SURGICAL INFECTIONS

Volume 19, Number 0, 2018 Original Article

ª Mary Ann Liebert, Inc.
DOI: 10.1089/sur.2018.134

Efficacy of Negative-Pressure Wound Therapy for Preventing

Surgical Site Infections after Surgery for Peritonitis
Attributable to Lower-Gastrointestinal Perforation:
A Single-Institution Experience

Katsuki Danno,1 Chu Matsuda,2 Susumu Miyazaki,1 Takamichi Komori,1 Megumi Nakanishi,3
Masaaki Motoori,1 Masaki Kashiwazaki,1 and Kazumasa Fujitani1
Downloaded by UNIV OF LIVERPOOL from www.liebertpub.com at 09/09/18. For personal use only.

Abstract

Background and Purpose: For patients at high risk, such as those with lower-gastrointestinal perforations, it is
important to establish a preventive method that reduces the incidence of surgical site infections (SSIs) sig-
nificantly. We applied negative-pressure wound therapy (NPWT) as part of a delayed primary closure approach
to prevent SSIs. This study evaluated the value of this technique.
Methods: We included prospectively 28 patients undergoing abdominal surgery for peritonitis caused by a
lower-gastrointestinal perforation between May 2014 and November 2015. Historical controls comprised ret-
rospective data on 19 patients who had undergone primary suturing for managing peritonitis incisions for a
lower-gastrointestinal perforation from January to December 2013.
Results: We found a significant association between the SSI incidence and the type of incision management
(10.7% with NPWT and delayed closure vs. 63.2% with primary suturing; p < 0.001). There was no significant
difference between the groups in the length of the hospital stay (22 days for NPWT and delayed closure vs.
27 days for primary suturing; p = 0.45). No severe adverse events were observed related to NPWT.
Conclusion: The use of NPWT and delayed primary closure was an effective measure for preventing SSI in
patients undergoing abdominal surgery for peritonitis caused by lower-gastrointestinal perforation.

Keywords: delayed primary closure; lower-gastrointestinal perforation; negative pressure wound therapy;
prevention; surgical site infection

C ontaminated surgical incisions develop into surgi-

cal site infections (SSIs) [1]. The National Nosocomial
Infectious Surveillance System of the U.S. Centers for Dis-
excess cost in the United States [6]. These infections also are
associated with a higher risk of incisional hernia [7,8]. Al-
though several risk factors for SSIs have been described and
ease Control and Prevention (CDC) introduced the concept of are being corrected, the incidence remains disturbingly high
SSIs in 1992 [2]. Incisions used for lower-gastrointestinal [9,10]. Therefore, particularly for patients at high risk, such
perforation surgery are designated class III/IV surgical sites as those with lower-gastrointestinal perforations, it is im-
according to CDC guidelines [3]. The incidence of incisional portant to establish a preventive method that can reduce the
SSIs associated with contaminated abdominal incisions is SSI incidence significantly.
>40% [4] . Furthermore, the incidence of incisional SSIs is Delayed primary closure (DPC) is a commonly used
>50% after colon-perforation surgery [5]. method for treating dirty incisions. The method is applicable
Surgical site infections are responsible for increasing to either contaminated or dirty sites; it allows soft tissue fluids
morbidity in patients who undergo abdominal surgery. to drain from the open areas, and the skin is closed a few days
Moreover, SSIs incur additional hospital care costs through later. Although a previous meta-analysis failed to provide
longer stays, which amount to more than US$1.5 billion in definitive evidence of DPC efficacy [11], some randomized

1
Department of Gastroenterological Surgery and 3Department of Nursing, Osaka General Medical Center, Osaka, Japan.
2
Department of Gastroenterological Surgery, Osaka University Graduate School of Medicine, Suita, Osaka, Japan.

1
 2 DANNO ET AL.

studies on contaminated abdominal incisions showed that the institutional or national research committee and with the
DPC reduced the SSI incidence compared with conven- 1964 Helsinki Declaration and its later amendments or
tional primary closure (PC) [4,12]. Additionally, negative- comparable ethical standards. The study was approved by the
pressure wound therapy (NPWT) has been used for managing Institutional Review Board of Osaka General Medical Center
open abdominal incisions [13,14] and treating infected ab- (approval no. 29-S0901). For this type of study (retrospective),
dominal walls [15]. informed consent is not required.
Several recent studies have reported that, compared with
conventional incision dressings, NPWT reduced the risk of Demographics
SSI in patients with a primary closed abdominal surgical
The following data were collected for all patients: Age,
incision after a high-risk repair [16-18]. However, the quality
sex, body mass index, bacteria cultured from ascites fluid,
of the evidence was low in those observational studies.
time to epithelialization (time to achieve complete wound
Moreover, the World Health Organization emphasized that
closure), NPWT device application time, length of post-
the devices used for NPWT are expensive, and prioritization
operative hospital stay (LOS), and the presence of SSIs. We
should be considered carefully [18].
also recorded underlying medical conditions that could
Therefore, we focused on peritonitis surgery, with the in-
contribute to infectious complications, including diabetes
fection resulting from lower-gastrointestinal perforations.
mellitus, obesity, malnutrition, steroid use, and the presence
This scenario is characterized by a particularly high SSI rate.
of a stoma.
In our institution, NPWT has been applied as part of the DPC
Downloaded by UNIV OF LIVERPOOL from www.liebertpub.com at 09/09/18. For personal use only.

Epithelialization was defined as a state in which the wound

approach to create an optimal incision-healing environment.
was covered with the epithelium, no wound exudate was
This prospective cohort study was performed to evaluate the
present, and gauze or wound dressings were unnecessary.
effect of NPWT on the rate of SSIs after abdominal surgery
Surgical site infections were defined according to the National
for peritonitis secondary to lower-gastrointestinal perfora-
Nosocomial Infectious Surveillance System of the CDC [2].
tion. We hypothesized that the application of NPWT as part
Treatment-related adverse events (AEs) were evaluated ac-
of the DPC approach would reduce the incidence of SSIs.
cording to the Common Terminology Criteria for Adverse
Events (CTCAE) version 4.0.
Patients and Methods
Study design and patient treatment Historical controls
This prospective cohort study included all patients who We assessed whether NPWT reduced the SSI incidence,
underwent abdominal surgery for a lower-gastrointestinal shortened the time to epithelialization, and lowered the LOS
perforation in our hospital between May 2014 and November compared with conventional PC. For controls, we retrieved
2015. Starting in May 2014, we implemented the NPWT historical records of patients who underwent conventional
treatment (described below) for all patients as a basic policy, PCs to manage peritonitis incisions occasioned by lower-
and prospective data were recorded. The extent of peritonitis gastrointestinal perforations from January to December 2013.
was identified by surgeons at the time of surgery. We targeted During that year, all patients who underwent abdominal
patients with peritonitis for NPWT, and any patient with a surgery for lower-gastrointestinal perforations in our hospital
localized abdominal abscess was excluded. For the NPWT had conventional PCs. All patients received peri-operative
procedure, the peritoneum, muscle, and fascia were closed intravenous antibiotics with anaerobic coverage until they
in one layer with polydiaxanone sutures, but the skin and achieved normalizations of body temperature, white blood
subcutaneous tissue incisions remained open. Open subcu- cell count, and gastrointestinal function.
taneous tissues were packed with gauze soaked in normal
saline. All NPWT placements of black GranuFoam dressing Statistical analysis
(V.A.C. Therapy, Kinetic Concepts Inc. [KCI], San An-
Continuous data are presented as the median and range;
tonio, TX) were initiated within three days of surgery, and
categorical data are presented as the frequency and percent-
treatment continued for five to seven days. Dressing chan-
age. The w2 or Fisher exact tests and the non-parametric
ges were performed every 48 to 72 hours no fewer than three
Mann–Whitney U test were used to compare demographic or
times per week. Negative pressure was set at 100 mm Hg,
clinical characteristics between groups. A two-sided p value
with continuous suction. When detectable wound granulation
<0.05 was considered statistically significant. JMP 10 (SAS
had formed, and the wound appeared clean, it was closed
Institute Inc., Cary, NC, USA) was used for all statistical
with a vertical mattress suture under local anesthesia in the
analyses.
patient’s room (DPC). All patients received peri-operative
intravenous antibiotics with anaerobic coverage until they
Results
achieved normalizations of body temperature, white blood
cell count, and gastrointestinal function. Patients were ex- From May 2014 to November 2015 (19 months), 47
cluded when naked blood vessels were present in the wound patients underwent surgery for peritonitis secondary to a
or a fistula was found in the abdominal cavity; when an lower-gastrointestinal perforation at Osaka General Medical
uncontrolled organ-space SSI was present; when they ex- Center. Of these patients, 40 were found to have peritonitis,
hibited any contraindications to the V.A.C. therapy device but 11 were ineligible for the study (Fig. 1). Of those 11
[19]; or when they were judged unsuitable for this therapy patients, four had contraindications to the V.A.C. therapy
by an attending physician. device, and seven were judged unsuitable for this therapy by
All procedures performed in studies involving human an attending physician. One had a mental disorder; six could
participants were in accordance with the ethical standards of not start NPWT because of insufficient preparation of the
 NPWT PREVENTS SSI AFTER PERITONITIS 3
Downloaded by UNIV OF LIVERPOOL from www.liebertpub.com at 09/09/18. For personal use only.

FIG. 1. Flow chart of patient selection procedure. *Discontinuation because of protocol violation. NPWT = negative-
pressure wound therapy.

equipment. Of the remaining 29 patients, one was excluded formation of granulation tissue. Consequently, a total of five
by a protocol violation. Thus, we compared 28 patients who patients developed SSIs or failed the DPC (17.9%).
received NPWT+DPC with 19 patients who received PC. The PC group had 12 incision infections (63.2%). There
These groups were not significantly different with regard to was a significant association between the incidence of SSIs
age, sex, Body Mass Index, risk factors, or stoma formation and the type of incision management (NPWT+DPC: 10.7%
(Table 1). vs. PC: 63.2%; p < 0.001). However, the mean hospital stay
Bacterial cultures from the ascites fluid were positive in 39 was not significantly different, although the LOS of the pa-
patients (83.0%). Bacteroides spp., which are anaerobic or- tients having NPWT tended to be shorter (NPWT+DPC:
ganisms, were identified in 19 patients (40.4%). The most 22 days vs. PC: 27 days; p = 0.45). Moreover, there was no
common organisms cultured were Bacteroides, Escherichia significant difference between the groups in the time to epi-
coli, and various Enterococcus spp. (Table 2). There were thelialization (Table 3).
three SSIs in the NPWT+DPC group (10.7%). In addition, The incidence of treatment-related AEs from the time of
two patients with incisions unsuitable for suture failed to therapy initiation is shown in Table 4. No severe AEs related
complete the DPC. Their incisions were repaired by the to NPWT were observed. Two patients exhibited bleeding at

Table 1. Baseline Clinical Characteristics of Treatment Groups

NPWT+DPC PC
Characteristic (N = 28) (N = 19) p
Median age (y) (range) 72.5 (20-90) 64 (13-83) 0.09
Sex, N
Female 17 8 0.21
Male 11 11
Body mass index, kg/m2 (Median, Range) 23.6 (12.3-30.1) 21.6 (15.2-31.2) 0.14
Risk Factors, N (%)
Diabetes mellitus 6 (21.4) 4 (21.0) 0.98
Steroid Use 2 ( 7.1) 0 ( 0.0) 0.14
Cardiovascular disease 4 (14.3) 4 (21.0) 0.55
Obesity (body mass index >30 kg/m2) 2 ( 7.1) 2 (10.5) 0.69
Stoma formation, N (%) 6 (21.4) 7(36.8) 0.25
NPWT, negative pressure wound therapy; DPC, delayed primary closure; PC, primary closure
 4 DANNO ET AL.

Table 2. Bacteria Cultured from Ascites Fluid Table 4. Adverse Events Related
to Negative-Pressure Wound Therapy
NPWT+DPC PC
Organism (N = 28) (N = 19) CTCAE, version 4.0
Not submitted 1 1 Event Grades 1, 2 (%) Grades ‡3 (%)
No growth 4 2
Bleeding 2 (7) 0
Gram-negative organisms
Pruritus 1 (4) 0
Escherichia coli 11 6
Pain 2 (7) 0
Pseudomonas aeruginosa 3 2
Hematoma 0 2 (7)
Klebsiella spp. 2 3
related to DPC
Other 0 1
Gram-positive organisms CTCAE = Common Terminology Criteria for Adverse Events;
Streptococcus spp. 5 1 DPC = delayed primary closure.
Enterococcus spp. 8 8
Enterobacter spp. 1 2
Other 2 0
Anaerobic organisms reduced the incidence of SSIs in patients with primarily
Bacteroides spp. 14 5 closed abdominal incisions that were thought to be at high
Downloaded by UNIV OF LIVERPOOL from www.liebertpub.com at 09/09/18. For personal use only.

Other 2 2 risk of infection [16,18]. In contrast, a review of previous

NPWT treatments showed no benefit in patients with com-
DPC = delayed primary closure; NPWT = negative-pressure wound plicated contaminated hernias [22]. That finding might in-
therapy; PC = primary closure. dicate that the effectiveness of this intervention is limited
in highly contaminated surgical incisions. However, the
present study reported preliminary results from a case series
the time of changing the foam that was stopped with simple that comprised patients with incisions created during lower-
compression. Two patients experienced pain that did not gastrointestinal perforation surgery, which are highly con-
require analgesics. On the other hand, two patients exhibited taminated, being class III/IV according to the surgical
abdominal wall hematomas after the DPC, and both devel- incision classifications described in the CDC guidelines [3].
oped SSIs. Because of the highly contaminated nature of these sites, we
considered it inappropriate to apply the NPWT in closed
incisions. Moreover, in Japan, the Pharmaceuticals and
Discussion Medical Devices Agency approved NPWT in November
Negative-pressure wound therapy is a novel approach in 2009 only for the treatment of non-healing open sites. Since
site-healing management in gastrointestinal surgery, with then, the indications have not been expanded. Therefore, the
several benefits that may contribute to healing. In particular, use of a NPWT device in closed incisions would be con-
open drainage evacuates interstitial fluids and cellular debris sidered off-label use.
continuously, which increases local tissue blood perfusion To our knowledge, this study is the first to evaluate the
and vascularization and stimulates the formation of granu- application of NPWT to incisions associated with lower-
lation tissue [20]. gastrointestinal perforation surgery for the prevention of
Moreover, incision tissues are deformed by the porous
open-cell sponge used in V.A.C. therapy. This stretching of
individual cells promotes cell proliferation in the incision
microenvironment [21]. In other words, it is important for the
porous open-cell sponge to maintain contact with the site
surface. Several authors have recently reported that NPWT

Table 3. Comparison of Treatment Results

Result NPWT+DPC PC p
Surgical site 3 (11) 12 (63) <0.01
infection (%)
Failure to complete 2 – –
DPC (%)
Infection+failure(%) 5 (18) 12 (62) <0.01
Median NPWT 6 ( 3–21) – –
duration (d) (range)
Median time to 18.5 (14–91) 35 ( 7–152) 0.40
epithelialization
(range) FIG. 2. Time to epithelialization for patients with perito-
nitis treated with negative-pressure wound therapy (NPWT)
DPC = delayed primary closure; NPWT = negative-pressure wound and delayed primary closure (DPC) or with conventional
therapy; PC = primary closure. primary closure (PC).
 NPWT PREVENTS SSI AFTER PERITONITIS 5

SSIs. The emergency status of the patients in this study in- Author Disclosure Statement
dicated that this group was at extremely high risk of SSIs; No competing financial interests exist.
thus, the efficacy of NPWT could be assessed most easily in
this group. Our comparison of the two study groups con- References
firmed that the NPWT was an effective preventive measure
against SSIs in patients undergoing abdominal surgery for 1. Cruse PJE, Foord R. The epidemiology of wound infection:
peritonitis resulting from lower-gastrointestinal perforation. A 10-year prospective study of 62,939 wounds. Surg Clin
North Am 1980;60:27–40.
One limitation of this study was the difficulty in analyzing
2. Horan TC, Gaynes RP, Martone WJ, et al. CDC definitions
whether the NPWT provided benefit in addition to the DPC
of nosocomial surgical site infections, 1992: A modification
in preventing SSIs. Bacterial contamination of the incision of CDC definitions of surgical wound infections. Am J
during surgery is the major risk factor in the development of Infect Control 1992;20:271–274.
SSIs. In the present study, we obtained bacterial cultures of 3. Mangram AJ, Horan TC, Pearson ML, et al., Hospital In-
the ascites fluid of 39 patients (83.0%; Table 2). The benefits fection Control Practices Advisory Committee. Guideline
of NPWT were most likely attributable to the constant for prevention of surgical site infection, 1999. Am J Infect
drainage of contaminated wound effluent, which can interfere Control 1999;27:97–134.
with incision healing. 4. Cohn SM, Giannotti G, Ong AW, et al. Prospective ran-
Extended hospital stays, considered an indirect predic- domized trial of two wound management strategies for
tor of the overall cost of treatment, were reported to be dirty abdominal wounds. Ann Surg 2001;233:409–413.
Downloaded by UNIV OF LIVERPOOL from www.liebertpub.com at 09/09/18. For personal use only.

an economic concern associated with the DPC method 5. Watanabe M, Suzuki H, Nomura S, et al. Risk factors for
[23]. We assumed that the time required to install the NPWT surgical site infection in emergency colorectal surgery: A
device and then close the skin definitively might be offset retrospective analysis. Surg Infect 2014;15:256–261.
by the reduction in SSI incidence. However, the LOS and 6. de Lissovoy G, Fraeman K, Hutchins V, et al. Surgical site
the time to epithelialization were not significantly different infection: Incidence and impact on hospital utilization and
in the two groups (Table 4). Many patients in the NPWT treatment costs. Am J Infect Control 2009;37:387–397.
group exhibited shortened times to epithelialization, but 7. Murray BW, Cipher DJ, Pham T, et al. The impact of
because of the small number of patients in this study, the surgical site infection on the development of incisional
difference between groups was not statistically significant hernia and small bowel obstruction in colorectal surgery.
(Fig. 2). Am J Surg 2011;202:558–560.
8. Veljkovic R, Protic M, Gluhovic A, et al. Prospective
We also showed that NPWT was associated with only a
clinical trial of factors predicting the early development of
few AEs. In several cases, skin itching, pain, and minor
incisional hernia after midline laparotomy. J Am Coll Surg
bleeding were noted at the site of application. However, these 2010;210:210–219
AEs resolved rapidly with local treatment and temporary 9. Anthony T, Murray BW, Sum-Ping JT, et al. Evaluating an
device discontinuance. Moreover, NPWT was convenient, evidence-based bundle for preventing surgical site infec-
because the incision dressing required replacement only two tion: A randomized trial. Arch Surg 2011;146:263–269.
or three times per week. 10. Stulberg JJ, Delaney CP, Neuhauser DV, et al. Adherence
This study had some limitations. The decision of whether to surgical care improvement project measures and the
to apply the NPWT was at the discretion of the attending association with postoperative infections. JAMA 2010;303:
physician, and not all attending physicians in the study ac- 2479–2485.
cepted NPWT. Although it was possible that the NPWT 11. Bhangu A, Singh P, Lundy J, et al. Systemic review and
might have been used in patients at lower risk of SSIs, in fact, meta-analysis of randomized clinical trials comparing pri-
all cases were considered to be at very high risk in this sit- mary vs delayed primary skin closure in contaminated and
uation. In one case, the hesitation in introducing the NPWT dirty abdominal incisions. JAMA Surg 2013;148:779–786.
was attributable to the patient’s mental disorder. In all other 12. Chiang R-A, Chen S-L, Tsai Y-C. Delayed primary closure
cases, the reasons for avoiding NPWT were related to the versus primary closure for wound management in perfo-
arrangement or technical problems with this instrument. rated appendicitis: A prospective randomized controlled
Other limitations of this study were those inherent in its non- trial. J Chinese Med Assoc 2012;75:156–159.
randomized nature, its single-center small cohort design, and 13. Roberts DJ, Zygun DA, Grendar J, et al. Negative-pressure
the comparison with historical controls. A randomized study wound therapy for critically ill adults with open abdominal
would strengthen the analysis, but the randomized design is wounds: A systematic review. J Trauma Acute Care Surg
2012;73:629–639.
not always feasible in an emergency surgical setting because
14. Bee TK, Croce MA, Magnotti LJ, et al. Temporary ab-
of ethical concerns. Currently, we are planning to conduct a
dominal closure techniques: A prospective randomized trial
prospective multi-center cohort study, which is expected to comparing polyglactin 910 mesh and vacuum-assisted
provide further information. To date, we have enrolled some closure. J Trauma Acute Care Surg 2008;65:337–344.
patients, and we have registered the trial in the University 15. Mees J, Mardin WA, Senninger N, et al. Treatment options
Hospital Medical Information Network (UMIN) system for postoperatively infected abdominal wall wounds heal-
(UMIN ID: 000020391) ing by secondary intention. Langenbeck’s Arch Surg 2012;
In conclusion, our data support the use of NPWT as 397:1359–1366.
part of the DPC approach. We showed that NPWT pre- 16. Bonds AM, Novick TK, Dietert JB, et al. Incisional nega-
vented SSIs in most patients undergoing abdominal sur- tive pressure wound therapy significantly reduces surgical
gery for peritonitis secondary to a lower-gastrointestinal site infection in open colorectal surgery. Dis Colon Rectum
perforation. 2013;56:1403–1408.
 6 DANNO ET AL.

17. Blackham AU, Farrah JP, McCoy TP, et al. Prevention of 22. Pauli EM, Krpata DM, Novitsky YW, et al. Negative
surgical site infections in high-risk patients with laparot- pressure therapy for high-risk abdominal wall reconstruc-
omy incisions using negative-pressure therapy. Am J Surg tion incisions. Surg Infect 2013;14:270–274.
2013;205:647–654. 23. Khan KI, Mahmood S, Akmal M, et al. Comparison of rate
18. WHO. Global Guidelines for the Prevention of Surgical of surgical wound infection, length of hospital stay and
Site Infection. 2016;186. Available at: www.who.int/gpsc/ patient convenience in complicated appendicitis between
ssi-prevention-guidelines/en/ primary closure and delayed primary closure. J Pak Med
19. V.A.C.Therapy System Safety Infromation and V.A.C. Assoc 2012;62:596–598.
Granufoam Bridge Dressing and V.A.C. Granufoam Bridge
XH dressing Application Instructions. ª2013 KCI Licen-
sing Inc 415192 RevA. San Antonio, TX. Kinetic Concepts, Address correspondence to:
Inc. July 2013. Katsuki Danno, MD, PhD
20. Morykwas MJ, Argenta LC, Shelton-Brown EI, et al. Department of Surgery
Vacuum-assisted closure: A new method for wound control Minoh City Hospital
and treatment: Animal studies and basic foundation. Ann 5-7-1 Kayano, Minoh
Plast Surg 1997;38:553–562. Osaka 562-0014
21. Saxena V, Hwang C-W, Huang S, et al. Vacuum-assisted Japan
closure: Microdeformations of wounds and cell prolifera-
tion. Plast Reconstr Surg 2004;114:1086–1096. E-mail: [email protected]
Downloaded by UNIV OF LIVERPOOL from www.liebertpub.com at 09/09/18. For personal use only.