Cost-Effectiveness of Sertindole Among Atypical Antipsychotics In

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/vhri

Cost-Effectiveness of Sertindole among Atypical Antipsychotics in the Treatment of Schizophrenia in South Korea Bo-Ra-Mi Kim, MPH1, Tae-Jin Lee, PhD1,*, Hye-Jae Lee, MPH1, Bo-Hyun Park, PhD2, Bong-Min Yang, PhD1 1Graduate School of Public Health, Seoul National University, Seoul, South Korea; 2Department of Nursing, Kimcheon Science College, Kimcheon, South Korea

ABSTRACT

Objectives: This study assessed the cost-effectiveness of sertindole without relapse was 1.90 years for all study drugs. The estimated qual- compared with existing atypical antipsychotics in the treatment of pa- ity-adjusted life-years were 1.27 for sertindole, followed by quetiapine, tients with schizophrenia in the South Korean setting. Methods: A risperidone, and olanzapine. Total costs were 10.51 million Korean won Markov model was developed to estimate the cost-effectiveness of (KRW) for sertindole, 12.86 million KRW for olanzapine, 8.38 million sertindole compared with risperidone, olanzapine, and quetiapine KRW for risperidone, and 8.91 million KRW for quetiapine. The incre- with a cycle of 6 months on a 5-year time horizon. Effectiveness was mental cost-effectiveness ratios showed that sertindole was dominant deﬁned as the length of time without relapse and quality-adjusted life- only over olanzapine and was not cost-effective compared with risperi- years. Parameter estimates including drug-induced adverse events, done and quetiapine. Various sensitivity analyses conﬁrmed the re- compliance rate, and relapse rate were based on published literature sults from the base-case analysis. Conclusions: Sertindole may be and clinical trial data. Resource utilization data were obtained from the considered a valuable treatment option for South Korean patients who 2010 National Health Insurance reimbursement data, and costs were have failed the therapy with other atypical antipsychotic agents. estimated from the health care system’s perspective. A discount rate of Keywords: antipsychotics, atypical, cost-effectiveness, schizophrenia, 5% was applied to both cost and effectiveness. One-way sensitivity sertindole. analyses and probabilistic sensitivity analysis were carried out to check Copyright © 2012, International Society for Pharmacoeconomics and the robustness of the base-case analysis. Results: The length of time Outcomes Research (ISPOR). Published by Elsevier Inc.

weight gain, metabolic disorders including diabetes mellitus, som- Introduction nolence, and sexual dysfunction. The tolerability profiles differ Schizophrenia is a mental illness with substantial short-term and between atypical antipsychotic medications, and drug-induced long-term consequences for individuals, their families, the health side effects have been suggested to be one of the main factors care system, and society. Schizophrenia is a relatively common contributing to treatment nonadherence [5]. illness and the most common form of psychotic disorder. The 12- Sertindole (Serdolect) is an antipsychotic drug with affinity for month prevalence of schizophrenia reported in a South Korean dopamine D2, serotonin 5-HT2A and 5-HT2C, and alpha1-adreno- epidemiological study was 0.3% in 2006 [1]. The economic burden receptors. In Europe, sertindole was approved and marketed in 19 of schizophrenia is high [2]. According to National Health Insur- countries from 1996. In the United States, it first applied for Food ance (NHI) claims data in South Korea (hereafter, Korea), the ex- and Drug Administration (FDA) approval. But this application was penditure on schizophrenia accounted for 0.7% of total health in- withdrawn in 1998 following concerns over adverse events, that is, surance expenditures in 2008 [3]. In schizophrenia, early onset, the increased risk of sudden death from QT prolongation [6].In persistent psychotic symptoms even with antipsychotic treat- 1999, however, it was revealed that the adverse event was not ment, adverse events, frequent failure of treatments, and pro- associated with increasing rates of cardiac arrhythmias and that longed functional impairment all contribute to making schizo- patients on sertindole had the same overall mortality rate as those phrenia a particularly costly illness [4]. on risperidone using the results of the Sertindole Cohort Prospec- In the late 1990s, progress was made in the management of tive study [7]. As of January 2012, the drug has not been approved schizophrenia following the introduction of atypical antipsychotic by the FDA for use in the United States [8]. medications that demonstrated marked improvements in tolera- Despite concerns about safety issues, sertindole has a good bility profiles when compared with typical antipsychotic medica- tolerability profile, which is likely to favor long-term treatment tions. But current pharmacological options still carry some limita- adherence, reduce relapse, rehospitalization, and suicide, and im- tions. Atypical antipsychotic medications are associated with prove overall functioning. Sertindole is prescribed to patients who various adverse events such as extrapyramidal symptoms (EPS), are intolerant to at least one other antipsychotic agent. All pa-

Conﬂicts of interest: The authors have indicated that they have no conﬂicts of interest with regard to the content of this article. * Address correspondence to: Tae-Jin Lee, Department of Health Policy and Management, Graduate School of Public Health, Seoul National University, Seoul, South Korea. E-mail: [email protected]. 2212-1099/$36.00 – see front matter Copyright © 2012, International Society for Pharmacoeconomics and Outcomes Research (ISPOR). Published by Elsevier Inc. doi:10.1016/j.vhri.2012.03.015 60 VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65

Fig. 1 – Markov model for cost-effectiveness analysis. Comp +, compliant; Comp −, noncompliant; EPS, extrapyramidal symptoms; M, Markov; Rel +, relapse; Rel −, nonrelapse. tients should be started on sertindole at 4 mg/d. The dose should patients are assumed to disrupt the antipsychotic treatment for be increased by increments of 4 mg after 4 to 5 days on each dose a cycle of 6 months, after which they can either return to treat- until the optimal daily maintenance dose, usually within the ment because of relapse or remain as dropouts. Patients who range of 12 to 20 mg, is reached [7]. Electrocardiogram (ECG) mon- remain on treatment are then at risk of experiencing different itoring is required before and during treatment. adverse events: EPS, weight gain (and associated diabetes), sex- Two head-to-head comparisons of sertindole and risperidone ual dysfunction, somnolence, and other adverse events. The showed equivalent efficacy on positive symptoms such as delu- degree to which patients comply is assumed to be the same sion, hallucination, hyperactivity, conceptual disorganization, across medication regimens administered but to differ accord- and so on measured by the Positive and Negative Syndrome Scale. ing to the side effects experienced. The patients may therefore For negative symptoms such as emotional withdrawal, difficulty be compliant or noncompliant. At the end of the 6-month period in abstract thinking, and poor rapport, one study obtained equiv- covered by the model, patients can be in one of two health alent effects [9], while the other study obtained superior effects of states: relapse and nonrelapse. The risk of relapse increases sertindole to risperidone [10]. Sertindole should not be used as a with decreasing compliance to treatment. The patients with first-line treatment for first-episode patients with schizophrenia relapse are assumed to receive inpatient care in hospitals while because of QT prolongation. However, it has a side-effect profile the patients with no relapse are assumed to continue outpatient that makes it a favorable alternative for many patients who do not care (Fig. 1). respond well to the initial choice of antipsychotic drugs [11]. The length of a cycle was 6 months, which was based on clin- According to IMS health data in 2009, the market size of antip- ical practice patterns and expert opinion. The decision to use a sychotics was almost 140 billion Korean won (KRW) in Korea. 6-month cycle was clinically justified, because it is currently ac- Among the antipsychotics, risperidone accounted for 26.24% of cepted that any deterioration in schizophrenia that occurs within the market share, olanzapine 25.53%, and quetiapine 14.45%. 6 months following a relapse should be considered as being part of Sertindole is not launched yet in Korea because of delayed ap- that relapse [13]. As is commonly required in pharmacoeconomic proval of US FDA. But once it is approved by the Korean FDA, it is analyses, a 5-year time horizon was employed [14] and a discount expected to compete with the other antipsychotic drugs in this rate of 5% was applied to both cost and effectiveness. market. Using a decision analysis model, therefore, this study aimed to examine the cost-effectiveness of sertindole compared with ris- Data peridone, olanzapine, and quetiapine in the treatment of schizophrenia in the Korean health care setting. Clinical inputs Clinical inputs of the treatment are based on the results of random double-blind comparative clinical trials. We performed systematic Methods reviews by searching electronic databases: PubMed, EMBASE, Co- chrane Library (Central register of controlled trial; CENTRAL), MEDLINE (OVID), Korea medicine database (KMBASE), and RISS Study design database (produced by Korea Education and Research Information This study is a cost-effectiveness analysis of atypical antipsy- Service) from 1990 to March 2011. The keywords were “schizo*,” chotic drugs for the management of schizophrenia. Sertindole “relapse,” “hospitalization,” “sertindole,” “risperidone,” “olanzap- was compared with three atypical antipsychotic medications ine,” and “quetiapine” (Fig. 2). The inclusion criteria were to only that had the highest average market share for 5 years in Korea accept schizophrenia or schizoaffective patients, flexible dose, according to IMS health data: risperidone, olanzapine, and que- and head-to-head trials between comparator and risperidone as tiapine [12] (Table 2). common reference. Consequently, we selected seven randomized Markov model for cost-effectiveness analysis is particularly controlled clinical trials [10,15–20] (Table 1). suitable for the evaluation of chronic diseases such as schizophre- Drug-specific input data on adverse events for each drug nia. The study population consisted of treatment-resistant pa- were obtained through indirect comparison. Using meta-analy- tients with schizophrenia requiring hospitalization. It was as- sis, relative risks between drugs were derived on the basis of sumed that patients entered into the model on experiencing percentage of patients experiencing adverse events from se- intolerance to their antipsychotic treatment during an episode of lected articles (Table 2). acute psychopathology after already having received a previous Non–drug-specific input data used in this study were based on antipsychotic treatment. published articles (Table 2). These included premature dropout After starting treatment on the recommended daily dose of a rate, compliance rates, relapse rates by compliance, and mortality given drug, patients can either die or remain alive at the first rate. Dropout rates were derived from data on flexible doses for chance node of the decision tree. Patients then enter either of patients with schizophrenia [21]. Compliance rates depended on two possible paths: drop out or remain on treatment. Dropout different adverse events (EPS, weight gain, somnolence, and sex- VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65 61

medication taken, but indirectly dependent on the adverse events experienced because they impact compliance rates. Different relapse rates were applied depending on compliance [15]. The mortality rate of schizophrenia was estimated by using the prevalence rate of schizophrenia from Korea national statistical database [24] and the mortality rates of the whole mental disorder from a Ko- rean epidemiological study [1].

Economic input Costs were estimated from the health care system’s perspective. Indirect costs such as productivity loss were not included because of limited availability of Korean data. Direct costs consist of health care costs and non–health care costs. Commonly, health care costs include the costs for physician visits, medication, hospitalization, laboratory tests, mental therapy, mental health day-care center, and adverse events. Health care costs were estimated by gross costing by using 2010 Health Insurance Review and Assess- ment Service (HIRA) data in Korea [3]. HIRA data are a claims data of NHI that include frequency of patient visit and total cost of health care utilization by disease classification. Annual costs of in-/outpatient care for schizophrenia were obtained from HIRA data and were divided by 2 to get 6-month costs for model input. The same costs of adverse events were applied for each comparator in the model, except for ECG monitoring cost added on sertindole. It was assumed that there was no difference in health care resource utilization among adverse events (EPS, weight gain, somnolence, and sexual dysfunction) except for medication costs. The drug costs of benzotropine for EPS, metformin for diabetes, and modafinil for somnolence were included while weight gain and sexual dysfunction were assumed to have no prescription. The drug costs included in the model were obtained from the weighted average annual drug price in HIRA [25] (Table 3). Prices were presented according to the mean daily dose for each drug by inpa- Fig. 2 – Flow of systematic review. RCT, randomized tients/outpatients. The mean daily doses for sertindole and com- controlled trial. parators were based on a previous study in Korea [27]. All the health care costs were adjusted for inflation in 2010 by health care inflation rate of national statistics [28]. ual dysfunction). And noncompliance to antipsychotic treatment Non–health care costs included time and travel costs of the is clearly associated with an increased risk of schizophrenic re- patients in the treatment of schizophrenia. Estimation of time lapse [22]. Relapse rates were independent of the antipsychotic and travel costs was based on the data from 2005 Korea National

Table 1 – Summary of included studies. Source Country Diagnosis Treatment Comparison F/U period Drugs Dose n Drugs Dose n (mg/d) (mg/d)

Conley and Mahmoud [19] United States Schizophrenia or Olanzapine 5–20 189 Risperidone 2–6 188 8 wk schizoaffective disorder (DSM-IV) Gureje et al. [20] Australia, Schizophrenia, Olanzapine 10–20 32 Risperidone 4–8 33 30 wk New schizoaffective Zealand disorder, or schizophreniform disorder (DSM-IV) Lieberman et al. [15] United States Schizophrenia (DSM-IV) Olanzapine 7.5–30 336 Risperidone 1.5–6.0 341 18 mo Quetiapine 200–800 337 Alvarez et al. [16] Spain Schizophrenia (DSM-IV) Olanzapine 10ϩ 120 Risperidone 3ϩ 115 12 mo Azorin et al. [10] France Schizophrenia (DSM-IV) Sertindole 12–24 98 Risperidone 4–10 89 12 wk Potkin et al. [17] United States Schizophrenia or Quetiapine 400–600 156 Risperidone 4–6 153 42 d schizoaffective disorder (DSM-IV) Zhong et al. [18] United States Schizophrenia (DSM-IV) Quetiapine 200–800 338 Risperidone 2–8 335 8 wk

DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; F/U, follow up. 62 VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65

Table 2 – Model input: Effectiveness data. Sales market share of atypical antipsychotics in Korea1 Value

Risperidone 26.24% Olanzapine 25.53% Quetiapine 14.45%

Drug-speciﬁc input2 Value Non–drug-speciﬁc input Value

Adverse event RR-EPS Dropout rate3 Sertindol vs. risperidone 0.41 (0.13-1.28) EPS 3.97% Olanzapine vs. risperidone 0.68 (0.51-0.91) Weight gain 3.97% Quetiapine vs. risperidone 0.57 (0.42-0.77) Somnolence 1.78% Adverse event RR-Weight gain Sexual dysfunction 4.86% Sertindol vs. risperidone 1.31 (0.71-2.44) Total 14.59% Olanzapine vs. risperidone 2.16 (1.57-2.98) Compliance rate4 Quetiapine vs. risperidone 1.08 (0.81-1.44) EPS 96.03% Adverse event RR-Somnolence Weight gain 96.03% Sertindol vs. risperidone 1.05 (0.40-2.77) Somnolence 98.22% Olanzapine vs. risperidone 1.02 (0.86-1.20) Sexual dysfunction 95.14% Quetiapine vs. risperidone 1.25 (1.05-1.49) Total 85.41% Adverse event RR-Sexual dysfunction Conditional probability5,* Sertindol vs. risperidone 4.59 (1.03-20.37) P(relapseϩ|compϩ)† 0.12 Olanzapine vs. risperidone 1.30 (0.99-1.71) P(relapseϩ|compϪ) 0.42 Quetiapine vs. risperidone 1.30 (0.99-1.71) P(relapseϪ|compϩ) 0.89 HRQOL6 P(relapseϪ|compϩ) 0.58 Healthy: 0.88 (Ϯ0.20), illness: 0.73 (Ϯ0.31) Mortality rate of schizophrenia7 2.82 per 10,000 person

Sources. 1IMS Health Data [12]; 2Azorin et al. [10] and meta analysis; 3Martin et al. [21]; 4Hansen et al. [22]; 5Lieberman et al. [15]; 6Seong et al. [23]; and 7Korea Statistical Information Service [24]. EPS, extrapyramidal symptoms; HRQOL, health-related quality of life; RR, relative risk. * Conditional probability p(x | y): Probability of x given that y. † P(relapseϩ|compϩ) means the probability of relapse(ϩ) occurring under compliant patient(ϩ), and P(relapseϪ|compϪ) means the probability of no relapse(Ϫ) occurring under noncompliant patient(Ϫ).

Health and Nutrition Examination Survey (NHANES) [26]. Time costs the normal distribution for all the costs according to the guideline were estimated by multiplying the average number of visits to hos- of pharmacoeconomic evaluation in HIRA [29]. pital or clinic obtained from NHANES by average wage and employ- All analyses were performed by using Tree-Age Pro 2009 (Tree- ment rate in 2010 [28]. The transportation costs of outpatients and Age Software, Williamstown, MA). inpatients obtained from NHANES were adjusted for inﬂation during the period between 2005 and 2010 by 11.8%. Results

Utility Based on the model, sertindole showed nearly equivalent out- To get quality-adjusted life-years (QALYs), utility values for each comes compared with its comparators. The length of TwR was 1.90 state experienced in the treatment pathways were employed. The years for all study drugs during 5 years on model projection. The utility weights were obtained from the published data of health- estimated QALYs were 1.27 for sertindole, followed by quetiapine, related quality of life of illness from the Korean population, which risperidone, and olanzapine. Total costs including in- and outpa- was measured by using EuroQol ﬁve-dimensional questionnaire tient care were 10.51 million KRW for sertindole, 12.86 million validated in Korean language [23]. The utility weight was multi- KRW for olanzapine, 8.38 million KRW for risperidone, and 8.91 plied by the time spent in that state for each branch of the model million KRW for quetiapine. The incremental cost-effectiveness and summed across all branches to get QALYs. ratios showed that sertindole was dominant only over olanzapine and was not cost-effective compared with risperidone and quetia- Analysis pine (Table 4). The main outcome measure was time without relapse (TwR). In Various one-way sensitivity analyses showed that sertindole addition, QALYs were calculated on the basis of utility values of was still a dominant alternative to olanzapine while it had very each health state in the model. The costs of treatment were high incremental cost-effectiveness ratios compared with risperi- assessed on the basis of typical resource use in Korea associated done and quetiapine. The compliance rate connected to EPS had to with each different treatment path. The cost per outcome, that be increased to 50% to make sertindole a cost-neutral alternative is, either TwR or QALY, was expressed by using incremental to comparators. The main driver in the difference in total costs cost-effectiveness ratio of sertindole versus comparators. between sertindole and comparators was inpatient care. When To check uncertainties for economic modeling for schizophre- the resource utilization of inpatient care was reduced by 50%, nia, we conducted one-way sensitivity analysis and probability sertindole was still the dominant treatment strategy compared sensitivity analysis (PSA). The sensitivity analyses were carried with olanzapine. Changes in the time frame of the analysis from 5 out on parameters such as compliance rate, relapse rate, drug cost years to 3 years and 10 years did not affect the results from the of sertindole, and discount rate. PSA was conducted on the as- base-case analysis. Neither did the changes in the discount rate sumption of the beta distribution for effectiveness parameters and from 5% to 3% and 7%. On the other hand, the results of PSA, VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65 63

Table 3 – Model input: Cost data. Direct health care cost Value

Outpatient Inpatient

Drug mean dose and cost per day1 Sertindole 16 mg/d 3,812 16 mg/d 3,812 Risperidone 4.1 mg/d 1,640 5.25 mg/d 1,936 Olanzapine 10.85 mg/d 5,381 14.93 mg/d 7,715 Quetiapine 300 mg/d 1,925 400 mg/d 2,994 Outpatient visit cost2 73,368/d Inpatient cost2 3,540,352/patient ECG examination2 19,040/y Adverse event treatment cost1 EPS 980 Diabetes 72,074 Sedation 53,228 Direct non–health care cost3 Transportation cost Outpatient per visit 8,830 Inpatient per visit 22,544 Time cost Travel, waiting, and treatment for outpatient visit 80,064

Sources. 1) It was calculated by HIRA formula of pharmacoeconomic guideline [29] and 2010 annual HIRA weighted average price [25]; 2) National Health Insurance Corporation [3]; 3) Korea Center for Disease Control [26]. Note. Unit: Korean Won, . ECG, electrocardiogram; EPS, extrapyramidal symptoms; HIRA, Health Insurance Review and Assessment Service. presented in cost-effectiveness scatter plots, showed that sertin- The purpose of this pharmacoeconomic analysis was to exam- dole had nearly the equivalent effectiveness to its comparators ine whether sertindole still fulfills the criteria for general reim- and that it was cheaper than only olanzapine among other atypi- bursement in the Korean NHI. The model was adapted to fit clinical antipsychotic medications (Fig. 3). cal practice, current reimbursement criteria, and treatment costs in Korea. Because of cardiovascular safety concerns, sertindole should be used only for those patients who are intolerant to at Discussion least one other antipsychotic agent. By performing the cost-effectiveness analysis, it was possible In schizophrenia, failure in therapy is particularly expensive to compare the different atypical antipsychotic agents in the treat- and is due to several factors such as lack of efficacy, side effects, ment of patients with schizophrenia. The results showed that nonresponsiveness to treatment, and repetitive hospitaliza- sertindole was able to slightly increase the length of TwR and tions. QALY compared with other atypical antipsychotics for patients When compared with typical antipsychotic medication, with chronic schizophrenia. With regard to total treatment costs, atypical antipsychotic medications have improved the treatment of schizophrenia because they are associated with signif- sertindole was cost-saving compared with olanzapine, but it was icantly fewer extrapyramidal side effects. This has, however, involved in higher costs compared with risperidone and quetiap- renewed attention toward other side effects such as weight ine. In other words, sertindole was dominant over olanzapine gain, diabetes, sexual dysfunction, and somnolence. Because of (greater effect at lower cost), but it was not a cost- effective alter- the individual patients’ response and tolerance of specific com- native to risperidone and quetiapine. pounds, many patients need to switch from one compound to This model was built on the hypothesis that the side effects another. Consequently, important unmet needs exist for treat- from antipsychotic therapy would influence the compliance rate ments that can fulfill individual patients’ needs in terms of re- of the antipsychotic medication in question. It is a well-known fact sponse and tolerability. that noncompliance to antipsychotic treatment is associated with

Table 4 – Results of base-case cost-effectiveness analysis. Item Sertindole Risperidone Olanzapine Quetiapine

Effectiveness TwR (y) 1.90 1.90 1.90 1.90 QALY (y) 1.27 1.27 1.27 1.27 Costs (KRW) 10.51 M 8.38 M 12.86 M 8.91 M Sertindole vs. comparator ICER (TwR) 4260 M Dominant 8000 M ICER (QALY) 710 M Dominant 1600 M

ICER, incremental cost-effectiveness ratio; KRW, Korean won; M, million; QALY, quality-adjusted life-year; TwR, time without relapse. 64 VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65

because there have not been many studies conducted on schizophrenia in Korea, this study is meaningful in that it is a pharmacoeconomic evaluation using the Markov model and the PSA.

Conclusions

Sertindole should not be used as first-line treatment for first-episode patients with schizophrenia because of the QT prolongation. How- ever, it has a side-effect profile that makes it a useful alternative for many patients who do not respond well to the initial choice of antipsychotic drug. Sertindole demonstrated nearly equivalent outcomes to other atypical antipsychotics and proved to be a cost-saving alternative to olanzapine. Therefore, it is concluded that sertindole may be considered as a treatment option for Korean patients who have failed the therapy with other atypical antipsychotic agents. Source of financial support: These findings are the result of work supported by Lundbeck Korea Co. Ltd. The views expressed in this Fig. 3 – Cost-effectiveness (CE) scatter plot. article are those of the authors, and no official endorsement by Lun- dbeck Korea Co. Ltd is intended or should be inferred. an increased risk of schizophrenic relapse [30]. The effectiveness of antipsychotic drug treatment in terms of TwR was therefore REFERENCES estimated in the model on the basis of drug-specific data (percent- ages of patients experiencing certain adverse events) derived from clinical trial data. Only five adverse events (EPS, weight gain, diabetes, somnolence, and sexual dysfunction) were included be- [1] Cho MJ, Ham BJ, Hong JP, et al. The Epidemiological Survey of cause those were the major concomitant diseases and had great Psychiatric Illnesses in Korea 2006. Seoul, South Korea: Ministry of effects on the compliance of schizophrenia patients. Health and Welfare in Korea and Seoul National University, 2007. [2] Fadden G, Bebbington P, Kuipers L. The burden of care: the impact of Treatment with sertindole is associated with an increased risk of functional psychiatric illness on the patient’s family. Br J Psychiatry QT prolongation. However, patients are closely monitored during 1987;150:285–92. treatment and have their ECG taken every third month. QT prolon- [3] National Health Insurance Corporation, Health Insurance Review and gation was not handled as a separate adverse event in the model, Assessment Service. National Health Insurance Statistical Yearbook 2010. Available from: http://www.nhic.or.kr/portal/site/main/ only with the inclusion of the ECG costs in the sertindole branch. menuitem.33f399436ee39a404bf15151062310a0. [Accessed July 1, 2011]. Adverse events defined as “other adverse events” included all events [4] Kuipers L. Family burden in schizophrenia: implications for services. except EPS, weight gain, diabetes, somnolence, and sexual dysfunc- Soc Psychiatry Psychiatr Epidemiol 1993;28:207–10. tion. Because of the broad spectrum of other adverse events, they [5] Kim YS. Pharmacological Treatment of Schizophrenia. Seoul, South Korea: Iljogak, 1998. were not associated with a cost in the model. Because sertindole had [6] World Health Organization. Regulatory actions: Sertindole—approval the lowest rate of “other” events among atypical antipsychotic application withdrawn. Pharmaceuticals Newsletter 1998, No. 03 and agents in the model, the strategy of not including the costs for these 04. Available from: http://apps.who.int/medicinedocs/en/d/Js2256e/ events was considered to be conservative. 1.12.html#Js2256e.1.12. [Accessed April 30, 2011]. [7] Peuskens J, Tangh P, Mittoux A. The Sertindole Cohort Prospective There are some limitations in this study. One of the limitations (SCoP) study: rationale, design and methodology. Pharmacoepidemiol was the absence of direct comparative clinical trial data between Drug Saf 2008;17:425–33. sertindole and its comparators, quetiapine and olanzapine, respec- [8] US Food and Drug Administration. Available from: http://www.fda tively. This lack of direct comparison was resolved by employing .gov. [Accessed January 10, 2012]. [9] Hale A, Azorin J, Kasper S, et al. Sertindole improves both the positive indirect comparisons of sertindole versus those two comparators and negative symptoms of schizophrenia: results of a phase III trial. with risperidone as common reference. Another limitation was lack Int J Psychiatry Clin Pract 2000;4:55–62. of the detailed empirical data of resource utilization of Korean pa- [10] Azorin J, Strub N, Loft H. A double-blind, controlled study of sertindole tients with schizophrenia. For example, such information as the versus risperidone in the treatment of moderate-to-severe schizophrenia. Int Clin Psychopharmacol 2006;21:49–56. number of hospital/clinic visits of individual patients, the type and [11] Lewis R, Bagnall A, Leitner M. Sertindole for schizophrenia. Cochrane amount of health services after relapse, and the type and cost of Database of Systematic Reviews. 2009. Available from: http:// treatment for adverse events was not available because of protection onlinelibrary.wiley.com/doi/10.1002/14651858.CD001715.pub2/abstract. of personal information. Instead of the microcosting method, there- [Accessed April 30, 2011]. [12] IMS Health Data. 2010 Total Market Share in Antipsychotics by Products in fore, this study used the gross-costing method based on aggregated Republic of Korea. 2011. Available from: http://www.imshealth.com/portal/ number of patients and hospital/clinic visits for 1 year and annual site/ims/menuitem.edb2b81823f67dab41d84b903208c22a/?vgnextoidϭ total reimbursement cost in schizophrenia. Besides, indirect costs dfa33cd32705f210VgnVCM10000071812ca2RCRD&vgnextfmtϭdefault. were not considered because of the lack of data on productivity loss [Accessed March 1, 2011]. [13] Frances A, Pincus HA, Firste MB, et al. Statistical Manual of Mental of the patients with schizophrenia in Korea. Disorders (4th ed.). Washington, DC: American Psychiatric Despite the limitations mentioned above, this study has some Association, 2000. implications in a few aspects. First, in terms of decisions on the re- [14] National Institute for Clinical Excellence. Guidance for Manufacturers imbursement in the Korean NHI, the study results suggest the possi- and Sponsors: Technology Appraisals Process Series No. 5. London: National Institute for Clinical Excellence, 2001. bility that sertindole could be used as another option of atypical [15] Lieberman J, Stroup T, McEvoy J, et al. Effectiveness of antipsychotic antipsychotic medication for the treatment of patients with schizo- drugs in patients with chronic schizophrenia. N Engl J Med 2005;353: phrenia who do not respond to initial medication. Second, it is antic- 1209–23. ipated that, though not yet launched, sertindole will be able to have a [16] Alvarez E, Ciudad A, Olivares JM, et al. A randomized, 1-year follow-up study of olanzapine and risperidone in the treatment of negative room for the market share and competition among atypical antipsy- symptoms in outpatients with schizophrenia. J Clin Psychopharmacol chotic drugs in Korea if its price is set at an appropriate level. Third, 2006;26:238–49. VALUE IN HEALTH REGIONAL ISSUES 1 (2012) 59–65 65

[17] Potkin S, Gharabawi G, Greenspan A, et al. A double-blind comparison [24] Korea Statistical Information Service. 2010 All Causes of Crude of risperidone, quetiapine and placebo in patients with schizophrenia Mortality Rate in Korea. 2011. Available from: http://kosis.kr/abroad/ experiencing an acute exacerbation requiring hospitalization. abroad_01List.jsp. [Accessed April 29, 2011]. Schizophr Res 2006;85:254–65. [25] Health Insurance Review and Assessment Service. The Weighted Price [18] Zhong K, Sweitzer D, Hamer R, et al. Comparison of quetiapine and of Drug by Component in the First Half Year of 2010. Seoul, South risperidone in the treatment of schizophrenia: a randomized, double- Korea: HIRA, 2011. blind, ﬂexible-dose, 8-week study. J Clin Psychiatry 2006;67:1093–103. [26] Korea Center for Disease Control. 2005 Korea National Health and [19] Conley R, Mahmoud R. A randomized double-blind study of Nutrition Examination Survey. Available from: http://knhanes risperidone and olanzapine in the treatment of schizophrenia or .cdc.go.kr/. [Accessed July 1, 2011]. schizoaffective disorder. Am J Psychiatry 2001;158:765–74. [27] Na C, Kim CW, Jeon WT, et al. Mental Illness Optimal Treatment [20] Gureje O, Miles W, Keks N, et al. Olanzapine vs risperidone in the Model for Each Type. Seoul: Ministry of Health and Welfare in Korea, management of schizophrenia: a randomized double-blind trial in 2001. Australia and New Zealand. Schizophr Res 2003;61:303–14. [28] Korea Statistical Information Service. Available from: http://www [21] Martin J, Perez V, Sacristan M, et al. Meta-analysis of drop-out rates in .kosis.kr. [Accessed July 3, 2011]. randomised clinical trials, comparing typical and atypical [29] Health Insurance Review and Assessment Service. The Guideline of antipsychotics in the treatment of schizophrenia. Eur Psychiatry 2006; Pharmacoeconomic Evaluation and Standard Costing. Seoul, South 21:11–20. Korea, 2006. Available from: http://www.hira.or.kr/dummy.do? [22] Hansen K, Launois R, Sapin C. Pharmacoeconomic positioning of pgmidϭHIRAA910001000000&cmsurlϭ/cms/information/04/01/05/05/ sertindole among atypical antipsychotics in the management of 1210820_13530.html. [Accessed September 30, 2011]. schizophrenia in Turkey. Presented in: iHEA 6th World Congress: [30] Bebbington P, Angermeyer M, Azorin J, et al., for the EuroSC Research Explorations in Health Economics. Copenhagen, Denmark, 2007. Group. The European Schizophrenia Cohort (EuroSC): a naturalistic [23] Seong S, Choi C, Sung Y, et al. Health-related quality of life using EQ- prognostic and economic study. Soc Psychiatry Psychiatr Epidemiol 5D in Koreans. J Korean Rheum Assoc 2004;11:254–62. 2005;40:707–17.