To evaluate the effectiveness of venous thromboembolism (VTE) prophylaxis with sequential compression devices (SCD) in medical ICU patients and the impact of VTE on ICU length-of-stay and mortality.
We evaluated 15985 patients treated at 92 ICUs participating on the Project IMPACT from 11/2001 till 9/2002. Patients were divided into 3 groups according to the choice of VTE prophylaxis: SCD, low-dose unfractionated heparin or low molecular weight heparin (HEP), or a combination of methods (COMBO). Propensity scores for using SCD versus other methods were constructed using logistic regression.
Ninety-seven percent (15,499) of patients received VTE prophylaxis and were included in the analysis: 39% received SCD, 47% HEP, and 14% COMBO. Females comprised 47% of patients. The overall mean age (±SD) was 62±18 year-old, and SAPS II score 37.5±16. Two hundred and fourteen patients had clinically apparent VTE (1.4%). There were 160 episodes of DVT, 50 PE, and 22 catheter-related thromboses. The incidence of VTE according to prophylaxis choice was: 1.3% SCD, 1.0% HEP, and 3.0% COMBO (p<0.0001). After adjusting for propensity scores, the number of central venous catheters, right heart catheterization, previous VTE, obesity, and ICU-stay ≥48h were found to be independently associated with VTE (p<0.01 for all), whereas the choice of prophylaxis was not. Patients with VTE had a longer ICU length-of-stay (median 12 vs. 3 days, p<0.0001) and mortality (20% vs. 11%, p<0.0001) than those without VTE. After controlling for predictors of ICU mortality, the hazard of dying was 2.3 fold higher in patients with VTE (95%CI 1.98 to 2.68, p<0.0001).CONCLUSIONS: 1) The odds of developing VTE were similar in patients receiving SCD or HEP; 2) VTE was independently associated with increased ICU length-of-stay and mortality.
SCD appears to be an effective method of VTE prophylaxis in medical ICU patients. Although the rates of VTE are low when patients receive prophylaxis, clinically apparent events negatively impact the length-of-stay and mortality of ICU patients.
A.T. Rocha, None.