Subhash Reddy, MD, Jonathan Grotts, MA, David Thoman, MD
Santa Barbara Cottage Hospital
Background
Adhesions account for 50-70% of all Small bowel obstructions in adults. Historically after conservative management fails a laparotomy is performed to enable lysis of adhesions (LOA). There are a growing number of small series demonstrating decreased morbidity and shorter stay with laparoscopic LOA. We sought to identify risk factors for adverse postoperative events after open or laparoscopic LOA by querying a large national database.
Methods
Under the data use agreement for the American College of Surgeons National Surgical Quality Improvement (NSQIP) Program Public Use File, and with the institutional review board approval, we reviewed perioperative variables of patients who underwent LOA with principal diagnosis of bowel obstruction from 2006 to 2010. Patients who underwent any additional procedure apart from LOA were excluded. The primary endpoint was mortality and secondary endpoints included post-operative adverse events. A multivariate model was used to control for pre-operative morbidity, age and BMI. Odds ratio (OR), Chi square, logistic regression were performed using SPSS 2011
Results
Of 3499 patients who had LOA, 768 (21.9%) were done laparoscopically. Overall, patients who had laparotomy were older and had higher ASA scores. There was no difference in total duration of surgery (P=0.12). Postoperative complications were more common in the laparotomy group and length of stay was significantly longer (9 day vs 4 days). Demographics and univariate analysis of postoperative adverse events are listed in the table with odds ratio.
Using multivariate regression analysis: perioperative variables like age > 70, respiratory, sepsis, renal and neurological comorbities (P<0.05) were independently associated with mortality and respiratory adverse events. For wound infection, BMI > 30 and diabetes were independently associated (P<0.05). Preoperative sepsis (OR 2) and steroid (OR 3) usage increased the risk of blood transfusions (P<0.05). Laparoscopic LOA had protective effect on all postoperative adverse events (P<0.05) with maximum benefit in regards to blood transfusions.
Variables |
Open n=2731(%) |
Laparoscopy n=768(%) |
p value |
Mean Age | 63 years | 58 years | p<0.001 |
Male | 1098(40.2) | 248(32.3) | p<0.001 |
Mean BMI | 26 | 27.3 | p<0.001 |
ASA 3,4,5 | 1717(63.4) | 359(46.7) | p<0.001 |
Previois Surgery within 30 days | 111(41.1) | 35(4.6) | p=0.6 |
Total Duration of Surgery (mean) | 60 minutes | 60 minutes | p=0.43 |
Total Length of Stay (mean) | 9 days | 4 days | p<0.001 |
Steroids | 116(4.2) | 20(2.6) | P<0.03 |
Postoperative Wound Infection | 180(6.6) | 16(2.1) |
OR 3.3 (P<0.001) |
Postoperative respiratory adverse events | 194(7.1) | 20(2.6) |
OR 2.9 p(<0.001) |
Postoperative cardiac adverse events | 16(0.6) | 4(0.5) |
OR 1.1 (p=0.8) |
Urinary Tract Infection | 137(5) | 13(1.7) |
OR 3.0 (p<0.001) |
Unplanned Return to OR | 137(5) | 13(1.7) |
OR 1.0 (p=0.9) |
Bleeding Requiring Transfusions | 48(1.8) | 2(0.3) |
OR 6.8 (p<0.001) |
Mortality | 109(4) | 9(1.2) |
OR 3.3 (p<0.001) |
Conclusion
Based on a large national database, lysis of adhesions performed laparoscopically results in significantly fewer postoperative complications and a much shorter hospital stay. However, the patients approached by laparotomy tended to be older with more comorbidities.
Session: Poster Presentation
Program Number: P629