Estimating the Clinical Outcomes and Cost Differences Between Standard Care With and Without Cadexomer Iodine in the Management of Chronic Venous Leg Ulcers Using a Markov Model

Empirical Studies

Estimating the Clinical Outcomes and Cost Differences Between Standard Care With and Without Cadexomer Iodine in the Management of Chronic Venous Leg Ulcers Using a Markov Model

Volume 62 - Issue 6 - June 2016 ISSN 1943-2720

Abstract

Chronic venous leg ulcers (VLUs) affect up to 1% of the adult population in the developed world and present a significant financial and resource burden to health care systems. Cadexomer iodine (CI) is an antimicrobial dressing indicated for use in chronic exuding wounds. The aim of this study was to estimate the cost utility of using CI + standard care (SC) — ie, high compression multicomponent bandaging including debridement — compared with SC alone in the management of chronic (>6 months’ duration) VLUs from a payer’s perspective.A Markov model was constructed to evaluate the cost and clinical benefits (healing and decreased infection rates) of the 2 treatment modalities over a 1-year period using data from 4 randomized, controlled clinical studies (RCTs) included in a recent Cochrane review and cost data from a recently published economic evaluation of VLUs. Costs were calculated using 2014 United States dollars; wound outcomes included complete healing in 212 patients reported in the Cochrane meta-analysis and quality-adjusted life years (QALYs), with utility values obtained from 200 patients with VLUs calculated using standard gamble. Treatment with CI over 1 year was $7,259 compared to $7,901 for SC. This resulted in a cost savings of $643/patient in favor of CI compared with SC. More patients treated with CI (61%) had their wounds healed compared to 54% treated with SC. Furthermore, patients treated with CI+SC experienced 6 additional ulcer-free weeks compared to persons treated with SC alone (ie, 25 ulcer-free weeks compared to 19 ulcer-free weeks, respectively). Overall, CI resulted in 0.03 more QALYs (ie, 0.86 QALYs compared to 0.83 for SC). The use of CI in addition to SC compared to SC alone over 52 weeks resulted in more wounds healed and more QALYs along with a decrease of overall costs The results of this study suggest CI is cost effective compared to SC alone in the management of patients with chronic VLUs. Prospective, controlled clinical studies are needed to elucidate the effect and cost effectiveness of CI on VLUs with and without signs of infection as compared to SC, other antiseptics, and more advanced topical treatment modalities.

Introduction

Chronic venous leg ulcers (VLUs) affect up to 1% of the adult population in the developed world; with an expected increase in the number of older people in future decades, a Cochrane review¹ anticipates a corresponding increase in age-associated medical problems. A review of the burden of VLUs on the United States’ economy² and a retrospective observational database study³ of 81,148 matched patients with or without VLUs concluded venous leg ulceration imposes a significant financial and resource burden to health care systems. According to a cost-effectiveness study by Jemec et al,⁴ VLUs are typically painful, malodorous, moderately to highly exudating, and have a significant negative impact on patients’ quality of life. Furthermore, a systematic review⁵ on the impact of leg ulceration on patients’ quality of life that included 24 original studies concluded patients had significantly more pain, more restrictions regarding social functioning, less vitality, and limitations with respect to emotional roles compared to the respective controls.

The prevalence of VLUs in the US is approximately 600,000/year. VLUs are estimated to result in the loss of 2 million working days per year, and the annual cost of treating VLUs is estimated at $2.5 billion to $3.5 billion in the US.² Compression therapy is considered the first-line treatment for uncomplicated venous ulcers. According to international clinical guidelines on VLUs,^6,7 including a US VLU guideline⁸ and a clinical evidence review on the management and diagnosis of VLUs,⁹ multilayer compression systems are considered more effective than low-compression garments or stockings for the management of VLUs.

High bioburden resulting in the development of biofilms, bacterial colonization, and ultimately wound infection are implicated in delayed wound healing.^10,11 Furthermore, the presence of biofilms in wounds is also increasingly linked to delayed healing in both porcine¹² and murine models¹³and inferred by their presence in chronic wounds,¹⁴ with a high proportion of chronic wounds reported to contain such structures.¹⁵ A microscopic evaluation of specimens from chronic (n = 50) and acute (n = 16) wounds by James et al¹⁵identified wound biofilms in more than 60% of chronic wounds (including VLUs, diabetic foot ulcers [DFUs], pressure ulcers [PUs], and other traumatic chronic wounds) compared to only 6% in acute wounds. The protective nature of a mature biofilm and the slow growth of bacteria within them, along with host components in wounds (ie, tissue protein, red blood cells, and serum), have been shown in a number of in vitro studies^16-21 to inhibit antimicrobial penetration and activity, leading to enhanced survival and persistence of biofilm communities.

Cadexomer iodine (CI; IODOSORB™, Smith & Nephew, Hull, UK) is a topical antimicrobial dressing used in the treatment of infected or critically colonized leg ulcers. Critical colonization has been classified clinically as a level of bacterial colonization in a wound that leads to a subtle state of localized “covert” infection.²² CI is one of the few treatments demonstrated to kill mature biofilms in an ex vivo porcine skin model; CI and polyacrylate silver gel significantly reduced biofilm bacteria following 72-hour treatment in this test.²³ Such evidence combined with clinical efficacy of CI on bioburden in DFUs and VLUs noted in noncomparative multicenter and RCTs, respectively,^24,25 suggests a role for CI in successful treatment of biofilms in chronic wounds.

A clinical review of iodine use in clinical studies¹¹ concluded CI promotes wound healing when compared to other dressings such as hydrocolloids, paraffin gauze, and compression therapy. A recent Cochrane meta-analysis¹ of 212 patients included 4 RCTs among patients with chronic VLUs: Lindsay et al²⁶ compared CI and dry sterile dressing with a nonadherent dressing plus support bandaging or stocking in 28 patients; Laudanska and Gustavson²⁷ compared CI to saline wet-to-dry compressive dressing in 60 patients; Ormiston et al²⁸compared CI to gentian violet and polyfax ointment in 61 patients; and Steele et al²⁹ compared CI in addition to compression therapy to compression therapy alone plus various topical agents including antibiotics, antiseptics, and hydrophilic agents in 60 patients. The meta-analysis¹ of these 4 studies found evidence (risk ratio 2.17, 95% confidence interval 1.30–3.60, P = 0.003) supporting increased healing following the use of CI. However, only 2 trial-based clinical studies were identified that examined the cost effectiveness of CI in VLUs compared to hydrocolloid dressings and paraffin gauze³⁰ and gentamicin solution, streptodornase/streptokinase, and dry saline gauze.³¹ Given the current health care environment, decision makers need to use all available clinical and cost-effectiveness information to help make policy decisions.

The objective of the current study was to estimate the cost-effectiveness and cost utility of CI plus standard care (SC) compared with SC alone in the management of chronic VLUs. Although the studies included in the Cochrane meta-analysis used different comparators, they all used 0.9% CI as the intervention.

Methods and Perspective

A Markov model was used to simulate the expected cost and benefits of the 2 stated strategies in patients with chronic VLUs using all current available clinical evidence from RCTs included in the 2014 Cochrane review meta-analysis.¹ A pragmatic decision was taken not to conduct a systematic literature search for clinical evidence because data used from the Cochrane review had involved a thorough literature search in 2013.

Markov models are recursive (repetitive) decision trees used for modeling conditions involving events that may occur repeatedly over time, as is the case with VLUs and wound infections.³² For the purpose of this study, chronic VLUs were defined as ulcers >6 months in duration in line with the clinical effectiveness data taken from the Cochrane review¹ where the mean ulcer duration ranged between 7–17 months. The model was developed in accordance with best practice standards for economic evaluations (ie, The Consolidated Health Economic Evaluation Reporting Standards [CHEERS])³² from the perspective of a US health care payer (ie, Centers for Medicare and Medicaid Services [CMS]).

SC was defined as the use of high-compression, multicomponent bandaging, providing compression of 23 mm Hg to 35 mm Hg.^1,6-9,33 In addition to compression therapy, other known barriers to healing such as infection, chronic inflammation, and necrotic or nonviable tissue and biofilms were managed by debridement to facilitate optimal wound healing. For the purpose of the study model, it was assumed a physician performed debridement in accordance with 2014 clinical practice guidelines of the Society for Vascular Surgery and American Venous Forum,³⁴ although it was acknowledged that clinicians do not always follow these guidelines.³⁵ The model assumed foam dressings were used as the primary dressing in both the CI and SC groups. The Markov model assumed the patient was always in 1 of a finite number of health states, which in this case were unhealed VLU, healed, infected, and dead.

Two (2) measures of benefits were used in the model: incremental ulcer-free weeks for the cost-effectiveness analysis and quality-adjusted life years (QALYs) for the cost-utility analysis. A QALY is a measure of disease burden, which takes into account the quantity or length of life and quality of life generated by health care interventions. QALY is used to assess the value (in health gains) for a given cost of a medical intervention. A year of perfect health is worth 1 QALY, and a year of less than perfect health is worth <1. Death is equivalent to 0.^36,37

Model structure and assumptions. The model used weekly cycles to examine the costs and outcomes associated with the 4 Markov health states: healed, unhealed, infected, and dead over the 12-month time horizon (see Figure 1). The principle of Markov modelling is that the disease of interest is divided into discrete states by which the progression of disease falls over a certain amount of defined time (termed cycle length); in this study, the cycle length for this model was 1 week to represent the typical change frequency of compression bandages.⁹ In the Markov cohort, it is assumed transitions happen at the end of each cycle; in reality, patient transition is a continuous process that may occur during any time in the cycle. To address this assumption, a half-cycle correction can be used, which assumes health state transitions occur, on average, half way through the cycle. This correction approach is important when evaluating health outcomes, particularly when the cycle length is very long. For the study model, the cycle length was considered to be small enough not to require a half-cycle correction³⁶; thus, the individual in the model is assumed to move between health states on a weekly basis. As such, the health benefits of the intervention (healing, reduction in infection rates) are evaluated together with the costs on a weekly basis for the duration of the model (1 year).

Neither costs nor health benefits (ulcer-free weeks and QALYs) were discounted because the time horizon was <1 year.³⁵ Discounting renders benefits and costs that occur in different time periods comparable by expressing their values in present terms (ie, how much future benefits and costs are worth today). This is not a requirement in models with a short timeline of 1 year. Baseline transition probabilities (expected events with SC) were taken from an RCT of 2 types of bandage for treating VLUs (the VenUS 1V trial³³). These baseline data then were adjusted to reflect the expected reduction in adverse outcomes such as infection and recurrence while showing expected increase in healing rates observed in the clinical studies of CI. Evidence for the effectiveness of CI was taken from the Cochrane review¹ of 4 RCTs for healing outcomes, which included a meta-analysis that evaluated 212 patients. Three (3) additional studies were identified in the Cochrane review that considered CI and SC; however, these studies were excluded in the meta-analysis because they did not report on the healing outcome. Two (2) of the studies^25,26 excluded from the meta-analysis of healing outcome provided information on the effect of CI on infection control, which was meta-analyzed. Of the studies identified by the Cochrane review, only these 2 studies provided comparative evidence on infection control. The treatment effect of CI was assumed to persist for the duration of included trials (ie, 12 weeks) and thereafter all outcomes were treatment-independent, the same as seen in SC.

Model description. All patients start in the stalled/nonhealing health state. This health state represents a wound of >6 months that has failed to improve after 4 weeks of treatment with adequate compression therapy. Within each week, wounds can either remain in the stalled/nonhealing health state or move to any of the other 3 health states (ie, healed, infected, or death). People who transition to the healed health state are exposed to the risk of dying, have a recurrent ulcer, or remain in the healed state. Once healing is achieved, the model assumes patients will be put on below-knee graduated compression hosiery to prevent recurrence of a VLU in accordance with published guidelines.^1,6-9

A proportion of ulcers (9.5% and 5% in the first 12 weeks and beyond 12 weeks, respectively) were assumed per Carter et al³⁸ to recur. Ulcer recurrence was defined as a wound that had healed and reopened (ie, moved from the healed health state to the unhealed health state). Once an ulcer recurred, 3 possible transitions were assumed (ie, move to healing health state, remain unhealed, or death). Recurrence data used were taken from the 4-layer bandaging arm of the VenUS 1V study³³ in accordance with international guidelines that recommend high-compression, multicomponent bandaging.^1,7,9 Infection was modeled with 2 possible transitions assumed for the infection health state: either the infection was resolved (ie, patients transitioned to the unhealed health state) or the patient died. However, the model did not allow for an infected ulcer to become healed over the course of 1 week. Data on infection were taken from Carter et al,³⁸ which examined the cost-effectiveness of 3 topically applied cellular/tissue-derived products (CTPs) used as adjunct therapies to SC in the management of VLUs. CTPs evaluated in the study were extracellular matrix, human skin equivalent, and living skin equivalent. The impact of infection in the model was to delay healing, because infection had to be resolved before patients could move to the healing health state.

Mortality is the absorbing health state; no further transitions occur once a person enters this state. The model does not capture any VLU-specific mortality — rather, the age-specific mortality of the general population has been modeled. An annual mortality rate of 2% was taken from the Centers for Disease Control and Prevention’s 2010 life tables³⁹ for the 70- to 75-year-old age group (a common age group for venous ulcer studies). The weekly probability was derived using a formula by Briggs et al,⁴⁰ and none of the interventions was assumed to have an impact on mortality. Baseline transition probabilities and the effect of CI are summarized in Table 1.

Health-related quality of life (HRQOL) and utility data used in the model. The ultimate goal of health care is to restore or preserve functioning and well-being related to health (that is, HRQOL). As such, HRQOL is subjective and multidimensional, encompassing physical and occupational function, psychological state, and social interaction. A review⁴¹discussing the developments in VLU management noted chronic leg ulcers significantly impact a person’s HRQOL because wounds make it difficult for patients to perform normal activities due to pain, exudate, and other issues. In a cross-sectional survey of 38 patients with VLUs, Hareendran⁴² identified the following VLU-specific HRQOL factors as concerning for patients: pain, itching, lack of quality sleep, and altered appearance. Furthermore, a cross-sectional study⁴³ among 30 patients with DFUs has shown HRQOL is correlated with wound severity (ie, open ulcers associated with poor HRQOL) compared to healed ulcers.

For the cost-utility analysis, data from a cost-utility analysis⁴⁴ of bio-electric stimulation therapy compared to SC in 21 elderly patients with chronic, nonhealing wounds >6 months’ duration was used for the different health states. Researchers assumed HRQOL is dependent on age, owing to evidence suggesting HRQOL is negatively correlated with age.⁴⁵ Hence, values are modeled to change as the patient ages. In order to demonstrate robustness of the model results, a sensitivity analysis was undertaken that assumed HRQOL was not related to age. Utility data used in the model are shown in Table 2.

Health care resource costs. Resource use costs were based on 2014 Medicare national average reimbursement rates and were calculated as the product of resources (quantity) used and the relevant unit costs. VLU medical diagnosis-related groups (DRGs) 571 and 593 were used to determine the cost of hospitalization of all unhealed ulcers. DRG data were taken from the Healthcare Cost and Utilization Project (HCUP) website (https://hcupnet.ahrq.gov). The rates of hospitalization were taken from Carter et al,³⁸ who estimated the cost effectiveness of 3 adjunct cellular/tissue-derived products used in the management of chronic VLUs. Estimated hospitalization rates for up to 12 weeks were 4.8%. No data were available for weeks 13 through 52; it was assumed the rate would be 50% less than the rate seen at 12 weeks (2.4%). This resulted in an estimated annual probability of hospitalization of approximately 10%.

For infection, DRGs 602 and 603 were used. The estimated annual probability of hospitalization for infection was 10%. It was further assumed that 80% of hospitalization for infection would be without complications.

Outpatient visits and home health care. Using data from Carter et al,³⁸ 25% of the patients were assumed to have additional home health care with a once-a-week visit to change dressings. The code C2F2S1, which denotes 1 visit per week for a compression dressing change with a 60-day episode of care, was used for Medicare. A proportion of patients was assumed to be debrided in the first 12 weeks with no debridement thereafter (according to clinical experts). The 12.5% weekly probability of debridement was taken from Clegg and Guest’s⁴⁴ cost-utility study of bio-electric stimulation therapy. When debridement occurs, the clinic cannot bill for compression in accordance with CMS reimbursement policy. The current researchers assumed 1 hospital outpatient visit per week in accordance with the cycle length. CPT codes 11042 for debridement and 99203 for the evaluation and management visit were used for week 1, and CPT code 97597 (an established clinic visit for debridement) for weeks 2–12 were used to generate costs. In addition, the model did not include the cost of primary dressings and use of additional adjunctive therapies such as negative pressure or low-dose ultrasound; these were assumed to be used at the same rate between the treatment groups and cost-neutral in the model.

Infection control drugs, pain medication, and intervention costs. It was assumed the cost of treating wound infection would be included in the DRG for infection. Because one of the most common side effects of nonhealing VLUs is pain,^1,9 the model incorporated the costs of managing pain; pain medications, their costs, and proportions of patients on each drug were taken from Carter et al.³⁸ Expert opinion (personal communication) suggested CI would be applied once per week when patients have their weekly appointments to change compression therapy, especially in patients with low to moderately exuding wounds. The model assumed 40 mg of CI was used and compression was changed once a week (see Table 3).

Cost-effectiveness analysis. The incremental cost-effectiveness ratio (ICER) is the added cost per additional unit of health — in this model, measured in QALYs and ulcer-free weeks. This was calculated as the difference between the expected costs of the 2 strategies divided by the difference between the expected probability of ulcer-free days or QALYs between the 2 strategies over 52 weeks. To determine if the intervention is cost-effective, the ICER is compared with the maximum amount that can be paid by the health care payer for an additional unit of health benefit — in this case, assumed to be the CMS. Unlike some national payers, the US government has been reluctant to consider cost-effectiveness evidence in coverage decisions and does not have an explicit willingness to pay threshold.⁴⁶ However, previous studies suggest the observed trend ranges from $50,000 to $100,000 per QALY for interventions that have been covered. Although some interventions have a significantly higher cost utility ratio,⁴⁶ for the purpose of this study a maximum willingness to pay threshold of $50,000 per QALY was adopted based on empirical precedent.⁴⁷ If the calculated ICER is less than this amount, intervention is deemed cost effective.

Sensitivity analysis. One-way sensitivity analyses were conducted by varying some of the parameters in the model to address the impact of possible uncertainty in the best information available on clinical benefits and treatment costs. Each key parameter was alternately assigned a low and high value and the deterministic cost-effectiveness results using this value were recorded. For deterministic sensitivity analysis, baseline values were varied ±20% if ranges were not reported in the literature in accordance with other published economic studies.³⁷ Furthermore, a probabilistic cost-effectiveness analysis was performed, which entails specifying a distribution for each model parameter to represent the uncertainty around the point estimate and then selecting values at random from those distributions using Monte Carlo simulation. Essentially, this means the uncertainty around multiple input assumptions can be tested simultaneously, distinct from one-way sensitivity analysis that allows uncertainty around single inputs to be tested individually. The lognormal distribution was implemented to capture the uncertainty surrounding the treatment effect; the gamma and beta distributions were used to capture the uncertainty in cost and utility values, respectively. The 2,000 trials or simulations conducted included the usual practice of running more than 1000 simulations.⁴⁸

Uncertainty regarding the ICER was assessed by Monte-Carlo simulation using 2000 iterations. To explore decision uncertainty regarding the cost-effectiveness of CI, the joint distribution of mean cost and mean outcomes was evaluated on a decision plane formally known as the incremental cost-effectiveness plane (see Figure 2). The decision plane is dived into 4 quadrants by the origin (ie, southeast [SE], northeast [NE], southwest [SW], and northwest [NW]), with each quadrant having a different implication for economic evaluation.⁴⁹ The horizontal axis divides the plane according to incremental cost (positive above, negative below), and the vertical axis divides the plane according to incremental effect (positive to the right, negative to the left). Cost-effectiveness acceptability curves (CEACs) (see Figure 3) then are calculated by plotting the proportion of cost and effects pairs that are cost effective for a given value the payer is willing to pay. Thus, the CEAC expresses the likelihood the cost-effectiveness estimate reflects a cost-effective intervention based on the existing evidence.⁴⁹ A scatter plot also is presented.

Results

Sixty-one (61%) percent of wounds are expected to heal with CI+SC compared to 54% treated with SC alone over a 52-week period. The model also predicts treatment with CI+SC will result in 6 additional ulcer-free weeks per patient over 52 weeks. In addition, the model suggests treatment with CI+SC is expected to lead to an overall health gain of 0.03 QALYs per patient over 52 weeks compared to a patient treated with SC. Treating patients with CI+SC rather than SC alone resulted in a cost savings of $643 per patient over 52 weeks. Consequently, CI+SC was found to be a dominant treatment for chronic VLU because it resulted in a reduction in health care costs (-$643) and an improvement in health benefits of 6 ulcer-free weeks or 0.03 QALYs (see Table 4 and Table 5).

One-way sensitivity analyses did not materially affect the outcomes. CI remained dominant when key parameters were varied. The only time CI was not dominant was when lower values of treatment effect were used, but it remained cost-effective at $50,000/QALY threshold, which is commonly reported in many cost-effectiveness studies⁴⁶ (see Table 6).

Uncertainty regarding the ICER is shown in Figure 3 which illustrates the probability any one strategy is cost-effective as a function of the willingness to pay. Given a willingness to pay $50,000/QALY, the probability CI is cost effective compared to SC is 96%. This is visually illustrated in Figure 2, which shows the majority of the samples are located in the SE and NE quadrants of the cost-effectiveness plane. The model estimates 83% of samples show CI is a dominant strategy, yielding better outcomes at a lower cost.

Discussion

The need for cost-effectiveness evidence has increased for many reasons, mainly due to limited financial resources and an increased demand for cost-effective health care technologies that consider acquisition and overall clinical and patient benefit. The Markov model estimated the 1-year clinical and cost effectiveness of adding CI to SC compared to SC alone from a US payer’s perspective in treating patients with VLUs. Using CI in addition to SC resulted in greater clinical benefit and lower treatment costs overall when compared to SC alone: using CI resulted in more wounds healed at 52 weeks (61% versus 54% wounds healed), more ulcer-free weeks (25 versus 19 ulcer-free weeks), and a cost savings of $643 per wound treated ($7259 versus $7901, respectively). Results of sensitivity analysis assigning high and low values of the clinical and cost data inputs remained robust (ie, the conclusions reached about the cost-effectiveness of CI were not sensitive to changes in the data input assumptions).

Economic evaluations are relatively rare in wound care, largely due to an absence of sufficient robust clinical evidence.⁵⁰ CI is one of the few wound care interventions with clinical evidence to support its efficacy (ie, the RCTs in the Cochrane review¹). Two (2) economic evaluations compared the cost effectiveness of CI. Hansson³⁰ compared the efficacy and cost effectiveness of CI in a multicenter RCT over a 12-week period with either a hydrocolloid dressing or paraffin gauze in 153 patients with exudating VLUs. The calculated mean ulcer reduction was 58%, 14.8%, and 45% for CI, hydrocolloid, and paraffin gauze, respectively. Total treatment costs over 12 weeks were calculated to be $517 for CI, $480 for hydrocolloid, and $582 for paraffin gauze. CI was found to be cost-effective with a lower average cost per percentage reduction in ulcers of $8.80 per percentage ulcer reduction compared with $32.50 and $21.90 per percentage ulcer reduction for hydrocolloid and paraffin gauze, respectively. A 12-week RCT by Apelqvist et al³¹ of 41 patients with exudating DFUs found the use of CI resulted in fewer dressing changes per week when CI was compared with gentamicin solution and dry saline gauze. The authors estimated that treatment with CI resulted in weekly treatment costs in Swedish Krona (SEK 1993) of 903 (524 to 1,697) versus 1,421 (428 to 2,679).

The results of the current study are consistent with 2 previously published studies showing CI is either cost effective or cost saving in managing chronic wounds compared to SC of hydrocolloid dressings and paraffin gauze in 1 study³⁰ and gentamicin solution, streptodornase/streptokinase, and dry saline gauze in another study.³¹ Unlike these studies, the current model extrapolated the cost and benefits over 1 year to reflect the actual costs of chronic ulcers, which in most cases are not healed by 12 weeks, making the results more relevant to policy makers.⁵¹ In addition, QALYs were considered regarding alternative treatment options. Although QALYs are not widely reported in wound care (again, largely due to an absence of robust clinical data), this is a concept widely used among health care payers and coverage bodies; it should make the findings of this study more widely relevant beyond the wound care community.

The Markov model was based on the results of 4 RCTs included in an independent Cochrane review of healing^26-29 and infection^25,26 outcomes. The use of Cochrane reviews and meta-analysis of RCTs is considered gold standard evidence for clinical and economic evaluations. Furthermore, the current results are robust because both one-way sensitivity and probabilistic sensitivity analysis were used to test the robustness of the model results and the CI remained cost effective in all cases of sensitivity analysis.

For instance, the model assumed treatment benefit of CI lasted for 12 weeks as determined by the duration of the trials and the usual treatment time for VLUs.³³ However, a sensitivity analysis was conducted assuming that treatment effect lasted for 6 weeks to reflect the follow-up time in 2 of the studies included in the meta-analysis,^27,29 and CI remained cost saving. The model was also conservative in that the cost of infection was excluded from the model because the cost of antibiotics was deemed to be negligible. However, CI resulted in better infection control compared to SC. Thus, if the costs of infection control were included, an even greater cost savings would have been recognized. Resource use data were based on a recently published cost-effectiveness study³⁸ that solicited input from a panel of chronic wound practitioners as opposed to protocol-driven data use obtained from clinical trials.

Limitations

An important limitation of the current evaluation is that it is a model that extrapolates clinical results from a 12-week trial out to a full year. Extrapolation was necessary to adequately capture all relevant health outcomes and costs associated with treatment. Another limitation is that for the clinical evidence review, not all possible databases were searched. However, given that the Cochrane review¹ used herein included a systematic review of literature published up to 2013, it seems unlikely that any new relevant literature that reported on outcomes of interest has been excluded from the analysis. A further limitation is the sample sizes in the included trials, which ranged from 28 patients in 1 study²⁶ to 66 in the biggest study.²⁷

The model used baseline data from the VenUS IV trial,³³ where the majority of patients included (72.6%) had ulcers of <6 months’ duration. However, the treatment effectiveness was taken from the Cochrane review,¹ which included patients with mean ulcer duration of at least 7 months. The authors are aware ulcer duration is a prognostic factor in ulcer healing; nonetheless, because effectiveness data were obtained from patients with longer ulcer duration, the results of the model are deemed conservative (shorter ulcer duration is associated with better healing rates).²⁵ In addition, the cost of primary dressings and other adjunctive therapies was assumed to be used at the same rate between CI and SC and hence were excluded in the model. This may not necessarily have been the situation and was likely to bias the results against the more effective intervention. In this scenario, the model is conservative (biased against CI) because CI was more effective in healing the wounds and therefore fewer wounds would have required the use of additional adjunctive therapies and primary dressings.

The authors are aware of concerns regarding cytotoxicity surrounding iodine use. Studies included in this analysis did not report on significant difference between CI and SC on any side effects; therefore, the current model did not include the management of this side effect and others except pain, which is widely reported in studies.^25,27,28 Two of the included studies reported initial iodine absorption; however, they noted the amount of iodine absorbed was not cytotoxic because it did not affect thyroid function.^25,27 One clinical review⁵² of iodine products in chronic wound care, including povidone-iodine solutions and CI, concluded that concerns about systemic toxicity are probably overstated in particular for the slow release CI. Furthermore, an in vivo study by Zhou et al⁵³ highlighted the minimal cytotoxicity of CI due to the sustained release preparation; in this study, skin biopsies of chronic exudative wounds being treated with CI demonstrated no evidence of cell necrosis and displayed reepithelialization. This evidence supports the faster healing rates highlighted in the Cochrane review¹ and associated clinical studies.^{26-28, 30}

All of the included studies did not mention whether patients were infected at study entry with the exception of the Lindsay et al study,²⁶ where the clinical signs of infection were monitored throughout the intervention. Nevertheless, high bioburden and infection are known to delay wound healing.^54,55 In the studies assessed for this Markov model, the presence of infection at baseline may have led to increased bias toward healing due to the antimicrobial effect of the CI.

Conclusion

Using the Markov model, the use of CI in addition to SC compared to SC alone in the management of VLUs over 52 weeks resulted in more wounds healed and more QALYs along with a decrease of overall costs of care. Patients treated with CI experienced 25 ulcer-free weeks and 0.86 QALYs compared to 19 ulcer-free weeks and 0.82 QALYs for those treated with SC. Using CI was found to be cost effective with a total cost per patient over 52 weeks of $7259 compared $7901, saving the health payer $643 per patient treated. Prospective, controlled, clinical studies are needed to confirm the results of this study and elucidate the effect and cost-effectiveness of CI on VLUs with and without signs of infection as compared to SC, other antiseptics, and more advanced topical treatment modalities.

Affiliations

Mr. Nherera is Health Economics Manager; Dr. Woodmansey is Scientific Communications Manager; and Mr. Trueman is Vice President, Market Access, Smith & Nephew Advanced Wound Management, Hull, UK. Dr. Gibbons is Medical Director, South Shore Hospital Center for Wound Healing, Weymouth, MA.

Correspondence

Please address correspondence to: Leo M. Nherera, Health Economics Manager, Smith & Nephew Advanced Wound Management, 101 Hessle Road Hull, HU3 2BN, UK; email: leo.nherera@smith-nephew.com.

References

1. O’Meara S, Al-Kurdi D, Ologun Y, Ovington LG, Martyn-St James M, Richardson R. Antibiotics and antiseptics for venous leg ulcers. Cochrane Database Syst Rev. 2014: CD003557. doi: 10.1002/14651858.CD003557.

2. Sen CK, Gordillo GM, Roy S, et al. Human skin wounds: a major and snowballing threat to public health and the economy. Wound Repair Regen. 2009;17(6):763–771.

3. Rice JB, Skornicki M, Parsons N. Burden of venous leg ulcers in the United States. J Med Econ. 2014;17(5):347–356.

4. Jemec GBE, Kerihuel JC, Ousey K, Lauemøller SL, Leaper DJ. Cost-effective use of silver dressings for the treatment of hard-to-heal chronic venous leg ulcers. PLoS ONE. 2014;9(6):e100582.

5. Herber OR, Schnepp W, Rieger MA. A systematic review on the impact of leg ulceration on patients’ quality of life. Health Qual Life Outcomes. 2007;5:44. doi:10.1186/1477-7525-5-44.

6. Scottish Intercollegiate Guidelines Network. Management of chronic venous leg ulcers. Edinburgh, UK: SIGN; 2010. (SIGN guideline No. 120). Available at: http://sign.ac.uk/pdf/sign120.pdf. Assessed May 7, 2015.

7. Royal College of Nursing (2006). Clinical Practice Guidelines: The Nursing Management of Patients with Venous Leg Ulcers Recommendations. London, UK: Royal College of Nursing. Available at: www.rcn.org.uk. Accessed May 7, 2015.

8. White-Chu EF, Conner-Kerr TA. Overview of guidelines for the prevention and treatment of venous leg ulcers: a US perspective. J Multidisciplinary Healthcare. 2014;7(63):111–117.

9. Collins L, Seraj S. Diagnosis and treatment of venous ulcers. Am Fam Phys. 2010:81(8):989–996

10. Vermeulen H, Westerbos SJ, Ubbink DT. Benefit and harm of iodine in wound care: a systematic review. J Hosp Infect. 2010;76:191e–199e.

11. Cooper RA. Iodine revisited. Int Wound J. 2007;4(2):124–137.

12. Roche ED, Renick PJ, Tetens SP, Ramsay SJ, Daniels EQ, Increasing the presence of biofilm and healing delay in a porcine model of MRSA-infected wounds. Wound Repair Regen. 2012;20(4):537–543.

13. Schierle CF, De la Garza M, Mustoe T. Staphylococcal biofilms impair wound healing by delaying reepithelialization in a murine cutaneous wound model. Wound Repair Regen. 2009;17(3):354–359.

14. Bjarnsholt T, Kirketerp-Møller K, Jensen PØ, et al. Why chronic wounds will not heal: a novel hypothesis. Wound Repair Regen. 2008;16(1):2–10.

15. James GA, Swogger E, Wolcott R, et al. Biofilms in chronic wounds. Wound Repair Regen. 2008;16(1):37–44.

16. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistent infections. Science. 1999;284(5418):1318–1322.

17. Nickel JC, Ruseska I, Wright JB, Costerton JW. Tobramycin resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter material. Antimicrob Agents Chemother. 1985;27(4):619–624.

18. Kapalschinski N, Seipp HM, Onderdonk AB, et al. Albumin reduces the antibacterial activity of polyhexanide-biguanide-based antiseptics against Staphylococcus aureus and MRSA. Burns. 2013;39(6):1221–1225.

19. Pitten F. Werner HP, Kramer A. A standardized test to assess the impact of different organic challenges on the antimicrobial activity of antiseptics. J Hosp Infect. 2003;55(2):108–115.

20. Stewart PS, Costerton JW. Antibiotic resistance of bacteria in biofilms. Lancet. 2001;358(9276):135–138.

21. Singh R, Ray P, Das A, Sharma M. Penetration of antibiotics through Staphylococcus aureus and Staphylococcus epidermidis biofilms. J Antimicrob Chemother. 2010;65(9):1955–1958.

22. World Union of Wound Healing Societies (WUWHS). Principles of Best Practice: Wound Infection in Clinical Practice. An International Consensus. London, UK: MEP Ltd, 2008. Available at: http://woundsinternational.com/article.php?contentid=127&articleid=31. Accessed May 8, 2015.

23. Phillips PL, Yang Q, Davis S, et al. Antimicrobial dressing efficacy against mature Pseudomonas aeruginosa biofilm on porcine skin explants. Int Wound J. 2013;12(4):1–15.

24. Schwartz J, Lantis JC, Gendics C, Fuller AM, Payne W, Ochs D. A prospective, non-comparative, multicenter study to investigate the effect of cadexomer iodine on bioburden load and other wound characteristics in diabetic foot ulcers. Int Wound J. 2013;10(2):193–199.

25. Skog E, Arnesjö B, Troëng T, et al. A randomized trial comparing cadexomer iodine and standard treatment in the out-patient management of chronic venous ulcers. Br J Dermatol. 1983;109(1):77–83.

26. Lindsay G, Latta D, Lyons KGB, Livingstone ED. A study in general practice of the efficacy of cadexomer iodine in venous leg ulcers treated on alternate days. Acta Therapeutica. 1986;12:141–148.

27. Laudanska H, Gustavson B. In-patient treatment of chronic varicose venous ulcers. A randomized trial of cadexomer iodine versus standard dressings. J Int Med Res. 1988;16(6):428–435.

28. Ormiston MC, Seymour MTJ, Venn GE, Cohen RI, Fox JA. A randomized comparison of cadexomer iodine and a standard treatment in out-patients with chronic venous ulcers. Br Med J (Clinical Research Edition). 1985;291(6491): 308–310.

29. Steele K, Irwin G, Dowds N. Cadexomer iodine in the management of venous leg ulcers in general practice. Practitioner. 1986; 230(1411):63–68.

30. Hansson C. The effects of cadexomer iodine paste in the treatment of venous leg ulcers compared with hydrocolloid dressings and paraffin gauze dressing. Int J Dermatol. 1998;37(5):390–396.

31. Apelqvist J, Ragnarson Tennvall G. Cavity foot ulcers in diabetic patients: a comparative study of cadexomer iodine ointment and standard treatment. An economic analysis alongside a clinical trial. Acta Derm Venereol. 1996;76(3):231–235.

32. Husereau D, Drummond M, Petrou S, et al on behalf of the CHEERS Task Force. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. Value in Health. 2013;2(16):231–250.

33. Ashby RL, Gabe R, Ali S, et al. Clinical and cost-effectiveness of compression hosiery versus compression bandages in treatment of venous leg ulcers (Venous leg Ulcer Study IV, VenUS IV): a randomised controlled trial. Lancet. 2014;383(9920):871–879.

34. O’Donnell FT, Passman MA, Marston WA, Ennis WJ, et al. Management of venous leg ulcers: clinical practice guidelines of the Society for Vascular Surgery and the American Venous Forum. J Vasc Surg. 2014;60(2):3S–59S.

35. Gibbons GW, Orgill DP, Serena TE, et al. A prospective, randomized, controlled trial comparing the effects of noncontact, low-frequency ultrasound to standard care in healing venous leg ulcers. Ostomy Wound Manage. 2015;61(1):16–29.

36. Briggs A, Sculpher M. An introduction to Markov modelling for economic evaluation. Pharmacoeconomics. 1998;13(4):397–409.

37. Gold MR, Siegel JE, Russel LB, Weinstein MC (eds). Cost-effectiveness in Health and Medicine. New York, NY: Oxford University Press;1996.

38. Carter MW, Waycaster CR, Schaum, K, Gilligan AM. Cost-effectiveness of three adjunct cellular/tissue-derived products used in the management of chronic venous leg ulcers. Value in Health. 2014;17(8):801–813.

39. Arias E. United States Life Tables, 2010. National Vital Statistics Reports. Centers for Disease Control and Prevention. Available at: www.cdc.gov/nchs/data/nvsr/nvsr63/nvsr63_07.pdf. Accessed May 7, 2015.

40. Briggs AHE, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation. Oxford, UK: Oxford University Press;2006.

41. Anderson I. Developments in venous leg ulcer management. Nurs Times. 2011;107(35):14–18.

42. Hareendran A, Bradbury A, Budd J, et al. Measuring the impact of venous leg ulcers on quality of life. J Wound Care. 2005;14(2):53–57.

43. Jaksa PJ, Mahoney JL. Quality of life in patients with diabetic foot ulcers: validation of the Cardiff wound impact schedule in a Canadian population. Int Wound J. 2010;7(6):502–507.

44. Clegg JP, Guest JF. Modelling the cost-utility of bio-electric stimulation therapy compared to standard care in the treatment of elderly patients with chronic non-healing wounds in the UK. Curr Med Res Opin. 2007;23(4):871–883.

45. Fryback DG, Dunham NC, Palta M, et al. US norms for six generic health-related quality-of-life indexes from the National Health Measurement Study. Med Care. 2007;45 12):1162–1170.

46. Weinstein MC. How much are Americans willing to pay for a Quality-Adjusted Life Year? Med Care. 2008;46(4):343–345.

47. Neumann PJ, Cohen JT, Weinstein MC. Updating cost-effectiveness — the curious resilience of the $50,000-per-QALY threshold. N Engl J Med. 2014;371(9):796–797.

48. Claxton K, Sculpher M, McCabe C, et al. Probabilistic sensitivity analysis for NICE technology assessment: not an optional extra. Health Econ. 2005;14(4):339–347.

49. Fenwick E, Byford S. A guide to cost-effectiveness acceptability curves. Br J Psychiatr. 2005;187(2):106–108.

50. Carter MJ. Economic evaluations of guideline-based or strategic interventions for the prevention or treatment of chronic wounds. Appl Health Econ Health Policy. 2014;12(4):373–389.

51. Drummond M. Experimental versus observational data in the economic evaluation of pharmaceuticals. Med Decision Making. 1998;8(2):S121–S128.

52. Leaper DJ, Durani P. Topical antimicrobial therapy of chronic wounds healing by secondary intention using iodine products. Int Wound J. 2008;5(2):361–368.

53. Zhou LH, Nahm WK, Badiavas E, Yufit T, Falanga V. Slow release iodine preparation and wound healing: in vitro effects consistent with lack of in vivo toxicity in human chronic wounds. Br J Dermatol. 2002;146(3):365–374.

54. Edwards R, Harding KG. Bacteria and wound healing. Curr Opin Infect Dis. 2004;17(2):91–96.

55. Menke NB, Ward KR, Witten TM, Bonchev DG, Diegelmann RF. Impaired wound healing. Clin Dermatol. 2007;25(1):19–25.