Cost-Effectiveness of Salmeterol Xinafoate/Fluticasone Propionate Combination Inhaler in Chronic Asthma

By Doull, Iolo Price, David; Thomas, Mike; Hawkins, Neil; Et al

Key words: Asthma – Cost-effectiveness – Fluticasone/salmeterol – Quality of life ABSTRACT

Objective: To determine where in the treatment steps recommended by the British Thoracic Society and Scottish Intercollegiate Guidelines Network (BTS/SIGN) Asthma Guideline it is cost-effective to use salmeterol xinafoate/fluticasone propionate combination inhaler (SFC) (Seretide*) compared with other inhaled corticosteroid (ICS) containing regimens (with and without a long acting beta-2 agonist (LABA)) for chronic asthma in adults and children.

Research design and methods: Meta-analyses of percentage symptom- free days (%SFD) were used within a cost-effectiveness model. Time spent in two asthma control health states, ‘symptom-free’ and ‘with- symptoms’ was used as the measure of differential treatment effectiveness. SFC was compared with varying doses of fluticasone propionate (FP) and beclometasone dipropionate (BDP) with or without a separate salmeterol inhaler, and with the budesonide/formoterol combination inhaler (BUD/FORM) (Symbicort[dagger]). Drug costs, nondrug costs and quality adjusted life years (QALY) were incorporated into the analyses. Results are presented as cost per QALY ratios and uncertainty explored using probabilistic sensitivity analysis.

Results: Compared with an increased dose of FP in adults, SFC either ‘dominates’ (i.e. cheaper and more effective) FP or the cost per QALY is Pounds 6852. The cost per QALYs estimated in sensitivity analyses using BDP costs range from Pounds 5679 to Pounds 15997. For children the cost per QALY for SFC 50 Evohaler* compared with an increased dose of FP is Pounds 15739. SFC is similarly clinically effective in improving %SFDs compared with FP plus salmeterol delivered in separate inhalers (mean differences for each dose comparison of -3.9 (low) (with a 95% confidence interval (CI): – 12.96; 5.16); 4.10 (medium) (95% Cl: -3.01; 11.21); -0.4 (high) (95% Cl: -8.88; 8.08)) and BUD/FORM (mean difference of 0.40 (95% Cl – 3.69; 4.49)) in adults, and a cheaper SFC option is available at all doses (annual cost savings range from Pounds 18-Pounds 427 per patient). SFC was similarly effective compared with FP plus salmeterol in separate inhalers in children under 12 and also resulted in annual cost savings of between Pounds 47 and Pounds 77. A number of other comparisons were also made and the results are available as electronic supplementary data.

Conclusions: This is the first analysis to estimate the cost- effectiveness of SFC in chronic asthma compared with multiple comparators and based on a systematic identification of relevant trials and data on %SFDs. The findings suggest that for adults and children uncontrolled on BDP 400 [mu]g/day or equivalent It Is a cost-effective option to switch to SFC (at an equivalent ICS dose) compared with increasing the dose of ICS. For adults and children aged 12 years and over who have passed this point and are uncontrolled on BDP 800[mu]g/day or equivalent, switching to SFC remains a cost-effective approach. Where an adult or child requires an ICS and a LABA to be co-prescribed, SFC is a cost-effective option compared with FP or BDP plus salmeterol delivered in separate inhalers. In adults who require combination therapy, SFC is a cost- effective option compared with BUD/FORM.

Introduction

Asthma is characterised by inflammation of the airways and by bronchoconstriction causing wheezing and shortness of breath. It is one of the most common chronic diseases, affecting an estimated 300 million adults and children throughout the world’. In England and Wales it is estimated that 3.7 million adults and 1 million children (under the age of 16 years) have asthma, causing considerable impact on health care resources2. Overall, health care costs for asthma patients have been found to be twice as high as for patients without asthma3. Survey and prospective observational data estimate that poorly controlled asthma patients cost at least three times that of wellcontrolled patients4-7.

The British Thoracic Society and Scottish Intercollegiate Guidelines Network (BTS/SIGN) Guideline8 describes the treatment of asthma as a series of steps chosen dependent on disease severity and response to current treatment. In adults and children aged 5 and over, step 1 is the intermittent use of an inhaled short-acting beta- 2 agonist (SABA) as necessary, step 2 is the addition of an inhaled corticosteroid (ICS) at 200-800 [mu]g/day beclometasone diproprionate (BDP) or equivalent (200-400 [mu]g/day for children 5- 12) and step 3 is the addition of an inhaled long-acting beta-2 agonist (LABA).

A combination inhaler containing salmeterol xinafoate (salmeterol), a LABA, and fluticasone propionate (FP), a corticosteroid, (SFC) (Seretide*) is licensed in the UK for the regular treatment of asthma where use of a combination product is appropriate, i.e. in patients not adequately controlled with ICSs and ‘as needed’ inhaled SABA or in those already adequately controlled on both ICS and LABA administered as separate devices.

Two important questions regarding the use of LABAs are left unanswered within the BTS/SIGN Guideline. Specifically, at what dose of ICS should a LABA be added to therapy and should it be added as a separate inhaler device or as a component of a combination device? This paper addresses these questions through a series of cost- effectiveness analyses for both adults and adolescents (aged 12 and over) and children (under 12). The analyses presented here differ from previously published cost-effectiveness analyses developed by Paltiel et al.9 and Price and Briggs10, as they are based on a systematic identification of a range of clinical trial data to address the cost-effectiveness of treatment options within a stepwise treatment pathway.

Methods

The question of when to add a LABA to ICS therapy was addressed by a series of cost-effectiveness analyses of SFC introduced at different points in the dose escalation of inhaled steroids. Four comparisons were evaluated using dose bands and ICS dose equivalences in line with current guidelines1’8 and reviews of the literature11 (see Table 1):

1. For patients failing to achieve adequate disease control on their current dose of ICS alone, SFC is compared with the current dose of FP. As patients might improve spontaneously due to the episodic nature of asthma or adhere better with background therapy, it is important to compare SFC with optimising current treatment.

2. For patients failing to achieve adequate disease control on their current dose of ICS alone, SFC is compared with an alternative treatment strategy of increasing the dose of FP.

3. For those patients requiring a LABA, SFC is compared with FP and salmeterol delivered in separate inhalers.

4. For those patients requiring a combination inhaler, SFC is compared with a combination inhaler containing budesonide and formoterol (BUD/FORM) (Symbicort, AstraZeneca).

Model overview

To estimate cost-effectiveness, a model with two asthma control health states was developed where patients were considered to be in either a ‘symptomfree’ or ‘with-symptoms’ state. Differential treatment effectiveness was incorporated based on a series of meta- analyses of the percentage of symptom-free days (%SFD) endpoint following a systematic review of clinical trials. This endpoint was selected as it directly reflects the patient’s experience of the condition and was widely reported. It was not possible to abstract comparable information on exacerbations across trials; therefore a distinct exacerbation state was not included.

Table 1. Comparisons for cost-effectiveness analyses of SFC introduced at different points in the dose escalation of inhaled steroids (ICS)

A number of key assumptions were made in building the model. As there is no curative treatment, ‘the aims of pharmacological management of asthma are the control of symptoms…, prevention of exacerbations and the achievement of best possible pulmonary function, with minimal side-effects’8. It was also assumed that treatments have no differential effects on mortality or toxicity12 and that the trial-based estimates are applicable to wider patient populations, such that the differential proportion of time patients spend in the ‘symptom-free’ state over their treatment lifetime would be the same as that observed during the trial period. There are two SFC devices, Accuhaler (GlaxoSmithKline) and Evohaler (GlaxoSmithKline), and both were used in the meta-analysis, thus assuming no differences in effectiveness13,14.

The estimated mean difference in %SFD between the treatments from the meta-analysis was taken to represent the difference in proportion of time spent in the ‘symptom-free’ compared with the ‘withsymptoms” state in the model. Therefore, definitions of ‘symptom-free’ used to estimate %SFDs were assumed to be consistent with the ‘symptom-free’ state defined using the health states in the Gaining Optimal Asthma controL (GOAL) trial15.

The difference between treatment arms in Other health service” costs (i.e., unscheduled health service costs other than the study therapies themselves) was estimated by multiplying the mean difference in %SFD by the difference between the two model states (‘symptom-free’ and ‘with-symptoms’) in annual Other health service’ costs. The incremental QALYssed to estimate costs and utilities. This 1-year, stratified, randomised, double-blind, parallel-group study of 3416 patients with uncontrolled asthma compared SFC with three doses of FP alone (200, 500 and 1000pg/day) in achieving two guideline-based measures of asthma control: totally- and well- controlled. These two measures are defined in electronic supplementary data (see Appendix). The modelled time horizon was nominally 1 year, corresponding to the duration of the GOAL trial. The 1 year duration made the GOAL trial particularly appropriate for this task as it ensured that costs and utilities estimates were not affected by any seasonal variation in the observed data. Within the model the proportion of time spent in the two model states was assumed to be constant over time and so costs and benefits accrue in a constant proportion. The estimated cost-effectiveness ratio would therefore be independent of the model time horizon and discounting was therefore not required. All costs are reported as 2006 Great Britain pounds (Pounds ). The R 2.2.1 statistical package16 was used initially to perform the cost-effectiveness analysis and as a validity test the analysis was fully replicated using a model built in Microsoft Excel.

Meta-analysis of clinical data

An extended search of available literature was undertaken on 22 February 2006, with the aim to identify all the relevant studies addressing the research questions. The databases searched and the search strategy used can be found as electronic supplementary data (see Appendix). A total of 740 records were retrieved from this process. An additional search was conducted in the GSK clinical trial register17 and other internal records to identify any trials for which no publications currently exist and to ensure identification of all the relevant publications for each trial.

Records were hand searched to fulfil the following inclusion criteria for the meta-analysis:

* Randomised controlled trials (RCTs)

and

* Patients aged 4 and over with chronic asthma (in line with SFC licence)

* %SFD data with standard deviation, standard error or percentiles reported

* Comparisons of SFC (in line with doses listed in Table 1) with:

_ FP alone (same and increased dose)

_ FP in combination with salmeterol, or

_ BUD/FORM

* Studies published in English language

In addition to estimating the cost-effectiveness of SFC against FP, estimates were also generated for BDP (in ICS alone and separate inhaler comparisons) and, where possible, budesonide (within BUD/ FORM) comparisons, assuming equal efficacy when using BDP or budesonide at double the FP dose”. As BDP and FP are not licensed above 400 [mu]g/day for children, only a low-dose comparison was made for this age group. Analyses against BUD/FORM were conducted using 100/6, 200/6 and 400/12 inhalers. The comparison of SFC with BUD/FORM was made on a fixed and equivalent dose basis, to allow comparison of efficacy rather than dosing strategy (Table 2). Therefore, the trials comparing SFC with BUD/FORM delivered using different dosing strategies were excluded18-20.

Table 2. Included studies in the meta-analysis of trial data

The meta-analysis of %SFD was conducted using the R 2.2.1 meta package. Both random and fixed treatment effect estimates were obtained.

Treatment acquisition costs

Annual costs for SFC were based on the unit prices for each inhaler device (Accuhaler and Evohaler). For FP and BDP, costs were input as average prices, calculated by weighting all licensed preparations capable of delivering the appropriate dose. For all steroids a reasonable number of puffs per day were assumed (two puffs twice daily (b.d.) for aerosols, one puff b.d. for dry powder inhalers). This was not possible for BDP2000 so the cost of BDP1000 was doubled. For children, licensed non-breath actuated metered- dose inhalers (MDIs) were assumed to be administered via a spacer device as recommended21’22. All calculations were based on July 2006 unit prices and were obtained from three sections of the online Drug Tariff3 with Part viii having priority over the ‘generics’ list which had priority over the ‘prescribing costs’ list. Prices unavailable from the Drug Tariff were taken from eMIMS24.

Estimation of utilities and other health service costs

The mean utilities and annual non-drug costs associated with the ‘symptom-free’ and ‘with-symptoms’ states were estimated using regression analyses of individual patient data relating control state to non-drug costs and utilities from the GOAL15 clinical trial. Compared with other potential data sources, the GOAL trial could provide non-drug cost and utility data that was most suited to the model design. In the economic analysis of the GOAL study25 patients’ asthma control during each week of the study was described as ‘totally-controlled’, ‘well-controlled’, ‘not wellcontrolled’ or ‘exacerbation’. For this analysis of the costs and utilities associated with the model states, patients in the ‘totally- controlled’ state, as defined in the GOAL study, were regarded as being in the ‘symptom-free’ state and patients in the other states were grouped together as the ‘with-symptoms’ state. Costs and utilities estimated for the ‘with-symptoms’ state would therefore represent a patient-week weighted average across ‘well-controlled’, ‘not wellcontrolled’ and ‘exacerbation’ states from the GOAL analysis.

Utility data were not directly collected in the GOAL study. A mapping algorithm derived from an external study in which both EQ5D utility scores and the asthma quality of life questionnaire (AQLQ) data were collected was used to calculate a utility score for patients in the GOAL study based on their AQLQ scores25. Similar mapping exercises have been carried out, although insufficient details were available to allow alternative utility estimates to be used in the model26. It was assumed in this study that the utility scores mapped from the AQLQ are representative of those that would have been obtained directly from patients in the two asthma control states in GOAL using the EQ5D.

Unscheduled non-drug Other health service’ costs were estimated from individual patient data on resource use collected in the GOAL study under three main categories:

* Secondary care visits: visits to emergency departments, length of stay (number of days) in intensive care, inpatient days, and outpatient visits

* Primary care visits: day and night general practitioner home visits, surgery visits, and telephone calls to primary care clinic

* Rescue medication use (per occasion cost)

Unit costs relating to each of these resources were taken from published sources for England and Wales, updating those used in the original GOAL analysis25 (Table 3).

Linear regression analyses were undertaken using weekly Other health service’ costs and utilities as the dependent variable and a dummy variable for the ‘symptom-free’ and ‘with-symptoms’ asthma control states, as the independent variable. A UK indicator variable was also included to adjust for differences in utility between the UK and other countries found in the GOAL analysis25, although the corresponding parameter was found not to be statistically significant in the cost regression, and therefore was not included in the regression analysis. Eicker-White robust standard errors were estimated, which take account of the repeated (weekly) measures at the patient level. The regression was conducted using the reg procedure and cluster option from the STATA software.

Model evaluation, decision rules and uncertainty

Standard cost-effectiveness decision rules were employed27. Uncertainty in the cost-effectiveness estimates was analysed probabilistically using Monte Carlo simulation. The mean treatment effects were sampled from a normal distribution; the Other health service’ cost and utility regression coefficients were estimated from a multivariate normal distribution. The results presented are based on the mean values from the distribution of Other’ costs and QALY results. Treatment costs were regarded as fixed. The uncertainty surrounding each of the cost-effectiveness estimates is expressed in terms of the probability of SFC being more cost- effective than its comparator against cost-effectiveness thresholds28. Three incremental cost-effectiveness ratio (ICER) thresholds were used – Pounds 0 (equivalent to presenting the probability that SFC is less costly), Pounds 20000 and Pounds 30000 per QALY gained. The latter two thresholds are indicated by the National Institute for Health and Clinical Excellence (NICE)29. Above Pounds 20 000/QALY factors such as decision uncertainty, innovation, and the nature of the condition are considered alongside cost-effectiveness. Above an ICER of Pounds 30 000/QALY the case for these factors needs to be increasingly strong.

Results

State-specific utilities and Other health service’ costs

The results of the regression analyses are shown in Table 4. The ‘with-symptoms’ state is associated with Pounds 78.26 higher annual health service costs and 0.12 lower utility per patient compared with the ‘symptomfree’ state.

Table 3. Unit costs of resources

Table 4. Other health service costs and utilities by model state (Pounds )

Meta-analysis of trial data

Fourteen studies retrieved by the search fulfilled the criteria for inclusion in the meta-analysis. A detailed description of these studies is presented in Table 2 (information on trials included for SFC versus the same dose of FP are available as electronic supplementary data (see Appendix)). A large number of trials were excluded as they did not include the %SFD endpoint or sufficient information required for the meta-analysis, or were not of the treatment comparisons of interest. Fixed effects estimates are presented and were used in the cost-effectiveness models. Using criteria described in the Cochrane Handbook30 there was no evidence of significant heterogeneity in all analyses where three or melow 50% and the chi^sup 2^ tests for heterogeneity did not reach significance at the p = 0.1 level. Table 5 shows that in adults, SFC results in a statistically significant higher %SFDs compared with an increased dose of FP. In paediatric trials, the differences are numerically higher with SFC (see Table 6) but not statistically significant at the 5% level for these comparisons. The differences between SFC and either FP plus salmeterol delivered in separate inhalers or BUD/FORM are small and variable and not statistically significant. This would suggest that they are clinically equivalent.

Table 5. Cost-effectiveness results for SFC in adults and children aged 12 and over

Table 6. Cost-effectiveness results for SFC in children under 12

Cost-effectiveness of SFC vs. increased dose FP alone in patients aged 12 and over

In Table 5 the cost-effectiveness of switching uncontrolled patients to SFC compared with the alternative treatment strategy of increasing the dose of FP is shown. As a result of the higher cost of increased doses of FP and lower effectiveness, SFC is shown to be cost-effective at both dose options, either cheaper and more effective than FP alone (referred to in health economics as ‘dominant’), or with an ICER of Pounds 6852 and therefore well below a Pounds 30000 threshold. Results for SFC versus the same dose of FP and other ICS doses can be found as electronic supplementary data (see Appendix).

Cost-effectiveness of SFC vs. separate FP plus salmeterol in patients aged 12 and over

The cost-effectiveness of SFC compared with the use of separate FP plus salmeterol is shown in Table 5. SFC is generally cheaper, with all of the six cost differences a saving (ranging from -Pounds 40 to -Pounds 303 per year); however, even though the differences in clinical effectiveness are small and uncertain, they were treated as ‘real’ differences in the analysis. As a result, the direction of differences in QALYs between the options varies, which means that the ICERs switch around in terms of whether they relate to SFC or comparator. Since the ICER calculation is highly sensitive to the magnitude of the denominator, two of the comparisons (low and high dose) result in an ICER for FP plus salmeterol as separate inhalers, since this is the (marginally) more effective and costly therapy. However, the resultant ICERs are appreciably higher than a Pounds 30000 costeffectiveness threshold with the exception of the comparison with the Accuhaler at the low dose. At the medium dose SFC is cheaper (annual saving of Pounds 175 per patient) and more clinically effective.

Cost-effectiveness of SFC vs. BUD/FORM in patients aged 12 and over

For the comparison of SFC and BUD/FORM, only one analysis could be informed by appropriate trial evidence (see Table 5). This is for medium dose SFC compared to BUD/FORM 200/6 for which SFC is shown to be a dominant treatment in cost-effectiveness terms. However, even though SFC is Pounds 18 cheaper per patient per year, the difference in %SFDs is small and statistically insignificant (0.40, 95% CI: – 3.69; 4.49). Where no trial evidence existed to conduct a full analysis, the costs of BUD/FORM and SFC were compared as a partial economic evaluation (see Table 7). In all but one of the eleven comparisons SFC is cheaper than BUD/FORM with annual cost savings ranging from Pounds 18-Pounds 427 per patient.

Table 7. Cost differences: SFC compared with BUD/FORM

Cost-effectiveness of SFC in children under 12

In children under 12, appropriate effectiveness data were available for all comparisons except versus BUD/FORM (data for SFC versus the same dose of FP are available as electronic supplementary data (see Appendix)). The trial data shown in Table 6 show that SFC has similar efficacy to both increased dose FP and FP plus salmeterol. For the comparison of low-dose SFC with an increased dose of FP (400) shown in Table 6, the cost per QALY for SFC Evohaler is well below the Pounds 30000 threshold (Pounds 15 739), but the Accuhaler ICER is above the threshold (Pounds 63 736). For the comparison of low-dose SFC against its components delivered in separate inhalers, SFC is cheaper and is the dominant option (Table 6).

Cost-effectiveness of SFC vs. BDP

Alternative model scenarios are also presented in Tables 5 and 6 showing the implications of using the acquisition cost of BDP rather than FP (at a clinically equivalent dose). These results show that SFC remains cost-effective for most comparisons. Exceptions for adults are the comparison of SFC with BDP plus salmeterol delivered in separate inhalers at the low and high doses, where the trials suggest there is little difference clinically between the options and at low dose both SFC devices are cheaper than the BDP option. However, in each case there is a more costeffective SFC device available. In children under 12, SFC is still dominant compared with BDP and salmeterol delivered in separate inhalers. There is no comparison of SFC with an increased dose of BDP (800pg/day), as it is not licensed in children.

Discussion

This study was conducted to determine where in the BTS/SIGN Asthma Guideline it is cost-effective to use the SFC combination inhaler in the treatment of chronic asthma in adults and children. This analysis is the first to undertake a systematic identification of the relevant trials and data on %SFDs, building a decision model to address questions relating to the cost-effectiveness of using inhaled LABA and ICS combinations in the treatment of chronic asthma. The results of this study indicate that in adults, SFC is a cost-effective alternative to increasing the dose of either FP or BDP. For children the cost per QALY for SFC 50 Evohaler compared with the increased dose of FP is well below the Pounds 30000 threshold. SFC is similarly clinically effective in improving %SFDs compared with FP plus salmeterol delivered in separate inhalers and BUD/FORM in adults, and a cheaper SFC option is available at all doses.

The main weakness of the analysis is that for a few of the comparisons examined, there was a lack of clinical trials that met the inclusion criteria for the metaanalysis and that therefore were able to support a costeffectiveness analysis. There is also little available data on the long-term consequences of poor asthma control and treatment outcomes, and a lack of consistent data allowing any differences in exacerbation rates to be compared across the treatments. Whilst randomised controlled trials are the gold standard in demonstrating clinical efficacy, the limited generalisability of these study designs could also be a weakness of this analysis31,32. Another limitation of the study was that EQ 5 D utility scores were not available directly from patients in each of the two asthma control states in the GOAL trial, and instead a mapping algorithm was used to translate AQLQ scores to EQ5D utility scores.

In the paediatric analyses few studies were available to make comparisons. This could be due to the difficulties of recruiting children into clinical trials. In addition, asthma is more variable and heterogeneous in children than in adults33, which may result in less severe patients being recruited into the trials than expected, and subsequently a reduced likelihood of finding clinically meaningful differences between the therapies being tested. An absence of specific data in children under 12 also required the use of the same utility and resource use data as in the adult population. This may underestimate treatment costs associated with more severe asthma states in children, as children may be more likely than adults to use NHS resources for the same level of symptoms. In the model the effect would be to underestimate the cost offsets for the more effective treatment in a comparison.

The model is also conservative with respect to SFC in a number of other ways. In comparing SFC with FP plus salmeterol in separate inhalers, there is evidence that combination inhalers are associated with improved adherence34″38, which may have outcome benefits35’37’39. These benefits may not be seen in double blind, double dummy study designs where all patients use two inhalers. Other benefits of combination inhalers such as ensuring patients take both their ICS and LABA, which is in line with the Medicines and Healthcare products Regulatory Agency and Commission on Human Medicines guidance40, and the potential synergy of both drugs when taken in one inhaler41 are relevant to this analysis also. The analysis also does not include any longer-term benefits related to avoiding routine use of higher dose ICS1.

Another potential limitation is the grouping of GOAL study health states which may over-estimate the average utility and under- estimate the average costs of being in the ‘with-symptoms’ state because this state includes patients who are ‘well-controlled’, and are therefore likely to have reported symptom-free days in the trial. This leads to a more conservative model with respect to the more active treatment.

Conclusions

The BTS/SIGN Asthma Guideline8 does not specify the ICS dose to add in a LABA and therefore at which to use SFC. However, the findings presented in this analysis show that for adults and children uncontrolled on BDP 400pg/day or equivalent it is a cost- effective option to switch to SFC compared with increasing the dose of ICS and therefore this should be the initial preferred therapeutic approach. For adults and children aged 12 and over who have passed this point and are uncontrolled on BDP 800pg/day or equivalent, switching to SFC remains a cost-effective approach. For children under 12, SFC Evohaler may have benefits by delivering similar efficacy at a lower steroid dose than an ICS alone and is a cost-effective approach.

Compared with FP plus salmeterol delivered in separate inhalers, SFC is at least as clinically effective and has other benefits such as adherence. In addition, SFC is genera adult requires an ICS and LABA to be co-prescribed. The most appropriate way to compare SFC and BUD/FORM on a like-for-like basis is to review head-to-head studies comparing equivalent dosing regimens. There was only one study available, but the data suggests that SFC and BUD/FORM achieve similar levels of efficacy. There is, however, a cheaper SFC option at all doses and only SFC is available as an MDI, the device recommended by NICE as the preferred option in the management of asthma in children21,22. Acknowledgements

Declaration of interest: The authors would like to acknowledge Dr Mark Sculpher for his valuable contribution in advising on the design and analysis of this study. Funding for the study was provided by GlaxoSmithKline.

Authorship and contributorship: All authors were involved in the conception, design and data interpretation as well as drafting the article and final approval for publication. In addition, NH was responsible for the statistical analysis of the data, with assistance from ES. TG and MT had lead responsibility for editing and finalising the manuscript.

CrossRef links are available in the online published version of this paper: http://www.cmrojournal.com

Paper CMRO-3813_4, Accepted for publication: 20 March 2007

Published Online: 18 April 2007

doi: 10.1185/030079907X187982

* Seretide, Accuhaler and Evohaler are registered trade marks of the GlaxoSmithKline group of companies

[dagger] Symbicort is a registered trade mark of AstraZeneca AB

* Seretide is the trade mark of the GlaxoSmithKline group of companies

References

1. Global Initiative for Asthma (GINA). Global strategy for asthma management and prevention: NHLBI/WHO Workshop Report. Bethesda: National Institutes of Health. National Heart, Lung and Blood Institute 2002; Publication No. 02-3659

2. Asthma UK. Where do we stand? Asthma in the UK today 2004

3. Lyseng-Williamson KA, Plosker GL. Inhaled salmeterol/ fluticasone propionate combination. Pharmacoeconomics 2003;21:951- 89

4. National Asthma Campaign. Out in the open. A picture of asthma in the UK today 2002

5. Hoskins G, McCowan C, Neville R, et al. Risk factors and costs associated with an asthma attack. PharmacoEconomics and Outcomes News 2000;253:8

6. Vervloet D, Williams A, Lloyd A, Clark T. Costs of managing asthma as defined by a derived Asthma Control TestTM score in seven European countries. Eur Respir Rev 2006; 15:17-23

7. Lloyd A, Price D, Brown R. The impact of asthma exacerbations on health-related quality of life in moderate to severe asthma patients in the UK. Prim Care Respir J 2007;16:22-27

8. British Thoracic Society. British Guideline on the Management of Asthma. Revised edn. November, 2005

9. Paltiel AD, Fuhlbrigge AL, Kitch BT, et al. Cost- effectiveness of inhaled corticosteroids in adults with mild-to- moderate asthma: Results from the Asthma Policy Model. J Allergy Clin Immunol 2001;108:39-46

10. Price MJ, Briggs AH. Development of an economic model to assess the cost effectiveness of asthma management strategies. Pharmacoeconomics 2002;20:183-94

11. Adams N, Bestall JM, Lasserson TJ, Jones PW. Fluticasone versus beclomethasone or budesonide for chronic asthma in adults and children. The Cochrane Database Syst Rev 2005; Issue 3: CD002310.pub3. DOI: 10.1002/14651858.CD002310.pubS

12. Ni Chroinin M, Greenstone H, Danish A, et al. Long-acting beta2-agonists versus placebo in addition to inhaled corticosteroids in children and adults with chronic asthma. Cochrane Database Syst Rev (Online: Update Software) 2007;1

13. Bateman ED, Silins V, Bogolubov M. Clinical equivalence of salmeterol/fluticasone propionate in combination (50/100 mg twice daily) when administered via a chlorofluorocarbon-free metered dose inhaler or dry powder inhaler to patients with mildto-moderate asthma. Respir Med 2001;95:136-46 (SFCB3022)

14. van Noord JA, LiIl H, Carillo Diaz T, et al. Clinical equivalence of a salmeterol/fluticasone propionate combination product (50/500 meg) delivered via a chlorofluorocarbon-free metereddose inhaler with the Diskus in patients with moderate to severe asthma. Clin Drug Invest 2001;21:243-55 (SFCB3023)

15. Bateman ED, Bousehey H, Bousquet J, et al. Can Guidelinedefined asthma control be achieved? The Gaining Optimal Asthma ControL Study. Am J Respir Crit Care Med 2004; 170:836-44 (SAM40027)

16. Hornik K. The R FAQ. http://CRAN.R-project.org/doc/FAQ/R- FAQ.html (accessed July 2006) ISBN 3-900051-08-9 2006

17. GSK. Clinical Trials Register. http://ctr.gsk.co.uk/Summary/ fluticasone_salmeterol/studylist.asp (accessed 30 April 2006) 2006

18. Aalbers R, Backer V, Kava T, et al. Adjustable maintenance dosing with budesonide/formoterol compared with fixed-dose salmeterol/fluticasone in moderate to severe asthma. Curr Med Res Opin 2004; 20:225-40

19. Fitzgerald J, Boulet L, Follows R. The CONCEPT trial: A 1- year, multicenter, randomized, double-blind, double-dummy comparison of a stable dosing regimen of salmeterol/fluticasone propionate with an adjustable maintenance dosing regimen of formoterol/budesonide in adults with persistent asthma. Clin Ther 2005;27:393-406

20. Vogelmeier C, D’Urzo A, Pauwels R, et al. Budesonide/ formoterol maintenance and reliever therapy: an effective asthma treatment option? Eur Respir J 2005;26:819-28

21. National Institute of Health and Clinical Excellence. Technology Appraisal Guidance No 10. Guidance on the use of inhaler systems (devices) in children under the age of 5 years with chronic asthma. http://www.nice.org.uk/download.aspx?o=TA 010guidance&template=download.aspx (accessed July 2006) 2000

22. National Institute of Health and Clinical Excellence. Technology Appraisal Guidance No 38. Inhaler devices for routine treatment of asthma in older children (aged 5-15 years). http:// www. nice.org.uk/page.aspx?o=TA038guidance (accessed July 2006) 2002

23. Drug Tariff, http://www.drugtariff.com (accessed July 2006)

24. eMIMS. http://www.emims.net (accessed July 2006) 2006

25. Briggs AH, Bousquet J, Wallace MV, et al. Cost-effectiveness of asthma control: an economic appraisal of the GOAL study. Allergy 2006;61:531-6

26. Tsuchiya A, Brazier J, McCoIl E, Parkin D. Deriving preferencebased single indices from non-preference based condition- specific instruments: Converting AQLQ into EQ5D indices. Discussion paper 02/1 http://www.shef.ac.Uk/content/1/c6/01/87/47/ DP0201 .pdf (accessed July 2006) 2002

27. Karlsson G, Johannesson M. The decision rules of cost- effectiveness analysis. Pharmacoeconomics 1996;9:120

28. Fenwick E, Claxton K, Sculpher M. Representing uncertainty: the role of cost-effectiveness acceptability curves. Health Econ 2001; 10:779-87

29. National Institute of Health and Clinical Excellence. Guide to the methods of technology appraisal. April 2004

30. Higgins J, Green S (editors. Cochrane Handbook for Systematic Reviews of Interventions 4.2.6 [updated September 2006]. In: The Cochrane Library, Issue 4, 2006. Chichester, UK: John Wiley & Sons Ltd

31. Bjermer L. Evidence-based recommendations or ‘show me the patients selected and I will tell you the results’. Respir Med 2006;100 Suppl. A:S17-21. Epub 2006 May 23

32. Herland K, Akelsen J, Skj0nsberg, Bjermer L. How representative are clinical study patients with asthma or COPD for a larger ‘real life’ population of patients with obstructive lung disease? Respir Med 2005;99:11-19

33. Chipps J, Spahn C, Sorkness L, et al. Variability in asthma severity in pediatrie subjects with asthma previously receiving short-acting beta2-agonist. J Pediatr 2006;148:517-21B

34. Stempel DA, Stoloff SW, Carranza R, Jr., et al. Adherence to asthma controller medication regimens. Respir Med 2005;99:1263-7

35. Stoloff SW, Stempel DA, Meyer J, et al. Improved refill persistence with fluticasone propionate and salmeterol in a single inhaler compared with other controller therapies. J Allergy Clin Immunol 2004; 113:245-51

36. Tews JT, Volmer T. Differences in compliance between combined salmeterol/fluticasone propionate in the Diskus device and fluticasone + salmeterol given via separate Diskus inhalers. Am J Respir Crit Care Med 2002;165:A188

37. Marceau C, Lemiere C, Berbiche D, et al. Persistence, adherence, and effectiveness of combination therapy among adult patients with asthma. J Allergy CHn Immunol 2006; 118: 574-81

38. O’Connor RD, Carranza Rosenzweig J, Stanford R, et al. Asthma- related exacerbations, therapy switching, and therapy discontinuation: a comparison of 3 commonly used controller regimens. Ann Allergy Asthma Immunol 2005;95: 535-40

39. Delea TE, Hagiwara M, Stanford R, Stempel DA. Utilization and costs of asthma-related care in patients initiating fluticasone propionate/salmeterol combination, salmeterol, or montelukast as add- on therapy to inhaled corticosteroids. Data in preparation 2006

40. MHRA. Commission on Human Medicines. Salmeterol (Serevent) and formoterol (Oxis, Foradil) in asthma management. cuit Probl Pharmacovigilance 2006;31:6

41. Nelson HS, Chapman KR, Pyke SD, et al. Enhanced synergy between fluticasone propionate and salmeterol inhaled from a single inhaler versus separate inhalers. J Allergy Clin Immunol 2003; 112:29-36

42. Bergmann KC, Lindemann L, Braun R, Steinkamp G. Salmeterol/ fluticasone propionate (50/250 microg) combination is superior to double dose fluticsone (500 microg) for the treatment of symptomatic moderate asthma. Swiss Med Wkly 2004; 134:50-8 (SAS40009)

43. Bateman ED, Britton M, Carillo J, et al. Salmeterol/ fluticasone propionate (50/100 mg) combination inhaler (Seretide). A new effective and well tolerated treatment for asthma. CUn Drug Invest 1998;16:193-201 (SFCB3017)

44. Chapman KR, Ringdal N, Backer V, et al. Salmeterol and fluticasone propionate (50/250 mg) administered via combination Diskus inhaler: As effective as when given via s eparate Diskus inhalers. Can Respir J 1999;6:45-51 (SFCB3018)

45. Aubier M, Pieters WT, Schlosser NJJ, et al. Salmeterol/ fluticasone propionate (50/500 meg) in combination in a Diskus inhaler (Seretide) is effective and safe in the treatment of steroid- dependent asthma. Respir Med 1999;93:876-84 (SFCB3019) 46. Dahl R, Chuchalin A, Gor D, et al. EXCEL: A randomised trial comparing salmeterol/fluticasone propionate and formoterol/ budesonide combinations in adults with persistent asthma. Respir Med 2006; 100:1152-62 (SAM40040)

47. Van den Berg NJ, Ossip MS, Hederos CA, et al. Salmeterol/ fluticasone propionate (50/100 [mu]g) in combination in a Diskus inhaler (Seretide) is effective and safe in children with asthma. Fed Pulm 2000;30:97-105 (SFCB3020)

48. Curtis L, Netten A. Unit Costs of Health and Social Care 2005. Personal and Social Services Research Unit, University of Kent. Available at htp://pssru.ac.uk/publications.htm (Accessed July 2006)

49. NHS Reference Costs. http://www.dh.gov.uk/assetRoot/ 04/10/ 55/61/04105561.xls. (accessed July 2006) 2004

Iolo Doull(a), David Price(b), Mike Thomas(c), Neil Hawkins(d), Eugena Stamuli(d), Maggie Jabberer(d), Toby Gosder(e) and Helen Rudge(e)

a Consultant Respiratory Paediatrician, Children’s Hospital for Wales, Cardiff, UK

b GPIAG Professor of Primary Care Respiratory Medicine, Department of General Practice and Primary Care, University of Aberdeen, UK

c Asthma UK Senior Research Fellow, Department of General Practice and Primary Care, University of Aberdeen, UK

d Oxford Outcomes Ltd, Oxford, UK

e GlaxoSmithKline Ltd, Uxbridge, UK

Address for correspondence: Neil Hawkins, Oxford Outcomes Ltd, Seacourt Tower, West Way, Botley, Oxford, OX2 OJJ, UK. Tel.: +44 1865 324930; Fax: +44 1865 324931; [email protected]

Appendix: Supplementary tables

The following tables are available as electronic supplementary data to the online version of this article (doi:10.1185/ 030079907X187991):

Table A1: Definitions of well-controlled and totally-controlled asthma based on global initiative for asthma/national institutes of health guideline aims of treatment

Table A2: Search strategy

Table A3: Included studies (SFC vs. same dose FP only)

Table A4: Meta-analysis results (SFC vs. same dose FP only)

Table A5: Cost-effectiveness of SFC compared with same dose ICS and additional doses of FP and BDP for other comparisons (adults & children aged 12 and over)

Table A6: Cost-effectiveness of SFC compared with same dose ICS (Children under 1

Copyright Librapharm May 2007

(c) 2007 Current Medical Research and Opinion. Provided by ProQuest Information and Learning. All rights Reserved.