Three randomised controlled trials (RCTs) (100 people) were included. All studies reported complete follow-up. One study reported adequate allocation sequence generation. Two trials failed to report on sequence generation. None of the trials adequately reported on allocation concealment.
There was no difference in risk of 30-day mortality between LVAD and IABP (RR 1.06, 95% CI 0.68 to 1.66; 24 deaths with LVAD and 20 with IABP; three trials). There was no evidence of heterogeneity.
Haemodynamic parameters were improved with LVAD compared to IABP. Higher cardiac index (WMD 0.35L/min/m2, 95% CI 0.09 to 0.61; three trials), higher mean arterial pressure (WMD 12.8mmHg, 95% CI 3.6 to 22.0; three trials) and lower pulmonary capillary wedge pressure (WMD -5.3mmHg, CI -9.4 to -1.2; three trials) were reported.
Adverse events: Incidence of leg ischaemia and of fever/sepsis were similar with both devices. Bleeding was more frequent with TandemHeart LVAD compared with IABP (RR 2.35, 95% CI 1.40 to 3.93; two trials). The requirement for both packed red blood cells and for fresh frozen plasma was higher with Impella LVAD compared with IABP. Haemolysis was higher with Impella LVAD (P<0.05; one trial).