192 Chapter 8 Information on diagnosis, location and treatment of the SBO are listed in Table 3. Out of those that developed a SBO, 94% (N= 83/88) did so following an open procedure. The other five SBOs developed following treatment by laparoscopy which means that SBO developed in 1% (N=5/367) of all laparoscopically treated patients. Surgery was needed to resolve the SBO in 93% (N=82/88) of the patients whilst conservative treatment was used in 4% (N=4/88). In 2% (N=2/88) surgical treatment was deemed necessary but not feasible due to the severe status of the patients, which eventually resulted in the death of both. In 41% (N=36/88) of the patients the SBO was diagnosed during the admission for the primary surgery, the other patients needed to be readmitted. Readmission occurred in a median of 211 days (range: 63 – 6185) and the duration of stay during readmission was a median of 13 days (range: 3 – 124). In 15% (N=13/88) a SBO reoccurred of which 11 underwent surgery and two underwent conservative treatment. Three patients experienced SBO a third time, of which two underwent redo surgery and one conservative treatment. During follow-up, death occurred in 11% (10/88) of the patients who experienced SBO, of which three were related to SBO. One patient deceased within a day due to intestinal ischemia following SBO after small intestinal atresia surgery. The other two patients, who experienced SBO following surgery for ARM and HD, died 15 and 12 days after the diagnosis of SBO due to bowel perforation and subsequent sepsis caused by the SBO. Multivariate analysis risk factors Cox-regression showed that a history of stoma (HR: 3.2, 95%-CI: 2.0-5.2), emergency operation (HR: 2.2, 95%-CI: 1.3-3.7) and post-operative infection (HR: 1.9, 95%-CI: 1.2-3.1) were significant risk factors for SBO development. Whilst congenital nature of a disease (p = 0.216), surgical approach (p = 0.634), the creation of an anastomosis during primary surgery (p = 0.924) and duration of surgery per hour (p = 0.994) were not statistically significant risk factors. Splitting the post-operative infections into categorical different types of infections did not show any significant difference in SBO hazard (superficial: p = 0.948, deep: p = 0.612, central line infection: p = 0.357) in the same model. During the Forward Wald selection, the inclusion of the variable total number of surgical procedures led to a decrease of >10% in the B-coefficient of the variable having a history of stoma. Therefore, confounding was assumed and the variable total number of surgical procedures was excluded from the primary analysis. Because of this confounding a separate subgroup cox regression was performed including only patients who did not
RkJQdWJsaXNoZXIy MTk4NDMw