Discriminant power and information content of Ranson's prognostic signs inacute pancreatitis: A meta-analytic study

Objective: To determine a meta-analytical definition of the discriminant po wer of Ranson's signs in the prediction of acute pancreatitis severity and outcome, and of their information content, also compared with clinical judg ment. Data Sources: Two hundred eleven studies since 1974, reporting any predicti ve system for acute pancreatitis (MEDLINE by various Medical Subject Headin gs in MEDLARS, Current Contents, Medscape, Virtual Hospital, and other on-l ine medical services). Study Selection: One hundred ten studies reporting clinical use of Ranson's signs were retained. A quality index was calculated for each study. A sele ction was made according to inclusion criteria, separately far prediction o f severity (19 studies; group S) and prognosis (10 studies; group P). Six o ther studies reporting clinical judgment results were also selected (group C). Data Extraction: Sensitivity and specificity values were extracted. Effect sizes were calculated and summarized by the inverse variance-weighted metho d. Categorical models were studied by analysis of variance. Publication bia s was sought by correlation test and analysis of variance. Summary receiver operating characteristic curves were drawn, and the corresponding false-po sitive rate (FPR) and true-positive rate were calculated for each group. Fr om the total true-positive rate and FPRs, the probabilities of illness for positive and negative results were calculated, for severe pancreatitis prev alence from 0 to 1. Last the area below the curve and the ratio between thi s and that of the "perfect test" were calculated as a measurement of inform ation content. Data Synthesis: Ranson's signs demonstrated poor discriminant power in bath predictions: "d" values were 1.200 (95% confidence interval, 1.083-1.318) and 1.302 (95% confidence interval, 1.046-1.559), respectively. The tack of homogeneity in group S (Q = 58.737; p = .0000032) can be explained by the presence of three outliers. The summary curves showed, for low FPRs, a high er sensitivity of clinical judgment; Ranson's signs reached useful sensitiv ity only for high FPRs. Mo differences between groups in the area below the information content curves were found. Conclusions: Ranson's signs showed a poor predictive power. The information content did not differ from that of clinical judgment.