|
Back to Annual Meeting Program
Adult Intussusception: A Decision Tree Analysis for Operative Management
Ali Ardestani, MD, MSc, Oliver Varban, MD, Dan Azagury, MD, Bela Kis MD, PhD, David Brooks, MD, Ali Tavakkolizadeh, MD
Brigham and Woman's Hospital, Boston, MA, Harvard Medical School, Boston, MA
Background: Adult intussusception (INT) is rare and often treated surgically because of the possibility of a pathologic lead point. We aimed to identify the factors that increase the probability of discovering a lead point and to develop an algorithm for management of INT.
Methods: We identified adult patients who were diagnosed by CT to have an INT and managed operatively between 1996 and 2010. Cases were classified based on intussusception site. Considering risk factors for having a pathological lead point, a decision tree was developed and tested to verify its impact on patient management.
Results: We had 9 (12%) colonic (C-INT), 41 (55%) small bowel (SB-INT) and 24 (33%) retrograde (R-INT) intussusception cases. All patients with C-INT had a pathologic lead point whereas none were found in R-INT cases. 46% of SB-INT cases had a lead point, with the relative risk of finding a lead point increased by: history of malignancy (RR=3.8; p<0.001), mass on CT (RR=3.0; p<0.001), and age >65 (RR=2.5; p<0.04). A decision tree algorithm was created and tested using our initial cohort, showing that absence of a lead point could have been predicted in 20% of cases that underwent bowel resection.
Conclusions: A pathological lead point is extremely common in cases of a C-INT, mandating surgical intervention, and absent in cases of R-INT. Using our treatment algorithm, developed based on identified risk factors for finding a lead point, we could have predicted lack of a concerning mass in 20% of cases and altered management.
Back to Annual Meeting Program
|
