The interaction effect between comorbidity and venous thromboembolism on mortality rates after venous thromboembolism: A 5-year cohort study

Abstract

Background: Comorbidity influences venous thromboembolism (VTE) mortality, but it is unknown whether this is due to comorbidity alone or whether biological interaction exists. Objectives: We examined whether comorbidity and VTE interact to increase VTE mortality beyond their individual effects. Methods: This register-based 5-year cohort study included all VTE patients ≥18 years during 2000–2016, and an age-, sex-, and comorbidity-matched comparison cohort of individuals without VTE. We computed age-standardized mortality rates and examined interaction on the additive scale via interaction contrasts (difference in rate differences). Results: After 30-day follow-up, the mortality rate per 1000 person-years among individuals with no comorbidity was 419 (95% confidence interval [CI]: 391–447) in the VTE and 16 (95% CI: 13–18) in the comparison cohort (rate difference: 403). The corresponding mortality rate increased to 591 (95% CI: 539–643) in the VTE cohort and 38 (95% CI: 33–44) in the comparison cohort among individuals with low comorbidity (rate difference: 553). The interaction contrast (150) showed that 25% (150/591) of mortality was explained by the interaction in individuals with low comorbidity. This percentage increased to 56% for moderate and 63% for severe comorbidity. Interaction effects were largest within 30-day follow-up, for provoked VTE, and in young individuals. Dose-response patterns between comorbidity and interaction effects were also observed after 31–365-day and >1–5-year follow-up (p<0.0001). Interaction effects varied between individual comorbidities. Conclusion: Biological interaction between comorbidity and VTE explains a substantial proportion of VTE mortality, and this interaction effect increases with comorbidity burden.