Impact of Hospital-Acquired Venous Thromboembolism on Surviving Medical Admission: Findings from the Medical Inpatient Thrombosis and Hemostasis Study

Introduction

Approximately 200,000 venous thromboembolism (VTE) events occur each year during or within 3 months after medical (non-surgical) hospitalizations in the U.S. The consequences of hospital-acquired VTE (HA-VTE, VTE occurring while hospitalized), such as likelihood of dying during admission, are unknown. The objective of this study was to examine the association between HA-VTE and risk of in-hospital mortality.

Methods

Adults admitted to medical services for at least 1 midnight at one of 5 hospital systems from 2016-2022 were included. Admissions for primary diagnosis of VTE were excluded. Each admission with HA-VTE was matched with up to 4 admissions without HA-VTE based on day of hospitalization, age (± 5 years), sex, race and ethnicity, hospital system, level of care (medical services versus intensive care unit [ICU]), calendar year, and COVID-19 status at admission. The primary outcome was in-hospital mortality. We used conditional logistic regression to estimate the association of HA-VTE with in-hospital mortality, adjusting for the matching factors, anticoagulation level at admission (therapeutic, prophylactic, or none), and the Elixhauser Comorbidity Index. We tested whether the association varied by COVID-19 status.

Results

The matched cohort included 2,061 medical admissions with HA-VTE and 7,908 medical admissions without HA-VTE. In the matched admissions, the mean age was 58 years (SD 16), 57.1% identified as male, 54.5% as White race, 26.6% as Black race, 24.0% were admitted to an ICU, and 10.8% were admitted with COVID-19. The mean Elixhauser Comorbidity Index score was 14.3 (SD 18.4) and 12.2 (SD 17.3) in admissions with and without HA-VTE, respectively. In-hospital mortality occurred in 12.4% of HA-VTE admissions and 5.1% of non-HA-VTE admissions. Compared to non-HA admissions, the adjusted odds ratio (OR) for in-hospital mortality with HA-VTE was 2.71 (95% confidence interval [CI] 2.23, 3.29). The association of HA-VTE with in-hospital mortality was similar for admissions with COVID-19 (OR 2.93; 95% CI 2.33, 3.70) and admissions without COVID-19 (OR 2.18; 95% CI 1.49, 3.18; p_interaction = 0.12).

Conclusions

HA-VTE was associated with a nearly 3-fold increased odds of death during hospitalization in a diverse patient cohort from 5 hospital systems. While increased illness severity may account for some of the increased mortality risk, HA-VTE itself or interventions to treat HA-VTE (anticoagulation) likely contribute. Efforts should be made to understand whether reducing HA-VTE can translate to decreased mortality.