Closed incision negative pressure therapy: international multidisciplinary consensus recommendations.

Surgical site occurrences (SSOs) affect up to or over 25% of patients undergoing operative procedures, with the subset of surgical site infections (SSIs) being the most common. Commercially available closed incision negative pressure therapy (ciNPT) may offer surgeons an additional option to manage clean, closed surgical incisions. We conducted an extensive literature search for studies describing ciNPT use and assembled a diverse panel of experts to create consensus recommendations for when using ciNPT may be appropriate. A literature search of MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials using key words 'prevention', 'negative pressure wound therapy (NPWT)', 'active incisional management', 'incisional vacuum therapy', 'incisional NPWT', 'incisional wound VAC', 'closed incisional NPWT', 'wound infection', and 'SSIs' identified peer-reviewed studies published from 2000 to 2015. During a multidisciplinary consensus meeting, the 12 experts reviewed the literature, presented their own ciNPT experiences, identified risk factors for SSOs and developed comprehensive consensus recommendations. A total of 100 publications satisfied the search requirements for ciNPT use. A majority presented data supporting ciNPT use. Numerous publications reported SSI risk factors, with the most common including obesity (body mass index ≥30 kg/m2 ); diabetes mellitus; tobacco use; or prolonged surgical time. We recommend that the surgeon assess the individual patient's risk factors and surgical risks. Surgeons should consider using ciNPT for patients at high risk for developing SSOs or who are undergoing a high-risk procedure or a procedure that would have highly morbid consequences if an SSI occurred.

Introduction

The World Health Organization estimated that surgeons performed over 234 million major surgeries (i.e., operative procedures involving significant risks to the patient) globally each year 1. In industrialised countries, major complications (i.e. those that are potentially life‐threatening and require hospitalisation and therapeutic intervention) occur in over 25% of inpatient surgical procedures 1. In the United States (US) alone, surgical site infections (SSIs) account for 36% of all health care‐associated infections, which are a major cause of morbidity, putting 8 million US patients at risk for developing an SSI annually 2, 3. Current standards of care for preventing SSI include preoperative prophylactic systemic antibiotics (for selected surgical procedures); preoperative antiseptic shower/bath; aseptic incision site surgical preparation; and sterile and meticulous surgical technique 4. Yet, the continued high SSI rates demonstrate the need for new preventative methods.

Traditionally, surgeons have closed surgical incisions with primary intention using sutures, staples, tissue adhesives, paper tape or a combination of these methods. However, negative pressure wound therapy (NPWT) has become a viable wound care option since its introduction two decades ago. For many different operative procedures, especially in the plastic surgery field, NPWT plays an integral adjunct treatment to enhance different interventions in the reconstructive pathway. Commercial negative pressure dressings are increasingly used in various clinical settings and for many types of acute and chronic open wounds.

Surgeons have recently discovered that foam‐based negative pressure dressings applied over closed incisions can also be beneficial in preventing incision complications. The term ‘closed incision negative pressure therapy’ (ciNPT) refers to any type of NPWT using foam‐based dressings over closed incisions.

Our goals were to investigate how ciNPT is beneficial in preventing wound incision complications and then to formulate recommendations for potential indications for its use. In December 2014, a multidisciplinary group of surgical and infectious disease experts met to discuss the following questions:

Is there evidence‐based data in the literature that reports any benefits from using ciNPT?

Which types of patients and closed surgical incisions are at greatest risk for postoperative complications in the different surgical specialty fields?

Can evidence‐based recommendations be formulated for ciNPT use?

Materials and methods

Search of literature and selection of studies

A review of the literature was performed searching computerised versions of MEDLINE (PubMed), EMBASE and the Cochrane library. We further expanded the potential evidence base using a ‘snowball’ system (i.e. continued searches in the references of the self‐researched publications). Search criteria included (i) publications in all languages, (ii) various study types [e.g. randomised clinical and experimental studies, systematic and non‐systematic reviews, meta‐analyses, expert opinions, case reports, experimental papers (animal and human studies)] and (iii) consensus conference reports. The authors received access to all publications in their full‐published versions.

Articles published in a peer‐reviewed journal that was considered relevant for the development and dissemination of medical knowledge [i.e. an Abridged Index Medicus (AIM) journal], supported the CONSORT statement, and a citation impact factor of >0·5 were used.

Search period and search keywords

The search covered papers published in the period from January 2000 to February 2016. The keywords included ‘prevention’, ‘NPWT’, ‘active incisional management’, ‘incisional vacuum therapy’, ‘incisional NPWT’, ‘incisional wound vacuum assisted closure’, ‘closed incisional NPWT’, ‘wound infection’ and ‘SSIs’.

An additional literature search was conducted to identify risk factors for SSI development. Keywords included ‘SSI’, ‘wound infection’, ‘general surgery’, ‘open abdomen surgery’, ‘hernia repair’, ‘plastic surgery’, ‘reconstructive surgery’, ‘orthopaedic surgery’, ‘open reduction and internal fixation’, ‘vascular surgery’, ‘vascular bypass’, ‘cardiovascular surgery’, ‘sternotomy’ and ‘amputation’.

Criteria of evidence‐based medicine

More than 50 different evidence level scales exist worldwide. For the purpose of this study, we selected the 2009 Oxford Centre for Evidence‐based Medicine (EbM) classification system 5.

Multidisciplinary consensus meeting

To formulate consensus guidelines, peer‐reviewed published literature focusing on ciNPT was used as the foundation for discussion and as evidence to support guideline statements. Using a modified consensus process, described below, panellists agreed on which patient risk factors and closed surgical incisions were at the highest risk of SSIs and created an algorithm for the use of ciNPT.

Selection of panellists

Leaders at Acelity (San Antonio, TX, USA), in conjunction with the academic lead authors (CW, VSR), selected the 12 panellists based on their peer‐reviewed publications on NPWT; clinical experience with negative pressure for incision management; and reputation for scholarly activity in their respective fields. To create a heterogeneous expert panel, we selected physicians who were from various geographic locations (US, Italy, Germany and Denmark), had diverse practice experience and represented several different surgical specialties (general, orthopaedic, trauma, plastic, cardiac, podiatric and vascular surgery) as well as clinical microbiology and infectious disease.

Developing the consensus recommendations

Before the meeting convened, all panellists reviewed the publications retrieved by the systematic literature review and were briefed on the process for consensus building. The one‐and‐a‐half day meeting was divided into four sections: (i) presentations (15–20 min) by each panellist reporting clinical experience with ciNPT; (ii) collection of comments to all distributed literature and evaluation/rating of the available literature on ciNPT; (iii) review of definitions of closed incisions at risk for complications and of patient‐related risks; and (iv) open discussion regarding appropriate use of ciNPT (i.e. algorithm). By digitally recording all comments, the lead authors ensured that all viewpoints were adequately captured and reviewed. Participants did not reach conclusive recommendations at this meeting; rather, they elected to reflect on definitions of closed incisions at risk in various fields of surgery and to participate in follow‐up discussions via electronic mail and a follow‐up teleconference 12 weeks following the meeting. The panellists received follow‐up documents, including a general manuscript outline and an assessment of ciNPT risk factors by surgical specialty, for review (i.e. agree or disagree) and comment via electronic mail. All participants reviewed comments made by other participants with the goal of reaching unanimous agreement, when possible, or consensus. The lead authors drafted a manuscript that was reviewed and commented on by all panellists. All panellists agreed upon the final manuscript prior to submission for publication.

Identifying risk factors and developing an algorithm

During the meeting, each panellist presented a list of risk factors considered to be important when assessing patients for ciNPT use. Each panellist also reviewed the resulting comprehensive list of risk factors and provided relevant supporting EbM literature, when available. Panel members recommended ciNPT use for risk factors with a reported odds ratio (OR) >2 or if the risk factor was present in multiple surgical fields. Once the panellists reached a consensus on risk factors, they created an algorithm to identify at‐risk scenarios in which ciNPT usage might be beneficial for incision management. All panellists reviewed and approved the algorithm.

Results

Type of ciNPT studies

A limited number of robust, prospective, randomised, comparative, controlled studies on ciNPT use over closed surgical incisions that might most benefit from this therapy exist. The literature search identified 100 publications that fulfilled the above mentioned criteria. Of these, 60 articles describe outcomes in a total of 2402 ciNPT‐treated patients following surgical procedures, including orthopaedic (n = 21 articles, n = 852 patients), general (n = 22 articles, n = 869 patients), cardiothoracic (n = 8 articles, n = 505 patients), plastic (n = 6 articles, n = 133 patients) or vascular (n = 6 articles, n = 95 patients). Three articles have more than one surgical specialty and patient population; thus, some patients and articles are counted twice. The remaining 40 publications were literature reviews including meta‐analyses, editorials, research articles or experimental model descriptions. Three articles were solely devoted to a health economic analysis 6, 7, 8, and three articles describe study protocols of future studies 9, 10, 11. A majority of the 100 publications reported data based on one manufacturer's system: n = 91, KCI, an Acelity company, San Antonio, TX, USA; n = 8, Smith and Nephew, plc, London, UK; n = 1, Daewoong Pharmaceutical, Co, Ltd., Seoul, South Korea.

Of the 100 publications, 51 (51·0%) had authors based in the US; 15 (15·0%) in Germany; 8 (8·0%) in Australia; 6 (6·0%) in Italy; 4 (4·0%) in UK/Ireland; 3 (3·0%) each in Canada, China and Spain; 2 (2·0%) in Turkey; and 1 (1·0%) each in Denmark, Poland, South Korea, South Africa and the United Arab Emirates.

Using the Oxford Centre for EbM evidence levels (Table 1) 5, 51 (51·0%) included papers received a level 4 or 5 (reviews, comparative historical studies, case series, case reports, economic studies) and 39 (39·0%) received an evidence level of 3 or higher (comparative studies, meta‐analyses). An additional 10 (10·0%) had no evidence level (research reports, technical information, editorials, study protocol, experimental study, etc.).

Table 1

Evidence levels for the available literature on the subject of closed incision negative pressure therapy

EbM level	Type of study	Number of studies	Percentage of studies (%)
No level	Research reports, technical reports, editorial, guidelines	10	10·0
1a	Systematic review of randomised controlled trials	6	6·0
1b	Individual randomised controlled trials (with narrow confidence interval)	2	2·0
1c	All‐or nothing result*	0	0
2a	Systematic review (with homogeneity) of cohort studies	2	2·0
2b	Individual cohort study [including low‐quality randomised controlled trials (e.g. with a follow‐up of < 80%)]	11	11·0
2c	‘Outcomes’ research, ecological study	0	0
3a	Systematic review (with homogeneity) of case–control studies	0	0
3b	Individual case–control studies	18	18·0
4	Case series (and poor‐quality cohort studies and case–control studies)	20	20·0
5	Expert opinion without explicit critical appraisal or based on physiology, bench research or ‘first principles’	31	31·0
Total		100	100·0

EbM, evidence‐based medicine

If all patients died before the therapy was available but now some survive, or if some patients died but now all survive. Classification provided by Centre for Evidence‐Based Medicine (March 2009) 5.

Main results of ciNPT studies

Preclinical studies evaluating ciNPT compared with standard wound care reported reduced scar thickness and narrower scar width, increased collagen at the incision site, increased mechanical properties and increased tensile strength in the ciNPT groups 12, 13. In addition, using Laser Doppler flowmetry, the peristernal perfusion after cardiac surgery was increased among the patients who underwent negative pressure therapy and decreased among the controls significantly 14. Mammary artery harvesting reduced peristernal perfusion by 25·7% in the controls, but negative pressure increased perfusion by 100% after mammary harvesting (P = 0·04). Thus, ciNPT increased perfusion relative to controls and compensated for reduced perfusion rendered by mammary artery harvesting, providing additional support in high‐risk patients 14.

Our review found a number of case studies, case series and non‐randomised controlled trials that described ciNPT use. These studies included high‐risk patients with one or more comorbidities who underwent various surgical procedures, including vascular bypass, sternotomy and caesarean section 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63. In 2013, Condé‐Green et al. reported that patients undergoing abdominal hernia repairs treated with ciNPT had a lower surgical site occurrence (SSO) rates (22% versus 63%, P = 0·02) and dehiscence (9% versus 38%, P = 0·014) compared with patients treated with wound dressings 17. In a retrospective study with a historical cohort by Gibbs et al., 34 after controlling for body mass index (BMI) and diabetes, wound complication rates in the ciNPT group (n = 103) were found to be equivalent to those in the standard dressing group (n = 867). Three other retrospective studies with a historical control group observed lower rates of SSI, SSOs, wound morbidity and re‐operation in the ciNPT group compared with the historical controls 16, 51, 63.Overall, a majority of these studies reported that ciNPT use was associated with decreases in wound complications, wound dehiscence, SSIs, haematoma/seroma formation and incisional drainage.

Since 2004, numerous randomised controlled trials and individual cohort studies have described ciNPT use (see Table 2). These studies encompass various wound types and surgical interventions, including traumatic injury repair, cardiothoracic surgery, lower extremity amputations, arthroplasty, hernioplasty and vascular surgery 44, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76. Enrolled patients often had comorbidities, including obesity (BMI ≥30 kg/m2), diabetes mellitus, peripheral vascular disease or chronic obstructive pulmonary disease 15, 67, 68, 69, 77. Two studies reported no differences in SSI rates or dehiscence between ciNPT and control (silver‐impregnated wound dressings or sterile gauze dressings) groups 69, 77. One study was stopped prematurely because of blister formation in a majority of ciNPT group patients 77. This adverse effect was most likely because of improper dressing configuration and too high tension when using the dressing as no other study reported adverse effects. The majority of randomised controlled trials reported uniformly decreased SSI incidence, wound dehiscence and seroma development in the ciNPT‐treated group versus the control groups 44, 63, 64, 65, 66, 67, 68, 71, 72, 78. Stannard et al. examined outcomes in 249 patients undergoing an orthopaedic procedure for blunt trauma, resulting in 263 tibial plateau, pilon or calcaneus fractures 66. Fractures randomised to receive ciNPT (n = 141), compared with standard of care (n = 122), had lower SSI rates (P = 0·049) and wound dehiscence (P = 0·044). Grauhan et al. reported a 4·5‐fold decrease in wound infection rates in the ciNPT group (n = 75) compared with the standard wound dressing group (n = 75; OR = 4·57; 95% confidence interval = 1·23–16·94; P = 0·0266) in obese patients (BMI ≥ 30 kg/m2) following cardiac surgery 67.

Table 2

Overview of published randomised controlled trials

Year	References	EbM level*	Number of patients	Type of wounds	Results	Conclusion
2015	Nordmeyer et al. 75	RCT level 1b	20 (10 ciNPT, 10 control)	Internal fixation of spinal fractures	Seroma day 5   ciNPT: 0 ml   Control: 1·9 ml Seroma day 10   ciNPT: 0·5 ml   Control: 1·6 ml Wound care time   ciNPT: 13·8 ± 6 min   Control: 31 ± 10 min Number of compresses   ciNPT: 11 ± 3   Control: 35 ± 15	ciNPT significantly reduced the development of seroma (day 5 P = 0·0007; day 10 P = <0·024), required wound care time (P = 0·005), and number of compresses (P = 0·0376)
2015	Gillespie et al. 74	RCT level 2b	75 (35 ciNPT; 35 standard dressings)	Elective primary hip arthroplasty	SSI   ciNPT = 2/35   Control = 3/35   (risk ratio = 0·67; 95%CI = 0·12‐3·7; P = 0·65)	Reduction of SSI suggests that a large RCT requires 900 patients per group.
					Wound complications   ciNPT experience more postoperative wound complications (risk ratio = 1·6; 95% CI = 1·0‐2·5; P = 0·04)	There is uncertainty in the benefit of ciNPT use following elective hip arthroplasty.
2014	Pauser et al. 71	RCT level 2b	21 [11 ciNPT (Group A); 10 control (Group B)]	Femoral neck fracture patients scheduled for hip hemiarthroplasty	Developed a seroma at 5 days   Group A 0·257 ± 0·75 cm3   Group B 3·995 ± 5·01 cm3 Duration of secretion   Group A 0·9 ± 1·0 days   Group B 4·3 ± 2·45 days Total time for dressing changes   Group A 14·8 ± 3·9 minutes   Group B 42·9 ± 11·0 minutes	Significant decrease in development of postoperative seroma, total wound secretion days, and time for dressing changes in ciNPT group (Group A, P < 0·05).
2013	Grauhan et al. 67	RCT level 2b	150 (75 ciNPT; 75 control)	Cardiac surgery in obese patients (BMI ≥30)	Wound infections   ciNPT: 3 (4%)   Control: 12 (16%)	Significantly reduced incidence of wound infection in ciNPT group (P = 0·0266; OR = 4·57; 95% CI = 1·23‐16·94).
					Wound infections with Gram‐positive skin flora   ciNPT: 1 (1·3%)   Control: 10 (13·3%)	Significantly lower incidence of wound infections with Gram‐positive skin flora in ciNPT group (P = 0·009; OR = 11·39; 95% CI = 1·42‐91·36).
2012	Stannard et al. 66	RCT level 1b	249 patients, 263 fractures (141 ciNPT; 122 control)	Blunt trauma with one of three high‐risk fracture types (tibial plateau, pilon, calcaneus)	Infection results   ciNPT: 1 (0·7%) acute 13 (9%) delayed   Control: 5 (4%) acute 18 (15%) late	Significantly lower rates of infection in ciNPT group (P = 0·049).
					Dehiscence results   ciNPT: 12 (8·6%) fractures   Control: 20 (16·5%) fractures	Significantly lower rates of total wound dehiscence in ciNPT fractures (P = 0·044).
					Discharge results   ciNPT: 2·5 days   Control: 3·0 days	No significant difference in time to discharge.
2012	Masden et al. 69	RCT level 2b	81 (44 ciNPT; 37 control)	Multiple wounds in high risk patients	Wound infections   ciNPT: 6·8% (n = 3)   Control: 13·5% (n = 5) Dehiscence   ciNPT: 36·4% (n = 16)   Control: 29·7% (n = 11)	No significant difference between ciNPT group and controls in wound infections (P = 0·46) or dehiscence (P = 0·54).
2012	Pachowsky et al. 70	RCT level 2b	19 (9 ciNPT; 10 control)	Total hip arthroplasty	Seroma mean volume day 5   ciNPT: 0·58 ± 1·21 ml   Control: 2·02 ± 2·74 ml Seroma mean volume day10   ciNPT: 1·97 ± 3·21 ml   Control: 5·08 ± 5·11 ml	Significant reduction of seroma mean volume at 10 days post‐surgery (P = 0·021)
2011	Howell et al. 77	RCT level 2b	51 patients, 60 total knee arthroplasties (24 ciNPT; 36 control) (9 bilateral)	Primary total knee arthroplasty in obese (BMI ≥30) patients	Time to dry wound   ciNPT: 4·3 days   Control: 4·1 days Postoperative infections   ciNPT: 1 individual   Control: 1 individual	No significant difference in days to a dry wound or number of postoperative infections The study was stopped prematurely when 15 knees (63%) treated with the ciNPT developed skin blisters.
2011	Atkins et al. 14	CC level 3b	20 (10 ciNPT; 10 standard dressings)	Sternotomy	Presternal perfusion Perfusion increased by 100% in ciNPT group and decreased by 25·7% in control group (P = 0·004).	ciNPT increased perfusion relative to controls and compensated for reduced perfusion resulting from mammary artery harvesting.
2006	Stannard et al. 64	RCT level 2b	Study A 44 (13 ciNPT; 31 control)	Study A Traumatic injury with subsequent surgical incision	Study A Wound drainage   ciNPT: 1·6 days   Control: 3·1 days Infection rate   ciNPT: 8%   Control: 16%	Study A Significantly reduced time of wound drainage in ciNPT group (P = 0·03). No significant difference for infection or wound breakdown.
			Study B 44 (20 ciNPT; 24 control)	Study B High‐energy trauma and calcaneus, pilon, and high‐energy tibial plateau fractures	Study B Wound drainage   ciNPT: 1·8 days   Control: 4·8 days	Study B Significantly reduced drainage time in ciNPT group (P = 0·02).

BMI, body mass index; ciNPT, closed incision negative pressure therapy; CI, confidence interval; OR, odds ratio; RCT, randomised controlled trial.

Classification produced by Bob Phillips, Chris Ball, Dave Sackett, Doug Badenoch, Sharon Straus, Brian Haynes, Martin Dawes (March 2009) 5.

Eight systematic reviews and meta‐analyses were identified in the literature search 58, 79, 80, 81, 82, 83. These studies have examined the potential effects of ciNPT in reducing SSI, seroma/haematoma formation and dehiscence as reported in the literature. Each systematic review used different methods for data comparisons; however, four reviews indicated that ciNPT use may help reduce rates of SSI 58, 79, 82, 83. ciNPT effects on seroma/haematoma formation and dehiscence rate were inconclusive because of inconsistent data reporting. Two reviews stated that while evidence is mounting, no definitive claims can be made as reported evidence is inconsistent 80, 81. A recent meta‐analysis evaluated the effectiveness of ciNPT in lowering the incidence of surgical‐site infections compared with standard incisional care 84. This study used a fixed‐effects model to assess between‐study and between‐incision location subgroup heterogeneity and effect size. The authors demonstrated reduced overall weighted average rates of SSI in the ciNPT (6·61% versus 9·36%). The relative reduction of SSI rate was 29·4%, with the odds of SSI rate decrease equalling 0·496 (P < 0·00001). Overall rates of dehiscence were also reduced in ciNPT versus control groups (5·32% and 10·68%, respectively). These results suggest that ciNPT can be a potentially effective method for reducing SSI and may be associated with decreased incidence of dehiscence. In total, despite the wide variety of surgical procedures and patient comorbidities included in the 35 comparative studies and analysed in eight systematic reviews, the majority reported that patients treated with ciNPT showed reduced SSI rates with the caveat that more large, randomised controlled trials are necessary.

Risk factors in different surgical fields

Based on the EbM literature review and panel member experience, the panel generated a list of risk factors for the development of SSI shown in Table 3. Among comorbidities, the most frequently cited are diabetes mellitus, American Society of Anesthesiologists (ASA) physical classification system score ≥3, advanced age, obesity (BMI ≥ 30 kg/m2), tobacco use, hypoalbuminaemia and corticosteroid use 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99. Most cited surgical incision risk factors for SSI development included incisions after prolonged surgical time, re‐operation or re‐exploration and emergency operation. In addition, incisions in the presence of ischaemia 91, high perioperative blood loss or high surgical tension also have an increased SSI risk. Panel members also designated high tension, open groin or sternotomy incisions as high‐risk incisions where ciNPT use is recommended.

Table 3

Top 25 Risk factors of surgical site infection ranked by number of articles, number of patients and number of surgical fields affected*

Risk factors	Number of articles	Number of patients	Supporting article(s)	Surgical field (GEN, PLA, CAR, ORT, VAS)	ciNPT recommended†
Diabetes mellitus	19	223 336	Imai et al. 85 Xue et al. 86 Harrington et al. 103 Pull ter Gunne et al. 114 Neumayer et al. 87 Martin et al. 99	GEN, CAR, ORT, VAS	X
ASA score ≥3	9	265 783	Berger et al. 88 Xue et al. 86 Si et al. 89 Ridgeway et al. 90 Neumayer et al. 87	GEN, PLA, CAR, ORT,VAS	X
Advanced age	8	231 813	Fahrner et al. 104 Baumeister et al. 105 Harrington et al. 103 Ridgeway et al. 90 Neumayer et al. 87	GEN, PLA, CAR, ORT, VAS	X
BMI > 30 kg/m2	12	151 935	Imai et al. 85 Xue et al. 86 Harrington et al. 103 Pull ter Gunne et al. 114 Turtiainen et al. 91	GEN, PLA, CAR, ORT, VAS	X
Prolonged surgical operation time	13	142 957	Imai et al. 85 Barber et al. 115 Simsek Yavuz et al. 106 Urquhart et al. 92	GEN, PLA, CAR, ORT	X
Active tobacco use	4	178 532	Neumayer et al. 87 Edmonston et al. 116	GEN, PLA, ORT, VAS
Hypoalbuminaemia	4	200 037	Shanmugam et al. 93 Neumayer et al. 87	GEN, VAS
Corticosteroid usage	2	166 026	Slaughter et al. 94 Neumayer et al. 87	GEN, CAR, VAS
Active alcoholism	2	163 624	Neumayer et al. 87 Aggarwal et al. 95	GEN, ORT, VAS
Re‐operation	9	23 825	Fahrner et al. 104 Xue et al. 86 Bryan et al. 96 Aggarwal et al. 95	GEN, PLA, CAR, ORT	X
Male	5	77 984	Imai et al. 85 Namba et al. 107	GEN, ORT
Renal disease/renal dialysis	4	85 004	Centofanti et al. 108 Bozic et al. 109	CAR, ORT
Local arterial insufficiency	2	83 081	Baumeister et al. 105 Bozic et al. 109	PLA, ORT
Chronic obstructive pulmonary disease	3	37 589	Shanmugam et al. 93 Diez et al. 97	GEN, CAR	X
Haematoma	2	38 177	Fahrner et al. 104 Xue et al. 86	GEN, PLA
Pedicled harvest using both internal thoracic arteries	1	126 235	Deo et al. 117	CAR
Hyperglycaemia	2	2351	Ata et al. 118 Richards et al. 119	GEN, ORT	X
Preoperative chemoradiation	2	3070	Xue et al. 86 Olsen et al. 110	PLA	X
Postoperative drainage	2	7463	Pessaux et al. 120 Xue et al. 86	GEN, PLA	X
High perioperative blood loss	1	4855	Sorensen et al. 98	GEN	X
Hypertension (blood pressure)	1	2745	Xue et al. 86	PLA
Malnutrition	2	64	Shinkawa et al. 121 Aggarwal et al. 95	GEN, ORT	X
Venous insufficiency	1	70	Baumeister et al. 105	PLA
High surgical incision tension	N/A	N/A	Panel experience	PLA	X
Thickness of lipodermis	N/A	N/A	Panel experience	PLA

ASA, American Society of Anesthesiologists physical classification system; BMI, body mass index; CAR, cardiothoracic surgery; COPD, chronic obstructive pulmonary disease; GEN, general surgery; N/A, not applicable; ORT, orthopaedic surgery; PLA, plastic surgery; VAS, vascular surgery.

Risk factor ranking was obtained by multiplying the number of articles, the number of patients and the number of surgical fields.

Based on odds ratio >2 or presence in multiple surgical fields.

Algorithm to use ciNPT

Based on the literature review and panel member experience, we developed an algorithm for when a surgeon might consider using ciNPT (Figure 1). In addition to the patient and surgical incision risk factors listed above, ciNPT use may also be appropriate for incisions where infection can cause high morbidity, such as sternotomy, open reduction and internal fixation with hardware or groin area vascular surgery (especially if accompanied by a synthetic graft or vascular graft inserted below the inguinal ligament). The group of authors decided against developing a score. Rather, the relevant risk factors for SSI are presented and must be considered in the light of each individual patient's situation.

Figure 1

Closed incision negative pressure therapy risk factors assessment. Blue indicates low risk for SSI while red indicates high risk for SSI. ciNPT use is recommended in patients with increased number of patient risk factors and incision risk factors. OB/GYN, obstetrics and gynaecology.

Discussion

In open wounds, negative pressure therapy helps promote a wound‐healing environment by reducing oedema, removing infectious materials and promoting perfusion and granulation tissue formation 100, 101, 102. Recently, surgeons are using negative pressure therapy over closed incisions (ciNPT) in a variety of clinical settings. ciNPT appears to manage the surgical incision by reducing incision line tension, decreasing oedema and providing an air tight seal. Using the results of the literature search and panel member experiences, we summarised potential evidence‐based benefits of ciNPT usage, identified both patients and incisions that could potentially benefit from ciNPT and created recommendations for the most appropriate use of this treatment.

Every surgical procedure has its own set of risks for SSIs. While many SSIs can be treated with antibiotics and/or superficial wound debridement, there are certain scenarios in which wound infection has disastrous consequences, such as in a lower extremity prosthetic bypass or joint replacement surgery. As with specific procedures, patients with certain comorbidities are at increased risk of developing SSIs. The most common patient and surgical operation risk factors identified by EbM and panel member experience were: obesity (BMI ≥30 kg/m2), diabetes mellitus (e.g. 50% higher risk of developing SSI following cardiac surgery), tobacco use, prolonged surgical time, ASA score ≥3 and corticosteroid use (Table 3) 85, 88, 92, 93, 95, 96, 103, 104, 105, 106, 107, 108, 109, 110. High‐risk incisions included those with specific characteristics (e.g. incisions that were re‐opened or under high tension) as well as those associated with specific surgical procedures (e.g. pelvic surgery incisions, sternotomy, extremity fractures, open reduction and internal fixation and vascular groin surgery in which synthetic grafts were used).

Using the above information, we created consensus recommendations for the most appropriate use of ciNPT (i.e. in patients with one or more comorbidities or in patients with a surgical incision that is historically at high risk for developing SSIs) (Figure 1).

Despite the small number of ciNPT studies, in comparison to the large number on NPWT, current literature supports its benefit in high‐risk patients and incisions. A majority of the 100 publications reported decreased rates of SSIs, dehiscence and haematoma/seroma formation 14, 15, 16, 17, 22, 23, 24, 25, 27, 28, 29, 30, 31, 32, 35, 36, 37, 38, 39, 40, 41, 42, 44, 58, 59, 63, 64, 65, 66, 67, 68, 70, 71, 111, 112, 113. A recent meta‐analysis reported a 50% reduction in the rate of SSIs in the ciNPT group compared with the control group (OR 0·564; P < 0·00001) 84. Groin incisions were excluded from the analysis. Nevertheless, this study further supports our consensus recommendation. Adverse effects with ciNPT use were only noted in one study (Howell et al.) 77, which was stopped prematurely because of skin blister development at the skin/dressing interface in 63% of the ciNPT group. This adverse effect was most likely because of improper dressing configuration (e.g. a lack of a non‐adherent film dressing or drape used to protect the skin from the foam dressing and too high tension when using the dressing). It is noteworthy that in this study, ciNPT was used for only 48 hours instead of the recommended 7 days. No other study reported any skin blistering or other adverse effects.

Treatment costs are an important issue in patient care. To date, three studies examined the cost of ciNPT use 6, 7, 8 and compared SSO rates and cost savings of ciNPT to routine incision care. Lewis et al. concluded that ciNPT may be a cost‐effective treatment for closed laparotomy incisions following removal of gynaecological cancers if it reduces SSO rates 6. Tuffaha et al. examined use of ciNPT in obese women following caesarean section. Here, ciNPT appeared to be cost‐effective compared with standard wound dressings, although the authors note the high uncertainty surrounding the decision to use ciNPT 7. Lastly, Echebiri and colleagues used a computer model to evaluate the potential economic benefit for prophylactic ciNPT after a caesarean section 8. The authors provided evidence suggesting that ciNPT in high‐risk patients following caesarean section could be cost‐beneficial 8. While these results are encouraging, large cohort studies examining cost savings in various surgical fields are needed.

Limitations exist in this study. The robustness of the consensus recommendations is highly dependent upon the knowledge experience, and objectivity of our panel members. These members were carefully selected based on their personal familiarity with the ciNPT system and their publications in the field. Each reviewed the full literature available on the topic. During the in‐person meeting, any potential panel members' biases were considered based on available evidence and vigorous debates of our medical practices. An additional limitation was the small number of prospective, randomised comparative studies identified in the literature search. Thus, the evidence‐based level of the available articles could skew the consensus guidelines because of a restricted evidence pool. Furthermore, we acknowledge the potential bias introduced by the meeting sponsor (the manufacturer of the PREVENA™ Incision Management System (KCI, an Acelity company, San Antonio, TX), one ciNPT device).

To our knowledge, this is the first consensus document attempting to better define the potential use of ciNPT to reduce the incidence of SSIs. The panel believes that data in the available literature, while limited, allows the surgeon to determine a patient's risk for a particular operative procedure. In high‐risk patients and high‐risk surgical procedures, ciNPT appears to have the potential to reduce surgical incision complications and surgical cost per patient up to $9000 15, 66, 67, depending on the type of incision and patient risk factors. With an estimated 8·7–58·2 million patients globally developing an SSI, use of ciNPT may substantially reduce these rates. As additional high‐level, peer‐reviewed publications become available, these consensus recommendations can be updated.