Use of home negative pressure wound therapy in peripheral artery disease and diabetic limb salvage

Abstract

Use of negative pressure wound therapy (NPWT) in peripheral artery disease (PAD) and diabetic limb salvage (DLS) improves wound healing by providing moist wound conditions, reducing exudate, controlling wound‐bed infection, and stimulating granulation. NPWT duration may take several weeks, and home‐based NPWT allows patient to recover in the community while minimising risks of prolonged hospitalisation. The aim of this study is to review the use and outcomes of home NPWT in PAD and DLS. The methodology is the retrospective review of patients who were discharged with home NPWT after in‐patient PAD revascularisation and DLS debridement or minor amputations. The results included a total of 118 patients who received home NPWT between January 2017 and December 2017. The mean age was 62.8 years with 66% male and 34% female patients. The study population comprised 25% smokers, 98% patients with diabetics, 35% with ischemic heart disease, and 21% with end‐stage renal failure (ESRF). Of which, 56% of patients required revascularisation while 31% of patients underwent foot debridement, 48% underwent toe amputations, and 20% underwent forefoot amputations. All patients received in‐patient NPWT for a week before being discharged on home NPWT for 4 weeks. Then, 62% received targeted antibiotics regime while 36% received empirical antibiotics on discharge; 60% of patients achieved wound healing on home NPWT, with 9% requiring split‐thickness skin graft; 4% required further surgical debridement, 16% required further minor amputation while 20% required major amputation. 9% required further home NPWT extension, with a mean length of 7.1 ± 4.7 weeks' extension. Overall survival of 1 year was 89%. Risk factors that predict the failure of home NPWT includes subjects with a background of ESRF and wet gangrene on presentation. Home NPWT is a useful adjunct in the management of PAD and DLS foot wounds.

1. INTRODUCTION

Diabetic foot wounds (DFW) are one of the major consequences of diabetic neuropathy and diabetic peripheral artery disease (PAD). DFW can mainly be classified into three types: infection (such as cellulitis, abscess, and osteomyelitis), ulceration, and gangrene. Patients with DFW will often require antibiotic treatment, limb revascularisation, and/or amputation. Major limb amputations are the most costly and feared consequence of DFW where 84% of non‐traumatic limb amputations are preceded by foot ulcers.1

Singapore has one of the highest rates of lower extremity amputations (LEA) in the world where public hospitals here conduct up to four amputation procedures a day.2 In 2013, the estimated number of major LEA and minor LEA secondary to diabetes is 13.3 and 13.9 per 100 000 in our population, respectively.2

The usage of negative pressure wound therapy (NPWT) after diabetic limb salvage (DLS) has shown substantial promise when compared to regular moist wound therapy. NPWT has been associated with fewer amputations3 and increased rate of wound healing,3, 4 with similar incidence of adverse events.

The main drawback of NPWT is the duration of inpatient treatment, which ranges from weeks to months. This increases the costs for patients, the government, and results in a strain on communal healthcare resources. Home NPWT mitigates this by reducing inpatient treatment duration, allowing majority of the wound care to be done outpatient.

The aim of this study is to review the use and outcomes of home NPWT in DLS and PAD.

2. METHODOLOGY

2.1. Data collection

A retrospective analysis of 118 patients who were treated with home NPWT from January 2017 to December 2017 was performed at a 1600 bed tertiary hospital in Singapore.

Data were collected on patient's demographics, co‐morbidities, pre‐/postoperative blood tests, radiological findings, length of NPWT treatment, initial wound type (abscess/ulcer/gangrene), and the type of initial surgical intervention.

2.2. Inclusion criteria

Inclusion criteria are patients who underwent DLS wound debridement or minor amputations and PAD revascularisation. NPWT was delivered in a continuous fashion via the vacuum‐assisted closure (VAC) therapy system (KCI, Texas), and all patients followed the Tan Tock Seng Hospital (TTSH) home NPWT treatment protocol.

2.3. Primary and secondary outcomes

Wound healing in the form of complete wound epithelialisation is the primary outcome of this study. Secondary outcomes include the number of subsequent major amputations, readmission rate, and death.

3. STATISTICS

All factors and outcomes investigated were evaluated using descriptive statistics. Percentages were used for categorical data and means with SDs were used for continuous data. Statistical analysis was performed using the IBM SPSS® Statistics software (Version 15, IBM).

Data were categorised based on the successful healing of the wound. Comparison between each categorical group was done using either chi‐square test or Fisher's exact test. For continuous variables, student's t‐test was used. Univariate analysis was performed to identify factors associated with home NPWT failure, where variables with a P value of <.1 were further included in the multivariate analysis using the binomial regression model to identify independent risk factors of home NPWT failure.

Kaplan‐Meier survival analysis was performed to determine the predicted healing rate of home NPWT. It was also done to compare the predicted healing rates between the different initial wound types and different types of initial surgery for sepsis control.

4. RESULTS

A total of 118 subjects who received home NPWT were included in this study (see Table 1). Of which, 78 (66%) males and 40 (34%) females underwent home NPWT with a mean age of 62.75 ± 11.77 years old.

Table 1.

Patient demographics and wound characteristics

Patient demographics	Patients with home NPWT (n = 118)
	N	%
Gender
Male	78	66
Female	40	34
Age	62.75 ± 11.77
Ethnicity
Chinese	69	59
Malay	27	23
Indian	20	17
Others	2	2
ASA
2	24	21
3 or 4	91	79
Smoker
Never	58	49
Current	11	9
Previous	19	16
Unsure	30	25
Past medical history
Type 2 diabetes mellitus	116	98
HbA1C% (mean ± SD)	7.33 ± 2.59
IHD	41	35
CVA	19	16
ESRF	25	21
Visual impairment	54	46
Previous vascular interventions
History of angioplasty	23	11
History of bypass	4	3
History of debridement	20	17
History of minor amputation	32	27
History of major amputation	3	3
Wound characteristics
Ulcer	27	23
Wet gangrene	44	37
Dry gangrene	14	12
Abscess	32	28
Admission laboratory tests
Raised WBC	75	64
Anaemia	68	32
Raised CRP	89	41
Hypoalbuminaemia	38	95
Microbiology
Bacterial growth
MSSA	21	18
No growth	19	16
Streptococcus agalactiae	12	10
Morganella morganii	10	8
Antibiotics
Targeted	73	62
Empirical	43	36
Duration (mean ± SD)	23.1 ± 14.2
Osteomyelitis on X‐ray	41	39
Index sepsis operation
Debridement	38	32
Ray amputation	57	48
Trans‐metatarsal amputation	23	19
Discharge laboratory results
Raised Tw	41	38
Anaemia	84	79
Raised CRP	48	86

Abbreviations: ASA, American Society of Anaesthesiologist classification; CRP, C‐reactive protein; CVA, cerebrovascular accident; ESRF, end‐stage renal failure; IHD, ischaemic heart disease; MSSA, methicillin‐sensitive staphylococcus aureus; NPWT, negative pressure wound therapy; Tw, total white cell count.

The study population consists of 91 (79%) patients with an American Society of Anaesthesiologists Classification (ASA) of 3 or 4. Almost all subjects—116 (98%) suffered from T2DM with a mean HbA1c% of 7.33% and 38 out of 40 (95%) subjects had hypoalbuminemia.

In terms of the initial presenting wound, the most common wound type is wet gangrene (37%), followed by abscesses (28%), ulcers (23%), and dry gangrene (12%). Of which, 41 (39%) patients have features of osteomyelitis on radiological imaging upon admission; 116 (98%) patients were treated with antibiotics where 73 (62%) were culture‐directed while the rest were given antibiotics empirically.

All subjects went through initial index operation for sepsis control. Of which, 57 (48%) patients had ray amputations, 38 (32%) patients underwent wound debridement only, and 23 (19%) patients had trans‐metatarsal amputations (TMAs).

Then, 72 (61.0%) patients achieved wound healing on home NPWT (see Table 2). Healing is described as 100% epithelialisation of the wound. Out of the 72 patients healed, 10 (8.5%) underwent secondary wound closure with split thickness skin grafting.

Table 2.

Summary of outcomes of patients on home VAC therapy

Endpoint	n	%
Ulcer outcomes
Healed	72	61.0
Not healed	46	39.0
Further treatment
Split skin graft	10/118	8.5
Debridement	10/118	8.5
Minor amputation	11/118	9.3
Major amputation	24/118	20.3
Deceased	1/118	0.9
Need for VAC extension	10	8.5
Mean ± SD (weeks)	7.10 ± 4.73
30‐Day unplanned readmission	32	14.8
1‐Year survival	105	89.0

Abbreviation: VAC, vacuum‐assisted closure.

For those who did not achieve wound closure, 10 (8.5%) patients required further wound debridement, 11 (9.3%) minor amputations, and 24 (20.3%) major amputations. Of which, 1 (0.9%) patient died from wound‐related infection and sepsis. For 10 (8.5%) patients, home NPWT regime was extended for a mean of 7.1 ± 4.7 weeks. 32 (14.8%) patients had unplanned admissions within 1 month from discharge, and 105 (89%) patients remained alive after a year.

Table 3 shows the analysis between the healed and non‐healed group. The results suggest that there is no statistical significance between the different demographic groups, types of previous treatment, basic blood results, or choice of antibiotic.

Table 3.

Statistical analysis of variables affecting outcomes of home VAC therapy

Patient variables	Healed (n = 72)	Not healed (n = 46)	Univariate analysis, P value	Multivariate analysis (odds ratio, 95% confidence interval)
Demographics
Female	24	16	0.871
Mean age	61.6	64.5	0.202
Ethnicity			0.121
Chinese	34	35
Malay	9	18
Indian	4	16
Others	1	1
Co‐morbidities
ASA 3/4	54	37	0.513
Smoker	13	17	N/A
HbA1c	9.4	9.2	0.618
IHD	19	22	0.04
CVA	7	12	0.01
ESRF	9	16	0.002	OR 4.064 (CI 1.382, 11.950) P = .011
Visual impairment	33	21	0.985
Previous treatment
Previous intervention (angioplasty or bypass)	15	9	0.867
Previous surgery			0.182
None	44	33
Local controla	27	11
Major amputation	2	1
Tissue loss			0.001	P = .007
Ulcer (comparison)	23	4
Dry gangrene	8	6
Wet gangrene	18	26		OR 9.005 (CI 2.396, 33.849) P = 0.001
Abscess	23	9
Blood markers
Change in Tw (×105)	4.3	3.6	0.507
Change in Hb (g/dL)	0.67	1.1	0.251
Change in CRP (mg/L)	101	78.3	0.451
Treatment
VAC duration (weeks)	4.3	6.2	N/A
Empirical antibiotic therapy	25	18	0.709
Antibiotic duration (days)	21.8	25.1	0.233
Length of stay (days)	13.8	16.2	0.259

Note: Only ESRF and wet gangrene significantly affected wound healing rates.

Abbreviations: ASA, American Society of Anaesthesiologist classification; CRP, C‐reactive protein; CVA, cerebrovascular accident; ESRF, end‐stage renal failure; Hb, haemoglobin; IHD, ischaemic heart disease; Tw, total white cell count; VAC, vacuum‐assisted closure.

^aLocal control = wound debridement, ray amputation, or minor amputations.

In terms of co‐morbidities, the analysis confirms that patients suffering from end‐stage renal failure (ESRF) is an independent predictive factor of poor wound healing (P = 0.011; 95% CI: 1.382, 11.950; OR: 4.064). Increased incidence of non‐healing wounds was also seen in patients with ischaemic heart disease and previous cerebrovascular accidents but were not independently significant.

The types of tissue loss also showed statistical significance (P = .007) in outcomes of wound healing. Wet gangrene was significantly shown to have reduced wound healing rates (P = .001; 95% CI 2.396, 33.849; OR: 9.005) compared with the other forms of tissue loss.

This is better illustrated in the Kaplan‐Meier estimator comparing the types of tissue loss in Figure 1. This demonstrates that for patients treated with home NPWT, wet gangrene has the lowest eventual healed percentage at 27%. This is followed by dry gangrene (50%), abscesses (58%), and finally, ulcers where up to 80% of patients are expected to achieve a fully epithelised wound. It is also of note that at each timepoint, wounds with wet gangrene has the lowest healed percentage. This indicates that undesirable outcomes were consistently higher for patients presenting with wet gangrene throughout time, while patients with ulcers have more favourable outcomes.

Figure 1.

The Kaplan‐Meier survival functions of the initial presenting wounds that patients can present with. It demonstrates that ulcers have the highest healing rate (80%), followed by abscesses (58%), dry gangrene (50%), and finally wet gangrene (27%). This result is statistically significant (P value = .011)

In Figure 2, the Kaplan‐Meier survival curve demonstrates that patients undergoing wound debridement only have higher healing rates followed by patients undergoing ray amputations and TMAs. Around 70% of patients who had wound debridement only had completely healed wounds at the end of the study, compared with only 40% of patients with ray amputation or TMAs.

Figure 2.

The Kaplan‐Meier survival function for the initial limb salvage method. It demonstrates that patients have the highest chance of healing after undergoing debridement (65%), followed by ray amputations (42%), and finally forefoot amputations (37%). However, this result is not statistically significant (P value = .243)

From the Kaplan‐Meier survival curve shown in Figure 3, 51% of patients who underwent the home VAC protocol are predicted to heal completely in the long run.

Figure 3.

The Kaplan‐Meier survival function of all the patients recruited in the study. It is predicted that the eventual healing rate in the long term will be 49%

5. DISCUSSION

Based on our literature review, this is the first dedicated home NPWT study done for diabetic patients with peripheral arterial disease undergoing limb salvage and the first to report the predictors of home NPWT failure.

In the literature of chronic wound management, NPWT has been used extensively for various types of chronic wound management besides arterial and diabetic wounds. A systemic review by Xie et al5 shows encouraging evidence for NPWT application on mixed wounds. Despite conflicting results for NPWT on pressure ulcers,5 some studies do demonstrate reduced hospitalisations and greater healing rates6, 7 Similar results are also seen in patients with chronic venous insufficiency.7 In the trauma setting, NPWT has also shown to reduce infections in patients who suffered from open fractures.8 Furthermore, it has also been demonstrated to significantly reduce the wound size and bacterial growth for various open musculoskeletal injuries while increasing its healing rate.9

The results of this study demonstrate that majority of the patients achieve wound healing with home NPWT therapy. This is comparable to the results from other multicentre randomised controlled trials that compared the efficacy of NPWT to AMWT by Blume et al in 20083 as well as by Armstrong and Lavery4 in 2005 (61% vs 43% vs 56%, respectively).

Home VAC allows patients to return to the community earlier.10 Prompt return to their normal lives, maintenance of mobility, and reduction in hospital‐acquired infections will improve the quality of life of the patients, whom are already subjected to lengthy healing process. However, the biggest advantage of portable NPWT is still the increase in cost effectiveness,11, 12 where home NPWT provides significant economic benefit of community‐based wound care to resource‐limited countries13 while reducing individual financial burdens.

Chronic wound care poses a challenge to any healthcare system, as it is economically and infrastructurally taxing. The mean cost for home NPWT for our study was SGD 3070.19 ± 1513.25 (USD 2222 ± 1095.29) with a calculated cost savings of SGD 2300.80 ± 2100.20 (1686.01 ± 1539.01). Similar amount of cost savings is seen in a study done by Payne and Edwards et al in 2014, comparing home NPWT with inpatient NPWT used in traumatic wounds.11 Others have also reported higher cost savings of home NPWT than those of regular wound dressings,12, 13 mainly due to faster healing rates and reduced complications. Length of hospital stay is also reduced by an estimate of 5.01 ± 2.47 weeks while the patient is receiving NPWT, freeing up healthcare resources for patients with more acute issues.

However, despite its benefits, there are still concerns surrounding the use of home NPWT as these patients are reviewed less by healthcare professionals, which might increase the risks of complications. Common complications such as bleeding, infection, necrosis, and pain are similar to patients receiving inpatient NPWT. Reports so far show no significant increase in complications between home and hospital NPWT,14, 15 which can be due to reliable home programmes within tertiary hospitals who offer home NPWT services.

Within our study cohort, binomial logistical regression identified wet gangrene and ESRF to be independent predictive factors of home NPWT failure. Patients with wet gangrene have infected dead, devitalised tissue, which pose a high likelihood of progressing to systemic sepsis.16 The delay of healing in wounds with high bacterial “burden” impedes healing17, 18, 19, 20, 21, 22 and explains the increased home NPWT failure rate. Patients with wet gangrene also often require multiple wound debridement and further amputations23 due to underlying osteomyelitis and extensive necrotic tissue.

In addition, this study also showed that ESRF independently predicts the failure of home NPWT therapy, even with adequate wound bed preparation. The effect of renal failure on delayed wound healing has been well studied in the literature with various reasons being proposed. in vivo study on mice demonstrated that uraemia delays wound healing by reducing re‐epithelialisation and formation of granulation tissue in wounds.24, 25 ESRF‐associated conditions such as anaemia, malnutrition, and depressed immune function also contribute to the poor wound healing. Furthermore, common co‐morbidities such as cardiovascular disease and diabetes place patients with ESRF at a higher risk of perioperative morbidity and mortality,26 which leads to failure of therapy.

In 1959, Stein and Wiersum concluded that renal dysfunction played a significant role in abdominal wound dehiscence in a retrospective analysis of 22 389.27 Studies involving lower extremity wounds provided similar results. Johnson et al (1995) demonstrated that patients with ESRF had high rates of foot salvage failure despite adequate surgical revascularisation.28 A study done in a similar population also demonstrated ESRF to be an independent predictor of below‐knee amputation (BKA) failure.29 In light of this, optimisation and compliance towards renal replacement therapy is the key. Acute kidney injury should also be treated promptly and if possible prevented before commencing NPWT treatment.

The type of initial foot wound affecting DLS is not a common predictive factor analysed in the literature. In similar Asian studies, it was not an independent predictive risk factor for the failure of TMA30 and BKA.29 Larger western data also failed to replicate this finding.31 The link with the usage of NPWT on previously treated infected wounds in the literature is still equivocal32 and could explain this study's results. Either way, greater vigilance towards the treatment of infected diabetic wounds can be done with better culture directed antibiotic treatment, adequate initial debridement/amputation, and closer monitoring of the wound for infection—before and after NPWT application—which might require a longer inpatient stay.

Overall, wounds secondary to diabetic foot and peripheral arterial disease continue to pose a significant challenge to individual physicians and a resource‐limited healthcare system. This is due to the complexity of wound care, underlying disease (diabetes, peripheral arterial disease) and accompanying co‐morbidities (such as ischaemic heart disease, cerebrovascular accidents, and ESRF).

Treatment for DFWs requires a multimodality, multidisciplinary approach, which involves debridement, treatment of any infection, revascularisation if indicated, and off‐loading33 is needed. Surgical wound debridement has been identified as one of the major components of the healing process as it enables removal of devitalised and necrotic tissue, which stimulates the healing response. NPWT works as an adjunct to facilitate the wound healing process.34 The importance is reflected in our home NPWT protocol where all patients received debridement or amputation for initial sepsis control before home NPWT initiation.

Kaplan‐Meier survival curve comparing the three types of initial sepsis control (debridement, ray amputation, and TMA) also suggest that patients who undergo initial forefoot amputation have an increased rate of home NPWT failure. However, this is not statistically significant.

Additional surgical interventions such as debridement and further amputations were still necessary during or after home NPWT therapy in a large proportion (46 out of 118) of our cases where one‐fifth eventually required major amputation. This is comparable to other Asian‐based study populations,35 but significantly higher when compared to the Armstrong and Lavery study,4 where only 2% of NPWT patients required further major amputation. This is, however, likely due to the shorter follow‐up period of 112 days. Furthermore, previous reports support the fact that primary healing rates of BKAs are notoriously unpredictable, with re‐amputations reaching up to 40%.30, 36, 37 More studies will have to be done in order to provide an explanation, which can include different surgical practices, inherent cohort differences, and study method variability. Importantly, the variability of results demonstrates the difficulty of balancing the need for a safe level of amputation by creating a viable, functional, residual limb to maximise mobility and independence in each patient.

Limitations of our study design will include its retrospective nature with its associated selection and information biases as well as the relatively short follow‐up period.

Limitations to data collection include our inability to determine the severity of tissue loss prior to any form of intervention or NPWT, which would influence wound healing. Another limitation is our significant lack of data points in ankle‐brachial pressure index/toe‐brachial pressure index to meaningfully determine how the severity of peripheral arterial disease would affect wound healing with NPWT. Secondary outcomes of wound healing such as wound size, depth, and estimated wound granulation at each visit could also have allowed for further in‐depth study of wound healing with home NPWT.

6. CONCLUSION

The use of home NPWT is effective in promoting wound healing in diabetic patients undergoing surgery for the management of DFWs. ESRF and wet gangrene are key independent predictors of home NPWT failure, and further considerations should be in place for these patients. Home NPWT is also cost effective and reduces the duration of inpatient stay.

CONFLICT OF INTEREST

All authors report no conflict of interest. This research did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors.

Lim K, Lim X, Hong Q, et al. Use of home negative pressure wound therapy in peripheral artery disease and diabetic limb salvage. Int Wound J. 2020;17:531–539. 10.1111/iwj.13307