ORIGINAL ARTICLE

A systematic review of the

effectiveness of negative
pressure wound therapy in
the management of diabetes
foot ulcers
Georgia Noble-Bell, Angus Forbes

Noble-Bell G, Forbes A. A systematic review of the effectiveness of negative pressure wound therapy in the
management of diabetes foot ulcers. Int Wound J 2008;5:233–242.

ABSTRACT
Foot ulcers are a common complication in patients with diabetes. Negative pressure wound therapy (NPWT) is
a wound care therapy that is being increasingly used in the management of foot ulcers. This article presents
a systematic review examining the effectiveness of this therapy. The review question is how effective is NPWT in
achieving wound healing in diabetes foot ulcers? The primary outcome for this study was the number of patients
achieving complete wound healing (secondary outcomes, other markers of wound healing, adverse events and
patient satisfaction). A systematic literature review and tabulative synthesis of randomised controlled trials
(RCTs). The review identified four RCTs of weak to moderate quality. Only one study examining NPWT in
postamputation wound healing reported data on the primary outcome. These data show a 20% improvement in
wound healing [odds ratios ¼ 2
0%, confidence interval (CI) 1
0 to 4
0] and number needed to treat ¼ 6 (CI
4–64). No serious treatment-related complications were reported by any of the studies. One study suggested
a reduction in the risk of secondary amputation (absolute risk reduction ¼ 7
9%, CI 0
5–15
43). Studies also
reported an increase in granulation and wound-healing rates in patients treated with NPWT therapy. No data on
patient satisfaction or experience were reported. While all the studies included in the review indicated that the
NPWT therapy is more effective than conventional dressings, the quality of the studies were weak and the nature
of the inquiries in terms of outcome and patient selection divergent. There is a strong need for larger trials to
assess NPWT therapy in diabetes care with different groups of patients and in relation to different clinical
objectives and parameters.
Key words: Diabetes mellitus • Foot ulcer • Negative pressure wound therapy • Systematic review • Vacuum-assisted closure

INTRODUCTION amputation and mortality (1). The prevalence of

Key Points
Foot ulcers are one of the most serious compli- foot ulcers ranges from 4% to 10% with the
cations of diabetes, being associated with both lifetime incidence being as high as 25% (2). Foot • foot ulcers in diabetes have a
ulcers in diabetes have a complex underpinning complex underpinning pathology
involving neuropathy, ischaemia,
Authors: G Noble-Bell, BSc, King‘s College London, The pathology involving neuropathy, ischaemia, in-
infection and deformity
Florence Nightingale School of Nursing & Midwifery, London fection and deformity (3). This complex pathol-
SE1 8WA, UK; A Forbes, PhD, King’s College London, The
• this complex pathology means
ogy means that foot ulcers have extended that foot ulcers have extended
Florence Nightingale School of Nursing & Midwifery, London
SE1 8WA, UK healing times, making them very costly to healing times, making them
Address for correspondence: Dr Angus Forbes, PhD, manage with intensive professional input and very costly to manage with
King‘s College London, The Florence Nightingale School of intensive professional input
a high use of expensive dressings.
Nursing & Midwifery, James Clerk Maxwell Building, 57 and a high use of expensive
Waterloo Road, London SE1 8WA, UK The management of diabetes foot ulcers is dressings
E-mail: angus.forbes@kcl.ac.uk multifaceted incorporating local and systemic

ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc 233
• International Wound Journal • Vol 5 No 2

Key Points
• there are now a number of
alternative therapies including
lava therapy, tissue engineering,
topical growth factors, hyper-
baric oxygen therapy and neg-
ative pressure wound therapy
(NPWT)
• the NPWT system is used to
manage extensive and very
complex wounds, often in con-
junction with surgical proce-
dures such as grafts and flaps
• the reported benefits of NPWT
therapy are that it provides
a closed moist wound-healing
environment; decreases wound
volume by drawing the wound
edges together; removes exu-
date; reduces infection rates;
reduces oedema at the wound
site, thereby increasing blood
flow and increases mitosis,
promoting granulation
• this article reports a systematic
review of the evidence relating
to the effectiveness of NPWT in
the management of diabetes
foot ulcers
• the primary outcome for this
study was the number of
patients achieving complete
wound healing (defined as
complete reepithelisation or
wound closure)
• in addition, results relating to
any adverse events associated
with NPWT and information on
the patients’ experience (satis-
faction and quality of life) of
the therapy were sought
234
Effectiveness of NPWT in the management of diabetes foot ulcers
management. Wound care is an important
element of this management, which has tradi-
tionally been based on the use of dressing
products (gauze, foam dressings, alginates and
hydrocolloids) and manual debridement. How-
ever, there are now a number of alternative
therapies including lava therapy, tissue engi-
neering, topical growth factors and hyperbaric
oxygen therapy. Negative pressure wound
therapy (NPWT) is one of these emerging
treatments.
NPWT was first introduced in North America
in 1995 (4). NPWT is an adjunctive therapy
consisting of a non invasive wound closure
system that uses controlled negative pressure to
promote healing (5). The NPWT system is used
to manage extensive and very complex wounds,
often in conjunction with surgical procedures
such as grafts and flaps. The therapy is targeted
to the needs of the individual wound and the
amount of negative pressure can vary from 50
to 175 mmHg.
The reported benefits of NPWT therapy are
that it provides a closed moist wound-healing
environment; decreases wound volume by
drawing the wound edges together; removes
exudate; reduces infection rates; reduces
oedema at the wound site, thereby increasing
blood flow and increases mitosis, promoting
granulation (6). No major disadvantages have
been associated with NPWT. However, the
therapy is quite intrusive and patients may be
attached to the NPWT device for up to 2 weeks,
although portable models suited to home-based
treatment are available. One recent study found
that NPWT had a negative effect on quality of
life (7). In addition, there is the issue of cost;
NPWT costs around £45 per day per patient (8).
While the therapy is expensive if it does reduce
healing times, substantial savings could be
made in both dressing costs and staff time.
Therefore, it is important to objectively
appraise the available evidence for the use of
NPWT to establish its contribution to diabetes
foot care. This article reports a systematic review
of the evidence relating to the effectiveness of
NPWT in the management of diabetes foot
ulcers.
METHODS
Aim and objectives
The aim of the review was to assess the clinical
effectiveness of NPWT in the treatment of
ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
diabetes-related foot ulcers. The review objec-
tives were
• to identify and retrieve primary studies
relating to the effectiveness of NPWT in
the management of diabetes foot ulcers
• to critically appraise the quality of the
selected studies
• to extract relevant information from the
selected studies
• to synthesise the data extracted to pro-
vide an overview of the evidence for the
use of NPWT.
Outcomes
The primary outcome for this study was the
number of patients achieving complete wound
healing (defined as complete reepithelisation or
wound closure). However, it was anticipated
that other facets of wound healing may also be
clinically important, such as the mean number
of days to wound healing and the speed of
reduction in wound size and tissue repair. In
addition, results relating to any adverse events
associated with NPWT and information on the
patients’ experience (satisfaction and quality of
life) of the therapy were sought.
Selection of studies
Comprehensive electronic searches of the Co-
chrane register of controlled trials, Medline,
Embase and CINHAL, were performed using
search terms for NPWT and diabetes mellitus
(Table 1). In addition, secondary references
were identified and citation searches were
undertaken. Hand searches of the following
journals (past 10 years issues) were also per-
formed: International Wound Journal, Wound Care
Journal, Diabetic Foot Journal and Ostomy Wound
Care. Advice was also sought from a clinical
specialist in diabetes foot care who was very
Table 1 Search terms (diabetes and NPWT)
Population Intervention
Diabetes mellitus Negative pressure wound therapy
Diabetes mellitus Vacuum assisted closure
(free text)
exp diabetes
mellitus (index)
VAC therapy
Negative pressure wound therapy
NPWT
Sub-atmospheric wound therapy
NPWT, negative pressure wound therapy.
 Effectiveness of NPWT in the management of diabetes foot ulcers
familiar with NPWT (Acknowledgement section).
Differences over the inclusion of studies were
resolved through discussions and consensus.
All the identified titles and abstracts were
examined and articles were selected indepen-
dently by two reviewers if they met the
following inclusion criteria:
• Diabetes mellitus: either type1 or type 2
diabetes
• Foot ulceration: any foot wound, chronic
or acute, including postoperative
wounds
• NPWT: vacuum-assisted closure with the
use of negative pressure was a central
component of the wound management
• Randomised controlled trials (RCT).
Data extraction
Data were independently extracted from the
selected studies by two authors (GN and AF)
using a standardised extraction sheet. The
standardised extraction sheet identified as
follows:
1. General information: paper details and
setting (hospital, primary care and
community).
2. Trial characteristics: design, duration,
randomisation method and allocation
concealment method (clearly blinding
of therapy was not possible).
3. Intervention details including mode of
intervention delivery and context.
4. Patients: details of patients considering
any differences at baseline and/or dur-
ing the course of the study between
intervention and control, other than
exposure to the NPWT.
5. Outcomes: data were extracted on all
outcomes (wound healing, adverse
events and patient satisfaction) and the
length of follow-up.
6. Results: for outcomes based on an
intention-to-treat analysis.
Synthesis
Given the small number of studies and the
heterogeneity of designs, statistical meta-analysis
was inappropriate. Therefore, a tabulative syn-
thesis was undertaken detailing the key charac-
teristics of the studies with their estimations
of effect. Absolute risk reduction (ARR), odds
ratios and number needed to treat were
calculated where data were available in the
ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
individual studies with 95% confidence inter-
vals (CIs).
Key Points
• only 20 studies were identified
Quality assessment through the different search
The quality of each trial was assessed in relation strategies and of these only 9
to the following criteria (9,10): met the inclusion criteria and
were subject to full assessment
1.Randomisation and concealment adequate. (see table 2)
2.80% follow-up.
3.Intention-to-treat analysis.
4.Outcome assessment blind – meaning
that the researchers assessing the
wounds at the pretest and posttest
screens did not know which patients
were in the intervention or control
groups. Clearly, wound assessment other
than complete healing can be subjective.
Therefore, consideration was given to the
validation methods chosen to establish
wound healing. In other types of com-
plex wound such as pressure sores and
fungating breast lesions, a range of
techniques have been suggested, such as
• validated wound assessment tools
• ultrasound imaging (to capture wound
depth)
• photographs (to capture wound surface
measurements) (11).
Based on these criteria, the studies were
defined as being strong (all quality criteria
met), low risk of bias; moderate (criteria 1 and
2 met), moderate risk of bias and weak (else),
high risk of bias.
RESULTS
Given the novel nature of the intervention and
its discrete application in the management of
diabetes-related foot ulcers, only 20 studies
were identified through the different search
strategies. Nine of these studies met the inclu-
sion criteria and were subject to full assessment.
Included studies
Four of the nine studies met the review inclusion
criteria and these studies are described in
Table 2. The studies were highly heterogeneous
with variations in sample, design and outcome
assessment.
The extracted data are summarised by out-
come below.
Wound healing
The primary outcome for the review was the
number of patients achieving complete wound
235
 Effectiveness of NPWT in the management of diabetes foot ulcers
Table 2 Summary of included studies
Study Study and grade of evidence
Armstrong and Multicentre RCT comparing a NPWT-
Lavery (12) treated group (n ¼ 77) to a control
group (n ¼ 85) treated with alginates,
hydrocolloids, foams or hydrogels in
postpartial foot amputation in pa-
tients with diabetes. The length of the
trial was 112 days. Nineteen patients
withdrew from both groups. Intention-
to-treat analysis. Grade of evidence ¼
moderate
McCallon Single-centre RCT comparing a NPWT-
et al. (13) treated group (n ¼ 5) to a control
group (n ¼ 5) treated with saline
gauze. Patients had a non healing foot
ulcer for more than 1 month, had
surgical debridement and had good
perfusion to the affected limb. The
length of the trial was 90 days. All
participants remained in the trial.
Grade of evidence ¼ weak
Eginton Two-centre RCT (cross-over design) com-
et al. (15) paring NPWT therapy to conventional
moist dressing (hydrocolloid) in the
treatment of large foot wounds in
diabetes. The sample comprised 10
patients (10 patients with 11 wounds)
with diabetes who had foot wounds not
expected to heal within a month,
adequate perfusion and debridement
prior to trial. The sample was divided
with each group being treated for 2
weeks on either therapy and then
switched. Trial lasted 4 weeks. Four
patients withdrew before the end of the
trial. Grade of evidence ¼ weak
Etoz et al. (14) Single-centre RCT comparing the effec-
tiveness of NPWT (n ¼ 12) to
conventional moist gauze dressing
(n ¼ 12) on foot wounds in diabetes
as a method to prepare the wound
bed for surgical closure. The maximum
length of treatment was 24 days.
Grade of evidence ¼ weak
ARR, absolute risk reduction; CI, confidence interval; NPWT, negative pressure wound therapy; NNT, number needed to treat; RCT,
randomised controlled trials.
236 ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
Outcomes
Wound healing
Primary outcome (number of patients achieving wound
healing): 43/77 in NPWT, 33/85 in control: odds
ratios ¼ 2
0% (CI 1
0 to 4
0), NNT ¼ 6 (CI 4–64)
Time to wound closure: median of 21 days shorter
in NPWT group (P , 0
01)
Wound healing: improved granulation in the NPWT-treated
group (P , 0
01)
Adverse events
Adverse events reported in 9 (12%) intervention and
11 (13%) control patients: ARR ¼ 1
3% (CI 8
87
to 11
37)
Second amputation necessary in two (3%) intervention and
nine (11%) control patients (P ¼ 0
06): ARR ¼ 7
9%
(CI 0
5–15
43)
Patients satisfaction
Not evaluated by the study
Wound healing
Primary outcome (number of patients achieving wound
healing): not evaluated by the study
Time to wound closure: mean of 20 days shorter (14
9)
in the NPWT group
Wound healing: 18
8% (7
4) greater decrease in wound
surface area in NPWT group compared with control
Adverse events
No serious adverse event reported (patients in NPWT group
report pain and minor bleeding at dressing change)
Patients’ satisfaction
Not evaluated by the study
Wound healing
Primary outcome (number of patients achieving wound
healing): not evaluated by the study
Time to wound closure: not evaluated by the study
Wound healing: the NPWT group showed advantages in
wound length, depth, width, surface area and volume
Adverse events
Treatment-related adverse events were not accounted for
in this study
Patients’ satisfaction
Not evaluated by the study
Wound healing
Primary outcome (number of patients achieving wound
healing): not evaluated by the study
Time to wound closure: not evaluated by the study
Wound healing: improved granulation and greater
reduction in wound size in the NPWT-treated group
Adverse events
No serious adverse event reported (patients in NPWT group
report pain and minor bleeding at dressing change)
Patients’ satisfaction
Not evaluated by the study
 Effectiveness of NPWT in the management of diabetes foot ulcers

healing. These data were only examined in one Patient satisfaction

study examining the effects of NPWT on
treating patients following a partial foot ampu-
tation (12). The data suggest that NPWT would
give a 20% better chance of achieving wound
healing in this type of patient and that one in
every six patients treated would benefit,
although the CIs for these data were large
(Table 2).
Secondary wound-healing outcomes included
time to wound healing and factors suggesting
progress in wound healing. Two of the studies
examined wound-healing time with both report-
ing quicker healing times in the NPWT group.
In the Armstrong and Lavery (12) study, the
average healing time was 27% quicker [median
time to wound closure in the NPWT group was
56 days (inter-quartile range 26–92) compared
with 77 days (IQR 40–112) in the control group],
and in the smaller McCallon et al. (12) study, it
was 53% quicker [wound healing was achieved
in a mean of 22
8 (17
4) days in the NPWT
group and 42
8 (32
5) days in the control
group]. Improved wound-healing characteris-
tics were reported in all of the studies (Table 3).
Adverse events
Overall, no major events were reported in
relation to NPWT treatment compared with
convention therapy, although the level of
reporting was poor in most of the studies. The
Armstrong and Lavery (12) study suggested
some risk reduction in the need for further
amputation in NPWT-treated patients (ARR ¼
7
9%, CI 0
5–15
43). Bleeding and pain at time of
dressing change with NPWT were reported in
two of the studies (13,14). The Eginton et al. (15)
study reported a withdrawal because of an
incorrect pressure setting (too low) for the
NPWT system.
None of the studies reported data on patient Key Points
satisfaction or quality of life. • none of the studies reported
data on patient satisfaction or
Excluded studies quality of life
Studies were excluded from this review for two
main reasons: a lack of relevance to the review
question and/or a failure to meet the inclusion
criteria (Table 4). While methodologically too
weak for inclusion, a few of these studies do
provide some (low level) evidence relevant to
this review. Armstrong et al. (16) undertook
a retrospective study of 31 patients having
NPWT postsurgical debridement of foot ulcers.
Prior to the initiation of NPWT, patients had
their ulcers for a mean of 25 weeks (standard
deviation 23). Only three of the patients
required further amputation, with the remain-
der showing clinically significant progress in
wound healing. Adverse findings included
periwound maceration (n ¼ 6), periwound
cellulitis (n ¼ 1) and deep space infection (n ¼
1). Page et al. (17) compared the records of 47
patients (66% of whom had diabetes) who were
treated with NPWT (n ¼ 23) or moist dressings
(n ¼ 25) and concluded that NPWT was
associated with reductions in hospitalisation
(80%), complications (83%) and the need for
surgical procedures (76%). No significant differ-
ences were found in wound-healing time,
although the NPWT group was older with
larger wounds. Another retrospective study of
70 patients with chronic wounds (50 postgraft-
ing) by Carson (18) found that NPWT was
associated with a high rate of closure (48 days),
with all of the skin-grafted patients achieving
complete healing. However, only 25 of 70 pa-
tients had diabetes and there are no differential
data. A recent study by Lavery et al. (19) com-
pared clinical record data (n ¼ 1135) of patients

Table 3 Reported changes in wound-healing factors

Study Factors Intervention NPWT Control

Eginton et al. (15) Wound length (% change) 4
3 (4
7) þ6
7 (11
5)
Wound width (% change) 12
9 (5
2) þ2
4 (7
5)
Wound depth (% change) 49 (11
1) þ7
7 (5
2)
Surface area (% change) 16
4 (6
2) þ5
9 (17
4)
Wound volume (% change) 59 (9
7) þ0
1 (14
7)
Armstrong and Lavery (12) Median time (days) to 70–100% granulation 42 (IQR 40–56) 84 (IQR 57–112)
McCallon et al. (13) Surface area (% change) 28
4 (24
3) 9
6 (16
9)
Etoz (14) Wound size decrease (mean cm2) 19
5 (11
1) 9
5 (4
1)

NPWT, negative pressure wound therapy; IQR, inter-quartile range.

ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc 237

Key Points
• while all the studies reported
superior wound healing with
NPWT, high levels of heteroge-
neity in the application of the
therapy mean that it is not pos-
sible to give any valid estimate
of the benefits of NPWT
• from the evidence presented,
the only group for which there
was a moderate level of evi-
dence for treatment are patients
with a post amputation wound
who have a good blood supply
• while assessing blood supply is
an important safety factor in
the use of NPWT, it also means
that many people with diabetes
related foot wounds will not
be suitable for the therapy as a
significant proportion of pa-
tients will have some level of
vascular compromise
• the danger is that such results,
in a very select patient pop-
ulation, distort the perceived
clinical benefits of the therapy
• a Cochrane review of NPWT in
other wound settings suggested
that the only valid comparison
in establishing the efficacy of
NPWT was a comparison with
an identical dressing system
without the negative pressure
component
• there have been suggestions
that NPWT is more appropriate
for the management of acute
wounds
238
Effectiveness of NPWT in the management of diabetes foot ulcers
Table 4 Excluded studies
Studies Rational for exclusion
Page et al. (17) Not a RCT and the study
population was not exclusive
to diabetes foot ulcers
Loree (20) Not a RCT and the study
population was not exclusive
to diabetes foot ulcers
Carson et al. (18) Not a RCT and the study
population was not exclusive
to diabetes foot ulcers
Armstrong et al. (16) Not a RCT
Ballard and McGregor (21) This was a case study with
only one patient
Lavery et al. (19) Not a RCT
RCT, randomised controlled trials.
treated with NPWT with data from a meta-
analysis of RCTs of traditional wound care. They
reported a 15% and a 13% greater level of
endpoint success at 12 and 20 weeks, respect-
ively, in the NPWT-treated sample. They also
suggest that the data show a significant reduction
in treatment costs. However, it is very difficult to
accept such a comparison as there is a high risk
of bias because of patient selection, treatment
heterogeneity and endpoint assignation.
DISCUSSION
This review set out to examine the effectiveness
of NPWT in relation to wound healing, adverse
events and patient satisfaction. The review
highlights a number of issues relevant to both
clinical practice and future research for each of
these outcome areas.
Wound healing
While all the studies reported superior wound
healing with NPWT, high levels of heterogene-
ity in the application of the therapy mean that it
is not possible to give any valid estimate of the
benefits of NPWT. Clinically, the most import-
ant question is which patients should be treated
with NPWT? From the evidence presented, the
only group for which there was a moderate level
of evidence for treatment are patients with
a postamputation wound who have a good
blood supply. This latter characteristic is impor-
tant in considering NPWT, as most of the studies
operated strict inclusion criteria to exclude
patients with peripheral vascular dysfunction
and cellulitis. While assessing blood supply is
an important safety factor in the use of NPWT,
ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
it also means that many people with diabetes-
related foot wounds will not be suitable for the
therapy as a significant proportion of patients
will have some level of vascular compromise.
Furthermore, it may be that the large improve-
ments observed with NPWT are in part ex-
plained by the purposive selection of patients
with a good blood supply. The danger is that
such results, in a very select patient population,
distort the perceived clinical benefits of the
therapy.
A related problem was the choice of compar-
ison. It was noted that two of studies compared
the NPWT with saline-moistened gauze. This is
a potentially weak comparison as there are
many newer wound-healing products in com-
mon use (22). Perhaps comparing NPWT with
other modern wound-healing options, such as
growth factors of hyperbaric oxygen therapy,
would provide a stiffer test for the therapy.
Indeed a Cochrane review of NPWT in other
wound settings suggested that the only valid
comparison in establishing the efficacy of
NPWT was a comparison with an identical
dressing system without the negative pressure
component (23).
An additional area of variation was in the
underlying purpose of the NPWT. While the
therapy was used to prepare the wound bed for
other surgical interventions such as skin graft-
ing in the Etöz et al. (14) study, it was used
specifically for wound closure in the Armstrong
and Lavery (12) and McCallon et al. (13) studies.
There was further variation, as in the Armstrong
and Lavery (12) study, the NPWT was specifi-
cally used in postamputation healing. This
variation raises important questions about the
function of this technology and its actual
contribution to wound management. Does the
technology work by directly manipulating
factors that improve wound healing (e.g. moist
wound healing or increasing blood flow) or is
it a tool for managing complex wounds that
minimises the range of factors that may impede
the wound-healing process (e.g. infections or
excess exudate)? Clear answers to these ques-
tions may help inform clinical decisions about
with whom, when and for how long this therapy
might best be deployed.
One factor relating to the decision to use
NPWT is whether the wound is chronic or acute.
There have been suggestions that NPWT is more
appropriate for the management of acute
wounds. In a follow-up analysis of the data in
 Effectiveness of NPWT in the management of diabetes foot ulcers
the Armstrong and Lavery trial, Armstrong et al.
(24) examined whether chronicity of wound
impacted on the observed effects of NPWT.
They examined a subgroup of patients with
a chronic wound (definition ¼ wound duration
.30 days). The number of patients in these
subgroups were 14 and 26 in the NPWT and the
control groups, respectively. While no differ-
ences in wound healing were reported, a signi-
ficant difference in wound-healing time was
observed in the NPWT-treated chronic sub-
group compared with the control. The authors
claim that this provides evidence for the
treatment of chronic wounds with NPWT.
However, as with most post hoc analyses,
caution needs to exercised, particularly in
relation to the definition of chronicity (30 days)
and the small number of patient observations.
Finally, there is the issue of measurement to
consider. There were inconsistencies in what
was measured between the studies making
common inferences problematic. While these
differences were partly related to the focus and
design of the studies, in the case of the wound-
healing studies, these differences were very
unhelpful. An additional factor acknowledged
in the Armstrong and Lavery study (12) is that
NPWT ‘has a rapid effect on wound appear-
ance’ (p. 1709). Thus, if healing outcomes are
not absolute in nature, then there is the uncer-
tainty of knowing whether the changes in
wound parameters (granulation and size) are
the product of the NPWT’s true healing effect
or a transformation of the shape of the wound
without any real progress in healing. Such an
effect could undermine both direct observation
and imaging techniques. It is noteworthy that
most of the relative improvements reported in
Table 3 were measures of wound size. The
adoption of a standard set of measures would
clearly aid future understanding of the benefits
of NPWT. While absolute outcomes are clearly
the most uncontroversial (number of patients
healed and/or healed by a specified time)
given that NPWT is used to treat complex
wounds (involving multiple systemic thera-
pies), finding a universally appropriate range
of measures is challenging. In such circum-
stances, the most useful outcome in wound-
healing terms would be the number of patients
achieving a pre-defined standard of wound
healing within a stipulated time frame. How-
ever, this would have to be blind assessment to
wound-healing stages rather than wound size
ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
(if size were used, it should be to a clinically
significant pre-determined level rather than
Key Points
mean or percentage differences in area). In • while no differences in wound
fact, the use of mean percentage changes as healing were reported, a signif-
presented in Eginton et al. (15) study is icant difference in wound-heal-
ing time was observed in the
questionable; if these were weighted averages,
NPWT-treated chronic sub-
this would inflate the values reported giving group compared with the con-
a positive bias to the results (Simpson’s trol group
paradox). • as with most post hoc analyses,
Unfortunately, finding generalisable solu- caution needs to exercised,
particularly in relation to the
tions to complex problems in individual wound
definition of chronicity (30
management will always be illusive. Another days) and the small number of
set of measures to consider are longer term patient observations
events and endpoints: number of further ampu- • there is the issue of measure-
tations, mortality and any significant clinically ment to consider; the adoption
of a standard set of measures
related morbidity requiring additional treat-
would clearly aid future under-
ment (e.g. use of antibiotics) or hospital admis- standing of the benefits of
sion. An alternative approach would be to NPWT
follow the methods used in other complex • none of the studies identified
wounds such as fungating breast lesions where any serious treatment-related
complications with the use of
measurement has focussed on the singular
NPWT
(progress/success in the context of the case) • the claim that NPWT decreased
rather than the general (25). bacteria colonization was not
considered by any of the
Adverse events studies, although no differen-
None of the studies identified any serious ces in infection levels were
treatment-related complications with the use reported between the treat-
ment and the control groups
of NPWT. However, some of the studies re-
• none of the studies measured
ported that bleeding and pain were associated
either patients’ satisfaction or
with NPWT (13,14). Therefore, caution should quality of life and the general
be applied when using NPWT in patients with importance of the patient expe-
anticoagulation problems. Given that the pain rience; future studies must
provide some account of these
was usually experienced on the application and
patient centered outcomes
removal of the NPWT dressing, prophylactic
analgesia prior to dressing changes should be
a consideration (14). The claim that NPWT
decreased bacteria colonization (26) was not
considered by any of the studies, although no
differences in infection levels were reported
between the treatment and the control groups.
One of the excluded studies, a large retrospec-
tive study, also found no significant differences
in wound infection between NPWT and con-
ventional wound dressings (16). A potentially
important finding in the Armstrong and Lavery
(12) study was that NPWT may reduce the need
for further amputations following primary
amputation.
Patient satisfaction
It was disappointing that none of the studies
measured either patients’ satisfaction or quality
of life. Given the previous study suggesting that
NPWT may impact negatively on quality of life
(7) and the general importance of the patient
239

Key Points
• the findings from the two
studies indicating faster healing
times in wounds treated with
NPWT suggest economic bene-
fits such as shorter hospital-
isation and savings on wound
treatments
• future studies must include
a full economic analysis as this
is clearly a vital consideration
in determining the use of this
therapy
• there is a need for much more
larger rigorous studies to be
undertaken and to further in-
vestigate the effectiveness of
NPWT in treating different
types of diabetes-related foot
ulcers, both as a monotherapy
and in conjunction with other
treatments
240
Effectiveness of NPWT in the management of diabetes foot ulcers
experience, future studies must provide some
account of these patient-centred outcomes.
Cost-effectiveness
Cost-effectiveness was another issue that was
not addressed by the studies. However, the
findings from the two studies indicating faster
healing times in wounds treated with NPWT
suggest economic benefits such as shorter
hospitalisation and savings on wound treat-
ments. Further savings are suggested by the
reduction in reamputation rates reported by
Armstrong and Lavery (12). Future studies
must include a full economic analysis as this is
clearly a vital consideration in determining the
use of this therapy.
Limitations of the review
The most significant limitation to the review
was clearly the paucity of the primary studies,
with heterogeneity in population, design, appli-
cation and outcome measurement being the
most important problems. Another potential
limitation of the review was that only English
language papers were included; although given
that this technology is most widely used in
Europe and North America, this is unlikely to
have significantly biased the review. A further
Table 5 The Lancet commentary on the Armstrong and Lavery study
Key issues raised
Greater than 50% difference in wound duration
in the control group at baseline
The possibility that the higher level of infection
(17% compared with 6%) in the intervention group
may be related to the NPWT’s sealed dressing system
There were no data on antibiotic use and
microbiology reports
A lack of detail on the concealment of
allocation procedure
Potential bias in deciding on surgical closure, as this
was not blind and it was only in this group that
significant differences were observed
Procedures for NPWT pressure adjustment were not explicit
The effect of the NPWT may be related to its immobilising
and pressure-relieving effect rather than any direct
effect on wound healing
Some differences in the numbers of patients reported
in discrete analyses
A lack of clarity in the wound categorisation system
NPWT, negative pressure wound therapy.
ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
potential bias in the primary studies was
a conflict of interest. It was noted that two
of the included studies had an affiliation with
KCI, the company that manufactures NPWT
equipment (12,13).
Additional limitations of the Armstrong and
Lavery study were identified in correspondence
in the Lancet about the study. The concerns
expressed in this correspondence (27–29) are
summarised in Table 5, together with the
authors’ response (30).
CONCLUSIONS
NPWT is becoming an increasingly popular
therapy in the management of diabetes-related
foot ulcers and wounds. The studies included in
this review generally report positive findings for
NPWT, without any major adverse events. The
findings from the strongest study included in
the review suggest that the NPWT may be
particularly beneficial in relation to wound
healing following partial foot amputation. How-
ever, most of the studies were small, heteroge-
neous and methodologically weak. There is
a need for much more larger rigorous studies
to be undertaken and to further investigate the
effectiveness of NPWT in treating different types
Armstron and Lavery‘s responses
There were no significant differences in wound size,
depth and grade of wound at baseline
There were no differences in infection rates between the
treatment and the control groups
No response
Allocation envelopes were opaque, sealed, sequentially number
and distributed in permuted blocks
The decisions to perform surgical closure were based on
clinical judgement
The decisions to adjust therapy were based on clinical
judgement in accord with the clinical parameters
recommended by the manufacturer
Offloading was standardised with identical pressure-relieving
devices; although they did not record data on activity levels
(activity is a confounder of wound healing)
Discrepancies in patient numbers were explained in relation to the
identification of ‘at-risk’ patients for the subgroup analysis
The wounds were blind assessed using digital photographs and
planimetric wound measurement
 Effectiveness of NPWT in the management of diabetes foot ulcers
of diabetes-related foot ulcers, both as a mono-
therapy and in conjunction with other treat-
ments. Such studies would do well to follow the
Medical Research Council’s Complex Evalua-
tion Framework to address the inherent com-
plexities of this therapy (31). Given the high cost
of the therapy, an economic assessment of the
therapy should be a core component of any
future trials. Finally, there is very little data
detailing the patients’ experience of NPWT,
given the intensive nature of the therapy this is
another area that needs consideration.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the
support of Mel Doxford (Diabetes Foot Practi-
tioner) and Dr Mike Edmonds (Consultant
Diabetologist) who provided clinical expertise
in conducting and writing the review.
REFERENCES
1 Boulton AJ, Vileikyte L, Ragnarson-Tennvall G,
Apelqvist J. The global burden of diabetic foot
disease. Lancet 2005;366:1719–24.
2 Singh S. Armstrong D, Lipsky, B. Preventing foot
ulcers in patients with diabetes. JAMA 2005;293:
217–28.
3 Geary N. The orthopaedic perspective. Prac Diab Int
2002;9 Suppl 5:S2–3.
4 Wolvos T. Wound instillation: the next step in
negative pressure wound therapy. Ostomy Wound
Manage 2004;50:56–61.
5 Sibbald RG, Mahoney J. Consensus report on the
issue of vacuum assisted closure in the chronic
difficult to heal wounds. Ostomy Wound Manage
2003;49:52–66.
6 Miller SM, Lowery CA. Negative pressure wound
therapy: ‘A rose by any other name’. Ostomy
Wound Manage 2005;51:44–9.
7 Mendonca DA, Drew PJ, Harding KG, Price PE. A
pilot study on the effect of topical negative
pressure on quality of life. J Wound Care 2007;
16:49–53.
8 Chan Y, Nichol I, Evans G, Stansby G. Managing the
diabetic foot with the use of vacuum-assisted
closure: a call for more studies. Int J Clin Pract
2006;60:256–7.
9 Jadad A, Moore R, Carroll D, Jenkinson C, Reynolds
J, Gavaghan D, McQuay M. Assessing the quality
of reports of randomized clinical trials: is blinding
necessary? Control Clin Trials 1996;17:1–12.
10 Schulz K, Chalmers I, Hayes R, Altman D. Empirical
evidence for bias: dimensions of methodological
quality associated with estimates of treatment
effects in controlled trials. JAMA 1995;273:408–12.
11 Reid J, Morison M. Towards a consensus: classifica-
tion of pressure sores. J Wound Care 1994;3:
157–60.
12 Armstrong D, Lavery L. Negative pressure wound
therapy after partial diabetic foot amputation:
ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc
a multicentre, randomised controlled trial. Lancet
2005;366:1704–10.
13 McCallon KS, Knight CA, Valiulus JP, Cunningham
MW, McCulloch, Farinas LP. Vacuum assisted
closure versus saline moistened gauze in the
healing of post operative foot wounds. Ostomy
Wound Manage 2000;46:28–34.
14 Etöz A, Ozgenzel Y, Ozcan M. The use of negative
pressure wound therapy on diabetic foot ulcers:
a preliminary controlled trial. Wounds 2004;16:
264–9.
15 Eginton MT, Brown KR, Seabrook GR, Towne JB,
Cambria RA. A prospective randomised evalua-
tion of negative pressure wound dressings for dia-
betic foot wounds. Ann Vasc Surg 2003;17:645–9.
16 Armstrong D, Lavery L, Abu-Rumman P, Espensen
D, Varques J, Nixon B, Boulton AJ. Out come of
atmospheric pressure dressing therapy on wounds
of the diabetic foot. Ostomy Wound Manage 2002;
48:64–8.
17 Page J, Newswander B, Schwenke D, Hansen M,
Ferguson J. Retrospective analysis of negative
pressure wound therapy in open foot wounds
with significant soft tissue defect. Adv Skin
Wound Care 2004;17:355–64.
18 Carson S, Overall K, Lee-Jahshan S, Travis E.
Vacuum assisted closure used for healing chronic
wounds and skin grafts in the lower extremities.
Ostomy Wound Manage 2004;50:52–8.
19 Lavery L, Boulton A, Niezgoda J, Sheehan P. A
comparison of diabetic foot ulcer outcomes using
negative pressure wound therapy versus historical
standard care. Int Wound J 2007;4:103–13.
20 Loree S, Dompmartin A, Penven K, Marel D, Leroy
D. Is vacuum assisted closure a valid technique for
debriding chronic leg ulcers. J Wound 2004;13:
250–5.
21 Ballard K, McGregor F. Use of vacuum assisted
closure therapy following foot amputation. Br J
Nurs 2001;10 Suppl:6–12.
22 Braakenburg A, Obdeijn MG, Feitz R, Van Griethuysen
AJ, Klinkenbijl J. The clinical efficacy and cost
effectiveness of vacuum assisted closure technique
in the management of acute and chronic wounds:
a randomised controlled trial. Plast Reconstr Surg
2006;118:390–97.
23 Evans D, Land L. Topical negative pressure for treat-
ing chronic wounds (Cochrane Review). Cochrane
Database of Systematic Review, Issue 1, Art. No.
0001898.
24 Armstrong D, Lavery L, Boulton A. Negative pres-
sure wound therapy via vacuum-assisted closure
following partial foot amputation: what is the role
of chronicity? Int Wound J 2007;4:79–86.
25 Grocott P, Cowley S, Richardson A. Solving meth-
odological challenges using a theory-driven eval-
uation in the study of complex patient care.
Evaluation 2002;8:306–21.
26 Argenta LC, Morykwas MJ. Vacuum assisted closure:
a new method for wound control and treatment:
clinical experience. Ann Plast Surg 1997;38:563–76.
27 Williams D. Negative pressure therapy in diabetic
foot wounds (correspondence Armstrong and
Lavery, 2005). Lancet 2006;367:725.
241
 Effectiveness of NPWT in the management of diabetes foot ulcers

28 Maegele M, Gregor S, Peinemann F, Sauerland S. 30 Armstrong D, Lavery L. Negative pressure therapy in

Negative pressure therapy in diabetic foot wounds
(correspondence Armstrong and Lavery, 2005).
Lancet 2006;367:725–6.
29 Chantelau, E. Negative pressure therapy in diabetic
foot wounds (correspondence Armstrong and
Lavery, 2005). Lancet 2006;367:726.
diabetic foot wounds (correspondence Armstrong
and Lavery, 2005). Lancet 2006;367:726–27.
31 Campbell N, Murray E, Darbyshire J, Emery J, Farme A,
Griffiths B, Guthrie F, Lester H, Wilson P, Kinmonth
A. Designing and evaluating complex interventions
to improve health care. BMJ 2007;334:455–9.

242 ª 2008 The Authors. Journal Compilation ª 2008 Blackwell Publishing Ltd and Medicalhelplines.com Inc