Severity scoring for prognostication in severe acute pancreatitis (SAP): Multivariate analysis of the ranson score (RS)
Introduction: SAP causes major morbidity and mortality, but an accurate definition of "severe" remains elusive. Clinical, radiologie, and biochemical scores have been described to quantitate pancreatitis, but no method has been proved superior. The RS was described 25 years ago as an 11-criteria, pancreatitis-specific system, but its continued validity has been questioned. We hypothesized that the RS and its individual components can predict the adverse outcomes of mortality (M), ICU stay>7 days (ULOS), and operative debridement of the pancreas and retroperitoneum. Methods: We prospectively studied RS of 58 patients with SAP in a tertiary surgical ICU. The 11 components of the RS (on admission: Age>55, WBG>16K, LDH>350, AST>250, Glucose>200), (within 48h: PaO2<60, Ca++<8, I>0≥6L, Base deficit<4, BUN ↑ ±5 or HCT ↓ 10 from baseline) were collected as raw data and dichotomous variables, as described originally. Any variables of possible significance (p<0.15 by ANOVA) for an outcome were included in a multivariate ANOVA. Demographic data: x±SEM, p<0.05. Results: Mean age 61±2 years, APACHE III 47±3, ULOS 12±3 days, M 12/58 (20.6%). RS was higher with M (5.3 vs. 3.7, p<0.05), and ULOS (5.4 vs. 3.5, p<0.01), but not before debridement. Net fluid balance (I>0) trended higher with M (9.2L vs 4.0L, p=0.07). Base deficit (p=0.04) predicted debridement by univariate ANOVA, so no multivariate model was constructed. RS variables absent from the table were insignificant by univariate analysis. (See Table) Conclusion: RS criteria on admission were poor predictors compared with those at 48h. BUN, Ca++, and to a lesser extent oxygenation predicted outcome in patients with SAP. ULOS > 7 Days Mortality Univariate p Multivtriate p Univariate p Multivariate p Base Deficit .09 < .05 BUN <.01 <.01 <.05 Ca++ <.01 <.01 <.05 I > 0 .09 .07 PO2 < .05 < .01 WBC .12.