cost-efficacy analysis of fluticasone propionate versus zafirlukast in patients with persistent...

Cost-Efficacy Analysis of FluticasonePropionate versus Zafirlukast inPatients with Persistent AsthmaRogelio Menendez1, Richard H. Stanford2, Lisa Edwards2, Christopher Kalberg2 andKathleen Rickard2

1 Allergy and Asthma Center of El Paso, El Paso, Texas, USA2 GlaxoSmithKline Inc., Research Triangle Park, North Carolina, USA

Abstract Objective:To compare the relative value of an inhaled corticosteroid, fluticasonepropionate 88μg twice daily, versus an oral leukotriene receptor antagonist,zafirlukast 20mg twice daily, in patients with persistent asthma currently receiv-ing short acting β2-agonists alone.Study design:Acost-efficacy analysis using resource utilisation and clinical dataobtained prospectively from a multicentre, randomised, double-blind, double-dummy, placebo-controlled 12-week clinical trial conducted in the US.Perspective: Third-party payor.Patients and methods: A total of 451 corticosteroid-naive patients with persist-ent asthma were treated with either fluticasone propionate 88μg twice daily orzafirlukast 20mg twice daily. All patients were given salbutamol (albuterol) to beused as rescue medication. Data were examined using intent-to-treat analysis.Results:Mean daily per person cost-efficacy ratios using improvement in forcedexpiratory volume in 1 second (FEV1) [≥12% increase from baseline] were$US3.47 for fluticasone propionate compared with $US7.81 for zafirlukast (1999values). The mean daily per person cost-efficacy ratios for symptom-free daysobtained were $US5.51 for fluticasone propionate compared with $US14.98 forzafirlukast. These cost-efficacy ratios remained in favour of fluticasone propio-nate after a robust sensitivity analysis.Conclusions: Treatment with fluticasone propionate 88μg twice daily was themost cost effective treatment compared with zafirlukast 20mg twice daily in this12-week clinical trial. This analysis supports the use of fluticasone propionate88μg twice daily as first-line treatment in patients with persistent asthma pre-viously treated with short-acting β2-agonist alone.

ORIGINAL RESEARCH ARTICLE Pharmacoeconomics 2001; 19 (8): 865-8741170-7690/01/0008-0865/$22.00/0

© Adis International Limited. All rights reserved.

Since 1978, the US and many other countrieshave experienced an increase in asthma mortality,prevalence and hospitalisations despite the numer-ous advancements in diagnosis and treatment of thedisease.[1-4] It has been suggested that the vast ma-jority of the morbidity and mortality may be due tothe undertreatment of the disease. Barnes and col-leagues[5] reported that approximately one-third ofthe direct cost of asthma is related to emergency roomvisits, hospitalisations and death that are most likelya result of undertreatment. It has been estimatedthat, in the US, the total annual cost of asthma careis $US5.8 billion (1994USdollars).[6] Direct expend-itures accounted for $US5.1 billion and indirect ex-penditures were valued at $US673 million, withhospitalisations accounting for more than half ofall expenditures.[7]Studies that examined the cost of asthma treat-

ment indicate that more aggressive treatment maytranslate into economic savings. Studies have shownthat the cost savings of treating uncontrolledasthma far outweigh the additional cost of the con-trol measures.[8,9] At present, controlling healthcarecosts is an issue of continuous focus in many coun-tries. With the increasing number of new treat-ments coming onto the market every year, deci-sions about their use should be based not only onthe safety and efficacy of the drugs but also on theirrelative overall value compared with alternativetreatments.Cost-effectiveness analysis is one of the meth-

ods used to evaluate the economic benefit of a pro-gramme or treatment. In this form of economicevaluation both the costs and consequences ofhealth programmes or treatments are examined.[10]This technique is designed to assist a decision-maker in identifying a preferred choice amongpossible alternatives.[11] Under strict definitions,economic analyses performed using data fromrandomised, controlled, clinical trials are consid-ered a cost-efficacy analysis (CEA) and not a cost-effectiveness analysis.Inhaled corticosteroids and leukotriene receptor

antagonists are 2 classes of drugs currently used forthe long term management of asthma. These drugs

are currently used as daily preventative therapy forpatients who suffer from persistent asthma symp-toms despite the use of as-needed short-acting β2-agonists. In the most recent asthma guidelines, in-haled corticosteroids are considered the mosteffective anti-inflammatory agents available forthe management of asthma.[12] However, thesesame guidelines list leukotriene modifiers as alter-natives to inhaled corticosteroids as first-line treat-ment for asthma.[12] Fluticasone propionate is aninhaled glucocorticoid with a high efficacy to safetyratio and a broad spectrum of anti-inflammatoryactivity.[13] The leukotriene receptor antagonistzafirlukast, is a selective, competitive antagonist ofthe leukotriene D4 and E4 (LTD4 and LTE4) recep-tors.[14]To date, there are few published data comparing

the cost efficacy of inhaled corticosteroids withthat of leukotriene modifiers in patients with per-sistent asthma.[15-17] Since leukotriene modifiersare considered alternatives to inhaled corticosteroidsfor the treatment of asthma, information on the costeffectiveness of these 2 treatment alternatives is ofvalue to medical therapeutic decision-makers andpharmacoeconomists. This CEA is based on a 12-week multicentre clinical trial with a cost perspec-tive of a third-party payor in the US.[18] The purposeof this CEA is to compare the economic value oftwice daily treatment with fluticasone propionateto twice daily treatment with oral zafirlukast in pa-tients with persistent asthma.

Methods

Study Design

The data utilised in the CEAwere collected froma multicentre, randomised, double-blind, double-dummy clinical trial which compared inhaledfluticasone propionate 88μg twice daily versus oralzafirlukast 20mg twice daily in patients who dem-onstrated signs of inadequate asthma control withshort-acting β2-agonist therapy alone. The com-plete safety and efficacy results for this study havebeen published separately.[18] Patients were eligi-ble for the study if they were ≥12 years of age and

866 Menendez et al.

© Adis International Limited. All rights reserved. Pharmacoeconomics 2001; 19 (8)

had a diagnosis of asthma as defined by the Amer-ican Thoracic Society, increased forced expiratoryvolume in 1 second [FEV1] of ≥12% over baselinewithin 30 minutes of salbutamol (albuterol) inha-lation, a medical history of asthma requiring as-needed short-acting β2-agonist bronchodilatortherapy and a baseline FEV1 between 50 to 80% ofpredicted normal value. Following an 8- to 14-dayrun-in period, eligible patients were randomly as-signed to 1 of the following treatment groups:fluticasone propionate 88μg twice daily or zafir-lukast 20mg twice daily. All patients were allowedto use salbutamol throughout the study on an as-needed basis for relief of breakthrough symptoms.Patients were treated for a period of 12 weeks. Inaddition to the collection of clinical data, facts onthe impact of asthma exacerbation on the overallcost of therapy were collected.

Cost-Efficacy Analysis (CEA)

This CEA was performed using the health out-come measures and the resource utilisation datacollected within the clinical study. The perspectivewas from a third-party payor and the efficacy pa-rameters used in this analysis were: improvementof ≥12% in morning FEV1 from baseline and meanpercent of symptom-free days over the 12-weektreatment period. All costs are reported in 1999dollars. The results of these analyses are reportedin terms of monthly cost per patient per unit ofefficacy.

Health Outcomes

The primary measurement of efficacy wasmorning premedication FEV1 obtained duringclinic visits. For the purpose of this analysis, aneffectively treated patient was defined as one whoachieved a ≥12% increase in FEV1 from baseline.In addition to FEV1, symptom data were collectedas a secondary outcome measurement. Efficacybased on symptom relief is reported as the percent-age of symptom-free days. A symptom-free daywas defined as a day in which the patient reportedno asthma symptoms (cough, wheeze, chest tight-ness or shortness of breath). Symptom data were

collected using daily diary cards on which patientswere asked to document their assessment of theirasthma symptoms prior to use of salbutamol. Symp-toms were measured using a 6-point scale where 0indicated ‘no symptoms’ and 5 indicated ‘symp-toms that caused discomfort and prevented normaldaily activities’. Patients who withdrew from thestudy for any reason were assumed to contributethe same proportion of symptom-free days thatthey had contributed prior to withdrawal.

Resource Outcomes

Asthma-related healthcare resources were col-lected during the study and included the following:• unscheduled physician visits due to asthma ex-acerbations

• emergency room visits due to asthma exacerba-tions

• outpatient clinic visits due to asthma exacerba-tions

• hospitalisations due to asthma exacerbations• amount of short-acting β2-agonist (salbutamol)consumed as rescue medication

• study asthma medications• medications given for treatment of drug-relatedadverse events.

Scheduled visits required by study protocol werenot included in the analysis. An asthma exacerba-tion was defined as any event that required treat-ment with oral or parenteral corticosteroids. Pa-tients who experienced an exacerbation werewithdrawn from the study. For each exacerbation,data on the setting of care and suspected cause ofexacerbation were collected. Patients recorded thenumber of puffs of rescue medication (salbutamol)consumed daily on their diary cards, and all adverseevents reported by patients were documented.[10,11]The adverse events that investigators reported asbeing potentially related to the study drug wereincluded in the economic analysis.

Unit Costs

Costs incorporated into the economic analysisincluded costs of study drugs, rescue medication(salbutamol), treatment costs for drug-related ad-

Cost-Efficacy of Fluticasone and Zafirlukast 867


verse events, unscheduled physician visits, emer-gency room visits and hospitalisations. All esti-mated costs are reported in 1999 US dollars and,when necessary, published average costs were ad-justed incrementally to reflect 1999 US dollars.Costs of drugs were based on the average whole-sale price.[19] The daily cost of fluticasone propio-nate 88μg twice daily was calculated to be $US1.50and the daily cost of zafirlukast 20mg twice dailywas calculated to be $US1.99. The cost of rescuesalbutamol use was based on $US0.11 per inhala-tion. In order to avoid incorporating bias from thevarying physician prescribing patterns into the eco-nomic analysis, the treatment cost for oral candidi-asis was based on a 5-day treatment regimen ofnystatin oral suspension (5ml four times daily); thecost was determined to be $US11.46 for each treat-ment. The cost of oral corticosteroids was not in-cluded in the analysis because of the very low costof oral prednisone (0.03 cents per 5mg tablet).Published estimates of costs were used for all

healthcare services. The cost of a physician’s visit($US40) was obtained from a 1998 national sur-vey of reimbursements for an office visit madeby an established patient to a family practi-tioner.[20] Using the published consumer price in-dex for healthcare, this cost was calculated to beapproximately $US43 in 1999 US dollars. Theaverage cost in 1999 US dollars for an emergency

room visit was determined to be $US270 while theaverage cost for an asthma hospitalisation was de-termined to be $US3100.[21] This published hospi-tal cost is an estimate of the cost to hospital andmay be lower than the actual cost to third-partypayors. However, to avoid underestimating the im-pact of hospitalisation costs in this analysis weused considerably higher values in the sensitivityanalysis.

Statistical Analysis

All analyses were performed on an intent-to-treat basis. All tests were 2-sided with a significancelevel of 0.05. No adjustments were made for mul-tiple comparisons ormultiple end-points. Cochran-Mantel-Haenszel tests controlling for investigatorwere conducted to test for differences in gender andethnic origin. Age, height, weight, and percent-pre-dicted FEV1 were compared between treatmentgroups with analyses of variance (ANOVA) ad-justed for investigator.The number of patients who experienced at least

a 12% increase in FEV1 were tabulated by treat-ment group, and a Cochran-Mantel-Haenszel testcontrolling for investigator was used to test fortreatment differences. The percentage of symptom-free days across the treatment period was calcu-lated for each patient, summary statistics werecomputed by treatment group, and ANOVAadjusted

Table I. Demographics and patient characteristics at baseline

FP 88μg bid (n = 231) ZA 20mg bid (n = 220)Mean age (y) [SD] 31.4 (13.1) 31.2 (12.9)Gendermale 119 (52%) 107 (49%)female 112 (48%) 113 (51%)

Ethnic originWhite 191 (83%) 183 (83%)Black 19 (8%) 17 (8%)Asian 1 (<1%) 4 (2%)Hispanic 15 (6%) 14 (6%)

Mean percent predicted FEV1 (SD) 67.2 (8.4) 68.0 (7.9)Percentage of SFD (SE) 7.4 (1.1) 5.1 (0.9)Salbutamol use (mean puffs/day) [SE] 4.6 (0.2) 4.8 (0.2)bid = twice daily; FEV1 = forced expiratory volume in 1 second; FP = fluticasone propionate; SD = standard deviation; SE = standarderror; SFD = symptom-free days; ZA = zafirlukast.

868 Menendez et al.


for investigator was conducted to compare the 2treatment groups.

Sensitivity Analysis

Estimates that use clinical trial data for a CEAare uncertain values; thus, it is necessary to per-form a sensitivity analysis of the results against theseestimates. In this study, a sensitivity analysis wasperformed on the common sources of uncertainty:the improvement in FEV1, the acquisition cost ofthe treatments, and the cost attributed to an asthmahospitalisation used in the analysis.

Results

Demographics

A total of 451 patients were randomly assignedto treatment after completing an 8 to 14 day run-inperiod. Table I shows the demographic and patientpopulation characteristics at baseline. Fewer zafir-lukast patients completed the study (87% in thefluticasone propionate group and 77% in the zafir-lukast group). A total of 81 patients withdrew fromthe trial (31 in the fluticasone propionate group and50 in the zafirlukast group); of these, 20 withdrewbecause of adverse events (8 fluticasone propio-nate group and 12 zafirlukast group) and 20 with-drew because they required oral corticosteroidsduring the study period (8 fluticasone propionategroup and 14 zafirlukast). Other reasons for with-drawal from the study were lost to follow up (7fluticasone propionate and 6 zafirlukast), consentwithdrawn (3 fluticasone propionate and 1 zafir-lukast), protocol violations (4 fluticasone propio-nate and 5 zafirlukast) and other (3 fluticasone pro-pionate and 12 zafirlukast).

Health Outcomes

The fluticasone propionate group had a signifi-cantly greater improvement in pulmonary function(mean change from baseline in FEV1) comparedwith the zafirlukast group. Fluticasone propionatetreatment resulted in a significantly greater propor-tion of successfully treated patients than the zafir-lukast group; 53 and 37%, respectively, had at least

a 12% increase in FEV1 (p = 0.001). The meanpercentage of symptom-free days across each pa-tient’s treatment period was 33.4% in the fluti-casone propionate group and 19.3% in the zafirluk-ast group (p < 0.001).

Resource Outcomes

Table II displays the number of healthcare re-sources utilised as a result of asthma exacerbationsduring the study period. The group treated withfluticasone propionate experienced fewer asthmaexacerbations than patients treated with zafirlukast(3.5% versus 6.4%). However, this was not statisti-cally significant. There were 2 asthma-related hospit-alisations in the zafirlukast group and none in thefluticasone propionate group; no emergency roomvisits due to asthma exacerbation were reported ineither treatment group. Most exacerbations re-quired a physician visit (10 zafirlukast group and8 fluticasone propionate group). It was assumedthat exacerbations that took place at home (2 zafir-lukast group and 0 fluticasone propionate group)did not incur any additional costs to the payor. Asjudged by the investigators, there was 1 potentiallydrug-related case of oral candidiasis in the fluti-casone propionate group. The daily mean numberof puffs of salbutamol at end-point was 3.33 in thezafirlukast group and 2.17 in the fluticasone pro-pionate group (p < 0.001).

CEA

The economic analysis was performed for allthe relevant costs and benefits of each drug and the

Table II. Healthcare resources utilised due to asthma exacerba-tions during the study period

FP 88μg bid ZA 20mg bidNumber of exacerbationsa 8 (3.5%) 14 (6.4%)Hospitalisation 0 2Emergency room visit 0 0Physician’s visit 8 10Home 0 2a Defined as needing either an oral or intravenous corticosteroid.bid = twice daily; FP = fluticasone propionate; ZA = zafirlukast.



total monthly costs were calculated for fluticasonepropionate and zafirlukast treatments. The cost-efficacy ratios were calculated using these costsand using each of the efficacy measurements ofFEV1 and symptom-free days. The mean daily costof asthma treatment was $US1.84 per patient forpatients in the fluticasone propionate group and$US2.89 for those in the zafirlukast group (tableIII). Most of the cost for both treatment groups wasattributed to the cost of asthma medication and res-cue medication. However, hospitalisation costswere approximately 14% of the total cost of zafir-lukast therapy.The cost-efficacy ratios reported as the mean

daily cost of treatment for each patient achievingat least a 12% increase in FEV1, were $US3.47 forthe fluticasone propionate group as compared with$US7.81 for the zafirlukast group (table IV). Usingsymptom-free day as an effectiveness parameteryielded amean daily cost-efficacy ratio of $US5.51per symptom-free day for the fluticasone propio-nate group and $US14.98 for the zafirlukast group.Incremental cost-efficacy ratios (ICERs) are gen-erated in routine cost-efficacy studies to determinethe additional cost that would be incurred if themore effective treatment were used. The incremen-tal ratio is calculated by dividing the difference incost between the 2 treatment groups by the differ-ence in their efficacy. Because fluticasone propio-nate is both more effective and less costly, an in-

cremental CEAwas inappropriate: the ‘incremental’costs would yield negative values for FEV1 andsymptom-free days. These negative incremental val-ues represent the potential cost savings of switch-ing patients currently on zafirlukast to fluticasonepropionate, since more patients would be symptom-free and would have improvement in lung functionon fluticasone propionate compared with zafir-lukast.

Sensitivity Analysis

The base-case lung function improvement of12% in FEV1 may not be considered clinically rel-evant by some clinicians or may be determined tobe too restrictive by others. In order to improvethe robustness of this analysis, lung function im-provements were varied between ≥10 and ≥15%improvement in FEV1. The difference betweentreatment groups (fluticasone propionate versuszafirlukast) continued to be statistically signifi-cant at all levels of efficacy (p < 0.001), indicatingthat the outcome was robust to changes in efficacy(fig. 1). Also, the cost-efficacy ratios generated usingthese various improvements in lung function re-mained constant with little change in the magnitudebetween groups.A sensitivity analysis was also performed on

varying the cost of an asthma-related hospitalisa-tion and the acquisition cost of zafirlukast. A best-and worst-case costs scenario was determined us-

Table III. Mean total daily cost of treatmenta (per person; 1999 USdollars)

Cost component Cost per day [$US] (% of total daily cost)FP 88μg bid ZA 20mg bid

Hospitalisations 0 0.40 (14%)Physician office visits 0.02 (1%) 0.02 (1%)Rescue salbutamol(albuterol)

0.26 (13%) 0.42 (13%)

Adverse events <0.01 (<1%) 0Drug treatment 1.56 (86%) 2.05 (72%)Total daily costs 1.84 2.89a Base-case using average wholesale price drug costs and

$US3100 asthma hospitalisation costs.bid = twice daily; FP = fluticasone propionate; ZA = zafirlukast.

Table IV. Base-case mean and incremental cost-efficacy ratios forzafirlukast (ZA) 20mg twice daily and fluticasone propionate (FP)88μg twice daily

Cost-efficacy ratio FP 88μg bid($US)

ZA 20mgbid ($US)

Cost per SFD 5.51 14.98Incremental cost-efficacyratioa

–n/e

Cost per 12% increasein FEV1

3.47 7.81

Incremental cost-efficacyratioa

–n/e

a Incremental analysis is the difference in cost/difference inefficacy (–n/e = nonevaluable negative values)bid = twice daily; FEV1 = forced expiratory volume in 1 second;SFD = symptom-free day.

870 Menendez et al.


ing the upper and lower ranges of asthma hospitalcosts ($US2195 and $US15 181) and varying theacquisition cost of zafirlukast in the best-case sce-nario until the incremental analysis approachedzero using symptom-free day and lung functionimprovement of ≥12% in FEV1 as the efficacy pa-rameters (table V). Under these assumptions, thecost of zafirlukast would have to be discounted by50% from the average wholesale price in the best-case scenario for the incremental cost-efficacy ra-tio to approach zero for both lung function im-provement and symptom-free day.

Discussion

This analysis evaluated the cost efficacy of 2treatments for mild asthma. Previous economicstudies have already shown that fluticasone propi-onate treatment is associated with significant costsavings when compared with other inhaled corti-costeroids, oral steroids, or leukotriene receptorantagonists.[17,22-26] This analysis found that pa-tients with inhaled corticosteroid-free asthma whostart on fluticasone propionate 88μg twice dailyhave better asthma symptom control and greaterimprovement in lung function at ≤50% the cost oftreating these patients with zafirlukast 20mg twicedaily. In order to achieve both improved efficacyand cost effectiveness in patients requiring initialcontroller therapy, fluticasone propionate may bea better choice than using a leukotriene receptorantagonist. The results were robust since zafirluk-ast would have to be discounted by 50% for theincremental ratios to approach zero for both im-provements in lung function and symptom-freeday using the lowest cost for an asthma-relatedhospitalisation. Incremental analysis can be ofbenefit to healthcare payors in determiningwhether the added effectiveness of a treatmentadds value. To date, little information is availabledescribing an acceptable threshold ICER forasthma therapy. Rutten van Molken et al.[9] foundthat an acceptable ICER for asthma therapy wasapproximately $US5 (1989 prices). The robustnessof this analysis is enhanced, since the incrementalvalues generated from this study are well below

$US5, even after discounting zafirlukast by 50%.In addition, these results are noteworthy becausetreatments that produce greater medical benefitsare usually associated with additional monetarycost. Thus, we found that fluticasone propionatedelivered greater asthma control at a lower costthan zafirlukast.To date, there is no universally accepted out-

comemeasure for asthma intervention studies witheconomic components. Asthma outcome measurescan be viewed from 2 perspectives: patient- andfamily-centred or organisationally based out-comes.[27] Patient and family-centred measuresconsider outcomes such as patient symptoms,functional status, quality of life, physiologicalfunction, and satisfaction with care or family out-

Mea

n pr

opor

tion

of p

atie

nts

(%)

60

50

70

40

30

20

10

010 11 12 13 14 15

Increase in FEV1 from baseline (%)

FP 88µg twice dailyZA 20mg twice daily

*

**

**

*

Fig. 1. Sensitivity analysis of improvement in lung function. Thethresholds for improvement in forced expiratory volume in 1second (FEV1) were varied in 1% increments from 10 to 15%with 12% being the base-case scenario. All differences betweentreatment groups were statistically significant (*p < 0.001) infavour of fluticasone propionate (FP). ZA = zafirlukast.



of-pocket medical expenditures. Organisationallybased measures include resource utilisation, asthma-related costs and population-based measures ofhealth status. These are usually examined fromthe perspective of the healthcare providers andpayor.Regardless of the perspective of the study, effi-

cacy must be measured by clearly defined out-comes. Since there is a lack of consensus on whichasthma outcome measure to include in economicevaluations, 2 different health outcomes measures(FEV1 and symptom-free days) were used as effi-cacy parameters in this economic analysis. Meas-ures of pulmonary function are essential for assess-ing the severity of asthma and are also useful formonitoring the patient’s response to therapy. Ab-normalities in pulmonary function are a measure ofthe degree of airflow obstruction and reflect theconsequence of asthma on airway mechanics. Un-fortunately, spirometry is not widely available toall patients because of its relatively high cost. Inaddition to assessing pulmonary function, symp-tom data is the most frequently captured outcomesmeasure for asthma and it correlates well with pul-monary measures.[27] Also, lower pulmonary func-tion correlates with more asthma symptoms. In aneffort to standardise asthma outcomemeasures, the

National Asthma Education Prevention Program(NAEPP) working group has proposed the conceptof symptom-free days as the principal outcomemeasure for CEA of asthma interventions.[27] Byconsidering both airflow and symptoms in the eco-nomic analysis, the results from this study may aida greater number of decision-makers in comparingasthma treatments.Since the manifestations of asthma are rarely

limited to 12 weeks, the duration of the trial limitsthis economic evaluation. Although 12-week asthmaclinical trials are quite common, this may not beenough time to capture all the symptomatic epi-sodes and resources utilised in this population ofpatients with asthma. In addition, the intent-to-treat approach used in this analysis only comparesthe resource use collected while the patient re-mained in the study. This is a conservative method,and assumes that all patients who withdrew fromthe study did not incur any additional costs uponwithdrawal and that they contributed the same per-cent of symptom-free days. Even in this short timeperiod and using such a conservative method, pa-tients treated with fluticasone propionate had fewerasthma exacerbations requiring intravenous or oralcorticosteroids than patients treated with zafirluk-ast. In addition, recent ‘real-world’ studies havedocumented the long term economic benefits of in-haled corticosteroids.[28-30] They have shown thatthe use of inhaled corticosteroids in an asthma pop-ulation reduces emergency room visits, hospitalis-ations and death from asthma. In contrast, the longterm economic benefit of using leukotriene antag-onists has not been established.Ideally, economic evaluations should incorpo-

rate clinical data on effectiveness rather than effi-cacy but these types of data are rarely available.Well-designed randomised clinical trials are cre-ated with a high degree of internal validity but maypossess a low degree of external validity and maypotentially lack generalisability to the real-worldsetting. This issue of internal versus external valid-ity is a main topic of debate among economic ana-lysts. Future research based on data from either along term open label trial or from a well-designed

Table V. Sensitivity analysis for cost-efficacy ratios and incrementalcost-efficacy ratios based on best-a and worst-caseb scenarios

Cost ofhospitalisation($US)

Mean daily cost perimprovement ($US)

Incrementalcost-efficacyratiocFP 88μg bid ZA 20mg bid

≥12% FEV1

Best 3.47 4.63 -0.81Worst 3.47 11.20 -n/e

Symptom-free dayBest 5.51 8.87 -0.92Worst 5.51 21.46 -n/ea Best-case is hospitalisation costs of $US2195 and ZA dis-

counted by 50%.b Worst-case is hospitalisation costs of $US15181.c Incremental analysis is the difference in cost divided by the differ-

ence in efficacy, (-n/e indicates non-evaluable negative values).bid = twice daily; FEV1 = forced expiratory volume in 1 second; FP= fluticasone propionate; ZA = zafirlukast.

872 Menendez et al.


database trial may provide results with greater ex-ternal validity and generalisability than thosebased on data from clinical trials. However, sinceclinical trials are the gold standard in determiningdifferences in efficacy between comparator prod-ucts, an intent-to-treat clinical trial may help pre-dict the economic outcomes that clinicians may ex-perience in their practice.

Conclusions

This analysis shows that fluticasone propionate88μg twice daily is more cost effective than zafir-lukast 20mg twice daily in patients previouslytreated with β2-agonist alone. With the shift intreatment decision-making moving away from in-dividual practitioners to large provider organis-ations, public and private healthcare decision-mak-ers are faced with the difficult decision of choosingamongmultiple treatment options for the treatmentand management of asthma. This economic analy-sis should provide needed information that will en-able third-party payors and physicians to make aninformed decision regarding the most cost-effec-tive treatment for patients with uncontrolled per-sistent asthma.

Acknowledgements

This study was funded by GlaxoSmithKline Inc., Re-search Triangle Park, North Carolina, USA.Presented in part at the Annual American Academy of

Allergy, Asthma and Immunology (AAAAI) meeting, SanDiego, California, USA, March 2000.Dr Menendez is a contracted clinical investigator and a

consultant to GlaxoSmithKline.

References1. Centers for Disease Control. CDC surveillance summaries;

1998 Apr 24. MMWR Morb Mortal Wkly Rep 1998; 47:SS-1

2. Evans III, R. Epidemiology and natural history of asthma, al-lergic rhinitis, and atopic dermatitis. In: Middleton Jr E, ReedCE, Ellis EF, et al. Allergy principles and practice. 4th ed. St.Louis: Mosby 1993; 1109-36

3. Arrighi MH. US asthma mortality, 1941-1989. Ann AllergyAsthma Immunol 1995; 74: 321-6

4. Jackson R, Sears MR, Beaglehole R, et al. International trendsin asthma mortality, 1970-1985. Chest 1988; 94: 914-9

5. Barnes PJ, Jonsson B, Klim JB. The costs of asthma. Eur RespirJ 1996; 9: 636-42

6. Weiss KB, Gergen PJ, Hodgson TA. An economic evaluationof asthma in theUnited States. N Engl JMed 1992; 326: 862-6

7. Smith DH, Malone DC, Lawson KA, et al. A national estimateof the economic costs of asthma. Am J Respir Crit Care Med1997; 156: 787-93

8. Rutten-Van Molken MPH, Van Doorslaer EKA, Jansen MC, etal. Cost and effects of inhaled corticosteroids plus broncho-dilator therapy versus monotherapy in children with asthma.Pharmacoeconomics 1993; 4: 257-70

9. Rutten-Van Molken MPH, Van Doorslaer EKA, Jansen MC, etal. Cost and effects of inhaled corticosteroids and bronchodi-lators in asthma and chronic obstructive pulmonary disease.Am J Respir Crit Care Med 1995; 151: 975-82

10. DrummondMF, O’Brien B, Stoddart GL, et al. Methods for theeconomic evaluation of health care programmes. 2nd ed. NewYork (NY): Oxford University Press, 1997

11. Bootman JL, Townsend RJ, McGhan WF. Principles of phar-macoeconomics. 2nd ed. Cincinnati (OH); Harvey WhitneyBooks Company, 1996

12. National Institutes of Health (National Heart, Lung, and BloodInstitute). Highlights of the expert panel report II: guidelinesfor the diagnosis andmanagement of asthma. Bethesda (MD),1997; NIH-97-4051A: 1-50

13. Flovent® product information. Glaxo-Wellcome Inc. Jul 199714. Accolate® product information. Zeneca Inc. Sep 199615. Westbroek J, Pasma HR. The effect of inhaled fluticasone pro-

pionate 100 mg bid compared with oral ZA 20mg bid onbronchial hyperresponsiveness [abstract]. Eur Respir J 1997;10 Suppl. 25: 243S

16. Westbroek J, Pasma HR, James MH, et al. Inhaled fluticasonepropionate and oral ZA in moderate asthma: a clinical andcost comparison [abstract]. Am J Respir Crit Care Med 1998;157 (3): A416

17. Mullen B, Driskel JE, Yancey SW. A statistical comparison ofthe efficacy of leukotriene modifiers and inhaled corticoste-roids. J Allergy Clin Immunol 1998; 101 (1): S232

18. Bleecker ER, Welch MJ, Weinstein SF, et al. Low dose inhaledfluticasone propionate vs oral ZA in the treatment of persist-ent asthma. J Allergy Clin Immunol 2000; 105: 1123-9

19. Medical Economics Company. Drug Topics 1998 Red Book.Montvale (NJ): Medical Economics Company, 1998

20. Crane M. Charge what you will: reimbursements are shrinking.Med Econ 1998 Sep 21: 128-48

21. Stanford R, Mclaughlin T, Okamoto LJ. The cost of asthma inthe emergency department and hospital. Am J Respir CritCare Med 1999; 150: 211-5

22. O’Reilly JF, Weir DC, Banham S, et al. Is high dose fluticasonepropionate via metered dose inhaler and volumatic as effica-cious as nebulised budesonide in adult asthmatics? RespirMed 1998; 92: 111-7

23. Price DB, Appleby JL. Fluticasone propionate: an audit of out-comes and cost-effectiveness in primary care. Respir Med1998; 92: 351-3

24. Williams J, Richards KA. Ease of handling and clinical efficacyof fluticasone propionateAccuhaler/Diskus inhaler comparedwith the budesonide Turbohaler inhaler in paediatric patients.UK study group. Br J Clin Pract 1997; 51: 147-53

25. Booth PC, Capsey LJ, Langdon CG, et al. A comparison of thecost-effectiveness of alternative prophylactic therapies in thetreatment of adult asthma. Br J Med Econ 1995; 8: 65–72

26. Barnes NC, Thwaites RMA, Price MJ. The cost-effectivenessof inhaled fluticasone propionate and budesonide in the treat-ment of asthma in adults and children. Respir Med 1999; 93:402-7



27. National Asthma Education and Prevention Program TaskForce on the cost effectiveness, quality of care, and financingof asthma care. National Institutes of Health (NationalHeart, Lung and Blood Institute). NIH Publication 1996;55: 807

28. Donahue JG, Weiss ST, Livingston JM, et al. Inhaled steroidsand the risk of hospitalization for asthma. JAMA 1997; 277:887-91

29. Dubois RW, Chawla AJ, Neslusan CA, et al. Explaining drugspending trends: does perception match reality? Health Aff(Millwood) 2000; 19: 231-9

30. Suissa S. Ernst P, Benayoun S, et al. Low-dose inhaled cortico-steroids and the prevention of death from asthma. N Engl JMed 2000; 343: 332-6

Correspondence and offprints: Dr Rogelio Menendez, Al-lergy and Asthma Center, 10470 Vista del Sol, El Paso, Texas79925, USA.E-mail: [email protected]

874 Menendez et al.


cost-efficacy analysis of fluticasone propionate versus zafirlukast in patients with persistent...

Documents