Personalised approach to venous thromboembolism prophylaxis in heavier critically ill patients (PROSPER): A randomised, parallel-group, electronic medical record (EMR)-embedded feasibility trial

Objective Data are limited regarding optimal low molecular weight heparin dosing for venous thromboembolism prophylaxis in heavier critically ill patients. This trial evaluated the feasibility of a weight-based dynamic dosing algorithm. Design, setting, participants Open-label, electronic medical record-embedded randomised feasibility trial. Patients, weighing ≥100 kg who required thromboprophylaxis were eligible. Enoxaparin was commenced with weight-based dosing and anti-Xa level-guided adjustment (target peak anti-Xa level 0.2–0.4 IU/mL) or usual care (40 mg twice daily). Main outcome measures The primary outcome was peak anti-Xa level at ‘steady state’ (after ≥3 enoxaparin doses). The feasibility threshold was set at 95% of steady state anti-Xa levels in the intervention group being less than the threshold for therapeutic anticoagulant effect (≤0.6 IU/mL). Results Fifty participants (intervention n = 17, control n = 33) received 331 doses of enoxaparin (median of 4 [interquartile range {IQR}: 3–7] doses per participant, and a dose of 50 [IQR: 50–60] mg (intervention) and 40 [IQR: 40-40] mg (control)). Thirty-two ‘steady state’ peak anti-Xa measurements were recorded from 24 patients (12 intervention and 12 control). The proportion of peak anti-Xa levels that were ≤0.6 IU/mL was: 15 of 17 (88%) vs. 15 of 15 (100%). There were fewer peak anti-Xa measurements below the prophylactic range (<0.2 IU/mL) in the intervention group compared with the control group (2 of 17 [12%] vs. 5 of 15 [33%]). Six patients developed venous thrombosis or pulmonary embolism (2 of 17 [12%] vs. 4 of 33 [12%]). Major bleeding occurred in one patient in the intervention group. Conclusions A weight-based dynamic thromboprophylaxis algorithm was implemented via the electronic medical record but did not meet the predefined feasibility threshold.