Rheumatology Advance Access originally published online on June 16, 2006
Rheumatology 2006 45(8):1029-1038; doi:10.1093/rheumatology/kel147
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© Published by Oxford University Press on behalf of the British Society for Rheumatology 2006.
Estimating the cost and health status consequences of treatment with TNF antagonists in patients with psoriatic arthritis
1 Health Economics and Decision Science, School of Health and Related Research (ScHARR), University of Sheffield S1 4DA, 2 Academic Unit of Musculoskeletal Disease, University of Leeds, LS3 9NZ and 3 Wyeth Pharmaceuticals, Taplow, Maidenhead, Berks, SL6 0PH, UK
Correspondence to: Nick Bansback, Health Economics and Decision Science, ScHARR, University of Sheffield, Regent Court, 40 Regent Street, Sheffield S1 4DA, UK. E-mail: n.j.bansback{at}sheffield.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Objectives. Tumour necrosis factor (TNF) has been shown to improve the outcomes in patients with psoriatic arthritis (PsA). We estimate the long-term impact on health status of prescribing the TNF antagonist etanercept, and evaluate the cost-effectiveness in a health economic model.
Methods. The relationship between disability (Health Assessment Questionnaire) and health state utility was explored to estimate the quality-adjusted life years (QALYs) gained from the TNF antagonist etanercept. A model was then used to compare sequences of treatments for PsA after failure of two conventional disease modifying anti-rheumatic drugs (DMARDs). One arm commences on etanercept therapy and this is compared with a strategy commencing with combination therapy of methotrexate and ciclosporin and another commencing with leflunomide. Individual patient data from Phase III etanercept trials is used to populate the model supported by published evidence from extensive literature searches. By incorporating a life table specific for a PsA population, and using a number of evidence- and expert opinion-based assumptions for disease progression, the model was extended beyond the trial duration to a 10-yr time horizon. Cost offsets were produced by avoiding surgery through delayed progression; drug and monitoring costs were also modelled.
Results. Over the 10 yrs, modelled etanercept treatment gave 0.82 more QALYs when compared with combination therapy with methotrexate and ciclosporin, and 0.65 more QALYs in comparison with leflunomide. This equates to a central estimate for the cost per QALY of £28 189 and £28 189 for ciclosporin and leflunomide, respectively. Sensitivity analyses demonstrated this could vary by as much as ±28%.
Conclusions. With limited data currently available, the potential cost-effectiveness of etanercept in DMARD failures for adults with PsA appears encouraging. The result for other TNF antagonists will depend on how their relative efficacy and drug price compares with etanercept. A number of limitations are described and priorities for further research suggested.
KEY WORDS: Tumour necrosis factor, Health status, Cost-effectiveness, Cost utility, Economic evaluation, Psoriatic arthritis
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
An important outcome of any clinical intervention is the change in the subject's own perceived state of health. This can be categorized as health-related quality of life (HRQL) or health state utility (preference-based health state). Health state utilities are particularly useful for quantifying the effectiveness of health care interventions in terms of quality-adjusted life years (QALYs) [1]. The QALY combines health state utility with the length of life into a single index. The incremental QALYs of an intervention in comparison with conventional therapy can be compared with the incremental costs to derive an incremental cost-effectiveness ratio (ICER). ICERs are used to inform priority setting in new health care interventions in a number of countries, including Australia, Canada and most recently the UK through the National Institute of Clinical Excellence (www.nice.org.uk).
Cost-effectiveness evaluations are particularly pertinent in the case of expensive yet effective therapies such as tumour necrosis factor (TNF). The effectiveness and efficiency of TNF antagonist therapy is well-established in subjects with rheumatoid arthritis (RA) [2–5]. Psoriatic arthritis (PsA) is an inflammatory joint disease classified as one of the spondyloarthropathies, with severe impact on patient's disability. If untreated, PsA can be a progressively disabling disease. The prospective follow-up of 180 patients with PsA for 8 yrs showed that 60% of them developed erosive arthritis [6]. Whilst exact prevalence figures for PsA are difficult to establish, psoriasis occurs in about 1–3% of the general population, and between 7 and 42% of the patients with psoriasis also suffer from arthritis [7–9]. Non-steroidal anti-inflammatory drugs (NSAIDs) are traditionally the first-line treatment with the second-line treatment generally provided by the disease modifying anti-rheumatic drugs (DMARDs) such as sulphasalazine and methotrexate. Use of these drugs is, however, based on limited evidence and has resulted from benefits seen in RA.
Whilst the impact on HRQL in PsA patients is considered to be equivalent to RA patients [10], there is a scarcity of research and evaluations into the economic burden of the disease and efficiency of new treatments [11]. The aim of this study was to first estimate the potential long-term benefits on health status of the TNF antagonist etanercept in PsA. Then, by combining costs, the long-term cost-effectiveness of TNF antagonists in comparison with conventional therapies can be forecast.
|
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Overview
A suitable preference-based utility instrument was not measured in the clinical trials of etanercept. Instead, we used the Health Assessment Questionnaire Disability Index (HAQ-DI) as a surrogate for health state utility. The HAQ-DI is a validated and well-established measure of patient disability that has been shown to be a good predictor of HRQL, direct costs and mortality in patients with RA [12–15]. It has also been validated for use in patients with PsA [16]. It is measured in all the clinical trials of treatments for PsA, and previous studies in RA have found a strong relationship between HAQ-DI and health utility [17]. We re-examined this relationship in PsA patients using the EQ-5D, a preference-based instrument.
To estimate the long-term effects, we used a combination of evidence on HAQ-DI from Randomised Controlled Trials (RCTs), open label and observational data. This was combined with a clinical pathway algorithm, which determines the length of time a patient would remain on therapy and the type of subsequent treatments that would be received. A mathematical model is used to synthesize the effects on health state utility with mortality and costs into a framework, which can be used to inform the long-term implications of a reimbursement decision [18].
Data
The patient level data from the phase III randomized controlled trial of etanercept vs placebo in patients with PsA was analysed [19]. At the time of analysis, this was the only published phase III clinical trial, and the only trial that we had access to patient level data. The analysis therefore focuses on etanercept, but has implications for all the TNF antagonists. In the trial, patients with active PsA were randomized to either etanercept 25 mg twice weekly (n = 101) or placebo (n = 104). HAQ-DI was measured at 4, 12 and 24 weeks. Patients were continued on blinded medication until all patients had completed the trial, at which point they were invited to join an open-label extension in which all patients were treated with etanercept. Patients were 47 yrs of age (range), had a relatively long disease duration (9 yrs), had failed an average of 1.7 DMARDs and had a HAQ of 1.1 at study entry (Table 1).
|
A second dataset, from a cohort of patients based at the Academic Unit of Musculoskeletal Disease, University of Leeds, was used to examine health state utility and long-term progression in patients with PsA. These patients were initially entered into a randomized controlled trial of combination therapy of methotrexate with and without ciclosporin A (n = 38 and n = 34, respectively), but were later followed up by postal questionnaire [20]. Patients answered questions on HAQ-DI, and the EQ-5D, a generic measure of patient's health state utility. Patients were 47 yrs of age at study entry, had a baseline HAQ-DI of 1.0 and most importantly, previously they had an inadequate response to methotrexate (Table 1). This dataset represents a more similar patient cohort of PsA patients to those in the etanercept trial, i.e. long disease duration with a moderate-to-severe disease not previously controlled by methotrexate.
Clinical pathway
The typical pathway for patients eligible for TNF antagonist therapy was obtained from the provisional British Society for Rheumatology (BSR) guidance [21]. In this, the patients must have had an inadequate response on two previous DMARDs. It is uncertain as to which treatment would be next given if a TNF antagonist was not available. We have assumed that the two prior DMARDs would be methotrexate and sulphasalazine and therefore, the most relevant comparator would be either combination therapy of methotrexate and ciclosporin or leflunomide. Data for these treatments in PsA patients come from two trials, but are limited in their outcome measurement, reporting and, in the case of leflunomide, comparability to the etanercept patient group [20, 22] (Table 1).
Withdrawal from etanercept was considered to be subject to two constraints. First, the provisional BSR guidelines specify that if a patient has not achieved a PsA response criteria (PsARC) at 3 months, then they should be withdrawn from treatment. In the Mease et al. study, 71% of etanercept patients achieved the PsARC. Second, the long-term annual withdrawal rate for etanercept patients due to loss of efficacy or adverse events was assumed to be equivalent to that seen in patients with RA [23]. PsARC response rates for ciclosporin and leflunomide are not reported at 3 months, but it is expected that patients on these treatments that do not meet a PsARC response but show signs of benefit would still continue therapy. Therefore, long-term annual withdrawal rates from the literature of 34 and 42% were used [20, 24].
After withdrawal from etanercept, a patient is assumed to receive either combination therapy with methotrexate and ciclosporin or leflunomide. If this treatment fails, then we assume that there are no further effective therapies so a strategy of best standard care is introduced (Fig. 1).
|
Estimating benefit
We predicted the patients HAQ-DI over 3-month time intervals for up to 10 yrs. Patient-level data from the etanercept clinical trial were used to determine the improvement in HAQ-DI over the first 3 months. We assumed that the initial benefit achieved if using traditional DMARD such as ciclosporin would be equal to the effect seen in patients using methotrexate in the placebo arm of the clinical trial. Multivariate regression models were used to predict 4 and 12 weeks HAQ changes using treatment arm, type of response and other clinical and demographic variables as explanatory variables. Whilst the statistical significance of variables in prediction models is not of primary importance, we used stepwise elimination to limit the number of variables in the final models. The models allowed us to analyse the effect of treatment, response and clinical and demographic variables on the magnitude of HAQ-DI change. For leflunomide, where suitable evidence exists, we additionally inflate HAQ-DI improvement by adjusting for PsARC response.
Beyond 3 months, the long-term progression on HAQ-DI whilst remaining on etanercept was estimated using data on patients that continued onto open label extension of the clinical trial [n = 88 (87%) at 48 weeks follow-up] [25]. When a patient withdraws from a treatment, they were assumed to worsen instantaneously by the same magnitude as they initially improved. This ‘rebound’ assumption is based on data from RA patients used in a previous economic model of etanercept [4]. Evidence on progression for patients on best standard care came from using models fitted to the long-term HAQ-DI data from Leeds.
The relationship between health state utility and the HAQ-DI was examined by fitting linear regression models estimated by generalized estimating equation (GEE) algorithms to the Leeds dataset. We also explored the effect of the skin component of PsA on perceived health status by adding the Psoriasis Area and Severity Index (PASI) to the statistical model.
Whilst no differential mortality risk was assumed for the alternative treatments, a natural incidence of mortality would be expected. Age- and sex-related life tables were adjusted by a standardized mortality ratio of 1.60 for women and 1.66 for men [26].
Estimating resource use
We examined all direct costs attributable to patients with PsA. The cost of etanercept in the UK was calculated to be £9296 per annum as a monotherapy. This compares with an annual cost of £2114 for a combination of ciclosporin and methotrexate and £1132 for leflunomide. Monitoring (laboratory tests) and administration (mode of injection and infusions) costs were calculated in line with BSR guidelines using published unit costs (www.emims.net). The total treatment costs are larger in the first 3 months due to the additional monitoring costs at the start of the treatment (Table 2).
|
We also used a relationship between HAQ-DI and other direct costs such as hospitalizations to estimate the cost offsets by improving disability. A study in patients with RA was used to populate a linear regression [3] (annual direct cost = £358 x HAQ + £1182, R2 = 0.37).
Analysis and model development
An individual sampling model was used to simulate the costs and benefits of 1000 hypothetical patients over a 10-yr time horizon [27]. This is a similar technique to that used in models for RA [4, 5, 28]. The patient baseline characteristics is sampled from the demographics of the Mease et al. study (e.g. age = 47, HAQ-DI = 1.1, 49% of the patients on concomitant MTX). At each 3 monthly node, a decision about whether a patient remains on treatment is sampled using a randomly generated number drawn from a probability of an event occurring (i.e. if the probability a patient responds to etanercept is 72% and if the random number is between 0 and 72 then they will be classed as a responder and the therapy continued, but if the random number is between 73 and 100 they will be classed as a non-responder and withdrawn from treatment). In line with published guidelines, we discounted the costs and benefits 3.5% per annum [29].
Sensitivity analyses are performed because there is always uncertainty in decision analysis. They ensure that our results are applicable throughout a wide range of variability in different patient groups. In our analysis, we used 2 methods to characterize this uncertainty: one-way sensitivity analysis and Monte Carlo simulation. One-way sensitivity analysis is a procedure in which each variable in the model is varied in succession between different values. The model results are then recalculated with the new inputs and recorded. The process is repeated for all central variables in the model. One-way sensitivity analyses alter only a single model input at a time and inform readers as to which variables are most important in the model, but do not provide information on the overall uncertainty associated with the model.
To explore the overall uncertainty, we used a Monte Carlo simulation to create a probabilistic sensitivity analysis, which is a method of varying all of the decision variables simultaneously to assess the overall variability in the model [30]. With this method, the probability distributions associated with model variables are defined. For a single Monte Carlo simulation run, values for each parameter are randomly selected and a new cost-effectiveness ratio is calculated. This process is repeated a pre-determined number of times, and the results are used to illustrate or explore the model variability. Table 2 describes all the input parameters and probability distributions.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Health state utility
Table 3 shows the results from the multivariate regressions predicting the HAQ-DI at 4 and 12 weeks. As expected, the HAQ-DI at baseline had the highest predictive value for HAQ-DI at both 4 and 12 weeks. The patient's age, whether the patient had received etanercept and whether a PsARC response was achieved were also predictive. In the 12-week analysis, disease duration, whether the patient had polyarthritis and whether the patient was taking concomitant methotrexate were found to be additional predictive variables. The explanatory power of the models indicates that the covariates explain the HAQ-DI at the different time points well (adjusted R2 of 0.68 and 0.63 at 4 and 12 weeks, respctively).
|
The average EQ5D utility in the Leeds sample (total observations = 202) was 0.52 (range–0.48–1) whilst HAQ-DI was similar to the baseline values in the Mease et al. study (mean = 1.02, range 0–0.275). The derived relationship between HAQ-DI and EQ-5D utility is shown in Fig. 2. This was used to estimate health state utility values from the modelled HAQ-DI scores (EQ-5D utility = –0.31 x HAQ-DI + 0.82). The effect of PASI was very small and not statistically significant. This may have been due to the limited range in PASI scores in the dataset. Many patients had a PASI of 0 and most had a PASI <4 (PASI ranges from 0–72). Using the 6 month HAQ-DI results, and the relationship between HAQ-DI and health status, we estimated the mean improvement in health state utility from the clinical trial at 6 months to be 0.01 for the placebo arm and 0.14 for the etanercept arm.
|
In the postal questionnaire of the patients who had entered the ciclosporin trial, 24 responses were received (70% of the original 35 patients recruited in Leeds). When these patients finished the trial, their mean HAQ was 1.13 and EQ-5D was 0.49. Some 4.2 yrs later, these had changed to 1.4 and 0.4, an annual change of 0.07 in HAQ-DI and –0.02 in EQ-5D. These figures are comparable with patients with severe RA (HAQ-DI = 0.066) [4]. This annual progression rate was assumed for patients who had failed etanercept, ciclosporin or leflunomide and were on best standard care. To estimate the progression in patients responding to etanercept, open-label data were analysed that suggested that there was no worsening in HAQ-DI. For patients on leflunomide or ciclosporin, a progression rate of 0.028 was assumed, taken from a dataset of 47 patients with PsA measured in the period before TNF antagonists were available [10].
This information was combined in the model along with rates of withdrawal and PsARC response rates. The predicted 10-yr trends in quality of life are shown in Fig. 3. The discounted QALYs for the etanercept arm are estimated to be 0.29, 0.63, 2.71 and 4.49 at 6 months, 1 yr, 5 yrs and 10 yrs, respectively. This compares with 0.24, 0.52, 2.24 and 3.67 for the ciclosporin arm and 0.25, 0.54, 2.34 and 3.84 for the leflunomide arm at the same time points. At 10 yrs, only 13% of the patients originally on etanercept remain on treatment, whilst 97% of the patients remain alive.
|
Costs
The total cost of the etanercept arm over 10 yrs was estimated to be £51 122. Direct costs account for 13% of the total cost of treatment in the etanercept arm at 6 months, and 23% at 10 yrs. For the comparator groups, the total costs were estimated to be £28 010 (ciclosporin), £26 822 (leflunomide) at 10 years. Of this 29% was estimated to be from direct costs at 6 months and 48–49% at 10 yrs.
Cost-effectiveness
In the model, the cost-effectiveness of etanercept improves over time (Table 4). At 6 months, etanercept is predicted to give a patient 0.04 additional QALYs at an additional cost of £3000. This indicates an ICER of close to £70 000 per QALY. By 10 yrs, the benefit increases to 0.65 and 0.82 QALYs vs leflunomide and ciclosporin, respectively, at an incremental cost of between £23 000 (ciclosporin) and £24 000 (leflunomide). This calculates to an ICER of £28 000 per QALY for the comparison against ciclosporin and £38 000 per QALY for the comparison with leflunomide. The ICER improves over time since patients slowly withdraw from etanercept adding smaller costs each year, but the delay in progression increases the QALYs gained during this period.
|
Sensitivity analysis
By using regression equations, the patient characteristics can be adjusted in a sensitivity analysis to determine the impact on long-term costs and QALYs. The one-way sensitivity analyses explored the impact of different patient groups on the ICER, along with parameter and structural uncertainties. Table 5 shows the results of this analysis. Of the patient group variables, the ICER was most sensitive to changes to the baseline HAQ. If patients had a mean baseline HAQ of 0.5, the ICER would be 23% higher at £35 000 per QALY. In the key parameters, the annual HAQ progression of biologicals was sensitive as well as the annual HAQ progression of patients on best standard care. Interestingly, when using the 95% confidence intervals of the regression between HAQ-DI and utility, the total number of QALYs at 10 yrs varied between 0.64 and 0.91. The structural assumptions in the model were examined by adjusting the cost and efficacy of the comparator arm, and by using a different management decision for deciding which patients to be removed from the treatment.
|
The results of the 1000 samples in the probabilistic sensitivity analysis are shown in Fig. 4 (results shown are for etanercept vs ciclosporin). Each symbol in each time horizon indicates a separate estimate of the mean ICER having sampled from the parameter distributions. Therefore, the ellipses represent the region in which we believe the true ICER will lie. The 6-month ellipse is small representing the relative certainty we have at this stage due to the data in the Mease et al. trial. All these points lie above the £30 000 per QALY threshold. As we extrapolate the analysis, a portion of the ellipses move under the £30 000 per QALY threshold. However, the size of the ellipses increase reflecting the increase in uncertainty we have in the effectiveness of treatments beyond the 6-month trial data. By 10 yrs, 58% of the symbols lie below the £30 000 per QALY threshold.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
This study is the first attempt to estimate the potential of TNF antagonists in improving the health status of patients with PsA. We also examined the economic implications and derive an ICER, useful for informing priority setting. The conclusions of this analysis fall broadly into two themes: (1) whether TNF antagonists are potentially a cost-effective use of resources and (2) the paucity of data for making judgements on new interventions for treating PsA. We examine these issues in turn.
TNF antagonists like etanercept are more expensive than comparator non-biological treatments and so it is important to understand whether this additional cost is an efficient use of health resources. We estimated how a TNF antagonist might improve a patient's health status in comparison with more conventional DMARD treatments. If a short-term perspective is adopted, it would appear difficult to justify these additional costs. However, if TNF antagonists can prevent the progression of disease, then a long-term perspective is necessary to appreciate the full benefits of the treatment. Using this perspective, we estimate the cost-effectiveness of etanercept to be between £28 000 and £38 000 per QALY. The other TNF antagonists infliximab and adalimumab have recently reported the results of clinical trials [31, 32]. If these results demonstrate similar benefits to etanercept in terms of response measures and improvements in disability then the comparable results will depend on the variation of drug price and dosing patterns.
Whether a treatment with such an ICER is indeed a cost-effective use of resources depends on the perspective of the healthcare payers. Whether a more effective yet more expensive intervention is cost-effective is dependent on the decision makers’ willingness to pay for additional benefit (e.g. QALYs). The value of this threshold should reflect a societal judgement, but in practice this is difficult to quantify. In the UK, the National Institute for Health and Clinical Excellence (NICE) informally uses values of between £20 000 and £30 000 per QALY [29]. Even when a treatment is demonstrably cost-effective, affordability may limit its use.
Our review of the epidemiological and economic literature highlighted that there are limited data available on PsA patients. Many assumptions in our modelling had to use information from studies of RA patients that is not always appropriate, such as the relationship between disease severity and direct costs, which does not include psoriasis-associated costs. Other sources of data came from studies of limited patient numbers with inadequate reporting. Health care agencies such as NICE in the UK, are required to make decisions based on existing data [33]. The objective of this analysis was to provide a synthesis of available evidence, to bring this together within a decision model, and to quantify the uncertainty associated with the decision about the cost-effectiveness of etanercept. The review has highlighted a number of deficiencies that will require further research to enable better decisions regarding the use of these treatments (Table 6).
|
The efficacy of treatments in RCTs does not necessarily reflect real clinical practice, and we made assumptions about the efficacy of treatment beyond a time frame for which data are available. The clinical trials of TNF antagonists focus on the efficacy of treatment whereas we are interested in their effectiveness vs conventional DMARDs such as ciclosporin and leflunomide. Only long-term registries and observational studies will enable us to answer such questions. Until then, we can perform sensitivity analysis to look at the effect of altering parameters to other plausible values on the ICER. For example, we adjust the control arm of the etanercept trial to estimate the comparative effectiveness of ciclosporin and leflunomide. This seems justifiable, as before adjustment for MTX and PsARC response, the improvement in HAQ-DI seen in the Mease et al. placebo arm (9.1%) is similar to that seen in the ciclosporin arm of the Fraser trial (10%) [20], and only 8.5% smaller than the leflunomide arm of the Kaltwasser trial [22].
The sensitivity analysis demonstrated that some of the most important parameters concern the progression of HAQ-DI. Whether the HAQ-DI is the best instrument for measuring progression of disease is not confirmed [34]. Despite this, it was the only instrument for which both short- and long-term data existed. We assumed that the natural progression for patients would depend on whether they were using etanercept, DMARD or on best standard care. We used a small and potentially biased sample of patients recruited using a postal questionnaire. However, this provided the only available evidence. Then, there is the issue of the rebound in disability when treatment is terminated. When a patient is withdrawn from treatment, they are lost to follow-up in most studies, hence we have virtually no understanding of the magnitude or time delay of the relapse. We are perhaps overly pessimistic that it would happen instantaneously, but since the half-life of etanercept is a matter of days we believe this is appropriate.
Other important items could have been included that would have improved the results of etanercept further. For example, in the Mease et al. trial, etanercept was shown to improve the patients' psoriasis, measured by the PASI, in comparison with placebo [19]. This could have an additional impact on both the patients' quality of life, and the drug-related costs associated with psoriasis. To be able to analyse this, information is needed on the multiplicative effect of HAQ-DI and PASI on the quality of life and the relationship between the severity of PASI and medication costs. Importantly, it would require patients with a range in severity of skin disease similar to that in the etanercept trial. Similarly, the benefit of TNF antagonists in improving enthesitis and dactylitis is not included. If we had taken the full societal perspective rather than that of the NHS, the impact on productivity and the quality of life of the primary carer would further improve the results for TNF antagonists.
The sequential or early use of TNF antagonists will inevitably become the standard of care in the future. However, the limited data that we have on the TNF antagonists is in patients with severe disease and relatively long disease duration. Cheap conventional therapies such as methotrexate are more efficacious earlier in the disease, so the comparable efficacy of TNF antagonists will diminish. Cost-effectiveness will depend on whether the treatments can alter the course of radiographic progression in such a way that long-term disability and the need for joint replacements is reduced. The use of prognostic algorithms to identify which patients would most benefit from these treatments would also help the economic argument.
This study demonstrates that etanercept has a benefit on patients' quality of life and describes a process for analysing the long-term benefits and costs of TNF antagonists. Treatment with TNF antagonists is effective albeit costly and policy makers will have to decide whether this group of patients is a priority given a limited budget. A number of limitations are described. Further studies should be designed so that they can help answer not only clinical questions but health economic considerations.
Pete Conway and Alan Reynolds are employees of Wyeth. The other authors have declared no conflicts of interest.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
- Torrance GW. Measurement of health state utilities for economic appraisal. J Health Econ 1986;5:1–30.[CrossRef][Web of Science][Medline]
- Wong JB, Singh G, Kavanaugh A. Estimating the cost-effectiveness of 54 weeks of infliximab for rheumatoid arthritis. Am J Med 2002;113:400–8.[CrossRef][Web of Science][Medline]
- Kobelt G, Jonsson L, Young A, Eberhardt K. The cost-effectiveness of infliximab (remicade) in the treatment of rheumatoid arthritis in Sweden and the United Kingdom based on the ATTRACT study. Rheumatology 2003;42:326–35.
[Abstract/Free Full Text] - Brennan A, Bansback NJ, Reynolds A, Conway P. Modelling the cost-effectiveness of etanercept in adults with rheumatoid arthritis in the UK. Rheumatology 2004;43:62–72.
[Abstract/Free Full Text] - Bansback NJ, Brennan A, Ghatnekar O. Cost effectiveness of adalimumab in the treatment of patients with moderate to severe rheumatoid arthritis in Sweden. Ann Rheum Dis 2005;64:995–1002.
[Abstract/Free Full Text] - Torre Alonso JC, Rodriguez PA, Arribas Castrillo JM, Ballina GJ, Riestra Noriega JL, Lopez LC. Psoriatic arthritis (PA): a clinical, immunological and radiological study of 180 patients. Rheumatology 1991;30:245–50.
[Abstract/Free Full Text] - Lebwohl M. Psoriasis. Lancet 2003;361:1197–204.[CrossRef][Web of Science][Medline]
- Gladman DD.Psoriatic arthritis: recent advances in pathogenesis and treatment. Rheumatic Diseases Clinics of North America 1992;18:247–56.[Web of Science][Medline]
- O’Neill T, Silman AJ. Psoriatic arthritis. Historical background and epidemiology. Baillieres Clin Rheumatol 1994;8:245–61.[CrossRef][Web of Science][Medline]
- Sokoll KB, Helliwell PS. Comparison of disability and quality of life in rheumatoid and psoriatic arthritis. J Rheumatol 2001;28:1842–6.
[Abstract/Free Full Text] - Javitz HS, Ward MM, Farber E, Nail L, Vallow SG. The direct cost of care for psoriasis and psoriatic arthritis in the United States. J Am Acad Dermatol 2002;46:850–60.[CrossRef][Web of Science][Medline]
- Fries JF, Spitz PW, Kraines RG. Measurement of patient outcome in arthritis. Arthritis Rheum 1980;23:137–45.[Web of Science][Medline]
- Wolfe F, Michaud K, Gefeller O, Choi HK. Predicting mortality in patients with rheumatoid arthritis. Arthritis Rheum 2003;48:1530–42.[CrossRef][Web of Science][Medline]
- Michaud K, Messer J, Choi HK, Wolfe F. Direct medical costs and their predictors in patients with rheumatoid arthritis: a three-year study of 7,527 patients. Arthritis Rheum 2003;48:2750–62.[CrossRef][Web of Science][Medline]
- Tijhuis GJ, de Jong Z, Zwinderman AH et al. The validity of the Rheumatoid Arthritis Quality of Life (RAQoL) questionnaire. Rheumatology 2001;40:1112–9.
[Abstract/Free Full Text] - Blackmore MG, Gladman DD, Husted J, Long JA, Farewell VT. Measuring health status in psoriatic arthritis: the Health Assessment Questionnaire and its modification. J Rheumatol 1995;22:886–93.[Web of Science][Medline]
- Hurst NP, Kind P, Ruta D, Hunter M, Stubbings A. Measuring health-related quality of life in rheumatoid arthritis: validity, responsiveness and reliability of EuroQol (EQ-5D). Rheumatol 1997;36:551–9.
[Abstract/Free Full Text] - Buxton MJ, Drummond MF, Van Hout BA et al. Modelling in economic evaluation: an unavoidable fact of life. Health Econ 1998;7:741–2.[CrossRef][Web of Science][Medline]
- Mease PJ, Kivitz AJ, Burch FX et al. Etanercept treatment of psoriatic arthritis: safety, efficacy, and effect on disease progression. Arthritis Rheum 2004;50:2264–72.[CrossRef][Web of Science][Medline]
- Fraser AD, van Kuijk AW, Westhovens R et al. A randomised, double blind, placebo controlled, multicentre trial of combination therapy with methotrexate plus ciclosporin in patients with active psoriatic arthritis. Ann Rheum Dis 2005;64:859–64.
[Abstract/Free Full Text] - Kyle S, Chandler D, Griffiths CE et al. Guideline for anti-TNF-alpha therapy in psoriatic arthritis. Rheumatology 2005;44:390–7.
[Free Full Text] - Kaltwasser JP, Nash J, Gladman D et al. Efficacy and safety of leflunomide in the treatment of psoriatic arthritis and psoriasis: a multinational, double-blind, randomized, placebo-controlled clinical trial. Arthritis Rheum 2004;50:1939–50.[CrossRef][Web of Science][Medline]
- Crnkic M, Petersson I, Saxne T, Geborek P. Infiximab, etanercept and leflunomide in rheumatoid arthritis. Clinical experience in southern Sweden. Rheumatology 2001;40:S82
- Aletaha D, Stamm T, Kapral T et al. Survival and effectiveness of leflunomide compared with methotrexate and sulfasalazine in rheumatoid arthritis: a matched observational study. Ann Rheum Dis 2003;62:944–51.
[Abstract/Free Full Text] - Mease PJ, Gottlieb AB, Yu EB, Dunn M, Woolley JM. Psoriatic Arthritis patients treated with etanercept experience sustained improvements in physical function. Ann Rheum Dis 2005;64(Suppl 3):322.
- Wong K, Gladman DD, Husted J, Long JA, Farewell VT. Mortality studies in psoriatic arthritis: results from a single outpatient clinic. I. Causes and risk of death. Arthritis Rheum 1997;40:1868–72.[Web of Science][Medline]
- Barton P, Bryan S, Robinson S. Modelling in the economic evaluation of health care: selecting the appropriate approach. J Health Serv Res Policy 2004;9:110–8.
[Abstract/Free Full Text] - Jobanputra P, Barton P, Bryan S, Burls A. The effectiveness of infliximab and etanercept for the treatment of rheumatoid arthritis: a systematic review and economic evaluation. Health Technol Assess 2002;6:1–110.[Medline]
- National Institute for Clinical Excellence. Guide to the Methods of Technology Appraisal. (2004). Reference NO515. Available at http://www.nice.org.uk/pdf/TAP_Methods.pdf
- Briggs AH. Handling uncertainty in cost-effectiveness models. Pharmacoeconomics 2000;17:479–500.[CrossRef][Web of Science][Medline]
- Antoni C, Krueger GG, de Vlam K, Birbara et al. Infliximab improves signs and symptoms of psoriatic arthritis: results of the IMPACT 2 trial. Ann Rheum Dis 2005;64:1150–7.
[Abstract/Free Full Text] - Mease PJ, Gladman DD, Ritchlin CT et al. Adalimumab Effectiveness in Psoriatic Arthritis Trial Study Group. Adalimumab for the treatment of patients with moderately to severely active psoriatic arthritis: results of a double-blind, randomized, placebo-controlled trial. Arthritis Rheum 2005;52:3279–89.[CrossRef][Web of Science][Medline]
- Symmons DPM. Anti-tumour necrosis factor
therapy: can we afford it?. Ann Rheum Dis 2005;64:969–70.[Free Full Text] - Husted JA, Gladman DD, Cook RJ, Farewell VT. Responsiveness of health status instruments to changes in articular status and perceived health in patients with psoriatic arthritis. J Rheumatol 1998;25:2146–55.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. BARRA, J. E. POPE, and M. PAYNE Real-World Anti-Tumor Necrosis Factor Treatment in Rheumatoid Arthritis, Psoriatic Arthritis, and Ankylosing Spondylitis: Cost-Effectiveness Based on Number Needed to Treat to Improve Health Assessment Questionnaire J Rheumatol, July 1, 2009; 36(7): 1421 - 1428. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Olivieri, S. de Portu, C. Salvarani, A. Cauli, E. Lubrano, A. Spadaro, F. Cantini, M. S. Cutro, A. Mathieu, M. Matucci-Cerinic, et al. The psoriatic arthritis cost evaluation study: a cost-of-illness study on tumour necrosis factor inhibitors in psoriatic arthritis patients with inadequate response to conventional therapy Rheumatology, November 1, 2008; 47(11): 1664 - 1670. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Bravo Vergel, N. S. Hawkins, K. Claxton, C. Asseburg, S. Palmer, N. Woolacott, I. N. Bruce, and M. J. Sculpher The cost-effectiveness of etanercept and infliximab for the treatment of patients with psoriatic arthritis Rheumatology, November 1, 2007; 46(11): 1729 - 1735. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D L Scott, B Khoshaba, E H Choy, and G H Kingsley Limited correlation between the Health Assessment Questionnaire (HAQ) and EuroQol in rheumatoid arthritis: questionable validity of deriving quality adjusted life years from HAQ Ann Rheum Dis, November 1, 2007; 66(11): 1534 - 1537. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Ara, A. V. Reynolds, and P. Conway The cost-effectiveness of etanercept in patients with severe ankylosing spondylitis in the UK Rheumatology, August 1, 2007; 46(8): 1338 - 1344. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||