Back to Annual Meeting Program

Background: Ileostomy creation has perioperative and postoperative complications, including re-hospitalization for fluid and electrolyte abnormalities. Although several studies have identified predictors of this morbidity, readmission rates remain high.

Methods: Retrospective chart review was performed on patients who had an ileostomy created for any surgical disease by two board-certified colorectal surgeons at a single tertiary institution January 2008-June 2011.

Results: 154 patients were included in this study; 71 (46.1%) were female. The mean age and BMI were 49 (range 16-91 years) and 26.9 (13-52), respectively. The most common indications for ileostomy creation were cancer (39.6%) and inflammatory bowel disease (48.1%). 115 (74.7%) patients had loop ileostomies constructed; 80 (51.9%) were performed laparoscopically and 7.8% were created emergently. The readmission rate for fluid and electrolyte abnormalities was 20.1%, which was 43.7% of total re-hospitalizations. Gender, older age, and BMI were not associated with readmission. Laparoscopy, loop ileostomy and emergency surgery were also not predictive. Inflammatory bowel disease and specifically Crohn’s disease were not significant, nor was previous intestinal resection. Cancer was strongly associated with readmission (X2=4.73, p=0.03) as was neoadjuvant therapy (X2=9.20, p=0.01); after logistic regression, only neoadjuvant remained significant. Examination of potential predictors showed preoperative use of narcotics, fiber, stool softeners, laxatives and anti-diarrheals were not predictive. Preoperative and discharge renal function, sodium and magnesium levels were not significantly associated; however, increased mean potassium level upon discharge trended toward significance (4.21 versus 4.05, p=0.089). Stoma and urine output on the day of discharge were not associated with readmission; number of days with ileostomy output >1500ml/24 hours was also not significant. Length of hospitalization, postoperative ileus, obstruction or sepsis was not predictive. Postoperative chemotherapy and radiation were not statistically significant. 52 (34%) patients were given anti-diarrheals and 22 (14.4%) were given fiber supplements; neither correlated with readmission. 103 (66.9%) patients had stoma reversal; 4 had ileostomy closure early, but only 2 of these patients were readmitted for fluid and electrolyte abnormalities.

Conclusions: Our results show that only neoadjuvant therapy was significantly associated with hospitalization for fluid and electrolyte abnormalities. Therefore, this morbidity does not appear to be preventable. Our study implies that home regimen and follow-up are the main determinants of readmission. Prospective studies focused on diligent stoma monitoring by patients and physicians may be the key to decreasing readmission rates.