Cost Minimization Analysis of Antiepileptic Drugs in Newly Diagnosed Epilepsy in 12 European Countries

Transcription

1 ~ Epilepsiu. (Suppl. ):S-S, Llpplncon Williams & Wilkins, Inc., Philadelphia International League Against Epilepsy Cost Minimization Analysis of Antiepileptic Drugs in Newly Diagnosed Epilepsy in European Countries *D. C. Heaney, *S. D. Shorvon, *J. W. A. S. Sander, tp. Boon, *V. Komarek, *P. Dravet, IIE. Perucca, J. Majkowski, **J. Lopes Lima, tts. Arroyo, **T. Ried,' k. van Donselaar, sae. Eskazan, ***P. Peeters, '"P. Carita, '". Tjong-a-Hung, ttte. Myon, and tttc. Taieb *National Hospital for Neurology and Neurosurgery, London, UK; University Hospital Ghent, Belgium; $Clinic of Child Neurology FN Motol, Prague, Czech Republic; $Centre Saint Paul, Marseille, France, "University of Pavia, Pavia, Italy; Forlowski State Clinical Hospital Nr. I, Warsaw, Poland; **Hospital de Santo Antonio, Porto, Portugal; "Hospital Clinito y Provincial, Arda, Barcelona, Spain; "Karolinska Hospital, Stockholm, Sweden; # Schweizerische Epilepsie-Klinik, Zurich, Switzerland; ""Sint Clara Ziekenhuis, Rotterdam, Netherlands: qqcerrapasa Faculty of Medicine, Istanbul, Turkey; ***Quintiles SA, Levallois-Perret Cedex, France and fttsanofi-synthelabo, Paris, France Summary: A recent United Kingdom cost minimization analysis (CMA) of four antiepileptic drugs (AEDs) used to treat newly diagnosed adult epilepsy demonstrated that a new drug, lamotrigine (LTG), incurred higher costs than carbamazepine (CBZ), phenytoin (PHT), and valproate (VPA), whose costs were similar. This analysis took account of each drug's sideeffect and tolerability profile. The present analysis investigated the costs of treatment with LTG, CBZ, PHT, and VPA in European countries. Data were derived from published sources and from a panel of locally based experts. When no published data were available, estimates were obtained using expert opinion by a consensus method. These data were incorporated into a treatment pathway model, which considered the treatment of patients during the first months after diagnosis. The primary outcome considered was seizure freedom, Randomized con- trolled trials demonstrate that the drugs considered are equally effective in terms of their ability to achieve seizure freedom, and thus the most appropriate form of economic evaluation is a CMA. These trials provided data on the incidence of side effects, dosages, and retention rates. The economic perspective taken was that of society as a whole and the analysis was calculated on an "intent-to-treat'' basis. Only direct medical costs were considered. In each country considered, LTG was twofold to threefold more expensive than the other drugs considered. A sensitivity analysis demonstrated that varying each of the assumptions (range defined by expert panels) did not significantly alter the results obtained. Key Words: Epilepsy-Cost minimization-direct costs-societal perspective-european. Over the past years, several new antiepileptic drugs (AEDs) have been licensed by European regulatory bodies as treatment for epilepsy (). Clinical trials have demonstrated that these drugs are as effective as established AEDs in terms of seizure control. They have differing pharmacokinetic and side-effect profiles, but their longterm safety has not been proved. These new AEDs are more expensive than established therapies and their role in the treatment of epilepsy has been questioned (). Epilepsy is a very common neurologic condition, with an incidence of -/, throughout Europe (). Although more than two-thirds of patients with newly Address correspondence and reprint requests to Dr. Dominic Heaney at Regional Neuro-Rehabllitation Unit, Homerton Hospital NHS Trust, Homerton Row, London E9 SR, UK. I Deceased June,. diagnosed epilepsy become seizure-free on firstline therapy (l), most patients with epilepsy require longterm treatment to maintain control of their seizures. The acquisition cost of AEDs represents a large proportion of the total cost of treating patients with epilepsy (). The use of new AEDs could have a large impact on the resources available to treat epilepsy in European countries, especially if savings resulting from improved clinical outcomes do not offset the cost of the drugs. A cost minimization analysis (CMA) examining the use of new AEDs in newly diagnosed epilepsy in the U.K. showed that the use of lamotrigine (LTG) incurs costs between 9 and f per patient higher in the first year of treatment than those resulting from treatment with carbamazepine (CBZ), phenytoin (PHT), and Valproate (VPA) (). s

2 S D. C. HEANEY ET AL. RATIONALE AND AIMS We sought to examine whether this CMA would produce similar results in other European health-care sy~tems in which the use of LTG, CBZ, PHT, and VPA was compared. The aim was to establish the direct medical costs that were likely to result from the use of different agents as first-line treatment in adults with newly diagnosed epilepsy in different European countries. METHODS Country selection and health-care systems review A spectrum of countries was selected in Western and Central Europe to include a wide variety of health-care systems. Only countries in which all four drugs were licensed for use were considered. Each health-care system was reviewed, and particular consideration was given to systems of referral and to the way in which services were financed. Randomized controlled trials of AED monotherapy A literature review revealed that, since 9, seven randomized controlled trials have been published comparing these drugs as first-line therapy for newly diagnosed partial and generalized epilepsy (-,,). Five trials contained data that allowed a CMA study (-,,). No trial compared all four drugs directly, which meant that the cost consequences were assessed separately for the five individual trials, in each European country. Time span The time span considered for this economic appraisal was the first year of treatment. The clinical trials on which this study is based evaluated treatment for periods ranging from weeks to years. Only data relevant to the first year of treatment were considered for this analysis. Types of costs In health economics (HE), it is essential to distinguish the three types of cost: direct, indirect, and intangible. Intangible costs include the human and psychological costs linked to stress, anxiety, pain, and quality of life. Indirect costs include the productivity losses in the workplace, absenteeism, and unemployment resulting from illness. The economic impact of indirect and intangible costs is difficult to assess. There are few data relating to the differential effects of AEDs on these economic outcomes, and these difficulties led to our decision to omit these costs. Types of economic analysis The five main types of economic analysis are as follows: cost of illness; cost minimization (comparison of two treatments that are equally effective in terms of clinical outcomes); cost effectiveness (comparison of two treatments that lead to a common outcome which differs in magnitude); cost benefit (comparison of a treatment with a range of other medical intervention); and cost utility (comparison of two treatments that differ in the way in which they influence the patient s quality of life). A recent review of the literature from 99 to 99 () bears witness to the limited number of HE publications in the epilepsy area. In the case of newly diagnosed epilepsy, comparative trials have demonstrated that there is no significant difference among AEDs in terms of the primary outcome of interest, seizure freedom. As a result, a CMA of economic evaluation is appropriate and the cost differences resulting from the use of each drug were calculated in this study. Perspective The four main perspectives that health economic analysis can consider are that of society as a whole, that of the third-party payer (e.g., a health insurance company or other such collective organization), that of the individual patient and, finally, that of the hospital. Although only direct medical costs were incorporated into this appraisal, the perspective of society as a whole was chosen to overcome and minimize the actual consequences of the differences in health systems among the participating countries. These variations mean, however, that comparisons between countries must be considered cautiously. Costs addressed in this exercise included the acquisition cost of drugs, medical consultations, medical investigations, and hospitalizations. Other costs, particularly indirect and intangible costs, were not considered. Sensitivity analysis In the absence of published information, this economic appraisal made several assumptions about clinical data and resource use, based on the results from a consensus panel convened in each participating country. A sensitivity analysis was performed to test whether and how changing any of the assumptions affected the overall results. Key assumptions that were considered in the sensitivity analysis are the frequency of hospitalization (tested from. to times original assumptions). the cost and dose of drugs used (tested from. to. times original assumptions), the frequency of follow-up visits (tested from a minimum of two fewer visits to a maximum of two extra visits to GPs and specialists), and the proportion of patients being treated with polytherapy. DATA SOURCES Data for treatment pathways were derived from published trials and databases. Resource use was modeled using consensus data obtained from a panel of experts in each country (). The panels addressed issues such as the use of add-on therapy, laboratory safety monitoring, medical follow-up in primary care and specialist clinics, Epilepsia. Vol., Suppl.,

3 COST MINlMIZATlON ANALYSlS OF AEDs s9 TABLE. Summary of randomized controlled trials comparing monotherapy in adults with newiy diagnosed epilepsy Trial (duration) Carbamazepine Lamotrigine Brodie Trial size (n) 9 (9 months) Average dose (mglday) Withdrawal rate (%) 9 Sideeffect rate (%) Reunanen Trial size (n) ( months) Average dose (mglday) or Withdrawal rate (%). 9 ( mglday) Steiner * Sideeffect rate (%) Trial size (n) ( mglday) ( months) Average dose (mg/day) I Withdrawal rate (%) Side-effect rate (%) Richens Trial size (n) ( and months) Average dose (mglday) Withdrawal rate (%) Sideeffect rate () Hellerg Trial size (n) ( months) Average dose NS Withdrawal rate (%) la Side-effect rate (%) NS Callaghan Trial size (n) 9 (- months)b Average dose (mgkglday).9 Withdrawal rate (%) Sideeffect rate (%). Ramsay 9 Trial size (n) ( months) Average dose (mgkglday) 9. Withdrawal rate (%) Side-effect rate (%) Trial size (n) ( months) Average dose Withdrawal rate (%) Side-effect rate (a) TumbuII, z (at m).9 (at m) Half NS Phenytoin 9 ~~ mglday 9 NS Valproate 9 9 ( m) 9. ( m) NS NS. NS NS Half NS NS 9 NS NS NS, not stated in published hial data a Percentage of patients withdrawn due to adverse events. The total percentage of patients withdrawn is not stated. Median follow-up varied among CBZ (partial seizures months, generalized months), PHT ( months and months), and VPA ( months and months). Mean dose stated for the group of patients who suffered no major side-effects and in whom the efficacy of a single medication could be evaluated. assessment of AED levels, and approximate rates of hospitalization and admission to accident and emergency departments. Unit costs were drawn from official sources. The detailed handling of these main three data areas is considered in turn. Clinical data and resource use Side-effect prevalence, drug tolerability, and the average dose of drug prescribed to patients were derived from clinical trials (Table ). A treatment pathway model was devised that categorized patients according to whether they tolerated AEDs and whether add-on drugs were used (Fig. ). The panel of experts in each country defined the amount of medical services that patients were likely to receive in each arm (Fig. ; Tables -) and set estimates of the other clinical data, including dose ranges in add-on therapy and the proportion of patients treated with polytherapy. It was assumed that % of patients suffered from partial epilepsy. The CMA incorporated any cost savings that might result from each drug s different tolerability and side-effect profile. Few data are available concerning the cost of side effects from AEDs. In this study, we have assumed that patients who develop side effects would consult their GP start: % Path A -X% Path D Pdh A Continw tst I!m mcwttmrapq path B Fan iiw mommenpy. mat wim nd tino monoth.npy. continue nd th. mnoth.n~~ Pdh C Fail. lm nwmnhnpy. mt whh nd lma ronomonpy. hit nd itm, bmt mth polymnapy Pam o ~aii tst IIW momttmmpy, treat mm wmpy FIG.. Treatment pathway for adults with newly diagnosed epilepsy. Epilepsia. Vol.. Suppl.,

4 S D. C. HEANEY ET AL. I I I GP Specialist Lab. AED No. of AE test levels days in admissions hospital Medical service FIG.. Medical services received in first year. a basis for the calculation. Government-recommended charges were used for medical consultations and services. RESULTS The CMAs demonstrated that treating a patient with LTG as first-line therapy is between twofold and fourfold more expensive than treatment with CBZ, PHT, or VPA, which share similar costs, over the first-year period (Fig. ). These results are consistent in all the countries considered, despite variations in the medical management of newly diagnosed epilepsy. (or specialist, in the case of Turkey) and, if receiving CBZ, PHT, or VPA, would have AED levels checked. In contrast to previous economic appraisals considering AEDs (-9), we have not estimated the cost of symptom treatment, because the majority of side effects are treated by changing the dose or withdrawing the AED. In an ongoing study of the cost of side effects in the U.K., very few instances of symptom treatment have been recorded. We have not considered rare, serious adverse reactions, such as hepatic failure or Stevens-Johnson syndrome. When a patient was withdrawn from a drug and started on an alternative therapy, we assumed that both GP and specialist would be consulted. The cost of investigations performed to establish the diagnosis of epilepsy was not addressed in this appraisal, because these tests are performed no matter which AED is prescribed. Unit costs Unit costs were derived from published sources and charging schedules (Table ). The method by which these costs were ascertained have been described previously (). In this study, a societal perspective of direct medical costs was used. When different formulations of drugs were available in a country, the charge made for the most commonly prescribed formulation was used as SENSITIVITY ANALYSIS The sensitivity analysis revealed that although the results were sensitive to changes in frequency of hospitalization, drug dose, and acquisition cost, frequency of follow-up visits, and the proportion of patients being treated with polytherapy, the overall conclusions of the CMA were not altered. Various assumptions that increased the cost of polytherapy (treatment paths C and D; see Tables and ) increased the absolute cost of prescribing all four drugs, thereby reducing the ratio of cost differences among the drugs. DISCUSSION Validity of the consensus approach When clinical and resource use data were not available, we used a consensus method approach to provide estimates for feeding the treatment pathway model (). Consensus methods have been criticized for generating quantitative estimates that might lead a casual observer to place greater reliance on the results than may be warranted. All the assumptions made in this study, however, were tested in a wide-ranging sensitivity analysis. The conclusions reached remained constant, even across a broad range of European health-care systems. TABLE. Resource use: path A Lab tests, Annual frequency of Average number of Specialist annual serum AED Patients hospitalized Duration of A&E admissions per Country GP visits visits frequency determination per year (%) hospitalization (days) patient per year Be g i u m Czech Republic France Italy Netherlands Poland Portugal Spain Sweden Switzerland... I. I..... Turkey U.K.. I Epilepsia. Vol., Suppl.,

5 COST MINIMIZATION ANALYSIS OF AEDs S TABLE. Resource use: path B" Lab tests, Annual frequency of Average number of Specialist annual serum AED Patients hospitalized Duration of A&E admissions per Country GP visits visits frequency determination per year (%) hospitalization (days) patient per year Belgium. Czech Republic. France... Italy I.. Netherlands. Poland.. Portugal.... Spain 9 Sweden.... Switzerland... Turkey U.K. a LTG levels not routinely measured. All percentages are rounded to the nearest %. Choice of clinical trials Five randomized, controlled clinical trials comparing the use of AEDs in newly diagnosed epilepsy were used to provide data on drug efficacy, side-effect profile, tolerability, and average dosage for the CMA. Each of these studies used a different protocol, and they also differed in their inclusion and exclusion criteria. To provide more meaningful results, we analyzed each trial separately, leading to five distinct CMAs performed for each country. There was only slight variation in the results observed among trials, which tends to confirm the robustness of our conclusions. Variations in health-care systems This study revealed several differences and similarities in the treatment of epilepsy throughout Europe. The management of epilepsy is comparable in terms of choice of drug for monotherapy, basic treatment pathway, and rates of consultation, but differs in terms of the proportion of patients treated with polytherapy, the rate of hospitalization, choice of drug in polytherapy, and the rate of drug titration. Further variations were observed in the unit costs of items relevant to the treatment of epi- lepsy. Acquisition costs for AEDs differed among countries, and marked price variation was observed in the management of epilepsy among each of the countries. Although, in this study, the perspective taken was that of society as a whole, differences in health service organization could cause the economic impact of epilepsy treatment to be borne by different societal groups. For example, in some health-care systems, individuals consult epilepsy specialists directly, without the involvement of a GP and, in many cases, patients are charged a proportion of the fees relating to medical consultations and treatment received. In the U.K., medical services are provided without any patient co-payment, and people with epilepsy are exempt from prescription charges. Funding for health-care services is achieved through a nationalized system of insurance. Therefore, from the patients' perspective, there is no difference between the cost of each of the four AEDs considered, and the perspective of the National Health Service is similar to that of society as a whole. In contrast, patients with epilepsy in Turkey are likely to consult an epilepsy specialist directly and, in many cases, are charged a proportion of the costs of both medical consultations and the drug prescribed. In such a TABLE. Resource use: path C Lab tests, Annual frequency of Average number of Specialist annual serum AED Patients hospitalized Duration of A&E admissions per Country GP visits visits frequency determination per year (%) hospitalization (days) patient per year Belgium Czech Republic. France Italy. Netherlands. Poland.... Portugal.. Spain. 9 Sweden.. Switzerland. Turkey. U.K.. Epilepsia, Vol., Suppl..

6 ~ ~~ S D. C. HEANEY ET AL. TABLE. Resource use: path D Lab tests, Annual frequency of Average number of Specialist annual serum AED Patients hospitalized Duration of A&E admissions per Country GP visits visits frequency determination per year (%) hospitalization (days) patient per year Belgium Czech Republic France IdY Netherlands Poland Portugal Spain Sweden Switzerland Turkey U.K I a LTG levels not routinely measured. All percentages are rounded to the nearest %. a a I..... country, the economic impact of prescribing expensive treatments is spread between the patient and the health insurance systems. Further studies are being performed to consider the costs of treatment from these different perspectives. Appropriate outcome measures in epilepsy This study dealt only with direct medical costs. Consideration of broader issues of outcomes, such as productivity loss, quality of life, and other intangible benefits, may reveal differences among the drugs that were not captured in this study. However, at present there is no evidence to suggest that the drugs considered might have a differential effect on these outcomes in newly diagnosed epilepsy. This study aims to provide a simple economic evaluation of the direct medical costs that might be anticipated when patients with newly diagnosed epilepsy are treated with different monotherapies. Consideration of such broader outcomes, as listed above, would address wider social and political considerations that were beyond the scope of this work. Role of the pharmaceutical industry This study was supported by Sanofi-Synthelabo, the makers of Depakine, Epilim, and Ergenyl, which are for- mulations of valproate. Pharmacoeconomic appraisals performed with the assistance of the pharmaceutical industry have been criticized as being open to unacceptable bias (.). In this study, the authors maintained editorial independence. The method and results are clearly stated and the assumptions made have been tested in a sensitivity analysis, complying with guidelines for publication of pharmacoeconomic analyses (). Overall validity The external validity of this study can be confirmed only by comparing its results with a prospective economic appraisal of the use of AEDs in naturalistic settings. Very few cost-of-illness studies have been performed that consider the cost of epilepsy in the first year. In the U.K., a community-based study that examined the use of medical resources of patients with incident epilepsy demonstrated that the average cost of treating such patients was likely to be f in the first year (). This figure is broadly comparable to the costs obtained in the pharmacoeconomic model used in this study. It should be noted, however, that none of the above patients was treated with new AEDs, very few of them being available at the time this survey was performed. Unfortunately, no TABLE. Choice of drugs in mono- and polytherapy Proportion treated with nd-line Partial epilepsy Generalized epilepsy monotherapy after first-line monotherapy failure Country st choice nd choice st choice nd choice (%o) Belgium CBZ VPA VPA CBZ Czech Republic VPAlPHT CBZ PRIWHT VPA France CBZ VPA VPA PB/LTG IdY CBZ PB VPA PB Netherlands CBZ VPA VPA CBZ so Poland CBZ VPA VPA CBZ Portugal CBZ VPA CBZ VPA Spain CBZ PHT VPA PHT Sweden CBZ VPA VPA CBZ 9%. LTG % Switzerland CBZ VPA VPA LTG Turkey CBZ VPA VPA CBZ 9 U.K. CBZ VPA VPA PHT 9 Epikpsia VoL, Suppl.,

7 ~ ~ COST MINIMIZTION ANALYSIS OF AEDs s TABLE. Costs and charges in European countries (US $) Exchange rate ($ = local currency) Visit to GP Visit to specialists A&E attendance Daily cost of hospitalization Lab tests of biochem. and hematology CBZ check PHT check VPA check Drug costslweek CBZ mg/day LTG mg/day PHT mg/day VPA lo mgtday Czech Belgium Republic France Nether- Italy lands Poland Portugal Spain Sweden Switzerland Turkey U.K. oooo. NA similar cost-of-illness studies of incident epilepsy have been performed in other European countries. CONCLUSION Throughout Europe, direct costs of prescribing lamotrigine to patients with newly diagnosed epilepsy are likely to be two- to fourfold higher than costs incurred with CBZ, PHT, and VPA, which are similar to one another. This conclusion applies to all the healthcare systems considered in this study, despite variations in unit costs and medical practice. This is not surprising because, for most patients who develop epilepsy, the acquisition cost of AEDs represents a large proportion of their total treatment cost. The differences in side-effect profile and tolerability observed among the four drugs are too small to affect the direct medical costs of treatment in a substantial manner. Choosing to manage such patients with expensive new AEDs, the clinical benefits of which are not clear, is likely to significantly increase the overall cost of treatment to both the health-care system and to society as a whole. FIG.. Unit cost of medical services. REFERENCES. Marson AG, Kadir ZA, Chadwick DW. New anti-epileptic drugs: a systematic review of their efficacy and tolerability. BMJ 99; : I 9-.. Sander JWAS. New drugs for epilepsy. Curr Opin Neurol 99; :-.. Sander JWAS, Shorvon SD. Epidemiology of the epilepsies. J Neurol Neurosurg Psychiatry I99; I :.. Cockerell OC, Hart YM, Sander JWAS, Shorvon SD. The cost of epilepsy in the United Kingdom: an estimation based on the results of two population-based studies. Epilepsy Res 99;:9-.. Heaney DC, Sander JWAS, Shorvon SD. An economic appraisal of carbamazepine, lamotrigine, phenytoin and valproate as initial treatment for adults with newly diagnosed epilepsy. Epilepsia 99;9(~~ppl ): 9-.. Brodie MJ, Richens A, Yuen AWC. A randomised open multicentre comparative trial of lamotrigine and carbamazepine. Lancer 99;:-9.. Reunanen M, Mogens D. Yuen AWC. A randomised open multicentre comparative trial of lamotrigine and carbamazepine. Epilepsy Res 99;:9-.. Richens A, Davidson DLW, Cartilidge NEF, Easter DJ. A multicentre comparative trial of sodium valproate and carbamazepine in adult onset epilepsy. J Neurol Neurusurg Psychiatty 99::-. 9. Heller AJ, Chesteman P, Elwes RDC, et al. Phenobarbitone, phenytoin, carbamazepine or valproate for newly diagnosed adult epilepsy: a randomised comparative monotherapy trial. J Neurol Neurusurg Psychiatry 99;:-.. Caflaghan N, Kenny RA, O'Niell B, Crowley M. Goggin T. A prospective study between carbamazepine, phenytoin and valproate as monotherapy in previously untreated and recently diagnosed patients with epilepsy. J Neurol Neurusurg Psychiatry 9;:9.. Ramsay RE, Wilder BJ, Berger JR, Brunt J. A double-blind study comparing carbamazepine with phenytoin as initial seizure therapy in adults. Neurology 9;:9&.. Turnbull DM. Howel D, Rawlins MD. Chadwick D. Which drug for the adult epileptic patient: phenytoin or valproate? BMJ 9; Steiner TJ, Yuen AWC. Comparison of lamotrigine and phenytoin monotherapy in newly diagnosed epilepsy [Abstract]. Epilepsiu 99;(suppl ):.. Levy P. Pharmacoeconomic evaluation of medical treatment of epilepsy: where do we stand? Acra Neurol Belg 999;99:9.. Jones J, Hunter D. Qualitative research consensus methods for medical and health services research. BMJ 99;:-. Epilepsia, Vol., Suppl.,

8 s D. C. HEANEY ET AL Shakespeare A, Simeon G. Economic analysis of epilepsy treatment: a cost minimisation analysis comparing carbamazepine and lamotrigine in the UK. Seizure 99;:9-. Navarro RP, Ashraf T. Cost analysis of carbamazepine, phenytoin and valproate for use in complex partial seizures. Med interface 99;: -. Hughes D, Cockerell OC. A cost minimisation study comparing vigabatrin lamotrigine and gabapentin for the treatment of intractable epilepsy. Seizure O Niell B, Trimble MR, Bloom DS. Adjunctive therapy in epi- lepsy: a cost-effectiveness comparison of alternative treatment options. Seizure 99;:.. Neumann PJ. Paying the piper for pharmacoeconomic studies. Med Decision Making 99; :-.. Johannesson M, O Brien BJ. Economics, pharmaceuticals and pharmacoeconomics. Med Decision Making 99;: -.. Pausjenssen AM, Detsky AS. Guidelines for measuring the costs and consequences of adopting new pharmaceutical products: are they on track? Med Decision Making 99; : 9-. Epilepsiu, Vof.. SuppL, W