SAN DIEGO – Pre- and postsurgical infections top the list of factors in putting patients at risk of wound dehiscence after laparotomy, a database study has found.
Before surgery, a contaminated or dirty wound and sepsis doubled the risk of a post-laparotomy dehiscence, Anam Pal*, MD, said at the Association for Academic Surgery/Society of University Surgeons Academic Surgical Congress.
After surgery, a deep wound infection raised the risk by more than four times, and a superficial wound infection almost tripled the risk, said Dr. Pal, a second-year surgical resident at Hofstra Northwell School of Medicine at Staten Island University Hospital Program, New York.*
Dr. Anna Pal
“Since infections are the strongest predictors, we need more aggressive efforts to prevent surgical site infections in these patients,” she said. Any patient who displays these risk factors should have retention sutures placed during closing as an extra measure of precaution against the potentially devastating complication.
Dr. Pal said the time is right for a new risk model of wound dehiscence after abdominal laparotomy. The existing predictive tool is almost 20 years old and was validated in the Veterans Affairs Surgical Quality Improvement Program database.
“This risk score was created using patient data gathered from 1996 to 1998 on the VA population. We know that this group is older and sicker than the general population,” she said. In fact, she ran that calculation on her own dataset and found that it “grossly overestimated” the risk of wound dehiscence in a general population. “This raises questions about the generalizability of that score.”
Among the 18,306 exploratory laparotomies in Dr. Pal’s dataset, there were 275 cases of wound dehiscence, for a rate of 1.5%.
There were striking baseline differences between the patient groups, she noted. Generally, patients with wound dehiscence were sicker and frailer than those without. “There was significantly more smoking, chronic obstructive pulmonary disease, diabetes, pneumonia and ventilator placement, obesity, and disseminated malignancy.”
She also noted significantly higher rates of wound infection and steroid use. Patients with dehiscence were significantly less likely to have lost weight during the 6 months before their laparotomy as well.
They were more likely to have sepsis or septic shock, to present emergently, and to have had a surgery within the 30 days prior. Functionally, they were significantly more likely to be rated as “totally dependent.”
A multivariate analysis identified six preoperative and four postoperative risk factors:
Preoperative
• Contaminated/dirty wound – odds ratio 2.00.
• Sepsis/septic shock – OR 1.85.
• Totally dependent status – OR 1.8.
• Male gender – OR 1.6.
• ASA class 3 or greater – OR 1.4.
• Smoking – OR 1.3.
• Weight loss protective – OR 0.44.
Postoperative
• Deep wound infection – OR 4.25.
• Superficial wound infection – OR 2.76.
• Reintubation – OR 2.38.
• Deep space infection – OR 1.67.
The investigators then split the data randomly into a 75% training cohort and 25% validation cohort. A receiver operator curve analysis determined that both cohorts had an AUC of around 0.70, meaning that the model was a moderate-good predictor of wound dehiscence.
“Our predictive model is just as good as the one that was developed 20 years ago,” and potentially, more appropriate for a general population, Dr. Pal concluded.
She had no financial disclosures.
On Twitter @Alz_Gal
*An earlier version of this article misstated Dr. Pal's name and affiliation.