Economic and clinical contributions of an antimicrobial barrier dressing: a strategy for the reduction of surgical site infections

Abstract

Objective:

In patients at risk of surgical site infection (SSI), there is evidence that an antimicrobial barrier dressing (Acticoat) applied immediately post-procedure is effective in reducing the incidence of infection. The objective of this study was to assess when it is appropriate to use an antimicrobial barrier dressing rather than a post-operative film dressing, by evaluating the net cost and budget impact of the two strategies.

Methods:

An economic model was developed, which estimates expected expenditure on dressings and the expected costs of surgical site infection during the initial inpatient episode, based on published literature on the pre-discharge costs of surgical infection and the efficacy of an antimicrobial barrier dressing in preventing SSI.

Results:

At an SSI risk of 10%, an antimicrobial barrier dressing strategy is cost neutral if the incidence of infection is reduced by at least 9% compared with a post-operative film dressing. At 35% efficacy, expenditure on dressings would be higher by £30,760 per 1000 patients, and the cost of treating infection would be lower by £111,650, resulting in a net cost saving of £80,890. The break-even infection risk for cost neutrality is 2.6%.

Limitations:

Although this cost analysis is based on published data, there are limitations in methodology: the model is dependent on and subject to the limitations of the data used to populate it. Further studies would be useful to increase the robustness of the conclusions, particularly in a broader range of surgical specialities.

Conclusions:

A strategy involving the use of an antimicrobial barrier dressing in patients at moderate (5–10%) or high (>10%) risk of infection appears reasonable and cost saving in light of the available clinical evidence.

Key words:

• Antimicrobial barrier dressing
• Cost
• Infection
• Silver
• SSI

Introduction

The increasing financial burden of Healthcare Associated Infections (HCAIs) to global health services is a major challenge. HCAIs include community and hospital-acquired pneumonias, urinary tract infections, bacteraemias (with meticillin resistant Staphylococcus aureus, MRSA, being the principal concern), Clostridium difficile infection (CDI) and surgical site infections (SSIs). The annual cost of HCAIs to the National Health Service in the United Kingdom (NHS) has been estimated at £1 billion1. Patients who develop an HCAI are approximately 7 times more likely to die in hospital than uninfected patients1. Up to 5,000 patients die each year in the UK as a direct result of HCAIs, and this is a factor involved in another 15,000 deaths1. Approximately 20–25% of HCAIs are SSIs and incidence studies have shown that 5% of all operations are followed by an SSI, detected before discharge from hospital, with an associated high mortality rate in high-risk patients2,3. The annual direct costs of SSIs in the UK are likely to be around £700 million4. (According to NICE4 at least 5% of 4 million surgical procedures annually will give rise to a surgical site infection. At an average cost of £3,500 per infection4, the national cost is around £700 million) Additionally there are unrecorded, indirect costs which are likely to include loss of productivity, reduced quality of life and expensive litigation.

SSIs are classified as being superficial, deep or in an organ/space with superficial SSIs comprising the majority of the total5. However, current methodologies which measure the incidence and prevalence (and even the definition of SSIs), are not universally accepted6. The most widely-used definition currently is that from the Centers for Disease Control and Prevention (CDC) in the United States: this definition includes presence of a purulent discharge, wound separation or the need for drainage, presence of Celsian signs of inflammation, or if organisms are isolated from the wound7. The accuracy of SSI data capture is also related to the adequacy of post-discharge surveillance and the major drawback of these systems is that they are recorded during the increasingly short post-operative hospital stay8,9. Therefore the financial impact of SSIs, the majority of which present later in primary care, is underestimated.

Guidelines have been issued to reduce HCAIs4,10–12. Although antibiotics have a part to play in prophylaxis and treatment, resistance (such as MRSA) and emergence (such as Clostridium difficile infection) are largely related to antibiotic misuse, and it is clear that good antibiotic stewardship is another key to the reduction of HCAIs13. Since the incentive to produce new antibiotics by the pharmaceutical industry is likely to decline, alternative antimicrobial strategies will increasingly have to be considered.

Control of bacterial bioburden in acute wounds through the use of alternative topical antimicrobials at the time of a breach in the skin barrier, may have an important role in prevention of infection in incisional and traumatic wounds. This principle could also be used for incisional wounds left open after an infection, or traumatic wounds electively managed to heal by secondary intention. This strategy might help reduce the need for, or inappropriate use of topical and systemic antibiotics.

The use of an active antimicrobial barrier dressing (Acticoat*) for the prevention and management of infection in acute wounds offers a potential solution. Antimicrobial agents need to be locally released and replenished at sufficient concentrations to be effectively bactericidal over the time the dressing is in place. One such antimicrobial is silver, which has been used medically for centuries14. Published outcomes of specific relevance to bacterial control with silver have demonstrated a rapid speed of kill and lowering of bacterial levels in wounds, with reduced infections15. Using the advanced Silcryst system of silver delivery16, as part of a preventative strategy, Childress et al.17 found that the rate of SSIs and associated complications was significantly reduced after peripheral vascular surgery.

The resource impact of SSIs on acute hospital providers is well-documented3,18,19. SSIs lead to a higher mortality risk, prolonged length of stay, a higher risk of readmission, and higher hospital costs. In the UK, the presence of an SSI is estimated to add an average of 11 days to inpatient length of stay3. In the United States, patients with SSIs were found to have a higher mortality (relative risk (RR) 2.2); higher rate of readmission (RR 5.5), higher postoperative length of stay (mean of 12 days per patient), and higher associated costs (mean $5038 per patient at 1991–1995 prices)18.

The National Institute for Health and Clinical Excellence (NICE) clinical guideline on the prevention and treatment of surgical site infection recommends that all surgical incisions anticipated to heal by primary intention should be covered postoperatively with an adhesive film dressing, with or without a central absorbent pad4. Film dressings act as a barrier to possible external bacterial contamination and to prevent cross contamination to other patients. In patients at risk of developing an SSI, there is evidence that application of an active antimicrobial barrier dressing immediately post-procedure is effective in reducing SSIs and antibiotic use17,20. The introduction of an antimicrobial barrier dressing requires clear guidance regarding when the film dressing should be stepped-up to the active dressing.

In order to address this question, the net cost and budget impact of a prevention strategy using antimicrobial barrier dressings in procedures and/or patients who are at medium or high risk of SSI were estimated, compared with the use of a film dressing with an absorbent pad.

Methods

An economic model was constructed to estimate the costs and outcomes associated with two alternative prevention strategies, in order to determine the appropriate use of an antimicrobial barrier dressing in patients at risk of surgical site infection. The two strategies were to cover surgical wounds post-operatively with (a) an adhesive film dressing with an absorbent pad, or (b) an antimicrobial barrier dressing. The model estimates expected expenditure on dressings and the expected costs of surgical site infection to a hospital during the initial inpatient episode, based on published literature on the pre-discharge costs of surgical infection and the efficacy of an antimicrobial barrier dressing in preventing SSI. The model inputs are:

• Cost of dressings

• Incidence of pre-discharge SSI

• SSI-attributable costs to the hospital pre-discharge

• Effectiveness of the antimicrobial dressing in reducing the incidence of SSI

Cost of dressings

Dressing prices were taken from the NHS Supply Chain Catalogue (April 2009)21. The prices of dressings are based on a standard 10 × 10 cm wound contact area (or closest available size) and 12 days of treatment. Because the focus of the analysis is on budget impact, no cost is included for nurse time to apply or change dressings. A specific film dressing (OpSite Post-Op) was used as representative of an adhesive film dressing with an absorbent pad. The unit price in the UK (12 × 10 cm) is £0.80. The cost for 12 days of application, at a wear time of 6 days, would be £1.60. In the absence of a recommended wear time for the adhesive film dressing the assumption in the NICE guidance (Appendix G) of the wear time for a film dressing (5–7 days) was applied4. The unit price of the antimicrobial barrier dressing is £8.09. The cost for 12 days at a wear time of 3 days is £32.36. The wear time of the antimicrobial barrier dressing is the manufacturer’s recommendation.

Incidence of pre-discharge SSI

In England, the Nosocomial Infection National Surveillance Service (NINSS) monitors patients undergoing one of 13 surgical procedures in 247 participating hospitals. Almost 240,000 surgical procedures, carried out in England between October 1997 and September 2005, resulted in 7194 hospital-detected SSIs22. The overall incidence was 3%, with a range between 1% (knee prosthesis) and 13.1% (limb amputation). The incidence of surgical infection is stratified by the US National Nosocomial Infection Surveillance (NNIS) risk index, which ranges from 0 to 3 according to the presence of three major risk factors: likely microbial contamination of the wound; physical status of the patient; and the duration of the operation. A patient with all three of the risk factors is expected to be in a high-risk category for developing an SSI. The cumulative rate of SSI in England is shown in Table 1 by surgical procedure and patient risk index for a sample of 11 procedures reported in the surveillance period 1997–2005²². The authors have estimated the net cost and budget impact of a preventive strategy involving the use of antimicrobial barrier dressings on patients and/or procedures at medium (5–10%) or high (>10%) risk of SSI. The analysis is based on an average risk in the intervention group of 10%.

Table 1.  Rate of SSI (%) by category of surgical procedure and risk index*.

Low rate (0–5%)	Medium rate (5–10%)	High rate (>10%)
Gastric surgery (0)	Vascular surgery (1)	Vascular surgery (2/3)
Knee prosthesis (0–3)	Large bowel surgery (0/1)	Limb amputation (1–3)
Hip prosthesis (0–3)	Hip hemiarthroplasty (2/3)	Bile duct, liver, pancreatic surgery (0/1)
Abdominal hysterectomy (0/1)	Small bowel surgery (0/1)	Large bowel surgery (2/3)
Coronary artery bypass surgery (0/1)	Gastric surgery (1)	Small bowel surgery (2/3)
Open reduction of long bone fracture(0/1)	Open reduction of long bone fracture (2/3)	Gastric surgery (2/3)
Vascular surgery (0)	Coronary artery bypass surgery (2/3)
Hip hemiarthroplasty (0/1)	Abdominal hysterectomy (2/3)
	Limb amputation (0)

*Surgical categories are shown with the NNIS risk index in parentheses (0–3). Adapted from HPA (2006)22.

Cost of surgical site infection

The hospital cost of treating SSI was estimated at £3190 per patient episode, based on an average infection-attributable excess length of stay of 11 days combined with an average cost of £290 per hospital day. Coello et al. estimated infection-attributable postoperative length of stay from the details of 2832 infections following 67,410 surgical procedures carried out in England between 1997 and 2001³. The mean additional length of stay was 11 days per patient, with a range between 3.3 days (abdominal hysterectomy) and 21 days (limb amputation). An average inpatient cost of £290 per day was calculated by taking a weighted average of daily rates imputed from 2005 to 2006 National Reference Costs for non-elective inpatients in Healthcare Resource Groups (HRG) for major and minor skin infections (codes J42 and J45)23. The authors’ estimate of £3,190 per episode is similar to the figure of £3,500 used by NICE in their cost analysis of surgical site infection4.

Effectiveness of the antimicrobial barrier dressing in the reduction of SSIs

Few clinical studies have investigated the efficacy of antimicrobial barrier dressings as part of a coordinated preventative strategy. The studies which have been published are consistent with a reduction in SSIs of between 60% and 100% using the antimicrobial barrier dressing as a postoperative dressing. The average rate of SSI in cardiothoracic surgery in one report was reduced from 4.7% to 0% in the 2 years following implementation of a protocol including the use of antimicrobial barrier dressings20. Childress et al. report on a comparison of patients undergoing open, lower leg revascularisation whose surgical wounds were dressed with conventional dressings (typically sterile cotton gauze in combination with an occlusive dressing; n = 118) or with an antimicrobial barrier dressing and an occlusive dressing (n = 130)17. The rate of wound complications (dehiscence or wound infection) was reduced by 64% in the group treated with the antimicrobial barrier dressing (14% vs 5%; p = 0.016). In this analysis it was conservatively assumed that the antimicrobial barrier dressing would reduce the incidence of SSIs by 35%.

Results

The overall clinical and budget impact of an antimicrobial strategy depends on the baseline risk of SSI and on the efficacy of an antimicrobial barrier dressing in reducing the rate of infection. At a baseline risk of 10% in the intervention group, an antimicrobial barrier dressing strategy is cost neutral, so long as the incidence of infection is reduced by at least 9% (i.e., to 9.1% or less). In this case, switching from a film dressing to an active antimicrobial barrier dressing in medium- to high-risk patients reduces the incidence of infection by 9 cases per 1000 operations, and reduces infection-attributable bed-days by 99 days per 1000 (saving £28,710 at £290 per day). Spending on dressings would be higher by £30,760 but total costs remain approximately the same. At an efficacy greater than 9%, the antimicrobial barrier dressing strategy improves patient outcomes and reduces overall costs.

The potential clinical and budget impacts of antimicrobial barrier dressings in medium- to high-risk patients are illustrated (Table 2) per 1000 operations for an efficacy of 35% (switching from film dressings to antimicrobial barrier dressings reduces the incidence of infection by 35%, e.g. from 10% to 6.5%):

• Spending on dressings (budget impact) is higher by £30,760 per 1000 operations

• The number of patients with an SSI is reduced by 35, from 100 to 65

• Infection-attributable bed-days are reduced by 385 (35 episodes × 11 days per episode)

• The cost of treating infection is reduced by £111,650 (385 days × £290 per day)

• The net cost saving overall is £80,890 (£111,650 to £30,760)

• The number needed to treat (NNT) to avoid one infection is 28.6 patients at an antimicrobial barrier dressing cost of £925 to save £3190 per infection.

Table 2.  Clinical and cost impact of film dressing vs. antimicrobial barrier dressing (per 1000 operations), at an efficacy of 35%.

	Postoperative film dressing	Antimicrobial barrier dressing	Incremental cost (saving)
Cost of dressings	£1600	£32,360	£30,760
Surgical site infections	100	65	(35)
Infection-attributable bed-days	1100	715	(385)
Hospital cost of treating infection	£319,000	£207,350	(£111,650)
Total cost	£320,600	£239,710	(£80,890)
Cost per procedure	£320.6	£239.7	(£81)

Some patients may require more than one dressing per application because of the size of their surgical incision. Assuming two dressings per patient, spending on dressings (budget impact) is higher by £61,520 per 1000 operations, and total cost is lower by £50130 with the antimicrobial barrier dressing strategy. NNT is 28.6 at a dressing cost of £1850 to save £3190.

The analysis is based on an average risk of 10% in the treatment group. An alternative approach is to estimate the break-even baseline risk at which the antimicrobial barrier dressing is cost neutral. Assuming 35% efficacy, the break-even risk is 2.6%. Antimicrobial barrier dressings are cost effective compared with film dressings in any patient/procedure in which the risk is perceived to be greater than 2–3%. At a baseline risk of 2.6%, the number of infections avoided is 9.1 (35%) per 1000 operations, at a cost saving of £29,000 (9.1 × £3190) compared with the additional cost of antimicrobial barrier dressings of £30,760. On this basis, a strategy involving antimicrobial barrier dressing use in patients at moderate (5–10%) or high (>10%) risk of infection is conservative.

Discussion

There is considerable motivation to reduce SSIs in acute hospitals in the NHS. In this analysis the authors have demonstrated that even on conservative assumptions about efficacy, a topical, antimicrobial barrier dressing may reduce both the incidence and costs of SSIs. The analysis shows that in patients and/or procedures at medium–high risk of wound infection, reducing the incidence of SSI by just 9% makes switching from a film dressing with an absorbent pad to an antimicrobial barrier dressing cost neutral. At any reduction more than 9%, the antimicrobial barrier dressing strategy is cost saving. This appears plausible in light of the clinical and other evidence on the effectiveness of the antimicrobial barrier dressing17,24,25. In relatively low risk patients and/or procedures, the use of a film dressing in line with NICE guidelines is likely to be more cost effective. To the extent that the analysis is restricted to the costs of surgical infection occurring before hospital discharge, the true clinical and cost benefits may be underestimated because many infections are detected and treated in primary care. This approach may also reduce the need for antibiotic prophylaxis and treatment of infected wounds, with subsequent reductions in microbial resistance and emergence.

This analysis is based on an average infection treatment cost of £3190 per episode. In practice, even a low-risk procedure may give rise to considerably higher costs3. For example, peripheral vascular surgery, in which the infection of a vascular graft, often with MRSA, can be followed by catastrophic haemorrhage or the need for prolonged infection control nursing care; infection of a hip prosthesis, which may also entail prolonged care with multiple courses of antibiotics, only to be resolved finally by revisional surgery; SSIs after hysterectomy or intestinal surgery which often present in primary care and result in a large demand on resources and a possibility of repeated inappropriate antibiotics; and sternal dehiscence and infection after cardiac surgery, which involves extensive multidisciplinary care. There are also consequences to an infection after breast cancer surgery, where there may be the emotional impact of waiting for healing prior to chemotherapy or radiotherapy. Many of these infections could be reduced by the use of topical antimicrobials, implemented as part of an institution-wide infection control policy.

Even though the analysis is based on published data, there are inherent limitations in this methodology. To an extent these can be addressed by sensitivity analysis, but the fact remains that the model is dependent on, and subject to the limitations of, the data used to populate it. Further studies would be useful to increase the robustness of the conclusions, particularly in a broader range of surgical specialities. It will be important to undertake research and audit following the introduction of topical antimicrobials to ensure that expected outcomes are reflected in clinical practice26–28.

Conclusions

A modelling approach such as the one described in this article can be a useful aid to decision making. It suggests that an antimicrobial barrier dressing may reduce both the incidence and cost of SSIs in patients and/or procedures at medium–high risk of wound infection. If the incidence of SSI is reduced by more than 9%, the antimicrobial strategy is cost saving. This appears plausible in light of the clinical and other evidence on the antimicrobial effect of the dressing.

Transparency

Declaration of funding

This study was funded by Smith & Nephew.

Declaration of financial/other relationships

D.L. has disclosed that he received a grant from Smith & Nephew to conduct this study. J.N., C.R. and R.S. have disclosed that they are employees of Smith & Nephew.

Notes

*Trademark of Smith & Nephew Ltd, Hull, UK.

*Trademark of Smith & Nephew Ltd, Hull, UK.

*Trademark of Smith & Nephew.