Necrotising infection and Fournier's gangrene

• Novel 3-step "eso-endo" NPWT technique: two sponges bridged through the rectal fistula so suction runs cavity-to-rectum, promoting depth-to-surface granulation and preventing fistula enlargement.
• Case managed with early debridement, diverting blow-out colostomy, culture-directed antibiotics (Meropenem/Tigecycline/Fluconazole; Enterococcus raffinosus, Proteus mirabilis), NPWT for 32 days and concurrent HBOT; bilateral perineal wounds closed primarily, no residual rectal defect.
• Review consensus: WSES/SIS-E recommend NPWT after complete necrotic-tissue removal, but the authors state clinical evidence of NPWT's added benefit over conventional dressing is not yet proven.
• Reported NPWT advantages across included studies: fewer dressing changes, less pain, greater mobility, reduced tissue exudate/edema, and support for skin-graft/dermal-matrix take.
• Iacovelli et al. cohort (n=92) cited: NPWT gave faster healing at 10 weeks and higher 90-day survival in disseminated (not local) FG; some studies (e.g. Yucel) found longer stay/more debridements with NPWT reflecting worse baseline disease.

CAT: OCEBM 4 · There is no effect estimate in this paper. Twenty-eight articles were included — 21 case reports or series and 7 cohorts — and the authors state that meta-analysis was not possible because of clinical and methodological heterogeneity. GRADE therefore begins Low for a body of uncontrolled observational evidence and falls to Very low for risk of bias and imprecision, and it attaches to a description of practice, not to a treatment effect. The review's own verdict is the citable finding: 'the heterogeneity and the low quality of evidence of all of the included studies do not lead to reliable conclusions', and separately that although WSES and SIS-E recommend NPWT after complete necrotic-tissue removal, clinical evidence of its added benefit over conventional dressing is not proven. The review also contains direct evidence against NPWT that the abstract does not carry. Yucel 2017 (n = 25) found NPWT associated with LONGER hospital stay and MORE debridements, which the authors attribute to worse baseline disease rather than to the dressing. Gul 2021 (22 cases, NPWT in 12) found no statistically significant benefit for mortality or morbidity. Ozturk 2009 (n = 10, five per arm) found no difference in length of stay or time to therapy. The one supportive cohort is Iacovelli 2021 (n = 92), which reported faster healing at 10 weeks and higher 90-day survival in disseminated but NOT in local Fournier's — and which also reported LONGER length of stay in the NPWT arm in both groups. The case itself is confounded: hyperbaric oxygen was co-administered alongside 32 days of NPWT at minus 125 mmHg continuous, so the contribution of NPWT cannot be isolated, and the authors say as much.

Figures: Figures checked

• Sepsis/SIRS criteria were present in 78/86 (91%) of patients before NPWT — i.e. the cohort included septic patients (p.13 abstract; p.14).
• Wound burden was large: tissue necrosis in 84/97 (87%), infection in 86/97 (89%, culture-confirmed in 83/86, 97%); average wound size 619 cm² by area and 786 cm³ by volume (abstract; p.15).
• Wound location (Results, p.15): extremities 41/97 (42%); torso/head and neck 56/97 (58%).
• Mean time to wound closure was 17 days (median 10, mode 6) (abstract).
• Durability of wound closure was 73/79 (92%) (abstract).
• Deaths occurred in 6/86 (7%); no deaths appeared related to NPWT — supporting NPWT as safe and effective even in acute, contaminated, septic wounds, the largest reported cohort of this type (abstract; p.14).

CAT: OCEBM 4 · A single-arm series cannot support a comparative claim, and the title of this one promises 'safety and efficacy'. There is no comparator. Efficacy is not assessed; what is assessed is what happened to 97 wounds under one surgeon. GRADE starts Low and falls to Very low for very serious risk of bias - retrospective, unblinded, single-surgeon, with the treating surgeon adjudicating whether his own therapy caused each death. Why the series still matters: it is the largest published cohort of NPWT applied to genuinely septic, necrotic, contaminated wounds, and that is the setting clinicians are most reluctant to enter. SIRS or sepsis criteria were met by 78 of 86 (91 per cent) - SIRS 4, sepsis 24, severe sepsis 19, septic shock 31, none 8. Tissue necrosis in 84 of 97 (87 per cent), reaching fascia in 38 of 84, muscle in 21 of 84 and bone in 8 of 84. Infection in 86 of 97 (89 per cent), confirmed by purulence or culture in 83 of 86 (97 per cent). Mean time to closure 17 days, median 10, mode 6. Durability of closure 73 of 79 (92 per cent). Deaths 6 of 86 (7 per cent), none judged NPWT-related. Three things that do not reconcile. First, wound size: the abstract and Results print 619 square centimetres, the Discussion of the same paper prints 617, and the companion paper reveals the denominator this one omits - area was measurable in only 53 of 97 wounds and volume in 43 of 97. The headline is a mean of 53, not of 97. SECOND, LOCATION: the abstract says torso 41 of 97 and extremities 56 of 97, the Results text says the reverse, and the companion's table 1 settles it - upper extremity 8, lower extremity 48, torso 41, combined 2, so the ABSTRACT is right and the Results text is transposed. Third, the outcome denominator: durability is 73 of 79 while complications are 6 of 81, and the stated exclusions - 4 deaths before closure, 2 comorbidity delays, 7 secondary-intention closures, 2 lost to follow-up, 1 refused graft - total 16, leaving 81. Where 79 comes from is never explained.

Figures: Figures checked

Overall mortality 19 of 95 (20.0%). Independent prognostic factors: FGSI per point OR 1.28 (1.10–1.55), procalcitonin per ng/mL OR 1.95 (1.35–2.75); qSOFA ≥ 2 OR 2.85 (1.40–5.80) in the sensitivity model. AUC: FGSI 0.82, qSOFA 0.78, procalcitonin 0.80.

"Patients in the conventional group had higher FGSI and procalcitonin levels and were older, indicating more severe disease at baseline. Although adjusted analysis showed that NPWT/VAC was independently associated with lower mortality, residual confounding cannot be excluded."

CAT: OCEBM 4 (retrospective cohort, non-randomised allocation). GRADE VERY LOW for any mortality effect of NPWT: starts LOW, downgraded for risk of bias (confounding by indication acknowledged, allocation by surgeon) and imprecision (19 events; crude p = 0.17 on recomputation). ROBINS-I: critical (confounding) · Start low for a non-randomised comparison. Downgrade for risk of bias: the exposure was assigned 'at the discretion of the treating surgeon based on clinical judgment, wound characteristics, and device availability, without predefined criteria', and the group that did not get NPWT was older (65.3 vs 59.9), sicker (FGSI 7.6 vs 6.4) and more inflamed (procalcitonin 8.9 vs 6.3), each printed as p < 0.05 in Table 1 — confounding by indication in the direction that produces the result, stated by the authors themselves in the Discussion. Downgrade for imprecision: 19 deaths in total; the crude contrast 29.0% vs 15.6% is printed as p < 0.05 but recomputes to Fisher two-tailed p = 0.17 (chi-squared without correction 0.13, with Yates 0.21); the adjusted OR 0.42 (0.18–0.95) comes from a five-covariate model on 19 events, below four events per variable, where confidence intervals are unreliable in principle; and the paper's power statement — 74% for 30 per group at 29% vs 15.6% — recomputes to about 24%, with roughly 150 per group needed for 80%. Very low. What survives: the cohort's overall mortality, 20% (19 of 95), is in the range every Fournier's series reports, and the prognostic findings — FGSI per point OR 1.28 (1.10–1.55), procalcitonin per ng/mL OR 1.95 (1.35–2.75), qSOFA ≥ 2 OR 2.85 (1.40–5.80), all with AUCs near 0.8 — are consistent with the literature and are the usable content. The paper's own last sentence, that the effect on survival 'remains uncertain', is the grade in words. It sits beside Küçük 2026 (268 patients, same country, same design, same flaw) and Iacovelli 2020, and adds a third cohort in which NPWT is given to the patients most likely to survive.

Figures: Figures checked

% wound-area reduction, median (IQR), as stated in the abstract and text (the table headers are reversed):

• Pain: no difference at any time point.
• Positive cultures: 22/24 in each arm at baseline. After 2 weeks: 5/24 with HBOT vs 6/24 without; no difference.
• Population: 18/24 per arm necrotising fasciitis; diabetes 7/24 (NPWT alone) vs 6/24 (NPWT + HBOT).

CAT: GRADE very low (HBOT added to NPWT; surrogate area-reduction outcome) · Start high (RCT). −1 risk of bias: unblinded patients and treating teams, retrospective registration, mislabelled result tables, per-timepoint testing without a repeated-measures model, two losses per arm by day 12. −1 indirectness: % area reduction over 12 days is a surrogate; 'chronic' wounds were mostly debrided necrotising fasciitis; anticoagulant and antiplatelet users excluded, which removes most vascular patients. −1 imprecision: 48 patients. Very low.

Figures: Figures checked

• Simple/low-risk wounds (small, acute, no exposed vital structures, minimal devitalised tissue, ready for closure): NPWT recommended as needed; where primary closure is feasible, ciNPT over the closed incision as needed; ciNPT specifically recommended when the patient is at high risk of surgical site infection.
• Complex/large/high-risk wounds: after determining whether debridement is needed — if the debrided wound is clean, manage with NPWT; if it requires cleansing (infection, foreign body, or needs extensive granulation), NPWTi-d is preferred; NPWT can bolster a skin graft.
• NPWT with instillation and dwell time (NPWTi-d) recommended for both simple and complex wounds that require cleansing; compatible solutions include normal saline, hypochlorous acid, dilute sodium hypochlorite, acetic acid, and PHMB/betaine.
• Timing matters: NPWT started within 1-2 days of admission (trauma wounds) is associated with reduced length of hospital/ICU stay and lower per-patient cost; NPWTi-d within 1 day of admission reduces treatment length, debridements, time to closure and re-admissions.
• Reassess patient and wound at every dressing change; discontinue both NPWT and NPWTi-d if healing remains stalled after 4 weeks; step NPWTi-d down to NPWT once cleansing goals are met.
• Following closure, use NPWT or ciNPT as needed to supplement reconstruction per the reconstructive ladder.
• Economics: an economic model of hospitalised patients reported total potential savings with NPWTi-d of USD 33,338 / EUR 8,467 / GBP 5,626 versus control; ciNPT after breast reconstruction saved ~USD 218/patient; DIEP-flap ciNPT reduced per-patient costs by GBP 420.77 (complications) and GBP 446.47 (length of stay); a retrospective SA private-insurance claims analysis found NPWTi-d patients less likely to need sub-acute/rehabilitation care with significantly lower average cost.

AGREE II: Not recommended for use as a guideline; citable only as an industry-convened opinion document, with its provenance stated and its silence on harm supplied from elsewhere — Scope 56% · Stakeholders 33% · Rigour 19% · Clarity 61% · Applicability 46% · Independence 17%

Figures: Figures checked