Four RCTs (2,753 participants) were included.
In terms of study quality, scores ranged from 58 to 63 (out of a possible 91). The inter-rater agreement was high (intra-class correlations were between 0.92 and 0.98).
Adjusted-dose warfarin reduced the risk of all thrombotic events compared with low-dose warfarin; the pooled RR was 0.50 (95% CI: 0.25, 0.97, P=0.041). There was no statistically significant heterogeneity (P=0.56).
Adjusted-dose warfarin tended to decrease ischaemic stroke compared with low-dose warfarin. However, this result was not statistically significant (RR 0.46, 95% CI: 0.2, 1.07, P=0.071). No heterogeneity was found (P=0.41).
Although there was some trend for adjusted-dose warfarin to increase systemic embolism, this was not statistically significant (RR 1.18, 95% CI: 0.33, 4.31).
There was no difference between the two treatment groups in terms of vascular death; the RR was 1.1 (95% CI: 0.72, 1.67).
There was no difference in haemorrhagic death between the two treatments; the RR was 0.97 (95% CI: 0.27, 3.54). Although mini-dose warfarin tended to reduce the risk of major haemorrhage, this was not statistically significant (RR 1.23, 95% CI: 0.67, 2.27, P=0.51).
When only groups that did not include aspirin as part of the mini-dose regimen were compared with the adjusted dose, there was no difference in the risk of any thrombotic events; the RR was 0.63 (95% CI: 0.38, 1.04).