Surveillance or no surveillance for deep venous thrombosis and outcomes of critically ill patients A study protocol and statistical analysis plan

Objective: Surveillance ultrasounds in critically ill patients detect many deep venous thrombi (DVTs) that would otherwise go unnoticed. However, the impact of surveillance for DVT on mortality among critically ill patients remains unclear. Design: We are conducting a multicenter, multinational randomized controlled trial that examines the effectiveness of adjunct intermittent pneumatic compression use with pharmacologic thromboprophylaxis compared to pharmacologic thromboprophylaxis alone on the incidence of proximal lower extremity DVT in critically ill patients (the PREVENT trial). Enrolled patients undergo twice weekly surveillance ultrasounds of the lower extremities as part of the study procedures. We plan to compare enrolled patients who have surveillance ultrasounds to patients who meet the eligibility criteria but are not enrolled (eligible non-enrolled patients) and only who will have ultrasounds performed at the clinical team's discretion. We hypothesize that twice-weekly ultrasound surveillance for DVT in critically ill patients who are receiving thromboprophylaxis will have more DVTs detected, and consequently, fewer pulmonary emboli and lower all-cause 90-day mortality. Discussion: We developed a detailed a priori plan to guide the analysis of the proposed study and enhance the validity of its results.