Seventy-nine consecutive patients underwent elective laparotomy for sy
mptomatic hepatocellular carcinoma between 1986 and 1992. In the first
4 years, 51 patients received liver resection without intraoperative
ultrasonography. In the subsequent 2 years, ultrasonography was perfor
med during operation on 28 patients; three were found to have inoperab
le tumours. The technique located tumours in two patients and changed
the extent of liver resection in another two. Eight of 51 patients (16
per cent) who had the parenchymal transection plane determined clinic
ally had histological evidence of tumour infiltration of the resection
margins, compared with none of 25 in whom the plane was determined by
ultrasonography. This difference was significant (P = 0.034). The use
of intraoperative ultrasonography improved patient survival, although
this was not statistically significant. The median survival of patien
ts with adequate resection margins was 9.2 months, compared with 3.9 m
onths in those with inadequate margins. Intraoperative ultrasonography
is useful in surgery for hepatocellular carcinoma.