Preoperative CT findings and interobserver reliability of fournier gangrene

OBJECTIVE. The objective of our study was to delineate CT findings and anatomic areas of involvement of surgically proven Fournier gangrene (FG) and determine interobserver reliability. MATERIALS AND METHODS. This study was a single-center retrospective study of patients with FG who underwent CT before surgical débridement of FG during a 9-year period. Thirty-eight patients with FG, 17 male and 21 female patients, underwent preoperative CT. Two radiologists reviewed the CT studies and recorded findings and anatomic areas of involvement. CT findings were categorized according to a previously described CT scoring system for necrotizing fasciitis and included the presence or absence of fascial air, muscle or fascial edema, fluid tracking, lymphadenopathy, and subcutaneous edema. Cohen kappa was calculated for interobserver reliability. RESULTS. Mean body mass index (BMI [weight in kilograms divided by height in meters squared]) was 42, and 22 of 38 (58%) patients had diabetes. Mean BMI and proportion of patients with diabetes were significantly higher in female patients (mean BMI = 46; 16/21 with diabetes) than male patients (mean BMI = 36; 6/17 with diabetes). CT studies of most patients showed fascial air (36/38 [95%], both readers 1 and 2). Interobserver reliability was substantial to almost perfect for all CT findings except lymphadenopathy, for which it was fair (κ = 0.37). Genital, perineal, and ischiorectal involvement were seen in 87% (33/38), 87% (33/38), and 32% (12/38) of patients for reader 1 and 84% (32/38), 84% (32/38), and 26% (10/38) of patients for reader 2 (κ = 0.29, penis; κ = 0.65, scrotum; κ = 0.91, vulva and labia; κ = 0.68, perineal involvement; κ = 0.80, ischiorectal involvement). CONCLUSION. Most CT findings of FG and anatomic areas of involvement showed good interobserver reliability. A high proportion of female patients with FG were observed in this study compared with prior series.