Comparison of the effectiveness of brand-name and generic ...

SCHRES-07352; No of Pages 7 Schizophrenia Research xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Schizophrenia Research journal homepage: www.elsevier.com/locate/schres

Comparison of the effectiveness of brand-name and generic antipsychotic drugs for treating patients with schizophrenia in Taiwan Chih-Wei Hsu a, Sheng-Yu Lee b,c,d, Liang-Jen Wang d,f,c,e,⁎ a

Department of Psychiatry, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Kaohsiung, Taiwan Department of Psychiatry, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan Department of Psychiatry, School of Medicine, and Graduate Institute of Medicine, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan d Department of Psychiatry, College of Medicine, National Yang-Ming University, Taipei, Taiwan e Department of Child and Adolescent Psychiatry, Kaohsiung Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Kaohsiung, Taiwan f Department of Chinese Medicine, College of Medicine, Chang Gung University, Taoyuan, Taiwan b c

a r t i c l e

i n f o

Article history: Received 13 January 2017 Received in revised form 9 June 2017 Accepted 11 June 2017 Available online xxxx Keywords: Brand-name Generic Antipsychotics Psychosis Effectiveness Cohort study

a b s t r a c t The purpose of this nationwide population-based study is to compare the long-term effectiveness of brand-name antipsychotics with generic antipsychotics for treating schizophrenia. We identified patients with schizophrenia who were prescribed antipsychotics from a random sample of one million records from Taiwan's National Health Insurance database, observed between January 1, 2000 and December 31, 2012. Only those with no prior use of antipsychotics for at least 180 days were included. We selected patients who were prescribed brand-name risperidone (N = 404), generic risperidone (N = 145), brand-name sulpiride (N = 334), or generic sulpiride (N = 991). The effectiveness of the treatments researched in this study consisted of average daily doses, rates of treatment discontinuation, augmentation therapy, and psychiatric hospitalization. We found that compared to patients treated with generic risperidone, those treated with brand-name risperidone required lower daily doses (2.14 mg vs. 2.61 mg). However, the two groups demonstrated similar rates of treatment discontinuation, augmentation, and psychiatric hospitalization. On the other hand, in comparison with patients prescribed generic sulpiride, those treated with brand-name sulpiride not only required lower daily doses (302.72 mg vs. 340.71 mg) but also had lower psychiatric admission rates (adjusted hazard ratio: 0.24, 95% confidence interval: 0.10–0.56). In conclusion, for both risperidone and sulpiride, higher daily doses of the respective generic drugs were prescribed than with brand-name drugs in clinical settings. Furthermore, the brand-name sulpiride is more effective at preventing patients from hospitalization than generic sulpiride. These findings can serve as an important reference for clinical practices and healthcare economics for treating schizophrenic patients. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Schizophrenia is a serious mental disorder, and schizophrenic patients often require inpatient treatment when the acute disease is exacerbated (van Os and Kapur, 2009). Patients with schizophrenia are often treated using antipsychotic drugs (De Oliveira and Juruena, 2006), which are usually classified as either first or second generation antipsychotics (Meltzer et al., 1989). Continuous antipsychotic treatment can successfully improve patients' psychotic symptoms and prevent them from relapse (Harvey and Keefe, 2001; Leucht et al., 2009; Lieberman et al., 2005; Masi and Liboni, 2011). Once the patents of the original antipsychotic compounds expired, corresponding generic formulations ⁎ Correspondence to: L-J. Wang, Department of Child and Adolescent Psychiatry, Kaohsiung Chang Gung Memorial Hospital, Kaohsiung city, Taiwan, No.123, Ta-Pei Road, Kaohsiung City, Taiwan. E-mail address: [email protected] (L.-J. Wang).

entered the market as competing prescription options (Borgheini, 2003). Therefore, understanding the long-term treatment effectiveness of both generic drugs and their original compounds for schizophrenic patients is crucial for clinical practice and health-care economics. Currently, several studies have measured blood concentration and hemodynamics between brand-name and generic antipsychotics (Boonleang et al., 2010; Chen et al., 1989; Elshafeey et al., 2009; Khorana et al., 2011; Liu et al., 2013; Mahatthanatrakul et al., 2008; van Os et al., 2007) and have generally demonstrated that generic drugs exhibit bioequivalence, similar tolerability, and comparable safety profiles with the original compound (brand-name drugs) (Frank, 2007; Gyorgy et al., 2008). However, these studies have small sample sizes, and the study populations were restricted to healthy subjects. Therefore, the results of such pharmacodynamics or pharmacokinetic studies may not be applicable to the field of clinical practice (Desmarais et al., 2011). In clinical studies, the findings related to potential differences in treatment effects between brand-name and generic antipsychotics

http://dx.doi.org/10.1016/j.schres.2017.06.020 0920-9964/© 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Hsu, C.-W., et al., Comparison of the effectiveness of brand-name and generic antipsychotic drugs for treating patients with schizophrenia in Taiwan, Schizophr. Res. (2017), http://dx.doi.org/10.1016/j.schres.2017.06.020

2

C.-W. Hsu et al. / Schizophrenia Research xxx (2017) xxx–xxx

have been mixed. Regarding second generation antipsychotic drugs, some studies have reported that generic clozapine had similar efficacy and safety profiles with generic clozapine, but the daily doses of generic clozapine were generally higher than the brand-name drug (Bobo et al., 2010; Healy et al., 2005; Italiano et al., 2015; Oluboka et al., 2010; Paton, 2006). (Gyorgy et al., 2008) indicated that both generic and original quetiapine demonstrated therapeutic equivalence. One study of New Zealand's Pharmaceutical Management Agency records found that for patients switching from brand-name to generic olanzapine, their clinical outcomes did not worsen (Lessing et al., 2015). However, switching from brand-name to generic risperidone has been reported to be associated with a loss of efficacy or the occurrence of side effects (Hardan et al., 2010). With regard to first generation antipsychotics, (Verster et al., 1998) observed no significant differences in the symptom changes between schizophrenia patients treated with brand-name and generic fluphenazine decanoate. Although manufacturers claim that generic drugs generate savings with regard to medical expenditures (Haas et al., 2005), whether generic antipsychotics provide long-term effectiveness equivalent with the brand-name drugs in the real world remains unclear. To fill the research gap, we used a claims database consisting of a representative nationwide sample to determine the long-term treatment outcomes of schizophrenic patients. The aim of this retrospective cohort study was to compare the effectiveness of generic antipsychotics and their original compounds (brand-name drug), using average daily doses, rates of treatment discontinuation, augmentation therapy, and psychiatric hospitalization as indicators of clinical effectiveness. 2. Methods 2.1. Data source Data for this study were obtained from the ambulatory claims database of the National Health Insurance Research Database (NHIRD). NHIRD comprises the reimbursement medical claims of the National Health Insurance (NHI) program in Taiwan, which started on March 1, 1995. The Bureau of NHI is the sole payer for healthcare services under the NHI program and covers 93% of all of Taiwan's healthcare providers. Participating medical care institutions are required to electronically submit monthly claim documents related to medical expenses by the 20th day of the following month. Such documents include such information as patient demographic data, diagnostic codes, medical institutions visited, dates of prescriptions, drugs prescribed, and claimed medical expenses. Individual and hospital identifiers are unique to the NHIRD and cannot be used to trace individual patients or medical care institutions. The reliability of diagnostic codes in the NHIRD has been proven by a previous study (Multhoff et al., 2014). The protocol for this study conformed to the Helsinki Declaration, and was approved by the Institutional Review Board (IRB) of Chang Gung Memorial Hospital (IRB No: 103-0637B). Patient records/information was anonymized and de-identified prior to analysis, and the need for written informed consent was waived by the IRB.

NHI claim records, including amisulpride, aripiprazole, clozapine, olanzapine, quetiapine, risperidone, and sulpiride. The remaining patients were classified according to the antipsychotic agent that was prescribed to them as of the index date. To eliminate the confounding effect of drug interactions, we established the following exclusion criteria: (1) patients who had been prescribed another antipsychotic agent within 180 days before using the selected antipsychotic; (2) the observation period after the initial prescription of the selected antipsychotic was b180 days; (3) patients aged b 18 years or ≥65 years at the index date of the selected antipsychotic prescription; (4) patients who were prescribed more than one antipsychotic agent at the index date; (5) patients who had used the long-acting injectable form of risperidone (only the brand-name drug but no generic drug with a long-acting injectable form was available in Taiwan); and (6) patients who were first prescribed risperidone prior to June 01, 2004 (the time at which the first generic risperidone appeared on the market in Taiwan). Using the aforementioned selection criteria, the case numbers of patients who had been prescribed at least one dose of antipsychotic drug with brand-name and generic forms are provided in the Supplementary Table 1. We found that patients treated with risperidone (N = 727) and sulpiride (N = 1916) were sufficient for comparing the effectiveness between the brand-name and generic drugs. Of the selected patients, 178 received both brand-name and generic risperidone prescriptions, and 591 patients received both brand-name and generic sulpiride prescriptions. To eliminate the cross-over effect and the confounding effect of the disease's course (e.g., most patients with mixed drug use were initially prescribed a brand-name drug and were subsequently switched to a generic drug), we excluded patients with mixed drug use. We ultimately obtained 404 patients who only used brand-name risperidone (risperidone-B group) and 145 patients who only used generic risperidone (risperidone-G group), 334 patients who only used brand-name sulpiride (sulpiride-B group), and 991 patients who only used generic sulpiride (sulpiride-G group) for further analysis. Fig. 1 is a flow chart of the detailed patient selection procedure. 2.3. Demographics and comorbidities In addition to age and gender, we listed the year of initial use of the selected antipsychotic agent, employed the Charlson Comorbidity Index (CCI) before the initial use of the selected drug to determine general health status (Deyo et al., 1992), and evaluated the comorbidity of psychiatric disorders before initial use of the selected medication. The CCI was calculated using diagnostic codes from outpatient records and discharge codes from hospitalization records, a method that is widely used for confounders in epidemiological research (Schneeweiss et al., 2001). The psychiatric comorbidities were defined as any ICD-9-CM from the same medical records and include alcohol use disorders (ICD-9-CM: 291.x, 303.x, 305.0, 357.5, 425.5, 535.3, and 571.0–571.3), substance use disorders (ICD-9-CM: 292.x, 304.x, and 305.2–305.9), mood disorders (ICD-9-CM: 296.2, 296.3, 300.4, and 311.x), anxiety disorders (ICD-9-CM: 300.x except 300.4), and sleep disorders (ICD-9-CM: 307.4 and 780.5).

2.2. Study subjects

2.4. Outcome variables

We included all patients in the NHIRD who had been diagnosed with schizophrenia-spectrum disorder and were prescribed at least one dose of antipsychotic drug between January 1, 2000 and December 31, 2012. A patient with schizophrenia-spectrum disorder was defined as a patient with at least one inpatient or outpatient record pursuant to the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code 295.x. After excluding patients whose gender was not recorded (N = 24), our study sample consisted of 9651 patients. The prescription of antipsychotic drugs whose brandname and generic forms are both available in Taiwan was traced from

In this study, the outcome measures of treatment effectiveness were average daily dose, medication discontinuation, augmentation therapy with another antipsychotic drug, and psychiatric hospitalization. All subjects were observed from the index date (on which the selected antipsychotic was initially prescribed) to the discontinuation date or December 31, 2012. The average daily dose was defined as the dose of antipsychotic on the day of medication discontinuation or the end of follow-up. Medication discontinuation was defined as cessation of risperidone-B, risperidone-G, sulpiride-B, or sulpiride-G for 60 days or longer. We defined augmentation therapy as when a patient treated with



3

Fig. 1. Flowchart showing the selection procedure of study subjects.

the selected antipsychotic had another antipsychotic added on for more than seven days. Psychiatric hospitalization was defined as a patient being admitted to a psychiatric ward while being treated with the selected antipsychotics. 2.5. Statistical analysis All statistical analyses were performed using the MedCalc Statistical Software version 16.8.4 (MedCalc Software bvba, Ostend, Belgium; https://www.medcalc.org; 2016). We applied the Kolmogorov-Smirnov test to verify the normal distribution of data. We used the chi-squared test, independent samples t-test, or Mann-Whitney U test (CCI scores and daily doses) to compare the data of brand-name drug groups and generic drug groups. A 2-tailed p b 0.05 was considered statistically significant. We adopted Kaplan–Meier curves for survival analysis. The time function was calculated as the number of days from the initial observation until drug discontinuation or December 31, 2012 (end of followup). Follow-up time was counted from the initial date of antipsychotic prescription to the date of the following outcomes (drug discontinuation, augmentation therapy, or psychiatric hospitalization). We used Cox regression models to determine the effectiveness of different antipsychotics, controlling for gender, age, year of the initial prescription of the selected antipsychotic, CCI scores, and psychiatric comorbidities. We calculated the hazard ratios of the treatment effectiveness of the antipsychotics, using patients treated with the generic drug as the reference group. Adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) were calculated. 3. Results Table 1 shows the characteristics of patients treated with brandname (B) and generic (G) forms of risperidone and sulpiride, respectively. Of the 549 patients treated with risperidone, the risperidone-B group was more likely to be recruited in the former study period and less likely to have an underlying sleep disorder in comparison to the risperidone-G group. Of the 1325 patients treated with sulpiride, the

sulpiride-B group was more likely to be recruited in the former study period and to have comorbid mood disorders when compared to the sulpiride-G group. We found no relevant difference in gender, age, medical comorbidities (CCI scores), or other psychiatric comorbidities between the brand-name drug and the generic drug groups. Table 2 demonstrates the treatment effectiveness of both the brandname drugs and generic drugs during the follow-up period. Regarding the average daily dose, the risperidone-B group required a significantly lower daily dose than the risperidone-G group (2.14 ± 1.49 mg vs. 2.61 ± 1.53 mg, p b 0.001). Of the patients treated with sulpiride, the daily dose of sulpiride-B group was lower than that of the sulpiride-G group (302.72 ± 244.30 mg vs. 340.71 ± 271.63 mg, p = 0.031). Among risperidone users, the risperidone-G group (4.8%) demonstrated a higher rate of receiving augmentation therapy within 180 days, compared to the risperidone-B group (1.5%). With the exception of this, we observed no significant differences in drug discontinuation, augmentation therapy outside the 180-day interval, and psychiatric hospitalization between the risperidone-B and risperidone-G groups. Of the patients treated with sulpiride, the likelihood of psychiatric hospitalization in the sulpiride-B group was significantly lower than that of the sulpiride-G group, which was consistent within various time periods (30 days, 180 days, 365 days, and 1095 days). The drug discontinuation and augmentation therapy rates did not differ between the sulpiride-B and sulpiride-G groups. The Kaplan-Meier event free survival curves (Supplementary Figure 1) display the drug discontinuation, augmentation, and psychiatric hospitalization during the overall study period. After controlling for the potential confounding effects of gender, age, year of initial recruitment, CCI scores, and psychiatric comorbidities (Table 3), the Cox regression model showed that the sulpiride-B group still had significantly lower rates of psychiatric hospitalization than the sulpiride-G group (aHR, 0.24; 95% CI, 0.10–0.56). However, we found no significant difference in drug discontinuation rates or augmentation therapy between sulpiride-B and sulpiride-G groups. Furthermore, no significant differences in drug discontinuation, augmentation therapy, or psychiatric hospitalization were discovered between the risperidone-B and risperidone-G groups.


4


Table 1 Characteristics of schizophrenic patients treated with brand-name and generic formula of risperidone and sulpiride in Taiwan, from 2000 to 2012. Characteristics

Sex Male Female Age, years Initiation year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 CCI Psychiatric comorbidities Alcohol use disorders Substance use disorders Mood disorders Anxiety disorders Sleep disorders

Risperidone (N = 549)

Sulpiride (N = 1325)

Brand-name (N = 404)

Generic (N = 145)

214 (53.0) 190 (47.0) 37.24 ± 11.88

83 (57.2) 62 (42.8) 38.83 ± 12.32

– – – – 65 (16.1) 87 (21.5) 53 (13.1) 59 (14.6) 50 (12.4) 31 (7.7) 23 (5.7) 21 (5.2) 15 (3.7) 0.14 ± 0.60

– – – – 0 (0.0) 0 (0.0) 4 (2.8) 21 (14.5) 26 (17.9) 19 (13.1) 33 (22.8) 31 (21.4) 11 (7.6) 0.17 ± 0.56

15 (3.7) 15 (3.7) 79 (19.6) 47 (11.6) 39 (9.7)

8 (5.5) 8 (5.5) 37 (25.5) 23 (15.9) 23 (15.9)

p-Value


Generic (N = 991)

155 (46.4) 179 (53.6) 36.93 ± 13.30

519 (52.4) 472 (47.6) 37.37 ± 11.97

0.371a

50 (15.0) 82 (24.6) 46 (13.8) 46 (13.8) 35 (10.5) 21 (6.3) 14 (4.2) 10 (3.0) 10 (3.0) 6 (1.8) 5 (1.5) 8 (2.4) 1 (0.3) 0.08 ± 0.35

54 (5.4) 123 (12.4) 128 (12.9) 128 (12.9) 98 (9.9) 97 (9.8) 95 (9.6) 59 (6.0) 58 (5.9) 56 (5.7) 55 (5.5) 28 (2.8) 12 (1.2) 0.12 ± 0.52

0.294a

0.353 0.353 0.132 0.191 0.043⁎

9 (2.7) 5 (1.5) 79 (23.7) 38 (11.4) 24 (7.2)

37 (3.7) 30 (3.0) 177 (17.9) 145 (14.6) 99 (10.0)

0.370 0.132 0.021⁎ 0.136 0.127

0.376

0.171 b0.001⁎⁎

p-Value 0.059

0.571 b0.001⁎⁎

Data were expressed as N (%) or mean ± SD. CCI indicates Charlson Comorbidity Index. ⁎ p b 0.05. ⁎⁎ p b 0.001. a p Value was used by Mann-Whitney test.

4. Discussion Our results showed that both the risperidone-G and sulpiride-G groups were prescribed significantly higher doses than the risperidone-B and sulpiride-B groups in clinical settings. Both the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA) recognize the following standard for generic drug products: the hemodynamic between the brand-name and generic drugs if the 90% CI for mean ratios of the maximum plasma drug concentration (Cmax) and the area under the plasma concentration-time curve

(AUC) are within the predetermined 80–125% equivalence range (Balkowiec-Iskra et al., 2015; Cessak et al., 2016). Various studies have previously found that risperidone-G met the bioequivalence criteria for generic drugs (Boonleang et al., 2010; Khorana et al., 2011; Liu et al., 2013). Another study has indicated that sulpiride-B and sulpirideG had similar bioavailability (Chen et al., 1989). In summary, past pharmacokinetic studies have indicated that brand-name and generic antipsychotics demonstrate bioequivalence (Frank, 2007; Gyorgy et al., 2008). However, the effectiveness of medication is also related to complex interplay of physiological and psychosocial aspects (e.g., doctor-

Table 2 Comparison of effectiveness in schizophrenic patients treated with brand-name and generic formula of risperidone and sulpiride at different time points. Effectiveness

Average dose, mg/day Discontinuation ≤30 days ≤180 days ≤365 days ≤1095 days Augmentation ≤30 days ≤180 days ≤365 days ≤1095 days Hospitalization ≤30 days ≤180 days ≤365 days ≤1095 days

Risperidone (N = 549)

Sulpiride (N = 1325)


Generic (N = 145)

p-Value


Generic (N = 991)

p-Value

2.14 ± 1.49

2.61 ± 1.53

b0.001⁎⁎a

302.72 ± 244.30

340.71 ± 271.63

0.031⁎a

185 (45.8) 322 (79.7) 356 (88.1) 383 (94.8)

60 (41.4) 109 (75.2) 121 (83.4) 131 (90.3)

0.360 0.255 0.153 0.060

175 (52.4) 283 (84.7) 312 (93.4) 330 (98.8)

519 (52.4) 824 (83.1) 898 (90.6) 962 (97.1)

0.994 0.500 0.116 0.080

1 (0.2) 6 (1.5) 9 (2.2) 9 (2.2)

2 (1.4) 7 (4.8) 7 (4.8) 7 (4.8)

0.113 0.023⁎ 0.111 0.111

4 (1.2) 7 (2.1) 7 (2.1) 7 (2.1)

7 (0.7) 26 (2.6) 32 (3.2) 33 (3.3)

0.392 0.593 0.290 0.255

27 (6.7) 33 (8.2) 34 (8.4) 35 (8.7)

4 (2.8) 6 (4.1) 8 (5.5) 8 (5.5)

0.079 0.105 0.260 0.227

2 (0.6) 4 (1.2) 5 (1.5) 5 (1.5)

36 (3.6) 61 (6.2) 67 (6.8) 70 (7.1)

0.0041⁎ 0.0003⁎⁎ 0.0002⁎⁎ 0.0001⁎⁎

Data were expressed as N (%) or mean ± SD. ⁎ p b 0.05. ⁎⁎ p b 0.001. a p Value was used by Mann-Whitney test.


C.-W. Hsu et al. / Schizophrenia Research xxx (2017) xxx–xxx Table 3 Effectiveness of brand-name and generic formula of risperidone and sulpiride for treating schizophrenic patients by Cox proportional hazard models.

Discontinuation Augmentation Hospitalization

Risperidone (brand-name vs. generic)

Sulpiride (brand-name vs. generic)

aHR

95% CI

p-Value

aHR

95% CI

p-Value

1.13 0.51 2.12

0.90–1.41 0.16–1.64 0.91–4.93

0.285 0.259 0.081

1.09 0.74 0.24

0.96–1.24 0.32–1.72 0.10–0.56

0.197 0.487 0.001⁎

Controlling for sex, age, year of initial prescription of selected antipsychotic, CCI scores, and psychiatric comorbidities. aHR indicates adjusted hazard ratios. ⁎ p b 0.05.

patient relationship, drug adherence, stigmatization and economical concern). Although we found that higher doses of generic risperidone and sulpiride may be needed to reach similar clinical effectiveness as their brand name compounds in the real world, such result may be confounded by psychological, therapeutic adherence, physiological, and pharmacological factors which we failed to control for. The defined daily doses of risperidone and sulpiride recommended by the World Health Organization (WHO) Collaborating Centre for Drug Statistics Methodology are 5 mg and 800 mg, respectively (WHO, 2016). The doses of risperidone and sulpiride identified in our study, whether brand-name or generic, were obviously lower than the amounts recommended by the WHO. Previous studies have indicated that Asian subjects require lower daily doses of psychotropic drugs than Caucasian patients to achieve an equivalent treatment response. One possible explanation for the ethnic difference may be the disparity in activities of Cytochrome P 450 enzymes, which are responsible for the metabolism of psychotropic drugs in humans (Bakare, 2008; Bertilsson, 2007). Furthermore, although the daily doses in this study were observed in realistic settings, they were not from a clinical trial. The dosing strategies in clinical trials are usually designed with a fixed dosage. In contrast, prescriptions in clinical settings are usually based on the clinicians' discretion. Clinicians generally prescribe antipsychotics for schizophrenic patients pursuant to the severity of patients' symptoms, as well as their age, body type, and past history; then the doses can be adjusted to properly improve psychotic symptoms. Notably, nearly half of patients discontinued their antipsychotic treatment in the first 30 days of administration (Table 2). The dosing habit of antipsychotics in clinical practice in Taiwan is start low and titrate up gradually. Because the antipsychotic usually need to be titrated in clinical practice, this implies that a certain part of newly treated patients might not have enough time to get their target doses. Therefore, the low doses of risperidone or sulpiride observed in this study may be related to the high premature dropout rate. In this study, the rates of psychiatric hospitalization did not differ between the risperidone-B and risperidone-G groups; however, the sulpiride-B group had considerably lower rates of admission than the sulpiride-G group. The psychiatric hospital admission rates in our study sample were b 10% throughout the study period, which may be related to the definition of admission (admission to a psychiatric ward during treatment with risperidone or sulpiride). The clinical meaning of psychiatric hospitalization in the current study represents a patient with acutely exacerbated schizophrenia even when receiving antipsychotic treatment. A previous nationwide population-based study indicated that the effectiveness of sulpiride was superior to that of haloperidol, olanzapine, and risperidone (Lai et al., 2013). Our findings indicate that patients treated with sulpiride-B may be steadier, as reflected in the lower admission rates, compared to those treated with sulpiride-G. However, psychiatric hospitalization may be related to various factors (i.e., patients' symptom severity, accessibility of medical resources, and chronology) (Chiang et al., 2016); therefore, the underlying reasons for the differences in admission rates need further investigation.

5

Previous evidence has found that continuous antipsychotic treatment is a “gold standard” for clinical practice and can effectively prevent schizophrenic patients from experiencing a relapse of their psychotic symptoms (De Hert et al., 2015). While no significant difference was found in the discontinuation rate between brand-name drugs and generic formulations in this study, the drug discontinuation rates in our study population were noticeably high (one-year discontinuation rate: 83.4% to 93.4%). A healthcare database from New Zealand showed that the average time to first discontinuation in new users of second-generation antipsychotics was 192.3 days (Ndukwe and Nishtala, 2016). Furthermore, a recent European study indicated that the continuation rate after initial antipsychotic prescription without switching or augmentation at 12-month-follow-up was approximate 39% (Landolt et al., 2016). In realistic clinical settings, antipsychotic treatments may be discontinued for various reasons (i.e., lack of efficacy, intolerable side effects, switching to other antipsychotics, lack of insight, or poor drug adherence of patients) (Haro et al., 2009). Moreover, the definition of discontinuation in this study was cessation of antipsychotics for 60 days or longer, which is relatively stricter than other previous studies. This represents that patients who were once non-compliant to prescription for 60 days or longer, no matter whether they reused antipsychotics or not, were labeled as discontinuation. Aforementioned reasons may have been related to the high discontinuation rates of antipsychotic treatment in our study. Augmentation with another antipsychotic drug is a treatment strategy that is usually considered when a single antipsychotic medication cannot achieve satisfying treatment effects. Our study indicated that the risperidone-G group exhibited a significantly higher rate of augmentation within 180 days than the risperidone-B group. In addition to the findings related to higher doses of risperidone-G, we found that the treatment effects of risperidone-G group may have a higher proportion of ineffective treatment response and may require adjunctive therapy. However, the Cox proportional hazard model did not reveal differences in augmentation rates for either the risperidone or sulpiride users during the overall study period. Moreover, the augmentation therapy rates in this study are low (one-year augmentation rate: 2.1% to 4.8%), compared to those of previous international studies (ranging from 10% to 25%) (Ganguly et al., 2004; Kreyenbuhl et al., 2006). A prior epidemiologic study indicated that the percentage of schizophrenic patients that received poly-pharmacy of antipsychotic treatments in Taiwan was lower than in other Asian countries (Xiang et al., 2012). Furthermore, our study recruited patients who were antipsychotic-free for at least 180 days. Therefore, our study population either suffered from their first episode of psychosis and was drug-naïve or had discontinued treatment for a long time. With such patients, clinicians may decide to switch to other antipsychotics rather than use augmentation strategy if the treatment response was poor (Hatta et al., 2014). This study had several limitations. First, the study used reimbursement data so many important clinical characteristics, such as the patients' socioeconomic status, residential area, symptom severity, family function, adverse effect, or drug compliance, were not available. Therefore, we were unable to determine whether or how the aforementioned factors affected patients' treatment outcome and could not examine the differences in treatment efficacy (improvement of psychotic symptoms or daily function) between brand-name and generic antipsychotics. Second, patients' allocation to brand-name or generic drugs in this study was through clinical judgment in real-world settings but not through random assignment. Therefore, we could not exclude selection bias from this study. In addition, imbalanced psychiatric comorbidities were found between the brand-name group and the generic group, possibly related to nonrandom assignment, which might play potential confounding effects in patients' outcomes. For example, sleep disorders were more prevalent in the risperidone-G group and mood disorders were more prevalent in the sulpiride-B group. Third, during the study period, there were different generic products from the same drug in use, but the effectiveness of the different


6


generic products was not analyzed individually in this study. Moreover, it might be difficult to generalize the findings to patients of other ethnicity, or to generics available in other countries. Forth, to ensure that previous antipsychotics did not interfere with treatment effectiveness, we excluded patients who had been prescribed other antipsychotics within 180 days before the studied medications. Furthermore, in this study, drug discontinuation was defined as 60 days, which may differ from the definition used in previous international studies. Finally, this study could only evaluate the effectiveness of risperidone and sulpiride because the case numbers of other generic antipsychotics were not sufficient for statistical analyses. Therefore, our results may not be applicable to other antipsychotic drugs. 5. Conclusion Using the nationwide-wide reimbursement data, we found that higher doses of generic risperidone and sulpiride may be needed to reach similar clinical effectiveness as their brand name compounds in Taiwan. Meanwhile, the brand-name sulpiride is more effective at preventing patients from hospitalization than generic sulpiride. These findings can serve as an important reference that generic drugs may not be able to provide the same effectiveness as their brand name counterparts and to save medical expenditures claimed by their manufacturers. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.schres.2017.06.020. Role of funding source This study was supported by the Chang Gung Memorial Hospital Research Project (CMRPG2A0341). The funding sources had no involvement in the study design, collection, analysis and interpretation of data, writing of the report or the decision to submit the article for publication. Contributors Chih-Wei Hsu participated in interpreting data, reviewing references, and drafting the manuscript. Sheng-Yu Lee participated in the design of the study. Liang-Jen Wang executed the statistical analysis and revised the manuscript. All authors read and approved the final manuscript and contributed to the drafting and revising of the paper. Conflict of interest The authors declare no conflicts of interest. Acknowledgements This study is based in part on data from the NHIRD-TW provided by the National Health Insurance Administration, Ministry of Health and Welfare and managed by National Health Research Institutes (Registered number 100196). The interpretation and conclusions contained herein do not represent those of National Health Insurance Administration, Ministry of Health and Welfare or National Health Research Institutes.

References Bakare, M.O., 2008. Effective therapeutic dosage of antipsychotic medications in patients with psychotic symptoms: is there a racial difference? BMC. Res. Notes 1, 25. Balkowiec-Iskra, E., Cessak, G., Kuzawinska, O., Sejbuk-Rozbicka, K., Rokita, K., MirowskaGuzel, D., 2015. Regulatory and clinical aspects of psychotropic medicinal products bioequivalence. Eur. Neuropsychopharmacol. 25 (7), 1027–1034. Bertilsson, L., 2007. Metabolism of antidepressant and neuroleptic drugs by cytochrome p450s: clinical and interethnic aspects. Clin. Pharmacol. Ther. 82 (5), 606–609. Bobo, W.V., Stovall, J.A., Knostman, M., Koestner, J., Shelton, R.C., 2010. Converting from brand-name to generic clozapine: a review of effectiveness and tolerability data. Am. J. Health Syst. Pharm. 67 (1), 27–37. Boonleang, J., Pipatrattanaseree, W., Tanthana, C., Mahatthanatrakul, W., 2010. Relative bioavailability and pharmacokinetic comparison of two 2-mg risperidone tablet formulations: a single dose, randomized-sequence, double-blind, 2-way crossover study in healthy male volunteers in Thailand. Clin. Ther. 32 (10), 1842–1853. Borgheini, G., 2003. The bioequivalence and therapeutic efficacy of generic versus brandname psychoactive drugs. Clin. Ther. 25 (6), 1578–1592. Cessak, G., Rokita, K., Dabrowska, M., Sejbuk-Rozbicka, K., Zaremba, A., Mirowska-Guzel, D., Balkowiec-Iskra, E., 2016. Therapeutic equivalence of antipsychotics and antidepressants - a systematic review. Pharmacol. Rep. 68 (2), 217–223. Chen, I.J., Yang, J.M., Ding, N.C., Jang, H.R., Tsay, Y.S., 1989. Comparative pharmacokinetics of two sulpiride formulations in healthy volunteers. Kaohsiung J. Med. Sci. 5 (4), 222–230.

Chiang, C.L., Chen, P.C., Huang, L.Y., Kuo, P.H., Tung, Y.C., Liu, C.C., Chen, W.J., 2016. Impact of universal health coverage on urban-rural inequity in psychiatric service utilisation for patients with first admission for psychosis: a 10-year nationwide populationbased study in Taiwan. BMJ Open 6 (3), e010802. De Hert, M., Sermon, J., Geerts, P., Vansteelandt, K., Peuskens, J., Detraux, J., 2015. The use of continuous treatment versus placebo or intermittent treatment strategies in stabilized patients with schizophrenia: a systematic review and meta-analysis of randomized controlled trials with first- and second-generation antipsychotics. CNS Drugs 29 (8), 637–658. De Oliveira, I.R., Juruena, M.F., 2006. Treatment of psychosis: 30 years of progress. J. Clin. Pharm. Ther. 31 (6), 523–534. Desmarais, J.E., Beauclair, L., Margolese, H.C., 2011. Switching from brand-name to generic psychotropic medications: a literature review. CNS Neurosci. Ther. 17 (6), 750–760. Deyo, R.A., Cherkin, D.C., Ciol, M.A., 1992. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J. Clin. Epidemiol. 45 (6), 613–619. Elshafeey, A.H., Elsherbiny, M.A., Fathallah, M.M., 2009. A single-dose, randomized, twoway crossover study comparing two olanzapine tablet products in healthy adult male volunteers under fasting conditions. Clin. Ther. 31 (3), 600–608. Frank, R.G., 2007. The ongoing regulation of generic drugs. N. Engl. J. Med. 357 (20), 1993–1996. Ganguly, R., Kotzan, J.A., Miller, L.S., Kennedy, K., Martin, B.C., 2004. Prevalence, trends, and factors associated with antipsychotic polypharmacy among Medicaid-eligible schizophrenia patients, 1998-2000. J. Clin. Psychiatry 65 (10), 1377–1388. Gyorgy, B., Robert, K., Laszlo, C., Andrea, J., Ketilept Vizsgalo, M., 2008. Generic quetiapine in the treatment of acute schizophrenia and schizoaffective disorder. Neuropsychopharmacol. Hung. 10 (5), 293–303. Haas, J.S., Phillips, K.A., Gerstenberger, E.P., Seger, A.C., 2005. Potential savings from substituting generic drugs for brand-name drugs: medical expenditure panel survey, 1997–2000. Ann. Intern. Med. 142 (11), 891–897. Hardan, A.Y., Fung, L.K., Amin, H., 2010. Risperidone: switching from brand name to generic. J. Child Adolesc. Psychopharmacol. 20 (5), 457–458. Haro, J.M., Novick, D., Suarez, D., Roca, M., 2009. Antipsychotic treatment discontinuation in previously untreated patients with schizophrenia: 36-month results from the SOHO study. J. Psychiatr. Res. 43 (3), 265–273. Harvey, P.D., Keefe, R.S., 2001. Studies of cognitive change in patients with schizophrenia following novel antipsychotic treatment. Am. J. Psychiatry 158 (2), 176–184. Hatta, K., Otachi, T., Fujita, K., Morikawa, F., Ito, S., Tomiyama, H., Abe, T., Sudo, Y., Takebayashi, H., Yamashita, T., Katayama, S., Nakase, R., Shirai, Y., Usui, C., Nakamura, H., Ito, H., Hirata, T., Sawa, Y., Group, J.s, 2014. Antipsychotic switching versus augmentation among early non-responders to risperidone or olanzapine in acute-phase schizophrenia. Schizophr. Res. 158 (1–3), 213–222. Healy, D.J., Taylor, S., Goldman, M., Barry, K., Blow, F., Milner, K.K., 2005. Clinical equivalence of generic clozapine. Community Ment. Health J. 41 (4), 393–398. Italiano, D., Bruno, A., Santoro, V., Lanza, G., Muscatello, M.R., Zoccali, R., Spina, E., 2015. Generic olanzapine substitution in patients with schizophrenia: assessment of serum concentrations and therapeutic response after switching. Ther. Drug Monit. 37 (6), 827–830. Khorana, N., Maphanta, S., Lohitnavy, O., Srichaiya, A., Sayasathid, J., 2011. Comparative pharmacokinetics and bioequivalence of two tablet formulations of 2 mg risperidone in healthy Thai male volunteers. Int. J. Clin. Pharmacol. Ther. 49 (6), 409–414. Kreyenbuhl, J., Valenstein, M., McCarthy, J.F., Ganoczy, D., Blow, F.C., 2006. Long-term combination antipsychotic treatment in VA patients with schizophrenia. Schizophr. Res. 84 (1), 90–99. Lai, E.C., Chang, C.H., Kao Yang, Y.H., Lin, S.J., Lin, C.Y., 2013. Effectiveness of sulpiride in adult patients with schizophrenia. Schizophr. Bull. 39 (3), 673–683. Landolt, K., Rossler, W., Ajdacic-Gross, V., Derks, E.M., Libiger, J., Kahn, R.S., Fleischhacker, W.W., Group, E.S, 2016. Predictors of discontinuation of antipsychotic medication and subsequent outcomes in the European First Episode Schizophrenia Trial (EUFEST). Schizophr. Res. 172 (1–3), 145–151. Lessing, C., Ashton, T., Davis, P.B., 2015. The impact on health outcome measures of switching to generic medicines consequent to reference pricing: the case of olanzapine in New Zealand. J. Prim. Health Care 7 (2), 94–101. Leucht, S., Corves, C., Arbter, D., Engel, R.R., Li, C., Davis, J.M., 2009. Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. Lancet 373 (9657), 31–41. Lieberman, J.A., Stroup, T.S., McEvoy, J.P., Swartz, M.S., Rosenheck, R.A., Perkins, D.O., Keefe, R.S., Davis, S.M., Davis, C.E., Lebowitz, B.D., Severe, J., Hsiao, J.K., Clinical Antipsychotic Trials of Intervention Effectiveness, I, 2005. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N. Engl. J. Med. 353 (12), 1209–1223. Liu, Y., Zhang, M.Q., Jia, J.Y., Liu, Y.M., Liu, G.Y., Li, S.J., Wang, W., Weng, L.P., Yu, C., 2013. Bioequivalence and pharmacokinetic evaluation of two formulations of risperidone 2 mg: an open-label, single-dose, fasting, randomized-sequence, two-way crossover study in healthy male Chinese volunteers. Drugs R&D 13 (1), 29–36. Mahatthanatrakul, W., Rattana, K., Sriwiriyajan, S., Wongnawa, M., Ridtitid, W., 2008. Bioequivalence study of a generic quetiapine in healthy male volunteers. Int. J. Clin. Pharmacol. Ther. 46 (9), 489–496. Masi, G., Liboni, F., 2011. Management of schizophrenia in children and adolescents: focus on pharmacotherapy. Drugs 71 (2), 179–208. Meltzer, H.Y., Matsubara, S., Lee, J.C., 1989. Classification of typical and atypical antipsychotic drugs on the basis of dopamine D-1, D-2 and serotonin2 pKi values. J. Pharmacol. Exp. Ther. 251 (1), 238–246. Multhoff, G., Wu, C.-S., Lai, M.-S., Gau, S.S.-F., Wang, S.-C., Tsai, H.-J., 2014. Concordance between patient self-reports and claims data on clinical diagnoses, medication use, and health system utilization in Taiwan. PLoS One 9 (12), e112257.


C.-W. Hsu et al. / Schizophrenia Research xxx (2017) xxx–xxx Ndukwe, H.C., Nishtala, P.S., 2016. Time-to-first discontinuation, adherence and persistence in new users of second-generation antipsychotics. J. Clin. Psychopharmacol. 36 (6), 649–657. Oluboka, O., Stewart, S., Landry, S., Adams, S., 2010. Does therapeutic equivalence follow bioequivalence? A randomized trial to assess clinical effects after switching from Clozaril to generic clozapine (gen-clozapine). J. Clin. Pharmacol. 50 (5), 531–535. van Os, J., Kapur, S., 2009. Schizophrenia. Lancet 374 (9690), 635–645. van Os, S., Relleke, M., Piniella, P.M., 2007. Lack of bioequivalence between generic risperidone oral solution and originator risperidone tablets. Int. J. Clin. Pharmacol. Ther. 45 (5), 293–299. Paton, C., 2006. Generic clozapine: outcomes after switching formulations. Br. J. Psychiatry 189, 184–185. Schneeweiss, S., Seeger, J.D., Maclure, M., Wang, P.S., Avorn, J., Glynn, R.J., 2001. Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data. Am. J. Epidemiol. 154 (9), 854–864.

7

Verster, G.C., Joubert, G., Stevens, M., van der Merwe, H., 1998. Generic substitution—comparing the clinical efficacy of a generic substitute for fluphenazine decanoate with the original product. S. Afr. Med. J. 88 (3), 260–262. WHO, 2016. Collaborating Centre for Drug Statistics Methodology. ATC/DDD Index. http://www.whocc.no/atc_ddd_index/. Xiang, Y.T., Dickerson, F., Kreyenbuhl, J., Ungvari, G.S., Wang, C.Y., Si, T.M., Lee, E.H., He, Y.L., Chiu, H.F., Yang, S.Y., Chong, M.Y., Tan, C.H., Kua, E.H., Fujii, S., Sim, K., Yong, M.K., Trivedi, J.K., Chung, E.K., Udomratn, P., Chee, K.Y., Sartorius, N., Shinfuku, N., 2012. Common use of antipsychotic polypharmacy in older Asian patients with schizophrenia (2001–2009). J. Clin. Psychopharmacol. 32 (6), 809–813.
