Incisional negative pressure wound therapy dressings (iNPWTd) in routine primary hip and knee arthroplasties

Article focus

• Wound complications in total hip and knee arthroplasty can lead to deep prosthetic infection, management of which can be difficult, expensive and with potential for disastrous outcomes.

• Several factors such as wound healing, prolonged wound discharge, early discharge from hospital and management superficial wound complications influence the outcome.

• Incisional negative pressure wound dressings may be a reasonable way of achieving improved predictability and safety in wound healing following hip and knee arthroplasty surgery.

Key messages

• Incisional negative pressure wound therapy dressing appears to predictably improve wound discharge, wound healing, minimise wound complications and improve length of stay following routine hip and knee arthroplasties.

• These dressings appear to be particularly beneficial in patients with an American Society of Anesthesiologists grade ⩾ 3 and a body mass index ⩾ 35 in reducing wound complications. The additional cost of these dressings is perhaps justifiable taking in to account all of these factors.

Strengths and limitations

• This is a very first randomised controlled trial studying the effects of an innovative wound dressing using the beneficial aspects of negative pressure wound therapy in primary hip and knee arthoplasty, without compromising early post-operative mobilisation.

• The limitations of this study being; this is a non-blinded study, the routine practice of changing wound dressings at the second day post-surgery and working towards expected date of discharge (four days post-surgery at the time of study period) was difficult to control, and the incidence of surgical wound complications were recorded rather than directly inspected.

Introduction

Total hip and knee arthroplasties (THA, TKA) are two of the most successful and predictable orthopaedic procedures. Although life changing,¹ they can be associated with serious complications, sometimes with devastating consequences. Prolonged incisional wound drainage has been correlated with superficial surgical site and deep prosthetic infection.^2-5 Incidence of post-operative wound complications is influenced by diabetes, obesity, quality of soft tissues, use of steroids, soft-tissue handling, and duration and complexity of surgery.^6-8 Evidence is sparse regarding nutrition and wound healing.⁹ United Kingdom surgical site infection (SSI) surveillance data indicate that higher body mass index (BMI), longer duration of surgery and higher American Society of Anesthesiologists (ASA) grading increase the incidence of SSI in hip arthroplasties. For knee arthroplasties, on the other hand, it has been shown that female gender, as opposed to BMI, increases the incidence of SSI.^10-12

With the changing economics in health care, length of stay (LOS) is increasingly seen as a preventable expense. Adoption of enhanced recovery protocols for hip and knee arthroplasties have reduced the LOS.^9,13-16 However, early discharge from the hospital has the potential for wound complications being managed in the community and an increase in readmissions to hospital. This was recognised by the SSI reports and National Institute for Health and Care Excellence (NICE) publication on SSI arthroplasties.^10,17 The economic and patient implications of superficial SSI leading to deep infection are well recognised.^18-21 Therefore it is reasonable to argue that ‘predictable wound healing’ is a key factor in achieving overall ‘predictable outcome’.

The United Kingdom NICE, in a publication on surgical site infection,¹⁷ highlighted the gap in published evidence on the use of alternate wound dressings for incisional surgical wounds^22-25 and suggested that chronic wound dressings could be used for surgical incisions to minimise the risk of SSI.

Negative pressure wound therapy (NPWT) is one such potential option. NPWT for chronic wounds was introduced in the early 90s;^26,27 its effectiveness in chronic wounds, pressure sores and diabetic foot ulcers is well established.^28-30 Over the last ten years the role of NPWT has been extended to acute wounds.^31,32 Its application as a prophylactic measure for incisional wounds, although suggested as long ago as 2002 by Webb,³³ remains a novel concept. In a recent review publication,³⁴ the mechanisms of action of incisional NPWT have been attributed to improved blood supply,^35-39 reduced lateral tension along wound edges,^40-42 decrease in oedema,^26,33,43-46 reduced subcutaneous haematoma and seroma,^41,47,48 improved lymphatic drainage⁴⁹ and prevention of wound contamination.⁴³

The routine use of incisional NPWT until recently was largely restricted by the size of the apparatus and the expertise required in application of the dressing. However, with the introduction of smaller battery-operated, ultra-portable, single-use and disposable NPWT devices and ready-made incisional wound dressings,^50,51 the concept of ‘incisional negative pressure wound therapy dressings’ (iNPWTd) has become more feasible. Currently there are two such devices marketed in the United Kingdom; one by Kinetic Concepts, Inc. (San Antonio, Texas) as Prevena, and the other by Smith & Nephew Healthcare Ltd (Hull, United Kingdom) as PICO. Both claim therapeutic effects equal to those of the larger traditional devices. PICO is a canister-less design with a sandwiched dressing component which allows negative pressure to be distributed across the skin and uses absorption and vaporisation to manage exudate from the wound (Fig. 1).⁵²

Fig. 1

The negative pressure wound therapy dressing used in the study, PICO.

With the positive experiences of using these dressings in complex revision hip and knee surgery (publication awaiting), it was hypothesised that iNPWTd could be adopted for primary THAs and TKAs as a means of controlling wound exudate levels and therefore influencing the length of stay without increasing post discharge complications. A randomised controlled trial (RCT) was designed to evaluate iNPWTd effectiveness in patients undergoing primary elective hip and knee arthroplasties at our institution.

Patients and Methods

The objective of this study was to assess the potential benefits of the portable disposable incisional negative pressure wound therapy dressing in the setting of routine elective THAs and TKAs. The primary outcome was to measure the impact of iNPWTd on wound healing and effect on LOS. The secondary outcomes were assessment of wound complications such as prolonged wound exudate and superficial wound infections, number of dressing changes and the overall cost-effectiveness of the dressing.

The Smith & Nephew PICO dressing was chosen for the following reasons: the dressing is versatile with various size options, is smaller and has a quieter pump which is better tolerated by patients as shown by our clinical experience to date.

This was a non-blinded single-centre randomised controlled parallel group study in patients undergoing routine elective THAs and TKAs at our elective orthopaedic hospital. The study was designed and reported in line with the Consolidated Standards of Reporting Trials statement.⁵³ The study was conducted between October 2012 and October 2013. The study period was from surgery to six week post-operative follow-up. Ethical approval was obtained from the National Research Ethics Service (NRES Committee South West – Exeter; REC No 12/SW/0094).

Three participating surgeons (SLK, NMG and RDB), experienced in performing primary and revision hip and knee joint arthroplasties took part in the study. All consecutive patients undergoing routine THAs and TKAs (for any indication) under these three consultant surgeons were approached by the surgeon or his registrar, and willing patients were recruited to the study, provided they met the study inclusion and exclusion criteria. Patients who had known allergies to dressings, were undergoing revision joint surgery, were unwilling to attend additional clinics, and those on warfarin were excluded. Patients on warfarin were excluded primarily because of the anticipated extended LOS to allow their prothrombin time and international normalised ratio (INR) to reach therapeutic levels before being discharged.

Patient consent and randomisation were undertaken by the research team upon admission. The randomisation was performed using sealed opaque envelopes with a block size of 20 shuffled envelopes with a 1:1 allocation. These envelopes were prepared in advance and were opened by the research personnel for the recruited patient on the day of surgery. The first patient was enrolled into the study on 5 October 2012. A total of 220 patients were recruited into the trial, with 110 in each group.

The dressings were applied in theatre immediately after the wound closure. The study group received PICO NPWT dressings and the control group received conventional dressings: either Mepore (Mölnlycke Health Care AB, Gothenburg, Sweden) or Tegaderm (3M Health Care Ltd, Loughborough, United Kingdom) as per the surgeon’s preferred practice. All patients received 1.5 g of cefuroxime antibiotic at induction and one dose post-surgery. Patients were operated on by either the consultant surgeon or by the trainee under direct supervision. Wounds were closed in a layered fashion with skin staples or subcuticular Monocryl (Ethicon Inc., Johnson-Johnson, New Brunswick, New Jersey) sutures to close the skin.

For the NPWT group the protocol was to leave the dressing in place for a week (design of PICO and pump is programmed to last seven days). The control group patients had their dressing changed to OPSITE Post-Op Visible dressing (Smith & Nephew Healthcare Ltd) on the second post-operative day as per the usual routine practice. For both groups the dressing was changed earlier or again if there was progressive exudate filling 50% or more of the dressing. All patients received enoxaparin (clexane; Sanofi-Aventis, Holmes Chapel, United Kingdom) post-surgery (four weeks for hips and two weeks for knees). In the hospital, patients were seen on a daily basis in the morning by the same team of two research personnel, in order to minimise errors and maintain consistency. The following information was recorded: level of wound exudate, any dressing change and reasons, discharge readiness from a wound perspective, LOS and reasons for delay in discharge.

Wound exudate level was graded on a five-point scale based on the fraction of the wound length stained by exudate as noted on the dressing (Table I). The length of the dressing was split into four quarters for this measurement. Where there was minimal staining along the length of the incision wound, this was assessed as either grade 1 or 2 based on lateral spread. No standardised-validated methodology exists for this assessment, but given the essentially one-dimensional nature of the wound, this methodology was deemed reasonable for the study. If the dressing was changed prior to morning inspection, the extent of exudate at the time of dressing change was captured as a part of data collection, as was the reason for changing the dressing.

From the wound perspective, the patient discharge criteria were set at dry dressing or non-progressive minimal staining on the dressing. All patients were subjected to an enhanced recovery protocol following surgery, and discharged home when medically safe and sufficient home/community care was in place.

Post-discharge, all patients were contacted by telephone at two weeks post-surgery, and in addition the study group were seen in a wound clinic (by the same wound care nurse) at one week post-surgery to ensure safety with the new dressing and removal of the iNPWTd. All patients were seen at six weeks post-surgery in the consultant-led follow-up clinic as per routine practice. Wound-related complications such as recurrence of wound exudate, surgical site infections, dressing-related issues, readmission to hospital and any further surgical interventions were documented at each encounter by the research personnel.

Results

The ITT population consisted of 108 patients in the control group (two patients in the control group had their surgery cancelled on the day) and 107 patients in the study group (two patients in the study group withdrew consent after randomisation and a third immediately after surgery) (Fig. 2).

Fig. 2

Consolidated Standards of Reporting Trials diagram for negative pressure wound therapy (NPWT) dressings.

The PP population consisted of 107 patients in the control group (one further patient received NPWT dressing by error) and 102 in the study group (one further patient withdrew two days post-operatively due to problems with the tubing on dressing, and four further patients due to protocol deviations as shown in Fig. 2).

Wound exudate

The majority of patients, as anticipated in a routine primary joint arthroplasty, had grade 0 to 2 peak level of exudate (Fig. 3, Table III). The level and number of peak exudates differed significantly between the two groups (Fisher’s exact test, p = 0.007), with more patients in lower grades and fewer patients in higher grades in the study group compared with the control group. Only 4% of the study group had grade 4 exudate as compared with 16% in the control group, a four-fold reduction. The number of patients who had grade 3 exudate was small in both groups. The difference in peak exudate level between the two groups did not depend significantly on the procedure (hip versus knee, multinomial regression, interaction term p = 0.38).

Fig. 3

Distribution of peak exudate levels between the two study groups.

Table III.

Distribution of peak exudate levels between the two study groups and between hip and knee arthroplasty (n, %)

Grade	All patients		Hip patients		Knee patients
	Study	Control	Study	Control	Study	Control
0	50 (49)	51 (48)	26 (49)	34 (54)	24 (49)	17 (39)
1	39 (38)	25 (23)	24 (45)	15 (24)	15 (31)	10 (23)
2	8 (8)	10 (9)	1 (2)	4 (6)	7 (14)	6 (14)
3	1 (1)	4 (4)	0 (0)	3 (5)	1 (2)	1 (2)
4	4 (4)	17 (16)	2 (4)	7 (11)	2 (4)	10 (23)

1. Differences significant for all patients (p = 0.007) and hip patients (p = 0.04) but not for knee patients (p = 0.09), all Fisher’s exact test.

Two confounding factors, use of drain and skin closure with staples, influenced wound exudate levels following surgery. Drains reduced peak exudate by a factor of 5 (grades 3 and 4), and staples reduced the overall exudate levels, but not grade 4. The usage of drains and skin staples between the two groups was similar (Table II).

Discussion

During 2013 and 2014, nearly 200 000 hip and knee arthroplasties were performed in the United Kingdom and the proportion of patients with higher ASA and BMI undergoing hip and knee arthroplasty has increased over the years.⁵⁷ Post-discharge wound complication rates reported by patients via PROMS (patient-reported outcome measures) are much higher (10% of hip arthroplasties and 13% of knee arthroplasties)⁵⁷ than those recognised by SSI reports.^11,12 Considering both data do not represent the whole population, the realistic SSI is probably somewhere in between.

Management of wound complications in the community can be variable. Eventual complications such as deep prosthetic infection requiring revision surgery are not only devastating to the patient, but are also a negative performance indicator for the treating surgeon. National joint registries in most countries now collect this information against the surgeon. Therefore it is not surprising that surgeons prefer to keep patients in hospital until the wound is dry to minimise post-discharge complications, which can be a limiting factor for a timely discharge from hospital. Consequently it is reasonable to argue that there is a need for a predictable way of managing surgical wounds following joint arthroplasties, and portable iNPWTd may be a way of achieving this.

In this first RCT studying the effects of iNPWTd on wound healing in primary hip and knee arthroplasties, we found improvements in all of the areas we set out to assess. Wound exudates were significantly lower in the study group (p = 0.007). This was particularly the case for grade-4 exudate, where there was a four-fold reduction for the study group. There was a reduction in LOS for the study group (0.9 days, 95% CI -0.1 to 2.5, p = 0.07). Although this reduction was influenced by the two outlier patients, both patients had in fact remained longer in hospital for wound-related complications and subsequent treatment. Even when capped at 13 days the difference in LOS was still shorter for the study group patients. Further analysis also indicates that the extreme values for LOS are significantly improved by iNPWTd (Moses test, p = 0.003), suggesting a better predictability in LOS with iNPWTd. There was a four-fold reduction in post-operative wound-related complications (OR 4.0, 95% CI 0.95 to 30), although this did not reach statistical significance (p = 0.06). Two of the patients in the control group had persistent wound exudate and required formal surgical washouts resulting in excessive LOS; this was not seen in the study group. Accepting that the complications noted in the community were recorded rather than inspected complications, these were again mainly in the control group, indicating better wound healing with iNPWTd. Our analyses also suggested that iNPWTd would be beneficial specifically for patients with an ASA ⩾ 3 and a BMI ⩾ 35 in reducing wound complications. As expected, there was a reduction in requirement for dressing changes (p = 0.002).The cost of iNPWTd is perhaps justifiable if one were to take into account the additional cost incurred due to increased wound complications, increased dressing changes and resultant increase in LOS in the control group.

There are a number of limitations to this study: a) this was a non-blinded study, given the obvious difference between the two dressings, both to research personnel as well as to the clinician; b) the routine practice of changing wound dressings at second day post-surgery and working towards expected date of discharge (four days post-surgery at the time of study period) was difficult to control, due to the number of nursing staff involved in patient care; c) the incidence of surgical wound complications was recorded rather than directly inspected as the information was obtained over telephone at the two-week telephone check; d) wound-related data were not collected following protocol failure, which meant an ITT analysis was only possible for LOS; and finally, e) we were unable to standardise the wound closure methodology, as this would have meant changing surgeons’ practice.

Skin staples have been implicated in the increased risk of wound complications.^58,59 An RCT looking at this specifically in hip and knee arthroplasties reports otherwise.⁶⁰ In this study, wound closure with skin staples was a confounding factor for wound complications, as was knee arthroplasty as opposed to hip arthroplasty, ASA ⩾ 3 and BMI ⩾ 35 (Table VII). On the other hand, interestingly, wound complications were far lower in the study group despite the higher usage of skin staples as compared with the control group (Table II). Two patients in the study group who had wound complications had a BMI of > 40 and had knee arthroplasties. The choice for skin staples for surgeons with mixed practice was influenced by the quality of the skin (poor or thin skin not amenable to subcuticular suture). Additionally, there were twice as many patients with BMI ⩾ 35 in the study group. Therefore, iNPWTd appears to reduce the risk of wound complication and enhance wound healing despite these confounding factors; this, we feel, is a result of reduction in wound oedema and stabilisation of wound edges.

As with any new device there are a few clinical and practical issues. The long tubing is prone to detachment and the pump lacks an integral clip for attachment to clothing (which resulted in one patient withdrawing from the study). This dressing is not suitable when there is excessive wound ooze and considerations must be given for a surgical washout. Blisters, as reported previously,⁶¹ are an issue with iNPWTd. These are minor, away from wound edges and minimised with careful application of the dressing.

In conclusion, based on the study findings, iNPWTd offers a safe and predictable control of wound exudate in incisional wounds following hip and knee arthroplasties. It could allow patients to be discharged early without an exudating wound or post-discharge wound complications being a limiting factor. Incisional iNPWTd may be a choice where there is a rapid turnover of patients and bed planning is crucial. Perhaps for these dressings to be truly versatile and further conducive to their use in hip and knee arthroplasty, the pump needs to be smaller and form an integral part of the dressing.