[Algorithm in hepatology. Blind biopsy or laparoscopic biopsy?].

Data from 1035 laparoscopic examinations were evaluated with 1% general morbidity and 0.1% death rate - Comparative diagnostic sensibility, specificity and predictive value between endoscopic and histologic findings from 748 cases (408 diffusse - parenchimatous liver diseases, 246 focal - nodular lesions, and 94 normal liver) were calculated. General sensibility (s) was similar for both laparoscopy and histology (s = 93) with different behavior for diffusse liver disease (DLD) and focal-nodular disease (FND) as follows: DLD: Laparoscopy s = 90 +/- 3 Vs. Histology s = 100. FND: Laparoscopy s = 98 +/- 1.8 Vs. Histology s = 85 +/- 4.8 Laparoscopic specificity (Sp) for FND was high (Sp = 100), but 10.6% false positive was diagnosed as DLD (Sp = 90 +/- 6.5). Non histologic false positive were found. High positive predictive value (p.p.v.) was found all cases: DLD: Laparoscopy ppv = 0.97 Vs. Histology p.p.v. = 1. FND: Laparoscopy ppv = 1 Vs. Histology p p.p.v. = 1. On the other hand, negative predictive value (npv) evidences well diferentiated results: DLD Laparoscopy n.p.v. = 0.69 Vs. Histology n.p.v. = 1. FND Laparoscopy n.p.v. = 0.94 Vs. Histology n.p.v. = 0.68. Histologic findings can not replace high fiability of macroscopic ones for FND diagnosis; then, blind liver biopsy is not recommended for this cases, but it could be very useful for DLD diagnosis.