Validation of the Fournier's Gangrene Severity Index in a Large Contemporary Series

Abstract

In this study we identified prognostic factors for survival and validated the accuracy of the Fournier's gangrene severity index in patients with Fournier's gangrene. We retrospectively reviewed medical records of patients diagnosed with Fournier's gangrene between 1996 and 2006. Fournier's gangrene severity index scores were assessed using a receiver operating characteristic curve. Using an outcome variable of inpatient mortality, univariate analyses were performed using the Mann-Whitney U, chi-square and Fisher exact tests. A total of 68 patients (79.4% male, mean age 55.8 +/- 15.2 years) diagnosed with Fournier's gangrene met the criteria for review. The inpatient mortality rate was 10% (7 patients). The mean Fournier's gangrene severity index score for survivors was 5.4 +/- 3.5 vs 10.9 +/- 4.7 for nonsurvivors (p = 0.006). Isolated Fournier's gangrene severity index and individual laboratory parameters associated with mortality included heart rate (p = 0.05), respiratory rate (p = 0.02), serum creatinine (p = 0.03), serum bicarbonate (p = 0.001), serum lactate (p = 0.001) and serum calcium (p = 0.03). Although mean total body surface area was only suggestive of an association (p = 0.169), abdominal wall (p = 0.004) or lower extremity (p = 0.005) involvement was associated with increased mortality. Using a Fournier's gangrene severity index score threshold of 9 (sensitivity 71.4%, specificity 90%) there was a 96% survival rate in patients with a Fournier's gangrene severity index of less than 9 and a 46% mortality rate in those with a Fournier's gangrene severity index of 9 or greater (p = 0.001, OR 22, 95% CI 3.5-139.7). The Fournier's gangrene severity index remains an objective and simple method to quantify the extent of metabolic aberration at presentation in patients with Fournier's gangrene. A Fournier's gangrene severity index threshold value of 9 is sensitive and specific for predicting mortality in this patient population.