Predictive Factors of Postoperative Ileus After Laparoscopic Colorectal Surgery

E Grzona, MD, F Carballo, MD, M Bun, A Canelas, MD, L Pereyra, MD, N Rotholtz, MD. Hospital Alemán de Buenos Aires

 

Background:

Postoperative ileus (PI) is defined as an interruption of bowel function after
surgery. It is the most important determinant of length of stay (LOS) after abdominal surgery, and thus has significant implications in hospital resource utilization. The pathogenesis is multifactorial. Predictive factors are not fully known. Although laparoscopic surgery reduces the recovery time of bowel transit it has also recorded PI rates.

Objective:

To evaluate the incidence and analyze predictive factors of PI after laparoscopic colorectal surgery.

Methods:

A retrospective study was performed using a prospective collected database of all patients who underwent a laparoscopic colorectal procedure between March 2000 and June 2011. PI was defined as a delay of the postoperative recovery time and bowel transit without a secondary etiology. The patients were divided in two groups: normal recovery (G1) and patients with PI (G2). As potential predictive factors for PI, demographic characteristics, surgery-related variables and disease-related variables were considered. Univariate analysis was performed to identify individual predictive risk factors for PI. Factors with p values below 0.05 were included in a regression model. The results were expressed as odds ratio (OR) and their 95% confidence intervals (CI).

Results:

A total of 869 patients were evaluated. 23 patients were excluded for secondary ileus. Median age was 58.7(15-92) years. 55.1% were men. The indications for surgery were: colorectal cancer 31.5%, diverticular disease 28.1%, polyps19.6%, inflammatory bowel disease 13%, others 7.8%. The procedures performed were: 399 (40%) left colectomies, 169 (20%) right colectomies, 120 (14.2%) proctectomies, 33 (3.9%) subtotal colectomies, 32 (3.7%) proctocolectomies, 25 (2.95 %) total colectomies, 18 (2.1%) segmentary colectomies, 16 (1.89%), Hartmann’s reversal, 15 (1.8%) ileocecal resections, 10 (1.2%) abdominoperineal resections and 3 (0, 35%) Hartmann’s procedures. 6% from all patients presented PI. There were no differences in demographics data between the groups (G1: 795; and G2: 51). Recovery parameters of bowel function were significantly shorter in G1 [Bowel sounds: 0.85 vs. 1.4 days (p <0.05), flatus 1.57 vs. 2.8 days (p <0.05), liquids intake 1.1 vs. 5.9 days, (p <0.05) solid intake 2.1 vs. 7.2 (p <0.05) days] and LOS [3.9 (2-8) vs. 9.43 (3-20) days (p <0.05)]. Surgical time >180 min [G1: 308/795 (38.7%) vs. G2: 28/51 (54.9%) OR 1.85 (CI: 1.05-3.25, p <0.05)], Conversion [G1: 73/795 (9.1%) vs. G2: 14/51 (27.45%) OR: 10.95 (CI 5.24-16.66, p <0.01)], left colectomy [G1: 386/795 (48.5%) vs. G2: 13/51 (25.5%) OR (-5.24) (CI: (-1.87) – (-8.61), p <0.01)] and total colectomy [G1: 20/795 (2.51%) vs. G2: 5/51 (9.8%), OR: 14.4 (CI: 3.33-25 .47, p <0.05)] showed association with PI in univariate analysis. In multivariate analysis, significant predictors of PI were total colectomy [OR: 13.5 (CI: 2.18-10.47)] and the conversion [OR: 3.84 (CI: 1.84-8.04)]; whereas left colectomy are a protector variable of PI [OR: 0.37 (CI: 18-75)].

Conclusion: Conversion to open surgery and total colectomy were independent predictive factors of PI, while left colectomy appears as an independent protective factor.

 


Session Number: Poster – Poster Presentations
Program Number: P111
View Poster

« Return to SAGES 2012 abstract archive