Four RCTs were included in the review (n=320 patients). The overall trial quality was reported to be moderate. None of the trials reported adequate allocation sequence generation, sample size calculation or clearly reported statistical methods. Other flaws included lack of reporting of withdrawals and drop-outs. Two trials compared portal triad clamping with selective hepatic vascular exclusion (n=150 patients. One trial compared portal triad clamping with a modified technique of hepatic vascular exclusion (n=118 patients). One trial compared portal triad clamping with total hepatic vascular exclusion (n=52 patients).
Morbidity and mortality: There was no significant difference in either morbidity or mortality between portal triad clamping and any of the techniques of hepatic vascular exclusion (four RCTs). There was significantly reduced overall morbidity for the portal triad clamping group compared with the total hepatic vascular exclusion group (OR 0.27, 95% CI 0.08 to 0.94; one RCT). No significant heterogeneity was found.
Intraoperative blood loss: Compared with the selective hepatic vascular exclusion group, there was significantly higher blood loss for both the portal triad clamping group (WMD 116.00mL, 95% CI 4.33 to 227.67; two RCTs) and the modified technique of hepatic vascular exclusion group (WMD 350.00mL, 95% CI 246.16 to 453.84; two RCTs).
Blood transfusion: Compared with the total hepatic vascular exclusion group, there was no difference in the numbers of patients in the portal triad clamping group requiring blood transfusion (four RCTs). Significantly fewer patients required blood transfusion from the hepatic vascular exclusion groups when portal triad clamping was compared with selective hepatic vascular exclusion (OR 2.86, 95% CI 1.31 to 6.22) and the modified technique of hepatic vascular exclusion (OR 5.66, 95% CI 2.29 to 14.00). This advantage of hepatic vascular exclusion was also present for all identified trials (OR 3.38; 95% CI 1.94 to 5.86; four RCTs; I2=6.4%).
Warm ischaemic time: Patients receiving portal triad clamping had a significantly reduced warm ischaemic time compared to total hepatic vascular exclusion patients (WMD −7.00 minutes, 95% CI −13.21 to −0.79; one RCT).
Aspartate aminotransferase level: Postoperative aspartate aminotransferase values were significantly lower in the portal triad clamping than the selective hepatic vascular exclusion group (WMD −100.00U/L, 95% CI −131.01 to −68.99; two studies).
Funnel plots suggested a low likelihood of publication bias.