccine breakthrough infections during the Omicron period had a risk of post hospital VTE that was not significantly above the risk for hospitalized SARS-CoV-2 negative patients, whereas unvaccinated patients with COVID-19 had a significantly higher risk. Increased incidence of post hospital VTE in patients with COVID-19 persisted when examining the subset of patients with low/moderate IMPROVE VTE risk scores. This finding was consistent across study periods in unvaccinated patients with COVID-19 but only in the Delta period among COVID-19 vaccine breakthrough patients.

Clinical risk factors associated with post hospital VTE in patients with COVID-19 include advanced age, previous VTE, intensive care unit stay, chronic kidney disease and cardiovascular disease.⁶ Our results suggest that SARS-CoV-2 variant and vaccination status may be independent risk factors for post hospital VTE in patients with COVID-19. Although there are no prospectively validated risk assessment models in patients with COVID-19, the IMPROVE VTE score has been externally validated in hospitalized patients with COVID-19.^7,8 The MICHELLE trial used the IMPROVE VTE score and D-dimer levels to identify high-risk patients with COVID-19 following hospitalization and randomized them to prophylactic oral anticoagulation or placebo for 35 days.⁹ The reduction in major and fatal thromboembolic events in patients on anticoagulation was driven by a reduction in asymptomatic, symptomatic, and fatal pulmonary emboli. However, the trial was conducted during the period before the SARS-CoV-2 Delta variant predominance and widespread COVID-19 vaccination.

There are limitations to our findings. Residual confounding may persist despite multivariable modeling given the retrospective nature of our analysis. In addition, we excluded patients with a history of VTE, potentially impacting post hospital VTE incidence. D-dimer testing was not routinely performed in patients admitted to the hospital; therefore, we may have not identified all subjects at high risk of post hospital VTE using a more dynamic risk assessment. The strengths of our study included evaluation of only symptomatic events and a large diverse cohort with a relatively high vaccination rate, allowing cross group comparisons of an infrequent event.

Our data support continued re evaluation of the risks and benefits of antithrombotic treatments with ongoing surges of novel SARS-CoV-2 variants. We await the publication of ongoing randomized clinical trials (XACT [NCT04640181], ACTIV-4c [NCT04650087], HEAL-COVID [NCT04801940] and others) to assess the benefit of extending thromboprophylaxis in patients with COVID-19 post-hospital discharge. Prospective clinical trials are needed to validate the effect of SARS-CoV-2 variant and vaccination status on the incidence of posthospital VTE, and whether this is reduced by extended thromboprophylaxis.

Acknowledgments: The authors thank Mei Lee and Wei Tao from the Kaiser Permanente Division of Research for their assistance in data collection.

Funding for this project was provided by the National Heart, Lung, and Blood Institute (R01HL126130). The funding agencies had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Contribution: N.R. and A.P. conceived and designed the study, analyzed the data, and wrote the first draft of the manuscript; and all authors participated in analyzing the data and critical revision of the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Ashok Pai, Kaiser Permanente Oakland Medical Center, 3701 Broadway, 3rd Floor, Oakland, CA 94611; e-mail: ashok.p.pai@kp.org.