Health technology assessment and reimbursement

"INESSS recommends discontinuing NPWT:

• as soon as healthy granulation tissue covers the wound and before it reaches the edges of the wound;
• as soon as the clinical objectives have been met;
• if there is evidence of necrotic tissue, hypergranulation, or stagnation;
• or if the wound is not progressing favourably after two weeks of NPWT;
• in the absence of any benefits related to the health care objectives and the pre-established clinical criteria;
• in a situation in which the use of NPWT is not appropriate;
• if there is no health care team or professional authorized to change the dressings;
• if a complication occurs that threatens the patient's life;
• if moderately severe adverse events occur, namely: excessive bleeding, severe wound or periwound infection, intense pain, an allergic reaction following application;
• when patients and/or their caregivers do not adhere to the treatment."

And the duration rules: "the maximum duration of the original prescription should be 30 days", with medical reassessment at two weeks and renewal only after prior reassessment. For an acute wound, closure should be achievable in "three to four weeks at most."

CAT: Expert consensus, ungraded for effect. AGREE II applied instead · GRADE is not applied because the document makes no effect claim to grade - it is a practice-guidance extract whose recommendations are labelled by basis rather than by certainty, and the great majority are expert opinion. Levels of evidence are cited for three chronic-wound indications - moderate for the non-ischaemic diabetic foot ulcer, low for the arterial or mixed ulcer and low for the stage III or IV pressure ulcer - but the extract does not define those levels, so they cannot be audited. The reason this page matters is not its evidence but its completeness on the questions the trial literature leaves open, and it corrects a claim this wiki has made repeatedly. Across candidate batches 9, 10 and 11 the library recorded that no published NPWT stopping rule exists. That was too strong. No study has ever compared one stopping criterion against another - that remains true and has been confirmed by three independent screening sweeps - but a written, itemised discontinuation rule does exist, and it is here. INESSS recommends discontinuing NPWT as soon as healthy granulation tissue covers the wound and before it reaches the edges; as soon as the clinical objectives have been met; on evidence of necrotic tissue, hypergranulation or stagnation; OR if the wound is not progressing favourably after two weeks; in the absence of benefit against pre-established criteria; where no authorised professional is available to change the dressings; on a life-threatening complication; on excessive bleeding, severe wound or periwound infection, intense pain or an allergic reaction; and where the patient or caregivers do not adhere. It also sets a maximum original prescription of 30 days with medical reassessment at two weeks. That is expert opinion, and any motivation quoting it must say so - but it is a written standard, it converges with the Edmonton criteria reproduced in the Ontario assessment, and it is defensible.

Figures: Figures checked

The 2004 Ontario Health Technology Advisory Committee decision, reproduced in the document's own Table 1:

"Do not provide additional funding for [NPWT], based on the dearth of existing evidence of effectiveness."

The 2006 conclusion, verbatim: "Based on the evidence, the clinical effectiveness of NPWT to heal chronic wounds is unproven. Furthermore, saline dressings are not the standard of practice for first line treatment in Ontario, thereby rendering the literature base irrelevant in an Ontario context. Nonetheless, despite the lack of methodologically sound studies, NPWT use has diffused across Ontario."

Funding context: "No Canadian province provides direct funding for NPWT to providers… There is no fee-for-service reimbursement for physicians under the Ontario Health Insurance Plan."

CAT: GRADE Moderate for the single pivotal trial and Low to Very low for the remainder; the Secretariat's overall verdict is that clinical effectiveness is UNPROVEN · The Secretariat assigned MODERATE to Armstrong 2005, LOW to Moisidis 2004, Moues 2004 and Wanner 2003, and VERY LOW to Ford 2002 and Joseph 2000. Its overall conclusion is that 'the clinical effectiveness of NPWT to heal chronic wounds is unproven', and the 2004 OHTAC decision reproduced in the document is blunter still: 'Do not provide additional funding for [NPWT], based on the dearth of existing evidence of effectiveness.' The most valuable content is not the verdict but two re-analyses of the pivotal trial, which run in opposite directions. On efficacy, Armstrong reported complete closure with or without surgery as 55.8% against 38.8%, P = .04; the Secretariat recalculated it excluding non-randomised surgical closure and obtained 40.3% against 29.4%, P = .15. On harm, the trial reported wound infection as 13 (17%) against 5 (6%) and called it non-significant; the Secretariat recalculated it as 'a statistically significant higher rate of wound infection in the NPWT group compared to the control group (P = .04, based on a 2-tailed Fisher's exact test)'. It also identified a baseline error in Joseph 2000 - initial wound volume 53 cc against 25 cc - and concluded 'the groups were not equal at baseline and the results of the RCT are unreliable'. The two recalculations are applied asymmetrically, and that is a defect: the efficacy recalculation is discounted as 'a post hoc analysis with low statistical power' while the harm recalculation is presented without the same caveat. Both are post hoc and both should carry it. Confidence is further limited by age - the search closed in March 2006, twenty years before this appraisal, and the evidence base has moved substantially since.

Figures: Figures checked

CMS Local Coverage Determination L33821, effective 1 July 2016: "an NPWT pump and supplies are covered when ulcers and wounds are encountered in an inpatient setting or in the home setting when the criteria are met. For ulcers and wounds in the home setting, the beneficiary has a chronic stage III or IV pressure ulcer, neuropathic (e.g., diabetic) ulcer, venous or arterial insufficiency ulcer, or a chronic (present for at least 30 days) ulcer of mixed etiology. A complete wound therapy program… has been tried or considered and ruled out prior to application of NPWT."

Aetna: covers NPWT pumps for inpatient or home use when criteria are met — but "considers NPWT experimental and investigational for the treatment of deep sternal wound infection, partial-thickness burns, tibial fractures, for use following surgical excision of pilonidal sinus disease… and all other indications." Also experimental and investigational: instillation with chemotherapeutic agents; "nonpowered (mechanical) NPWT devices"; and "single-use NPWT devices (PICO… Prevena Incision Management System)… for all indications because of insufficient evidence of their effectiveness."

Group Health Cooperative: covers "an initial 14-day course", with "a complete wound therapy program… tried for 30 days unless edema and/or exudate mandates NPWT." Non-powered devices not covered.

Oregon HERC Guideline Note 62: NPWT included "only for patients who: have wounds that are refractory to or have failed standard therapies; are not suitable candidates for surgical wound closure; or, are at high risk for delayed or non-healing wounds due to factors such as compromised blood flow, diabetic complications, wounds with high risk of fecal contamination, extremely exudative wounds, and similar situations."

Regence: "No published coverage policy for NPWT was identified."

CAT: GRADE Low for diabetic foot ulcer healing and for surgical wounds healing by secondary intention / GRADE Very low for arterial, venous and pressure ulcers / Insufficient for surgical wounds closed primarily · The certainty ratings are the review's own and are reproduced rather than re-derived. Low for complete healing in the diabetic foot ulcer, resting on Blume's randomised 43% against 29%, P = 0.007, together with Yao's adjusted hazard ratio 3.26 (2.21 to 4.83); LOW for time to closure in the same population; LOW for surgical wounds healing by secondary intention on Armstrong's 56% against 39%, P = 0.04; VERY LOW for arterial ulcers, HR 2.27 (1.56 to 3.78), for venous ulcers, HR 6.31 (1.49 to 26.6), and for pressure ulcers, where the risk difference is 3% with an interval from minus 18% to plus 25% and the hazard ratio 1.72 (0.43 to 6.95) crosses one; LOW for mixed ulcers, HR 2.63 (1.87 to 3.70); INSUFFICIENT for surgical wounds closed primarily; LOW for harms; and VERY LOW for effect modification by timing. SNaP against V.A.C. showed no difference in healing, p = 0.9620. Two structural problems depress confidence further than the ratings themselves. First, the diabetic foot conclusion rests substantially on a poor-quality retrospective cohort funded by the device manufacturer, while the one fair-quality randomised trial with a significant healing difference is itself rated only low. Second, and easily missed, the Yao 2014 cohort - the sole source of the arterial, venous, pressure and mixed-ulcer hazard ratios - excluded hiv-positive patients, so four of the eight population estimates carry no HIV-exposed population at all. That exclusion is decisive for South African transferability and is not flagged anywhere in the review's own summary.

Figures: Figures checked

"The Ministry of Health's MTAC has recommended subsidy for:

✓ Negative pressure wound therapy (NPWT) for wound management in patients with diabetic foot wounds and ulcers after debridement as suggested by a multidisciplinary foot care service team, taking into consideration of clinical assessment of the wound (i.e. depth, size, position and state), patient compliance, and use of devices and dressings appropriate to clinical circumstances.

✓ Where deemed clinically necessary, in addition to using NPWT post-surgical intervention, patients should be offered any of the following as adjunctive therapies e.g. offloading, antimicrobial therapy, and evaluation and management of vascular insufficiency. In special cases, NPWT may be used without prior surgical intervention.

✓ The multidisciplinary foot care service team may include, but is not limited to, podiatrists, orthopaedic surgeons, vascular surgeons, and wound care nurses."

Subsidy status, verbatim: "The use of NPWT is recommended for subsidy for the abovementioned indication(s) only. Single-use NPWT systems (e.g. PICO-NPWT) are not recommended for subsidy in the abovementioned indication(s)."

Exclusion, verbatim: "NPWT for these non-diabetic indications was not considered by the Committee for subsidy."

CAT: OCEBM 1 by document type / no certainty rating is assigned by the Committee and none is derived here · No GRADE OR OCEBM certainty is assigned, because the Committee assigns none and because the document cites no individual study by name, so its evidence base cannot be audited from the text. The page is held for the decision, not the evidence, and the decision is the strongest payer precedent in the library: a national government has recommended SUBSIDY for NPWT in diabetic foot wounds and ulcers after debridement, with the gate being A multidisciplinary foot care team decision rather than a device brand, a wound-size threshold or a failed-therapy interval - and that team is specified to include, among others, VASCULAR SURGEONS. Three features make it more useful than a guideline. First, it anticipates the commonest reviewer objection: 'In special cases, NPWT may be used without prior surgical intervention.' Second, it names adjunctive requirements rather than leaving them implicit - offloading, antimicrobial therapy, and evaluation and management of vascular insufficiency. Third, it sets out the five criteria the committee deliberated against, which are a ready-made structure for a submission: clinical need and the nature of the condition; overall benefit to the patient and the system; cost-effectiveness; estimated annual technology cost and the number of patients likely to benefit; and organisational feasibility including barriers to diffusion. And it declines single-use systems: 'Single-use NPWT systems (e.g. PICO-NPWT) are not recommended for subsidy in the abovementioned indication(s).'

Figures: Figures checked

Total cost of care was R109,110 higher per patient with NPWTi-d. The subacute saving of R9,086 per patient is real and significant, and it is one line of a total that runs the other way.

CAT: OCEBM 3 · Start LOW (non-randomised claims-based cohort). -1 risk of bias for selection, and the mechanism is stated by the authors themselves: the funder reimbursed NPWTi-d first-line ONLY where microscopy, culture and sensitivity was positive AND an antibiotic was prescribed, so infection severity was systematically greater in the NPWTi-d arm and was undercaptured by ICD-10 coding. The residual imbalance after matching is large and all in one direction - length of stay +5.3 days, theatre trips +0.6, intensive care +5.0 percentage points and index cost +R114,829, every one significant - and the abstract concedes that NPWTi-d patients 'were likely more complex'. Province was not a matching covariate and the arms remained significantly different on it after matching (Eastern Cape 7.4% against 14.6%, P = .0009; KwaZulu-Natal 17.7% against 10.8%, P = .0041; Limpopo 0.2% against 2.9%, P = .0021). -1 further for a selection step applied to the outcome variable itself: Outlier patients whose costs exceeded the upper interquartile range were removed before matching, in a study whose outcome is cost, with no count of patients removed given. Net very low. Two further observations that do not alter the rating but govern how the paper is quoted. First, the headline the abstract leads with is one line of a total that runs the other way: the subacute and rehabilitation saving is real and significant at R3,231 against R12,317 per patient, but total cost of care was R333,957 with NPWTi-d against R224,847 with NPWT, p < .001 - R109,110 higher per patient. Second, no clinical outcome was measured at all: no closure, no healing time, no amputation and no mortality, so the paper cannot support any claim that the more expensive therapy worked better.

Figures: Figures checked