Sixteen RCTs (6,055 patients) were included.
Efficacy (13 RCTs, 5,568 patients).
Recurrent VTE was significantly less common in patients treated with LMWH, compared with UFH; the OR was 0.66 (95% CI: 0.51, 0.86). A high absolute VTE recurrence rate in the UFH group (range: 0.0 to 8.2%) was associated with a higher relative efficacy of the LMWH compound (p=0.007). The relationship was stronger after adjusting for the proportion of patients with cancer (p=0.001). The year of publication and a diagnosis of deep venous thrombosis or pulmonary embolism did not influence this relationship.
Safety (16 RCTs, 6,055 patients).
Major haemorrhage was significantly less common in patients treated with LMWH, compared with UFH; the pooled OR was 0.56 (95% CI: 0.38, 0.83). No relationship was found between the absolute percentage in UFH groups (range: 0.0 to 8.65%) and the safety of the LMWH groups (plot of log OR against absolute percentage of major bleeding in the UFH group: beta = -0.015, p=0.845). Corrections for the frequency of cancer or the year of publication did not change this finding.
Mortality (12 RCTs, 5,364 patients).
Mortality was significantly less common in patients treated with LMWH, compared with UFH; the pooled OR was 0.68 (95% CI: 0.53, 0.88). There was a non-statistically significant trend towards a smaller treatment effect with higher mortality in the UFH group (range: 1.4 to 10.6%; plot of log OR against absolute mortality in the UFH group: beta=0.06, p=0.132). Corrections for the proportion of patients with cancer in the various studies, or the year of publication, did not change this association.
Comparison of different LMWH compounds.
The studies involved 7 different LMWH preparations with most compounds being studied in only one trial. No studies directly compared different LMWH compounds. In view of this, the authors stated that the interpretation of analyses comparing compounds was difficult.
Dalteparin (3 RCTs) appeared to be significantly less effective than other LMWH compounds (plot of log OR against absolute percentage of recurrent VTE in the UFH group: beta=0.730, p=0.30).
Dalteparin (3 RCTs) appeared to be significantly less likely to be associated with major haemorrhage than other LMWH compounds (plot of log OR against absolute percentage of recurrent VTE in the UFH group: beta = -0.592, p=0.008.