TECHNOLOGY ADVANCES

Impact of Hyperbaric Oxygen on More Advanced

Wagner Grades 3 and 4 Diabetic Foot Ulcers:
Matching Therapy to Specific Wound Conditions

William J. Ennis,1,* Enoch T. Huang,2 and Hanna Gordon3

1
Catherine and Francis Burzik Professor Wound Healing and Tissue Repair, University of Illinois at Chicago, Chicago, Illinois.
2
Hyperbaric Medicine and Wound Care, Legacy Emanuel Medical Center, Portland, Oregon.
3
Research and Informatics, Healogics, Inc., Jacksonville, Florida.

Objective: The goal of this research was to identify a population of diabetic

foot ulcer patients who demonstrate a significant response to hyperbaric ox-
ygen therapy (HBOT) using a large sample size to provide guidance for cli-
nicians when treating these complicated patients.
Approach: The effect of HBOT on diabetic foot ulcers, Wagner grades 3 and 4,
was evaluated using a retrospective observational real-world data set. The
study reported on the overall healing rate, (74.2%) at the population level, for
>2 million wounds.
Results: When a subgroup of patients of only foot ulcers with a Wagner grade
3 or 4 were considered, the healing rate was only 56.04%. The use of HBOT, William J. Ennis, DO, MBA, MMM,
without filtering for the number of treatments received, improved the healing CPE
Submitted for publication September 11,
rate to 60.01% overall. Healing rates for this same subgroup, however, were
2018. Accepted in revised form September 17,
improved to 75.24% for patients who completed the prescribed number of 2017.
hyperbaric treatments. *Correspondence: Catherine and Francis
Burzik Professor Wound Healing and Tissue Re-
Innovation: This observational study discusses the importance of reporting at
pair, University of Illinois at Chicago, 820-840 S.
the population level, specific wound etiology level, a risk-stratified level, and to Wood St. Suite 376, Third Floor, MC 958, Chi-
then overlay the effect of treatment adherence on those outcomes to provide cago, IL 60612 (e-mail: [email protected]).
clinicians with a comprehensive understanding of when to prescribe an ad-
vanced modality such as hyperbaric oxygen.
Conclusion: The authors provide healing outcomes data from several prior
HBOT studies as well as other advanced modalities that have been used in
diabetic foot ulcer care for comparison and context.

Keywords: diabetic foot ulcer, hyperbaric oxygen therapy, advanced wound

therapy, Wagner grade 3 or 4, adjunctive wound therapy

INTRODUCTION sician adoption were performed in

There has been much debate in hospital settings ensuring compliance
the literature surrounding the overall and thereby, not surprisingly, the re-
benefits of hyperbaric oxygen therapy sults did not translate to an outpatient
(HBOT) in wound care.1,2 Many of the clinic reality. Studies also reported
initial studies that resulted in positive various primary outcome objectives
outcomes, payment policies, and phy- making comparisons difficult while

ª William J. Ennis et al., 2018; Published by Mary Ann Liebert, Inc. This Open Access article is
distributed under the terms of the Creative Commons License (http://creativecommons.org/licenses/
by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the
original work is properly cited.

ADVANCES IN WOUND CARE, VOLUME 7, NUMBER 12

Mary Ann Liebert, Inc. DOI: 10.1089/wound.2018.0855
j 397
398 ENNIS, HUANG, AND GORDON

confusing clinicians when confronted with an indi- while confusing clinicians when confronted with an
vidual case.3–5 Even when outcomes were assessed for individual case.3–5 This observational study dis-
wounds of a single etiology, there was little effort to cusses the importance of reporting at the population
risk stratify the patients either for their overall clin- level, specific wound etiology level, a risk-stratified
ical condition or for the complexity of their specific level, and to then overlay the effect of treatment ad-
wound. Although population level wound-healing herence on those outcomes to provide clinicians with
rates have been reported, stratified outcomes data are a comprehensive understanding of when to prescribe
needed based on specific wound etiology to provide an advanced modality such as hyperbaric oxygen.
insight when making individual treatment decisions.
A modified intent-to-treat (mITT) healing rate for >1
million wounds was recently reported at 74.6%.6 This MATERIALS AND METHODS
article did not, however, describe the granular heal- The initial phase of this study was to review and
ing rates for individual wound etiologies. When dia- update the retrospective data on wounds, HBOT,
betic wound-healing rates are reported, they can be and the final clinical disposition from 682 outpa-
an aggregate of diabetic wounds of the leg or diabetic tient wound care centers nationwide between
foot ulcers of various Wagner grades. In many of these January 1, 2014, and April 28, 2018. The time
studies, when HBOT is given, the total number of frame for data inclusion was determined by the
treatments completed is rarely considered, making availability of aggregate data at the time of anal-
the impact of the therapy difficult to assess. ysis. The data were obtained from a proprietary
The primary focus of this study is to report on the clinical database and collected using a specialized
mITT outcomes of HBOT on diabetic ulcers limited wound data capture system that tracks wound-
to the foot and specifically only the more complex related treatments and patient outcomes. Nurses
Wagner grades 3 and 4 using the world’s largest and physicians document visits at the point of care.
wound care database. The goal of this research was A subset of centers document using paper-based
to identify a population of diabetic foot ulcer pa- forms, which were then entered into a central
tients that demonstrate a significant response to system at the end of each work day. Other centers
HBOT using a large sample size to provide guid- document visits on a fully electronic medical record
ance for clinicians when faced with treating these basis. The data used for the study were compiled
complicated patients. In addition, this study ex- into a deidentified research database table distinct
pands the previously reported mITT wound- from the enterprise data warehouse, before the
healing outcomes to provide continued updates on beginning of the analysis. All patient identifiers
the outcomes of patients who were still in treat- were removed from the research file. Deidentified
ment, and therefore excluded, at the conclusion of data were extracted using SQL software and ana-
the last study period. The wounds in the study are lyzed using Stata 14.1. The study was exempt from
an extension of the previously reported data and, IRB review by Quorum Review IRB (QR no. 33110).
therefore, any concerns about the final outcomes of All centers in the sample were managed by a
those patients still in treatment at the end of the wound care management company and staffed by a
last study period should be answered by a similar provider panel that consisted of a combination of
healing rate and the large sample size in the contract physicians in private practice and a subset
present trial. Finally, this report adds granularity of employed providers who practiced wound care
into specific wound etiology healing outcomes at a full-time. All providers whether contracted or af-
population level. Specifically, we studied diabetic filiated undergo a formal standardized course in
foot ulcers that were Wagner grade 3 or 4. wound healing that includes the management of
diabetic foot ulcers before providing care at a
wound center. All programs are hospital based and
CLINICAL PROBLEM ADDRESSED have program directors, managers, nurses, and
There has been much debate in the literature access to hyperbaric oxygen, and needed specialty
surrounding the overall benefits of HBOT in wound consultants. All providers who order and utilize
care.1,2 Many of the initial studies that resulted in HBOT have completed at a minimum a 40-h course
positive outcomes, payment policies, and physician approved by either the Undersea Hyperbaric
adoption were performed in hospital settings en- Medical Society or the American College of Hy-
suring compliance and thereby, not surprisingly, perbaric Medicine. All diabetic patients in either
the results did not translate to an outpatient clinic HBOT or standard of care only received care based on
reality. Studies also reported various primary evidence-based clinical practice guidelines that are
outcome objectives, making comparisons difficult used at all centers. All patients, whether offered
 IMPACT OF HBO ON ADVANCED WAGNER GRADES 3 AND 4 DFU 399

HBOT or not, undergo glycemic control, debridement elsewhere, but briefly represents the percentage of
as needed, off-loading, appropriate wound care all nonactive nonconsultation wounds, with great-
dressings, assessment and revascularization if nee- er than 7 days between first and last assessment,
ded, control of bioburden and, overall management of that were healed. In addition, population level
their comorbid medical conditions. All patients re- healing results were collected and reported by
gardless of wound etiology undergo a medical sur- wound etiology. This allows clinicians to put heal-
veillance review process every 4 weeks throughout ing rates in context with other types of wound care
the course of their wound care treatment to identify cases. Cases were then compared by the variable of
patients who are not healing along an anticipated either receiving or not receiving HBOT. Physicians
trajectory. Specifically, patients who are being con- in the centers contributing data in this study are
sidered for HBOT undergo pretreatment reviews of trained to provide the best standard of care for a
the medical record to ensure that standard of care was minimum of 30 days to assess for a positive healing
met and that patients did not show meaningful signs response. Those who fail to improve are considered
of improvement before starting their treatment. potential candidates for several advanced modali-
The next phase of the analysis was to create an ties, including HBOT. There would likely be more
analytic subsample of diabetic wounds. All Wagner heterogeneity in the DWLE population as diabetic
grade 3 or 4 diabetic ulcers that were located on the patients might have ischemic wounds, venous ul-
foot or toe were assessed. The decision to only uti- cers, or traumatic wounds located somewhere on
lize wounds located on the foot and toe was to focus the lower extremity, which might be coded as
the outcomes on purely diabetic foot ulcers and not having a diabetic etiology. The software used by
the broader category of diabetic wounds of the providers in this study has fields that allow for the
lower extremity (DWLE), also an approved HBOT documentation of both the primary wound etiology
indication. As previously stated, most of the initial and concomitant medical conditions that could
HBOT trials were limited to wound locations below contribute to the patients overall healing capacity.
the ankle, making more direct comparisons from To focus on the potential impact of HBOT on dia-
this study to the existing published literature betic foot ulcers, the study restricted the location of
possible. The sample was further limited to cases in the wound to below the ankle.
which a single wound was noted to ensure the
ability to accurately identify the ulcer for which
HBOT was prescribed, including those in active RESULTS
treatment at the time of study closure. The final During the study time frame, a total of 2,651,878
population size included 25,562 diabetic foot ul- wounds were evaluated (Table 1). The population
cers. The study reports retrospective observational level mITT healing rate was 74.2%, which is con-
data on healing and amputation outcomes using a sistent with the previously reported 74.6% based
mITT framework for outcomes measurement. Ad- on 1,006,690 wounds at the time of that publica-
ditional information regarding the mITT model for tion. There was variability in the specific wound
outcomes in wound healing has been reported mITT healing rates from 55.3% to 80.6%. Not sur-

Table 1. Wound healing rates by etiology and aggregate

mITT 2014–2018 Arterial Diabetic Pressure Venous All Wound Types

Total no. of healed wounds 34,745 328,158 190,832 296,219 1,408,871

Total no. of wounds 89,469 605,102 447,064 475,203 2,651,878
% Healed at population level 38.83 54.23 42.69 62.34 53.13
Exclude—no. of active treatments at study conclusion 4,516 8,544 32,406 8,331 87,098
% of total 5.05 1.41 7.25 1.75 3.28
No. of remaining wounds 84,953 596,558 414,658 466,872 2,564,780
% Healed at level 40.90 55.01 46.02 63.45 54.93
Exclude—no. of without wound documented 24 320 349 402 6,227
% of total 0.03 0.05 0.08 0.08 0.23
No. of remaining wounds 84,929 596,238 414,309 466,470 2,558,553
% Healed at level 40.91 55.04 46.06 63.50 55.07
Exclude—no. of consult and with days first to last assessment £7 days 22,049 133,350 116,073 99,078 658,735
% of total 24.64 22.04 25.96 20.85 24.84
Final—no. of remaining wounds 62,880 462,888 298,236 367,392 1,899,818
% Healed at level mITT 55.26 70.89 63.99 80.63 74.16
% Amputation at level mITT 2.99 2.42 0.5 0.11 0.94

mITT, modified intent-to-treat.

400 ENNIS, HUANG, AND GORDON

Table 2. Modified intent-to-treat healing rate and amputation Table 3. Modified intent-to-treat healing rate and amputation rate
rate: diabetic single wound Wagner grade 3/4 on foot or toe
HBO No HBO
All Single Wound 3/4
Total no. of wounds 6,616 18,946
Total no. of wounds 25,562 % Healed at population level 53.30 40.29
% Healed at population level 43.65 Exclude—no. of active at study conclusion 490 1,387
Exclude—no. of active at study conclusion 1,877 % of total 7.41 7.32
% of total 7.34 No. of remaining wounds 6,126 17,559
No. of remaining wounds 23,685 % Healed at level 56.58 42.83
% Healed at level 46.39 Exclude—no. of without wound documented 0 0
Exclude—no. of without wound documented 0 % of total 0 0
% of total 0 No. of remaining wounds 6,126 17,559
No. of remaining wounds 23,685 % Healed at level 56.58 42.83
% Healed at level 46.39 Exclude—no. of consult and with days first to 382 4,242
Exclude—no. of consult and with days first 4,624 last assessment £7 days
to last assessment £7 days % of total 5.77 22.39
% of total 18.10 Final—no. of remaining wounds 5,742 13,315
Final—no. of remaining wounds 19,057 % Healed at level mITT 60.01 54.33
% Healed at level mITT 56.04 % Amputated at level mITT 4.16 4.06
% Amputated at level mITT 4.09 9.47% delta

Hyperbaric oxygen vs. nonhyperbaric oxygen diabetic Wagner grade 3/4

single wound located on the foot or toe.
prisingly, arterial wounds demonstrated the lowest HBO, hyperbaric oxygen.
healing rates and venous leg ulcers healed at the
highest level. The overall healing rate for all ing rate (60.01%) than patients who did not receive
wounds classified as diabetic was 70.9% (328,158/ HBOT (54.33%), which results in a 9.47% delta.
462,888) at the population level. At this level of The mITT amputation rates were consistent be-
stratification, there does not seem to be a major tween the two groups with a 4.16% amputation
difference in the overall healing rate for the ag- rate in the HBOT sample and a 4.06% amputation
gregated overall population of wounds and the rate rate in the non-HBOT group. This improvement in
of healing specifically diabetic wounds using the healing rate, however, does not take into account
mITT method previously described. Only patients the actual amount of HBOT received or whether
with a single wound of Wagner grades 3 and 4 lo- the patients completed their entire overall clinical
cated on the foot or toe were included for additional course of care. Table 4 analyzes the patients’
study. The healing and amputation rates for the treatment based on whether the patients com-
full sample of Wagner grades 3 and 4 diabetic foot pleted their entire clinical treatment protocol di-
ulcers are reported in Table 2. Once the mITT ex- vided into those who did or did not receive HBOT.
clusions are applied, the sample is reduced to More patients who received HBOT went on to
19,057 ulcers with a 56.04% healing rate and a complete their entire wound care treatment pro-
4.09% amputation rate. This rate is comparable tocol. The wound treatment protocol refers to the
with the mITT population level healing rate for entire course of therapy that patients receive dur-
arterial ulcers. By comparison, the mITT healing ing their care at the wound center. For example, a
rate for all wound etiologies previously published patient might receive HBOT and still undergo
by Ennis et al. was 74.6%.6 The lower healing rate several more weeks of advanced wound care before
for Wagner grades 3 and 4 is an indication of the reaching a final discharge disposition. Patients
difficulty in healing these patients who often have who commit to such an intensive therapy such as
confounding medical comorbid conditions and em- HBOT, requiring every day treatments for up to 8
phasizes the importance of risk stratification when weeks, are likely to also be more committed to
reporting outcomes. The mITT subpopulation rep- completing the entire course of therapy. This cor-
resents 75% of the total population of diabetic foot relation could also, however, prove to be a con-
ulcer patients with the largest group excluded be- founder and represent a surrogate marker for
ing those still in treatment at the end of the study
time frame, which accounted for 18.1% of the total. Table 4. Admission marked as ‘‘completed treatment’’ by
Table 3 reports further granular outcomes for hyperbaric oxygen therapy status
patients who received at least one HBOT treat-
HBO No HBO Total (%) N
ment (6,616) compared with patients who did not
receive any HBOT (18,946). After the same mITT Admission—completed treatment 64.56 56.8 59.14 11,270
Admission—treatment incomplete 35.44 43.2 40.86 7,787
exclusions are applied, the patients who received Total 100 100 100 19,057
HBOT demonstrated a slightly higher mITT heal-
 IMPACT OF HBO ON ADVANCED WAGNER GRADES 3 AND 4 DFU 401

Table 5. Modified intent-to-treat by hyperbaric oxygen therapy Table 7. Summary statistics for hyperbaric oxygen therapy
course completion hyperbaric oxygen therapy group only sample by reason for incomplete treatment course,
hyperbaric oxygen therapy incomplete only
Healed (%) Not Healed (%) Total (%) N
Treatments Treatments Percentage Days First
Complete HBOT treatment course 75.24 24.76 45.23 2,597 Ordered Complete Complete to Last HBOT
Incomplete treatment course 47.44 52.56 54.77 3,145
Total 60.01 39.99 100 5,742 Death
Mean 36.47 14.72 0.40 25.53
HBOT, hyperbaric oxygen therapy. SD 12.90 14.26 0.32 28.60
Median 30.00 8.50 0.27 19.00
healthier patients or those with more adherence to Early resolution
their treatment plans. This is a descriptive retro- Mean 37.28 21.77 0.57 38.90
spective study using big data and although corre- SD 10.19 12.73 0.26 24.72
Median 40.00 20.00 0.57 36.00
lation may not equal causality, we further analyzed
Financial
the demographics and wound characteristics for
Mean 36.44 20.79 0.55 40.34
the two groups (HBO and non-HBO) to further SD 10.55 13.35 0.29 29.43
identify the potential impact for the therapeutic Median 39.00 20.00 0.60 38.00
intervention. Of the 5,742 patients who received Medical complication
HBOT, only 2,597 completed their hyperbaric Mean 36.24 16.10 0.43 29.40
SD 10.24 13.57 0.29 28.04
treatment (45.2%); however, of those patients who
Median 30.00 13.00 0.40 23.00
did receive a full HBOT course, 75.24% versus
Patient choice
47.44% were healed for a delta of 36.9% (Table 5). Mean 36.01 14.74 0.40 29.77
These data provide insight into why there are dif- SD 10.06 13.79 0.31 31.00
ferences in the literature describing efficacy out- Median 30.00 11.00 0.30 22.00
comes and how the therapy potentially might lose Wound progress plateaued
effectiveness in the outpatient clinic setting when Mean 40.95 36.99 0.88 62.87
SD 13.11 15.50 0.21 29.96
patients fail to adhere to the full treatment regi- Median 40.00 38.00 1.00 58.00
men. Patients who complete their HBOT received Total
89% of the prescribed treatments, whereas those Mean 37.65 22.51 0.57 40.66
marked as incomplete only received 57% of their SD 11.32 17.01 0.34 32.70
prescribed treatments (Table 6). The reasons for Median 40 20 1 36
N 3145 3145 3145 3145
incomplete HBOT are further detailed in Table 7.
These options are preloaded drop-down choices SD, standard deviation.
built into the software. The most common reason
for an incomplete treatment course was indicated HBOT treatment (Table 7). Overall, patients who
as ‘‘Patient Choice’’ followed by ‘‘Wound Progress were marked as having chosen to discontinue
Plateaued.’’ Patients who did not complete the or- treatment by their own request completed the
dered treatment course on average only completed lowest percentage of treatments (40%; SD 31),
57% (standard deviation [SD] 34) of the mean 38 whereas patients marked as ‘‘Wound Progress
treatments (SD 12) ordered and discontinued Plateaued’’ completed the highest percentage of
treatment an average of 40 days after the first ordered treatments with an average of 88% (SD 21)
of the mean 40 treatments ordered (SD 13). A hy-
Table 6. Summary statistics for hyperbaric oxygen therapy pothesis to explain this phenomenon is that a pro-
sample by hyperbaric oxygen therapy course completion vider might continue to treat a patient with the
goal of establishing a positive healing trajectory
Percentage Days First
Treatments Treatments Complete to Last HBOT and would, in that case, want to ensure a full course
of HBOT was administered before deeming the
Complete HBOT treatment course
Mean 40.14 36.44 0.89 61.91
treatment ineffective. Further information from
Std. 13.35 15.63 0.22 30.35 patients who choose to quit is needed for future
Median 40 35 1 57 studies. Another group that did not complete the
Incomplete treatment course course of HBOT are those in which the wound
Mean 37.78 22.62 0.57 40.66
Std. 12.00 17.51 0.34 32.70
healed during the treatment course. For obvious
Median 40 20 1 36 reasons, this group had no clinical reason to com-
Total plete their course of therapy. When treatment ad-
Mean 38.85 28.87 0.71 50.09 herence is not included (data separated by any
Std. 12.68 18.05 0.33 33.39
Median 40 30 1 48
HBOT vs. no HBOT), the previously noted 60.01%
healing rate was observed. Stated another way,
402 ENNIS, HUANG, AND GORDON

Healing Rates
100
90
80
70
60
50
40
30
20
10
0
mITT pop mITT db pop wag 3/4/foot wag 3/4/foot >1 wag 3/4/foot
HBO complete HBO

Figure 1. Healing Rates. mITT pop: modified intent-to-treat population level. mITT db pop: modified intent-to-treat diabetic population. wag 3/4/foot: Wagner
Grade 3 or 4 on foot. wag 3/4/foot>1 HBO: Wagner Grade 3 or 4 on foot incomplete HBOT. wag 3/4/foot complete HBO: Wagner Grade 3 or 4 on foot completed
HBOT treatment course. HBOT, hyperbaric oxygen therapy; mITT, modified intent-to-treat.

patients demonstrated a 23% improvement delta thermore, using the mITT framework to report
when HBOT is delivered as ordered and the overall healing outcomes allows for both transparency of
treatment plan is adhered to. Of course, this as- results and the ability to compare programs, indi-
sumes that the patient was medically stabilized, vidual centers, and ultimately providers. Although
revascularization was performed if indicated and the population level healing rate provides an
clinically possible, infection was controlled, off- overall picture of the effectiveness of wound care
loading provided, and the wound received de- centers in general, we also need to analyze results
bridement when indicated before, during, and after on more granular levels. Venous ulcer healing
HBOT was delivered. It is, therefore, imperative rates, for example, are frequently reported without
that future research includes patient adherence segregation into various clinical, etiology, anat-
information to fully appreciate the therapeutic omy, and pathophysiology (CEAP) classifications,
benefit achieved with HBOT. These data can also making it difficult to project an individual patients
help support patient engagement opportunities, potential for healing.7 Arterial ulcer healing rates
not unlike those employed in pharmaceutical in- rarely describe the level and extent of peripheral
dustry to assist providers in achieving the best arterial disease when reporting healing rates. In
possible outcomes when using HBOT or any other addition, the methods of establishing revasculari-
modality in which total dosing is important. These zation are often not a variable that is considered in
tables clearly identify two variables critical to con- the final analysis. Given the fact that HBOT is
siderations of the effectiveness of HBOT in real- approved for DWLE, a highly heterogeneous
world samples: first patients need to complete their group, separating wounds by anatomic location
overall care in the wound center as HBOT is only and Wagner grade may provide different results.
adjunctive to good care, and second, when HBOT is Variations in diabetic foot ulcer healing rates have
ordered, it is critical to complete the course of ther- been reported based on hospital designation, that
apy. Figure 1 identifies mITT healing rates for is, community versus tertiary academic center,
various subgroups and provides background for further complicating how results are interpreted.8
topics covered in the discussion section that follows. In that study, the same clinical team provided care
using the same protocols at two very different
hospital settings. The first, a small community 200-
DISCUSSION bed hospital and the second, a 700-bed level one
This retrospective study suggests that HBOT trauma tertiary setting. The noted difference in
can be effective for hard-to-heal Wagner grades 3 healing rates at these two centers (73.7% vs. 59.5%)
and 4 diabetic foot ulcers and demonstrates the achieved by the same clinicians sheds light on pa-
complexities of studying the therapy using obser- tient referral patterns and risk stratification.
vational real-world data. Specifically, the results There have been several randomized studies
underscore the importance of treatment adherence that have noted improvements in healing rates of
when analyzing the effectiveness of HBOT. Fur- diabetic foot ulcers. A few highly cited articles are
 IMPACT OF HBO ON ADVANCED WAGNER GRADES 3 AND 4 DFU 403

described herein. Löndahl et al. published a ran- healing studies need to include total closure as one of
domized single-center double-blinded placebo- the outcomes to ensure that early rapid healing does
controlled trial in 2010.9 The study was conducted not actually negatively impact the ultimate outcome
in an outpatient setting using a multiplace chamber. of healing.13 This is also why studies looking at al-
Patients received either oxygen or air at 2.5 at- ternative surrogate markers for healing need to en-
mospheres of pressure. The investigators did, sure that complete healing is in fact predicted by the
however, include Wagner grade 2 ulcers (24% of earlier time frame-based surrogate.14 Kalani et al.
the cases) in this study, making comparisons with reported on 38 patients with nonreconstructable
the present study more difficult. More patients vascular disease and diabetic foot ulcers for a 3-year
healed in the intent-to-treat group ( p < 0.03); period.15 HBOT-treated patients reported a 76%
however, this effect was improved in a per-protocol healing rate compared with 48% in the control arm.
subgroup, when >35 treatments were received Patients received between 40 and 60 treatments. In
( p < 0.009). This fact supports the findings of this all of these studies, there was a positive trend for
study in that total treatments received matters to HBOT when healing outcomes are measured. The
the overall outcome. The study was designed for a problem with all of the studies, however, is the var-
1-year time frame and significance was achieved at iation in HBOT treatment frequency, total number of
9 months. Kessler et al. randomized 28 patients HBOT treatments, small sample size, various sites of
with diabetic foot ulcers who were admitted to an care that impacted adherence, and variations in
inpatient hospital unit to receive HBOT or stan- major comorbid conditions such as infection and
dard of care. The patients were given two HBOT vascular status. The published outcomes of other
treatments per day, 5 days a week for 2 weeks.10 All advanced modalities, for example, ultrasound ther-
patients had normal vascular examinations before apy, have also been challenged due to various pro-
enrollment. There was a significant reduction in tocols, dosing regimens, and the use of various wound
wound area at the end of the 2 weeks in the HBOT etiologies without risk adjustment being used in a
patients, but upon discharge the significant im- single trial.16
provement was lost as both groups improved simi- Not all studies have found a positive correlation
larly. The HBOT impact in the first 2 weeks was with the use of HBOT and the healing of diabetic
significant given that both groups were receiving foot ulcers however. As with the literature pur-
intense inpatient management and the only differ- porting a positive impact of HBOT, the publica-
ential treatment was the use of HBOT. Questions tions that found no effect have limitations as well.
raised by this trial design include why was the Margolis et al. published a retrospective review of a
healing trajectory benefits of early HBOT lost once large database using propensity scoring.17 There
the HBOT was discontinued. Duzgan et al. demon- was a median of 29 treatments delivered in the
strated a 66% healing rate compared with 0% in an HBOT arm but no description of healing rates
inpatient setting treating infected diabetic foot ul- correlating with the number of treatments actually
cers.11 The protocol included two treatments per day received. This study on a cohort of 6,259 patients
followed by one treatment a day on a basis for 20– failed to demonstrate an improvement in healing
30 days. Many subsequent reviewers were troubled for nonischemic diabetic foot ulcers. An article
by the control group having no patients healed. The published in 2016 that used a double-blinded sham
inpatient setting likely had an impact on diet, off- protocol for diabetic foot ulcer treatment with
loading, medication adherence, and glucose control, HBOT found no statistical reduction in the recom-
all of which are more difficult to manage in the out- mendation for amputation.18 Surprisingly, these
patient setting. In addition, the patients were not in patients did not actually receive amputations, they
the hospital for >1 month, so it is not unreasonable were simply evaluated by a single surgeon through
that the healing rate for this site of care would not be photographs and a decision for amputation was
comparable with an outpatient trial, for example. created at that point. There was much disagreement
Abidia et al. studied nonreconstructable vascular with this study as documented by published letters
patients with a diabetic foot ulcer and found statis- to the editor requesting further clarifications.19
tically improved healing at 6 weeks and 1 year.12 The A Cochrane review also failed to support HBOT
protocol was daily HBOT at 2.4 air pressure absolute, but noted the positive trends in wound healing in
5 days a week for 30 treatments. Interestingly in this the short term but not the long term and re-
study even though the ulcer dimensions in the con- commended additional, higher quality studies to be
trol group became smaller, they did not go on to performed in the future.20
healing at the 1 -year mark. This concept has been The importance in selecting an appropriate delta
discussed by the FDA as a reason that wound- is critical when doing a power calculation to de-
404 ENNIS, HUANG, AND GORDON

termine the total number of patients needed for a tion in 1 month has repeatedly been found to be a
study. Given the paucity of consistent findings in useful surrogate for predicting those patients who
the HBOT literature, this poses a challenge to re- will likely go on to heal compared with those in
searchers conducting power calculations for stud- which advanced treatments should be consid-
ies of HBOT. As a result, studies that are often ered.24–26 The providers practicing in the centers
conducted using underpowered designs or clini- whose data comprise this study are all trained to
cal criteria for inclusion may be extended to pa- provide at least 1 month of standard of care, and to
tients who would not typically benefit from the monitor wound-healing trajectories before using
therapy to meet the needed sample size. Recent advanced therapies. All coverage and reimburse-
literature describes the risks of using random ment criteria include these standards as well. A
methods to assign delta values and the bias in- well-documented method of off-loading for diabetic
troduced by doing so. A recent trial conducted at foot ulcer patients is the use of total contact cast-
several centers in the Netherlands evaluated di- ing. Despite having strong evidence to support its
abetic foot ulcers with a study powered to an effectiveness, total contact casting is not frequently
anticipated delta of 12% improvement of limb used in many wound care centers.27,28 Once a pa-
salvage. When enrollment failed to meet targets, tient has been identified as being hard to heal, and
the delta was increased to a 25% limb salvage having failed standard of care, the clinician has
benefit and a 29% improvement in healing.21 By several options that have been studied for the
arbitrarily doubling the expected delta, the find- treatment of diabetic foot ulcers. Becaplermin, a
ings of the study are at substantial risk of bias recombinant platelet-derived growth factor prod-
and unlikely to detect a significant effect of the uct, was the first drug available for diabetic foot
therapy. Although the authors clearly articulate ulcers and underwent several prospective ran-
this limitation, its nuance may be overlooked by domized placebo-controlled trials and a meta-
clinical audiences. In addition to the underpow- analysis.29–31 In the meta-analysis, 50% of ulcers
ered study design, patients were combined and versus 36% with placebo gel healed at 20 weeks
crossed over due to patient preference, leaving (28% delta). The patients in these trials, however,
only 39 patients who completed HBOT. Despite were well perfused and were clinically assessed as
these shortcomings, the trends were all in favor Wagner grade 2, making it difficult to compare
of HBOT. It is of interest to note that in this with the findings from this study. Dermagraft, a
study, only 65% of patients undergoing HBOT human cellular-based product, produced a higher
were able to complete their course of treatment. A percentage (30%) of healed ulcers compared with
per-protocol analysis did show that the patients controls (18%) in a single-blinded RCT that en-
who completed HBOT did have statistically signif- rolled 314 patients (40% delta). Again in this trial,
icantly less amputations and higher amputation- well-perfused Wagner grade 2 wounds were eval-
free survival. This is consistent with the findings in uated.32 Apligraf, a human cellular bilayered con-
this report. struct, demonstrated a 56% versus 38% healing
Providers have other advanced treatment op- against controls at 12 weeks of therapy (32% delta).33
tions when caring for diabetic foot ulcer patients. Negative pressure wound therapy has also been
There have been few treatments, however, that studied with respect to diabetic foot ulcer healing.34
have undergone rigorous clinical trials at the ran- The studies evaluating negative pressure have
domized controlled trial (RCT) level. Guidelines looked at surgical diabetic wounds and chronic
have been proposed by several professional socie- wounds, whereas the advanced modalities have
ties for treatment of diabetic foot ulcers.22,23 Most focused primarily on more superficial well-perfused
guidelines recommend performing standard of care noninfected Wagner grade 2 ulcers. Although
for at least 1 month before considering any of the achieving significance in efficacy trials, many of
advanced modalities. The percentage area reduc- these methods have not performed as well in ef-

Table 8. Summary characteristics

HBOT No HBOT

Mean/% Standard Deviation Median Mean/% Standard Deviation Median

Wound area at first assessment 7.66 14.79 2.25 6.51 12.68 1.95
Wound duration at first assessment (days) 83.28 103.66 38 80.62 107.21 31
Patient age 61.16 12.36 61 62.59 13.37 62
Female (%) 30.44 33.31
 IMPACT OF HBO ON ADVANCED WAGNER GRADES 3 AND 4 DFU 405

fectiveness evaluations due to the heterogeneity of treatment. If we make the case that Wagner grade
the patients seen in a real-world clinical setting. 3 or 4 wounds on the foot that do not respond to
Needless to say, this study is a descriptive ob- standard of care can be healed at the overall rate
servational study and as such has limitations that for all wounds, we need to know how much we need
must be recognized. It does, however, double the to spend to achieve this clinical parity. What about
number of wounds available for healing outcomes, recidivism? Overall mortality? The results of this
and by doubling our prior work makes this the big data analysis identify a potential set of patients
largest study of its kind. The clinical procedures, for whom HBOT might provide a substantial im-
policies, and protocols at the clinical sites have now provement in wound healing. Other available ad-
consistently achieved similar repeatable healing vanced modalities have been proven to achieve
outcomes at an aggregate rolled up population le- significant improvements over controls for less se-
vel, which implies adherence to the agreed-upon vere cases and might offer a good option for those
clinical practice guidelines developed by the com- cases. Future studies should more completely ex-
pany. This is a retrospective analysis and as such plore questions related to patient adherence and
carries all the standard potential for bias that ob- possible incremental benefits of HBOT. Patients
servational studies are known to be vulnerable to. need to be engaged and adherent in order for this
However, the purpose of the study was to leverage treatment approach to work. We are now looking at
a large database and standard outcomes reporting gaining more granularity into the vascular status
framework to identify the directionality and mag- of these patients, their social determinants of
nitude of possible treatment effects. Using the health, cost of care, and their recurrence rates over
mITT method of outcomes reporting provides a time to continue to provide the most value-based
consistent comparable measure by which to com- wound care possible for this complex group of pa-
pare and analyze results and to more realistically tients. Ultimately, we need to use big data to help
calculate treatment effects that are achievable in create value-based algorithms of care along a
an outpatient setting. spectrum of various clinical complexities.
In Table 8, the overall wound-healing rate at the
population level is reported as 75.9% and this drops
to 72.3% when the data set is further modified to INNOVATION
include the condition of diabetes. When the more This retrospective study has clinical relevance
advanced Wagner grades 3 and 4 diabetic ulcers because it suggests HBOT can be effective for hard-
located on the foot are used as filters, the healing to-heal Wagner grades 3 and 4 diabetic foot ulcers
rate drops to 56.04%. The use of any HBOT brings and demonstrates the complexities of studying the
that value up to 60.01%, but when only completed therapy using observational real-world data. Spe-
HBOT cases are evaluated, the healing rate is cifically, the results underscore the importance of
75.24%. Although big data analysis identifies cor- treatment adherence when analyzing the effec-
relations, it does not imply causation, so one could tiveness of HBOT. Furthermore, using the mITT
argue that those patients who complete their framework to report healing outcomes allows for
HBOT are healthier, have less comorbid conditions both transparency of results and the ability to
that would make completing their HBOT more compare programs, individual centers, and ulti-
likely, or any number of other hypotheses. A ran- mately providers.
domized study by Faglia et al. found a significant
reduction in amputations when using HBOT, again ACKNOWLEDGMENTS
in a controlled inpatient environment where AND FUNDING SOURCES
treatment adherence was extremely high, provid-
ing further support for both the treatment and the This research received no specific grant from any
need to complete the course of therapy.35 For the funding agency in the public, commercial, or not-
purposes of this descriptive analysis, we stratified for-profit sectors.
ulcers using the Wagner scale. However, there are
currently a number of broader scoring systems that AUTHOR DISCLOSURE
warrant consideration for future studies as well as AND GHOSTWRITING
other clinical characteristics that should be mea- Hanna Gordon, PhD, and William J. Ennis, DO,
sured in subsequent analyses.36,37 are employees of Healogics, Inc. H.G. serves as the
A further consideration should be to balance the Executive Director of Research. W.J.E. is the Chief
cost of care relative to the likelihood of healing and Medical Officer of Healogics. Enoch T. Huang, MD,
the likelihood that the patient will complete MPH&TM, FUHM, FACEP, has no conflict of in-
406 ENNIS, HUANG, AND GORDON

terest to declare. The content of this article

KEY FINDINGS
was expressly written by the authors lis-
ted. No ghostwriters were used to write  HBOT can be effective for hard-to-heal Wagner grades 3 and 4 diabetic
this article. foot ulcers.
 Results underscore the importance of treatment adherence when ana-
ABOUT THE AUTHORS lyzing the effectiveness of HBOT.
William J. Ennis, DO, MBA, MMM,  Using an mITT framework to report healing outcomes allows for both
FACOS, is Catherine and Francis Burzik transparency of results and the ability to compare programs, individual
Professor of Wound Healing and Tissue centers, and ultimately providers.
Repair and Chief of the Section of Wound
Healing Tissue Repair at the University of Illinois University in New Orleans. He completed a fel-
at Chicago. Dr. Ennis is a founding board member lowship in Undersea and Hyperbaric Medicine at
and President of the American College of Wound the University of Pennsylvania. He is boarded in
Healing and Tissue Repair, a 501c3 nonprofit both Emergency Medicine and Undersea and
educational organization that has a mission of Hyperbaric Medicine. He is the past president of
bringing wound care to the level of a formal the Undersea and Hyperbaric Medical Society.
medical specialty. He also serves as Chief Med- Hanna Gordon, PhD, received her PhD from
ical Officer of Healogics, Inc., a wound care Florida State University in 2014. She currently
company based in Jacksonville, FL. Enoch T. serves as the Executive Director of Research and
Huang, MD, MPH&TM, FUHM, FACEP, re- Informatics at Healogics, Inc., a wound care
ceived his MD and MPH&TM from Tulane company based in Jacksonville, FL.

REFERENCES
1. Dauwe PB, Pulikkottil BJ, Lavery L, Stuzin JM, 8. Ennis WJ, Lee C, Vargas M, Meneses P. Wound 15. Kalani M, Jörneskog G, Naderi N, Lind F, Bris-
Rohrich RJ. Does hyperbaric oxygen therapy work outcomes from a single practice at a Sub-acute mar K. Hyperbaric oxygen (HBO) therapy in
in facilitating acute wound healing: a systematic wound care unit and 2 hospital based, outpatient treatment of diabetic foot ulcers. Long-term
review. Plast Reconstr Surg 2014;133:208e–215e. settings. Wounds 2004;16:165–172. follow-up. J Diabetes Complications 2002;16:
153–158.
2. Chuck AW, Hailey D, Jacobs P, Perry DC. Cost- 9. Löndahl M, Katzman P, Nilsson A, Hammarlund C.
effectiveness and budget impact of adjunctive Hyperbaric oxygen therapy facilitates healing of 16. Ennis WJ, Lee C, Gellada K, Corbiere TF, Koh TJ.
hyperbaric oxygen therapy for diabetic foot ulcers. chronic foot ulcers in patients with diabetes. Advanced technologies to improve wound healing:
Int J Technol Assess Health Care 2008;24:178– Diabetes Care 2010;33:998–1003. electrical stimulation, vibration therapy, and ul-
183. trasound—what is the evidence? Plast Reconstr
10. Kessler L, Bilbault P, Ortéga F, et al. Hyperbaric
Surg 2016;138(3S):pp.94S–104S.
3. Doctor N, Pandya S, Supe A. Hyperbaric oxygen oxygenation accelerates the healing rate of non-
therapy in diabetic foot. J Postgrad Med 1992;38: ischemic chronic diabetic foot ulcers: a prospec- 17. Margolis DJ, Gupta J, Hoffstad O, et al. Lack of
112–114, 111. tive randomized study. Diabetes Care 2003;26: effectiveness of hyperbaric oxygen therapy for the
2378–2382. treatment of diabetic foot ulcer and the preven-
4. Chen CY, Wu RW, Hsu MC, Hsieh CJ, Chou MC.
tion of amputation: a cohort study. Diabetes Care
Adjunctive hyperbaric oxygen therapy for healing 11. Duzgun AP, Satır HZ, Ozozan O, et al. Effect of
2013;36:1961–1966.
of chronic diabetic foot ulcers: a randomized hyperbaric oxygen therapy on healing of diabetic
controlled trial. J Wound Ostomy Continence Nurs foot ulcers. J Foot Ankle Surg 2008;47:515–519. 18. Fedorko L, Bowen JM, Jones W, et al. Hyperbaric
2017;44:536–545. oxygen therapy does not reduce indications for
12. Abidia A, Laden G, Kuhan G, et al. The role of
5. Kaur S, Pawar M, Banerjee N, Garg R. Evaluation hyperbaric oxygen therapy in ischaemic diabetic amputation in patients with diabetes with non-
of the efficacy of hyperbaric oxygen therapy in the lower extremity ulcers: a double-blind randomised- healing ulcers of the lower limb: a prospective,
management of chronic nonhealing ulcer and role controlled trial. Eur J Vasc Endovasc Surg 2003;25: double-blind, randomized controlled clinical trial.
of periwound transcutaneous oximetry as a pre- 513–518. Diabetes Care 2016;39:392–399.
dictor of wound healing response: A randomized
13. FDA Wound Healing Clinical Focus Group. Guidance 19. Huang E. Comment on Santema et al. Hyperbaric
prospective controlled trial. J Anaesthesiol Clin oxygen therapy in the treatment of ischemic
for industry: chronic cutaneous ulcer and burn
Pharmacol 2012;28:70–75. lower-extremity ulcers in patients with diabetes:
wounds—developing products for treatment. Wound
6. Ennis WJ, Hoffman RA, Gurtner GC, Kirsner RS, Repair Regen 2001;9:258–268. results of the DAMO2CLES multicenter random-
Gordon HM. Wound healing outcomes: Using big ized clinical trial. Diabetes Care 2018;41:112–119.
data and a modified intent-to-treat method as a 14. Driver VR, Gould LJ, Dotson P, et al. Identification Diabetes care 2018;41:e61.
and content validation of wound therapy clinical
metric for reporting healing rates. Wound Repair
Regen 2017;25:665–672. endpoints relevant to clinical practice and patient 20. Kranke P, Bennett MH, James MS, Schnabel A,
values for FDA approval. Part 1. Survey of the Debus SE, Weibel S. Hyperbaric oxygen therapy
7. Padberg FT, Jr. CEAP classification for chronic wound care community. Wound Repair Regen for chronic wounds. Cochrane Database Syst Rev
venous disease. Dis Mon 2005;51:176–182. 2017;25:454–465. 2015;CD004123.
 IMPACT OF HBO ON ADVANCED WAGNER GRADES 3 AND 4 DFU 407

21. Santema KT, Stoekenbroek RM, Koelemay MJ, randomized controlled trial. Diabetes Care 2005; the management of noninfected neuropathic dia-
et al. Hyperbaric oxygen therapy in the treatment 28:551–554. betic foot ulcers: a prospective randomized mul-
of ischemic lower-extremity ulcers in patients ticenter clinical trial. Diabetes Care 2001;24:290–
28. Armstrong DG, Short B, Espensen EH, Abu-
with diabetes: results of the DAMO2CLES multi- 295.34.
Rumman PL, Nixon BP, Boulton AJ. Technique for
center randomized clinical trial. Diabetes Care
fabrication of an ‘‘instant total-contact cast’’ for 34. Blume PA, Walters J, Payne W, Ayala J, Lantis J.
2018;41:112–119.
treatment of neuropathic diabetic foot ulcers. J Comparison of negative pressure wound therapy
22. Hingorani A, LaMuraglia GM, Henke P, et al. The Am Podiatr Med Assoc 2002;92:405–408. using vacuum-assisted closure with advanced
management of diabetic foot: a clinical practice moist wound therapy in the treatment of diabetic
29. Steed, D.L. Clinical evaluation of recombinant
guideline by the Society for Vascular Surgery in foot ulcers: a multicenter randomized controlled
human platelet-derived growth factor for the
collaboration with the American Podiatric Medical trial. Diabetes Care 2008;31:631–636.
treatment of lower extremity diabetic ulcers.
Association and the Society for Vascular Medi- Diabetic Ulcer Study Group. J Vasc Surg 1995;21: 35. Faglia E, Favales F, Aldeghi A, et al. Adjunctive
cine. J Vasc Surg 2016;63(2 Suppl):3S–21S. 71–78; discussion 79–81. systemic hyperbaric oxygen therapy in treatment
23. Huang ET, Mansouri J, Murad MH, et al. A clinical 30. Wieman TJ, Smiell JM, Su Y. Efficacy and safety of severe prevalently ischemic diabetic foot ulcer:
practice guideline for the use of hyperbaric oxy- of a topical gel formulation of recombinant human a randomized study. Diabetes care 1996; 19:
gen therapy in the treatment of diabetic foot ul- platelet-derived growth factor-BB (becaplermin) in 1338–1343.
cers. Undersea Hyperb Med 2015;;42:205–247. patients with chronic neuropathic diabetic ulcers. 36. Mills Sr JL, Conte MS, Armstrong DG, et al. The
24. Sheehan P, Jones P, Caselli A, Giurini JM, Veves A phase III randomized placebo-controlled double-
Society for Vascular Surgery Lower Extremity
A. Percent change in wound area of diabetic foot blind study. Diabetes Care 1998;21:822–827. Threatened Limb Classification System: risk
ulcers over a 4-week period is a robust predictor 31. Smiell JM, Wieman TJ, Steed DL, Perry BH, stratification based on wound, ischemia, and foot
of complete healing in a 12-week prospective Sampson AR, Schwab BH. Efficacy and safety of infection (WIfI). J Vasc Surg 2013;59:220–234
trial. Diabetes Care 2003;26:1879–1882. becaplermin (recombinant human platelet-derived e1–e2.
25. Lavery LA, Barnes SA, Keith MS, Seaman JW, growth factor-BB) in patients with nonhealing,
37. Lavery LA, Armstrong DG, Harkless LB. Classifi-
Armstrong DG. Prediction of healing for postoper- lower extremity diabetic ulcers: a combined
analysis of four randomized studies. Wound Re- cation of diabetic foot wounds. J Foot Ankle Surg
ative diabetic foot wounds based on early wound 1996;35:528–531.
area progression. Diabetes Care 2008;31:26–29. pair Regen 1999;7:335–346.

26. Snyder RJ, Cardinal M, Dauphinée DM, Stavosky J. 32. Marston WA, Hanft J, Norwood P, Pollak R. The
Acronyms and Abbreviations
A post-hoc analysis of reduction in diabetic foot efficacy and safety of Dermagraft in improving the
ulcer size at 4 weeks as a predictor of healing by healing of chronic diabetic foot ulcers: results of a DWLE ¼ diabetic wounds of the lower
12 weeks. Ostomy Wound Manage 2010;56:44–50. prospective randomized trial. Diabetes Care 2003; extremity
26:1701–1705. HBOT ¼ hyperbaric oxygen therapy
27. Armstrong DG, Lavery LA, Wu S, Boulton AJ. mITT ¼ modified intent-to-treat
Evaluation of removable and irremovable cast 33. Veves A, Falanga V, Armstrong DG, Sabolinski ML. RCT ¼ randomized controlled trial
walkers in the healing of diabetic foot wounds: a Graftskin, a human skin equivalent, is effective in