Incisional negative pressure wound therapy for prevention of surgical site infection

Authors: Patel S, Nischal SA, Kale KM, Dubb A, Prasad SK, Refai D

Affiliation: Thomas Jefferson University Hospital; University of Oxford; Emory University Hospital Midtown

Journal: JAMA Surgery, August 2026

PMID: 42616534

Key takeaways

• Incisional negative pressure wound therapy (iNPWT) reduced overall SSI from 14.4% to 9.6% (RR 0.64); NNT 21.

• Benefit extended to superficial and deep SSI, wound dehiscence, seroma, skin necrosis, and reoperation, but not organ-space SSI.

• iNPWT substantially increased skin blistering/reaction (RR 4.51; NNH 18) and device-related adverse events (RR 11.87; NNH 11).

• Evidence for SSI reduction appears mature, but benefit is not uniform across surgical populations; the authors favor selective rather than routine use.

• Only 4.2% of patients were in breast or plastic surgery, limiting reconstruction-specific conclusions.

Background

iNPWT is increasingly used prophylactically on primarily closed incisions as a nonantibiotic strategy to reduce surgical site infection (SSI). Prior studies have suggested benefit but differ substantially in surgical population, device, wound class, and reported harms.

Objective

Evaluate the efficacy and safety of iNPWT compared with standard dressings in adults undergoing surgery, including effects on SSI subtypes and wound/device complications.

Methods

• Design: Systematic review and random-effects meta-analysis of randomized clinical trials; prospectively registered with PROSPERO and conducted according to PRISMA/Cochrane guidance.

• Search: PubMed, Embase, and CENTRAL through March 7, 2026.

• Studies/patients: 85 RCTs; 16,980 patients; 8,463 iNPWT and 8,517 standard dressing.

• Population: Mean age 48.3 years; mean BMI approximately 31; 16% had diabetes. Most procedures were obstetric/gynecologic (33.5%) or abdominal (29.9%); breast surgery represented 2.7% and plastic surgery 1.4%.

• Primary endpoint: Overall SSI.

• Secondary endpoints: Deep, superficial, and organ-space SSI; dehiscence, seroma, necrosis, reoperation, readmission, mortality, blistering, and device-related adverse events.

• Statistics: Random-effects models using restricted maximum likelihood; risk ratios with 95% CIs; heterogeneity assessed with I². Sensitivity, leave-one-out, publication-bias, and trial sequential analyses were performed. Trial sequential analysis used α=0.05 and 90% power.

• Evidence assessment: Cochrane RoB 2 and GRADE; overall certainty was moderate.

Results

• Overall SSI: 800/8,338 (9.6%) vs 1,207/8,407 (14.4%); RR 0.64 (95% CI 0.57–0.72), p<0.001; NNT 21.

• Deep SSI: RR 0.66 (95% CI 0.51–0.86); NNT 74.

• Superficial SSI: RR 0.59 (95% CI 0.49–0.72); NNT 25.

• Organ-space SSI: No significant reduction; RR 0.90 (95% CI 0.67–1.20).

• iNPWT also reduced dehiscence (RR 0.73), seroma (RR 0.77), skin necrosis (RR 0.38), and reoperation (RR 0.80).

• No significant differences occurred in readmission, mortality, pain, serious adverse events, cellulitis, or wound bleeding.

• Skin blistering/reaction: RR 4.51 (95% CI 2.37–8.58); NNH 18.

• Device-related adverse events: RR 11.87 (95% CI 4.74–29.74); NNH 11.

• Trial sequential analysis supported firm evidence for overall, deep, and superficial SSI.

• Intended duration of therapy was not a significant effect modifier for overall SSI, nor were device pressure or surgical specialty; the optimal duration and pressure on closed incisions remain undetermined.

• Most trials used PICO (4325 pt) or PREVENA (3000), with the remainder using miscellaneous single-use or traditional NPWT systems.

Conclusion

iNPWT reduces overall, superficial, and deep SSI and several wound complications, but increases dressing-related adverse events. The authors conclude that current evidence supports selective use in patients at elevated risk of incisional wound complications rather than universal prophylactic application.

Strengths

• Largest RCT-only systematic review to date: 85 trials and nearly 17,000 patients.

• Prospective protocol, dual-reviewer methodology, RoB 2/GRADE assessment, sensitivity analyses, and trial sequential analysis.

• Reports absolute benefit and harm with clinically interpretable NNT/NNH.

Limitations

• Major clinical heterogeneity across operations, wound classes, devices, pressure settings, dressings, and outcome definitions.

• Only 9/85 trials were low risk of bias; 71 had “some concerns.”

• Blinding is difficult with a visible dressing intervention, and adverse-event reporting was inconsistent.

• Several large, rigorous individual trials reported null results, suggesting the pooled effect may overstate benefit in some settings.

• Subgroup analyses were trial-level and exploratory; direct breast/plastic surgery representation was small.

Clinical relevance

For reconstructive surgeons, these data suggest iNPWT in patients at increased risk for incisional wound complications. The meta-analysis supports an overall SSI benefit, but it does not establish routine use in breast or plastic surgery cases.

Editorial notes

This massive systematic review finds that iNPWT can reduce incision-related complications, but the headline RR of 0.64 should not be interpreted as a universal 36% SSI reduction for plastic surgery patients. Breast and plastic surgery accounted for only 705 of 16,980 patients (4.2%), and some of the largest individual RCTs in other specialties were negative.

The most useful result is the absolute tradeoff: one SSI prevented for every 21 patients treated versus one blistering event for every 18. Blistering is less consequential than an SSI, so these NNT/NNH values are not clinically equivalent, but they make indiscriminate use difficult to justify. For reconstructive practice, the unanswered question is not whether iNPWT works, but instead, which patients have a sufficiently high baseline wound risk for the benefit and cost to outweigh dressing-related morbidity.