|
Back to 2017 Posters
ARE SURGICAL SITE INFECTIONS REDUCED WITH MINIMALLY INVASIVE APPROACHES TO PANCREATICODUODENECTOMY? A NATIONWIDE PERSPECTIVE FROM THE PROCEDURE-TARGETED NSQIP DATABASE
Emily K. McCracken*, Leila Mureebe, Dan G. Blazer
Duke University Medical Center, Durham, NC
OBJECTIVE
Historically, pancreaticoduodenectomy (PD) has been associated with high rates of surgical site infection (SSI); rates are recorded between 11 and 21%. To date, published rates of SSI after minimally invasive PD (MIPD) are comparable or superior to open approach, but have been primarily limited to series from single institutions with high-volume experience. On a national level, rates of SSI after MIPD have not been reported. With the new availability of pancreatectomy-specific outcomes in NSQIP, we hypothesized nationwide SSI would also be reduced with MIPD.
METHODS
The procedure-targeted NSQIP participant user file (PUF) from January 1, 2014 through December 31, 2015 was queried for PD. Patients were then divided into open, laparoscopic, and robotic cohorts, excluding hybrid or converted procedures. For each group, demographics, treatment parameters, and surgical considerations known to affect SSI were evaluated. χ2 test was used to determine correlation of SSI with operative approach, and linear regression was used to determine correlation of clinical characteristics with infection by approach.
RESULTS
Over this time period, 296 patients underwent MIPD out of the total 6,882 PD evaluated. The majority of patients underwent open procedures (n=6,346, 92%). In contrast, 1.8% (n=125) were wholly laparoscopic and 2.5% (n=171) wholly robotic. The remaining 240 were hybrid or converted to open procedures; these patients were excluded from the analysis. In the open group, 24.2% (n=1,536) of the patients developed SSI. In the minimally invasive groups, SSI rates were 15.2% (n=19) in laparoscopic approach and 21.6% (n=37) in robotic approach. The overall rate of SSI in all three groups was 24.0%. Compared to laparoscopic approach, both robotic and open procedures had higher rates of infection (p=0.03 and 0.001, respectively). SSI were comparable between open and robotic approaches (p=0.6). Preoperative stenting (p=0.003), duct size (p=0.007), gland texture (p<0.001), and presence of drains (p=0.001) increased SSI in laparoscopic procedures. Events were too few to evaluate type of SSI by approach.
CONCLUSIONS
In this nationwide analysis, wholly laparoscopic approach to PD was associated with reduced rates of SSI. In accordance with previous single-center studies, stenting, duct size, gland texture, and drains increased SSI. This study, despite a greater overall rate of infection than previously reported in PD, even within NSQIP, suggests that outcomes in high-volume centers may be applicable nationwide. The comparable SSI rate between robotic and open procedures possibly reflects the former’s ongoing learning curve, and it is consistent with single-center experiences. It is unlikely due to the exclusion of converted and hybrid procedures, as there were fewer robotic converted/hybrid cases than laparoscopic (7.5% vs 29.0% of all attempted/hybrid MIPD).
Back to 2017 Posters
|
