P < 0.001), 1 thus reflecting a more advanced liver dysfunction at the time of surgery, the competing role of which, with regard to CSPH, cannot be assessed with the present meta-analysis. In the present study CSPH remains a determinant of surgical outcome because the investigators selected results that intrinsically bring a significant difference, regarding covariates, between the two groups. In individual series, this imbalance was handled using different statistical approaches, in particular, regression analyses, which cannot be fully captured with the present meta-analytic approach. The investigators indicated the observed (medium) heterogeneity as the consequence of the different methods used for diagnosing CSPH, 1 but a considerable source of heterogeneity may be caused by different covariate distributions among patients (clinical heterogeneity) with or without CSPH in each of the included studies. 3 In 2009, we outlined the role of such selection bias and how it could be reduced by applying a propensity-score match between patients with cirrhosis with or without CSPH. 2 Before match, CSPH patients had significantly worse preoperative liver function and survival rates in comparison to patients without CSPH. After one-toone propensity score matching, patients with and without CSPH (78 patients in each group) had the same preoperative characteristics (d-values: <|0.1|), experiencing the same intra-and postoperative courses. Thus, if this aspect was not taken into account in the individual series included in the meta-analysis, results would be inevitability affected by pooling togheter the selection bias. At this point, it is worth noting that survival rates extracted from our study, and reported in Table 3 of the article, are neither those of the whole study population nor those from the matched sample, but they were derived from the analysis of outliers after match (best compensated patients without CSPH vs. CSPH patients with worst liver function), thus not representative of our findings. 2 For these reasons, we are uncertain that findings from Berzigotti et al. can really help in clarifying the role of CSPH in the selection process of surgical candidates. The present results reinforce the need for well-conducted studies on this topic.