Abstract
OBJECTIVE
The purpose of this study was to evaluate the accuracy of MRI for the diagnosis of internal hernia (IH) in pregnant women who have undergone Roux-en-Y gastric bypass (RYGB).
MATERIALS AND METHODS
Fifteen consecutively registered pregnant women (eight with surgically proven IH, seven without IH) who had previously undergone RYGB underwent MRI to rule out IH between July 2011 and July 2016. Two blinded radiologists retrospectively evaluated MRI examinations for the presence or absence of 13 established CT findings of IH. The final diagnosis of IH was evaluated subjectively and with two previously validated CT models (model 1, mesenteric swirl or small-bowel obstruction; model 2, beaking of the superior mesenteric vein or small-bowel obstruction). Diagnostic odds ratio (DOR) and interobserver agreement were calculated for each feature, and the subjective and model-based diagnoses of IH were compared by chi-square test.
RESULTS
There were no statistically significant differences in patient age (p = 0.68), gestational age (p = 0.35), or time since RYGB (p = 0.55) between patients with and those without IH. The findings with best DOR and interobserver agreement were beaking of the superior mesenteric vein (reader 1 DOR, 39; reader 2 DOR, 39; κ = 1.00), mesenteric swirl (reader 1 DOR, 11; reader 2 DOR, 39; κ = 0.86), engorgement of mesenteric vessels (reader 1 DOR, 24; reader 2 DOR, 15; κ = 0.84), and mesenteric edema (reader 1 DOR, 11; reader 2 DOR, 3; κ = 0.73). The other findings had either low accuracy, poor interobserver agreement, or both. The overall sensitivity and specificity of the diagnosis of IH ranged from 75% to 88% and 86% to 100% for the two readers. There was no difference in diagnostic accuracy between the three methods (p = 0.93).
CONCLUSION
MRI is useful in the diagnosis of IH in pregnant women who have undergone RYGB. The specificity is comparable to that of CT, and the sensitivity approaches that of CT.
Collapse