Background: High rates of venous thromboembolism (VTE) have been reported with coronavirus disease 2019 (Covid-19). However, studies have predominantly focused on hospitalized patients. The risk of VTE in outpatients with Covid-19 is less known.

Methods: We conducted an observational study of adults (age ≥18 years) enrolled in two large, integrated healthcare delivery systems serving >9 million people in California. We identified from electronic health records all outpatients with laboratory-confirmed Covid-19 between January 1, 2020 and January 31, 2021. Follow-up continued until February 28, 2021. We identified all new diagnoses of acute VTE diagnosed in any setting (inpatient, emergency, or outpatient) using an algorithm based on diagnosis codes, radiology tests, and natural language processing. The goal was to describe the cumulative incidence of VTE after Covid-19 infection and identify risk factors for VTE.

Results: We identified 398,530 patients with laboratory-confirmed Covid-19 who were not hospitalized. The mean age was 43.8 years and 53.7% were women. Out of 292 VTE events, 168 (57.5%) were pulmonary emboli, 102 (34.9%) lower extremity DVTs, 15 (5.1%) upper extremity DVTs, and 7 (2.5%) other VTEs. The overall rate of VTE was 0.26 per 100 person-years (95% CI: 0.24-0.30), with the sharpest increase in risk observed during the first 30 days of infection (0.55 per 100 person years in the first 30 days vs 0.07 per 100 person years after 30 days). In a Cox multivariable regression model, factors associated with VTE risk included advancing age, with the greatest risk occurring in those ≥85 years (adjusted hazard ratio [aHR] 6.58 [3.06-14.13] vs. age ≤ 54), prior history of VTE (aHR 7.39 [4.24-12.86]), and increasing body mass index (BMI), with the greatest risk occurring in those with BMI ≥40.0 (aHR 3.21 [2.03-5.07] vs. BMI 18.5-25). Women had a lower risk for VTE than men (aHR 0.67 [0.51-0.87]).

Conclusions: The risk of VTE was low among non-hospitalized patients with Covid-19 and primarily observed during the first 30 days of infection. As a point of reference, the annual rate of VTE in the general population has been described in the literature as approximately 0.1 – 0.2 per 100 person-years. Increased VTE risk among patients in our study was associated with older age, being male, having a prior history of VTE, and high BMI.

IMAGE 1: Cumulative incident function plot of venous thromboembolism, adjusting for competing risk of death or disenrollment, among 398,530 outpatients with COVID-19