Antipsychotic prescription patterns in outpatient settings: 24-month results from the Intercontinental Schizophrenia Outpatient Health Outcomes (IC-SOHO) study

European Neuropsychopharmacology (2008) 18, 170–180

w w w. e l s e v i e r. c o m / l o c a t e / e u r o n e u r o

Antipsychotic prescription patterns in outpatient settings: 24-month results from the Intercontinental Schizophrenia Outpatient Health Outcomes (IC-SOHO) study Istvan Bitter a , Tamas Treuer b,⁎, Yulia Dyachkova b , Ferenc Martenyi b , Margaret McBride c , Gabor S. Ungvari d a

Department of Psychiatry and Psychotherapy, Semmelweis University, Budapest, Hungary Eli Lilly Regional Operations GmbH, Area Medical Center Vienna, Vienna, Austria c Intercontinental Information Sciences, Eli Lilly Australia Pty Ltd, Sydney, Australia d Department of Psychiatry, Chinese University of Hong Kong, Hong Kong SAR, China b

Received 26 April 2007; received in revised form 21 July 2007; accepted 9 August 2007

KEYWORDS Antipsychotic agents; Schizophrenia; Prescription patterns; Drug; Treatment

Abstract This report describes antipsychotic prescription patterns for outpatients with schizophrenia prescribed olanzapine (n = 3222), clozapine (n = 236), risperidone (n = 1117), quetiapine (n = 189) or haloperidol (n = 256) monotherapy at study entry and treated in a naturalistic, clinical practice setting over 24 months. Predictive factors associated with remaining on monotherapy were also identified. Olanzapine patients had significantly greater odds of remaining on their initial monotherapy compared to other treatment groups, while clozapine or risperidone recipients were more likely to remain on monotherapy, compared to haloperidol patients. Switching antipsychotic medication was more common than addition of another antipsychotic agent, and the most common reason for modifying treatment was lack of effectiveness. The odds of modifying antipsychotic prescription due to intolerability were lower for patients treated with olanzapine, compared to patients treated with risperidone or haloperidol (p ≤ .001). However, treatment modification due to patient request was significantly greater for olanzapine-treated patients, compared to risperidone-, quetiapine- or haloperidol-treated patients (p ≤ .001). © 2007 Elsevier B.V. and ECNP. All rights reserved.

1. Introduction ⁎ Corresponding author. Neuroscience Research, Area Medical Center Vienna, Eli Lilly Regional Operations GmbH, Koelblgasse 8-10, A-1030 Vienna, Austria. Tel.: +36 1 328 5127; fax: +36 1 328 5103. E-mail address: [email protected] (T. Treuer).

Standard clinical practice guidelines consistently reinforce the essential role of antipsychotic medication in the management of patients with schizophrenia (Lieberman et al., 2005). However, antipsychotics are often only

0924-977X/$ - see front matter © 2007 Elsevier B.V. and ECNP. All rights reserved. doi:10.1016/j.euroneuro.2007.08.001

Antipsychotic prescription patterns in outpatient settings partially effective, and frequently leave patients with a substantial residual symptom burden and distressing or harmful side effects (Davis et al., 2006). As a consequence, changing medications is a common feature of psychiatric practice as clinicians regularly choose to switch antipsychotic medications or resort to polypharmacy in an attempt to elicit a clinical response or to address tolerability concerns (Edlinger et al., 2005a; Faries et al., 2005; Ganguly et al., 2004). The resulting high rates of treatment discontinuation underscore the substantial limitations in the suitability of available antipsychotic treatments. The efficacy and safety of specific antipsychotic agents have been extensively studied in clearly defined patient populations with randomized, controlled trials (Leucht et al., 2003; Schooler, 2003). However, long-term naturalistic, observational studies are necessary to determine if these findings can be broadly generalized to the routine, ongoing care of patients with schizophrenia (National Institute of Clinical Excellence, 2003). Indeed, regulatory authorities have recently reinforced that observational studies in clinical practice settings are essential for assessing the optimal use of new drugs (U.S. Food and Drug Administration, 2003). Despite the knowledge that switching antipsychotic medications and antipsychotic polypharmacy are widespread in clinical practice, prospective studies investigating the prevalence of prescription change and the factors influencing prescription practices in real-life clinical settings are relatively scarce. Investigation into this topic is clinically relevant as illustrated by a recent re-analysis of data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study which demonstrated that patients who were randomly assigned to remain on their pre-study antipsychotic had a significantly longer time to treatment discontinuation during the study, when compared to patients who were randomized to switch to another antipsychotic medication at study entry (Essock et al., 2006). The purpose of this report was to describe prescription patterns for outpatients with schizophrenia treated with olanzapine, clozapine, risperidone, quetiapine or haloperidol monotherapy during normal clinical practice over a 24month period. Based on the data collected, factors associated with maintaining the originally prescribed antipsychotic monotherapy were also identified.

2. Experimental procedures 2.1. Study design The Intercontinental Schizophrenia Outpatient Health Outcomes (ICSOHO) study is a 3-year, global, prospective, observational study of health outcomes associated with antipsychotic treatment in outpatients with schizophrenia (Study code: F1D-SN-HGJR). The study was initiated in November 2000 and was conducted in 27 countries. The study methods for the IC-SOHO study have been described in detail previously (Dossenbach et al., 2004), and are summarized below. A similar study has been conducted in European countries (EU-SOHO) (Haro et al., 2003a), but due to anticipated differences to IC-SOHO in the clinical environments and prescribing practices of these countries, these results are not presented here. To ensure that the study reflected real-life clinical practice, all aspects of patient care were at the discretion of the participating psychiatrists. Treatment was open-label and included any available antipsychotic drug registered for the treatment of schizophrenia, so

171 the treatment options differed between participating countries. In an effort to avoid interference with normal clinical practice which would occur with a classic randomization study design, each treating clinician was instructed to first make the most appropriate treatment decision based entirely on their best clinical judgment and local treatment guidelines. As the initial objective of the ICSOHO study was to compare olanzapine with other antipsychotics, the study was designed to provide two patient cohorts of approximately equal size. As such, patients were enrolled using an alternating entry structure (until a block of ten patients was reached —i.e. five in each group) into one of the following two patient cohorts: (1) patients who were initiated on or switched to olanzapine therapy (monotherapy or in combination with other agents), or (2) patients who were initiated on or switched to any other antipsychotic therapy other than olanzapine. Hence, as a result of the study design, approximately half of enrolled patients were initially treated with olanzapine. This resulted in a stratified sample, with the olanzapine group as the ‘over-sampled’ stratum. In the analyses presented here, patients in the non-olanzapine cohort are separated into treatment groups based upon the antipsychotic prescribed at study entry. Patients were permitted to use concomitant psychotropic medications as clinically indicated.

2.2. Patients Patients were eligible to enter into the IC-SOHO study if they: (1) were diagnosed with schizophrenia according to either ICD-10 (World Health Organisation, 1992) or DSM-IV (American Psychiatric Association, 1994), (2) presented during the normal course of care, (3) were initiated on or were switched to antipsychotic therapy for the treatment of schizophrenia, (4) were at least 18 years of age, and (5) were not simultaneously participating in an interventional study. Patients from 27 countries throughout Africa, Asia, Central and Eastern Europe, Latin America and the Middle East participated in the study. The study was conducted in accordance with the local ethical and regulatory requirements of each country. All participants provided informed consent as required by local regulations.

2.3. Outcome measures Data were collected at study entry and at scheduled visits approximately 3, 6, 12, 18 and 24 months after enrolment. Data were not collected at outpatient consultations between the specified study visits. Details of antipsychotic treatment including dose, whether the patient remained on monotherapy, changed (added a new drug, switched to a new drug or removed the existing drug) or modified (categories as per ‘changed treatment’ but also included changed dose) treatment, and the reasons for treatment modification were recorded at every study visit. The treating clinician was permitted to cite more than one reason for treatment modification. Concomitant medication use was recorded at each scheduled visit. The clinical status of the patient was measured using the Clinical Global Impressions-Severity (CGI-S) rating scale, adapted to include four separate symptom domains (positive, negative, depressive, and cognitive) (Haro et al., 2003b; National Institute of Mental Health, 1976).

2.4. Treatment groupings and analyses Post hoc treatment groups were established based on the antipsychotic initiated or switched to at study entry. The most frequently prescribed monotherapy treatments were olanzapine (n = 3222), clozapine (n = 236), risperidone (n = 1117), quetiapine (n = 189) and haloperidol (n = 256) and this report will only describe results obtained from patients allocated to one of these treatment groups.

172 Table 1

I. Bitter et al. Treatments prescribed at study entry

Treatment

Sample size (n)

Percentage of total sample (%)

Olanzapine monotherapy Clozapine monotherapy Risperidone monotherapy Quetiapine monotherapy Haloperidol monotherapy Other atypical monotherapy Other typical monotherapy Combination of atypicals Combination of typicals Combination of typical/atypical Inadequate drug information a

3222 236 1117 189 256 104 710 185 421 952 266

42.1 3.1 14.6 2.5 3.3 1.4 9.3 2.4 5.5 12.4 3.5

a Includes patients who were not prescribed drugs or had missing drug information as well as patients who did not initiate or change antipsychotic treatment at study entry.

Antipsychotic dose could be altered and psychotropic concomitant medication was permitted during the 24-month observation period, but patients were included in the treatment group analyses only for as long as they remained on the antipsychotic monotherapy prescribed at study entry. While the design of the IC-SOHO study permitted the inclusion of data from patients who received combination therapies or antipsychotic agents other than those identified, these patients were not included in the analyses for this report. Statistical analyses were performed using SAS© Version 8.2 for Windows™ (SAS Institute, Cary, N.C.). Continuous variables were described using summary statistics such as means and standard deviations, and categorical variables were described using frequencies and percentages. Patients with missing data were excluded from relevant analyses. Differences across the treatment groups were tested using logistic regression for categorical variables and analysis of variance for continuous variables. To facilitate the interpretation of results, significant treatment group comparisons were also evaluated by the number needed to treat (NNT) to prevent one additional event. The NNT was calculated as the reciprocal of the absolute risk reduction (ARR), which, by convention, was calculated so that positive values indicated superiority for olanzapine, and negative values attributed superiority to the comparator. To adjust

for differences between the treatment groups at study entry, the following variables were identified a priori and used as covariates for post-baseline data: age, sex, duration of illness, overall baseline CGI-S score, prior use of typical depot antipsychotics or clozapine in the 6 months prior to study entry, and hospitalization in the 6 months prior to study entry. In these analyses, the value or status of the outcome variable at study entry was also included as a covariate in the statistical analyses. No additional covariates were included post hoc. In an attempt to identify patient and illness characteristics at study entry that may be associated with a change in antipsychotic treatment, the following variables were tested for associations with the odds of remaining on monotherapy: sex, age, body mass index, duration of illness, initiation or switching antipsychotic at study entry, monotherapy treatment group, use of clozapine, atypical antipsychotics or depot typical agents in the 6 months prior to study entry, first time antipsychotic use for the treatment of schizophrenia, hospitalization in the 6 months prior to study entry, overall CGI-S score (and positive, negative, depressive and cognitive symptom domain scores), presence of dystonia, akathisia or parkinsonism (extrapyramidal symptoms (EPS)), presence of tardive dyskinesia, use of mood stabilizers, previous treatment compliance, social activity, involvement with a spouse or partner, housing status, work status, previous suicide attempt, substance or alcohol abuse, exhibition of hostility, presence of sexual dysfunction, region and country. The variables that showed an association (p ≤ .05) with a change in antipsychotic medication were entered in the joint logistic regression model. Pairwise comparisons are reported for region and monotherapy treatment prescribed at study entry. Due to the large number of statistical comparisons undertaken in the overall analyses of the IC-SOHO data, the level required for statistical significance was defined, a priori, to be p ≤ .001. However, given the exploratory nature of this subgroup analysis, p values b.05 are also reported in all tables and figures. Statistical testing was not performed if comparisons included any treatment group with an insufficient number of patients.

3. Results 3.1. Patients In total, 7658 patients were enrolled in the IC-SOHO study from 27 countries: Algeria (n = 306), Argentina (n = 362), Chile

Table 2 Baseline characteristics of patients prescribed olanzapine, clozapine, risperidone, quetiapine or haloperidol monotherapy at study entry Characteristic

Olanzapine (n = 3222)

Clozapine (n = 236)

Risperidone (n = 1117)

Quetiapine (n = 189)

Haloperidol (n = 256)

Percentage of total sample (n = 7658) Sex, % women (n) Mean age, years (SD) Mean duration of illness, years (SD) First time antipsychotic use, % (n) Overall CGI-S⁎, mean score (SD) Mean body mass index, kg/m2 (SD)

42 45 (1438)C,r 34.9 (12.1)r 8.5 (9.7) 18 (565)c,q 4.36 (1.07)C,R 25.1 (4.5)q

3 31 (72) 34.2 (11.9) 9.1 (9.1) 9 (22)r,h 4.62 (1.01)R,q,h 25.1 (3.9)

15 49 (541)C 35.9 (12.2) 9.0 (10.0) 18 (194)q 4.23 (1.04) 24.9 (4.6)q

2 50 (93)C 34.4 (12.0) 9.0 (10.1) 10 (19)h 4.35 (1.07) 25.8 (5.2)h

3 47 (121)C 35.0 (11.5) 9.5 (9.8) 18 (44) 4.37 (1.07) 24.9 (4.7)

⁎ CGI-S, Clinical Global Impressions Severity rating scale (1–7). p ≤ .001 compared with clozapine. R p ≤ .001 compared with risperidone. c .001 b p b .05 compared with clozapine. h .001 b p b .05 compared with haloperidol. q .001 b p b .05 compared with quetiapine. r .001 b p b .05 compared with risperidone. C

Antipsychotic prescription patterns in outpatient settings (n = 167), Colombia (n = 201), Czech Republic (n = 489), Egypt (n = 198), Hungary (n = 200), Israel (n = 77), Lithuania (n = 110), Malaysia (n = 106), Mexico (n = 1067), Peru (n = 100), Poland (n = 626), Puerto Rico (n = 223), Romania (n = 140), Russia (n = 160), Saudi Arabia (n = 204), Slovakia (n = 302), Slovenia (n = 225), South Korea (n = 850), Taiwan (n = 300), Turkey (n = 694), Venezuela (n = 271) and Central America (Costa Rica, El Salvador, Guatemala and Honduras (n = 280)). Of these patients, 76% (n = 5834) were prescribed monotherapy at study entry (Table 1). As a consequence of the study design, the most frequently prescribed monotherapy was olanzapine (n = 3222), followed by risperidone (n = 1117), quetiapine (n = 189), haloperidol (n = 256) and clozapine (n = 236). Approximately one-fifth (20.3%, n = 1558) of patients were prescribed more than one antipsychotic (polypharmacy) at study entry. The most common form of antipsychotic polypharmacy was the use of both typical and atypical agents (12.4%, n = 952). No prescription data were available at study entry for 266 patients (3.5% of enrolled patients). The characteristics of the patients in the olanzapine, clozapine, risperidone, quetiapine and haloperidol monotherapy groups were broadly similar at study entry, although some statistically significant differences (p ≤ .001) were detected between treatment groups in the proportion of female patients and the overall CGI-S score (Table 2). In particular, the clozapine treatment group had significantly fewer female patients and significantly higher mean overall CGI-S score when compared with the other treatment groups. At the end of 24 months, prescription pattern data were available for a reduced number of patients in each treatment group (olanzapine (n = 2049), clozapine (n = 172), risperidone (n = 672), quetiapine (n = 124), haloperidol (n = 154)). Patients were not available for statistical evaluation if they discontinued the study before the 24-month visit (olanzapine: n = 983 (30.5%), clozapine: n = 58 (24.6%), risperidone: n = 387 (34.6%), quetiapine: n = 56 (29.6%) and haloperidol: n = 94 (36.7%)), or if prescription data were missing (olanzapine: n = 190 (5.9%), clozapine: n = 6 (2.5%), risperidone: n = 58 (5.2%), quetiapine: n = 9 (4.8%) and haloperidol: n = 8 (3.1%)).

3.2. Antipsychotic dose For the patients remaining on the antipsychotic monotherapy prescribed at study entry, the dose varied during the study period (Table 3). The mean dose of olanzapine and risperidone generally increased for the first 6 months, and then remained relatively constant until the 24-month consultation. In contrast, the mean quetiapine dose consistently increased throughout the 24-month period. The mean haloperidol dose increased during the early phase of the study and then decreased below the baseline dose at 24 months.

3.3. Maintaining antipsychotic monotherapy treatment Differences were detected between the treatment groups in the proportion of patients who remained on their original

173 antipsychotic monotherapy for the duration of the study period (Fig. 1). Among the patients with prescription data available, the proportion of patients to maintain the antipsychotic prescribed at study entry was highest in the olanzapine group (74.4%), followed by the clozapine (59.9%), risperidone (57.9%), quetiapine (42.7%) and haloperidol (38.3%) groups. Patients treated with olanzapine had significantly higher odds of staying on the originally prescribed medication compared to patients treated with other antipsychotics (p ≤ .001; Fig. 1). Patients treated with clozapine or risperidone had significantly higher odds of remaining on their original treatment, compared to patients treated with haloperidol (p ≤ .001). Approximately half of the patients who remained on monotherapy stayed on their baseline dose for the 24month period. The haloperidol group had the highest proportion (59.3%, n = 32) of patients remaining on a constant dose. This was followed by the olanzapine (55.9%, n = 822), quetiapine (46.2%, n = 24), risperidone (45.2%, n = 175) and clozapine (44.1%, n = 45) treatment groups. The odds of staying on a constant dose were significantly higher with olanzapine, compared to risperidone (p ≤ .001). The differences between haloperidol and the other treatment groups did not reach statistical significance. This may be due to the relatively small sample size of some treatment groups.

3.4. Changing treatments Of the patients who were prescribed monotherapy at study entry and changed their treatment within the 24-month period, switching to a new antipsychotic agent was more common than the addition of another antipsychotic to the existing medication (antipsychotic polypharmacy). The proportion of patients who switched medication was highest in the quetiapine group (42.9%, n = 57), followed by the haloperidol (37.7%, n = 61), risperidone (27.0%, n = 197), clozapine (19.7%, n = 35) and olanzapine (12.8%, n = 286) groups. The addition of another antipsychotic to the monotherapy prescribed at study entry was most common in the haloperidol group (13.6%, n = 22), followed by the clozapine (10.7%, n = 19), quetiapine (6.8%, n = 9), risperidone (6.4%, n = 47) and olanzapine (6.1%, n = 137) groups. Recognizing that clinicians could cite more than one reason for modifying treatment, the most common explanation for changing antipsychotic treatment was lack of effectiveness, regardless of the antipsychotic prescribed at study entry (Fig. 2). However, the odds of modifying treatments due to lack of effectiveness were significantly lower for olanzapine recipients, compared to patients treated with clozapine (olanzapine vs. clozapine, odds ratio [OR]: 2.5; 95% Confidence Interval [CI]: 1.5–4.25, p ≤ .001, NNT: 7.5), or quetiapine (olanzapine vs. quetiapine, OR: 4.5; 95% CI: 2.3–9.0; p ≤ .001, NNT: 5.5). Further, the odds of modifying antipsychotics due to intolerability were significantly lower for patients treated with olanzapine, compared to risperidone (olanzapine vs. risperidone, OR: 2.0; 95% CI: 1.6–2.5; p ≤ .001, NNT: 6.3) or haloperidol (olanzapine vs. haloperidol, OR: 2.2; 95% CI: 1.5–3.1; p ≤ .001, NNT: 5.6) recipients. The odds of modifying the antipsychotic due to patient request were significantly higher for patients treated with olanzapine, compared to patients who received

174

I. Bitter et al.

Table 3 Oral doses (mg/day) of antipsychotics prescribed at each visit for patients remaining on the monotherapy prescribed at study entry Time point and measure

Olanzapine

Clozapine

Risperidone

Quetiapine

Haloperidol

Baseline Sample size Mean (SD) Median

3133 9.7 (4.0) 10.0

235 151.5 (123.5) 100.0

1109 3.4 (1.7) 3.0

186 241.1 (165.4) 200.0

196 13.3 (9.2) 10.0

3 months Sample size Mean (SD) Median

2539 10.7 (4.5) 10.0

185 213.6 (133.2) 200.0

837 3.9 (1.9) 4.0

136 337.3 (203.5) 300.0

121 13.2 (8.8) 10.0

6 months Sample size Mean (SD) Median

2237 10.8 (4.7) 10.0

155 226.1 (141.2) 200.0

691 4.0 (2.0) 4.0

103 332.2 (188.9) 300.0

92 14.2 (9.4) 10.0

12 months Sample size Mean (SD) Median

1923 10.8 (4.8) 10.0

138 228.4 (138.2) 200.0

552 4.0 (2.2) 4.0

76 334.1 (199.1) 300.0

75 13.6 (8.6) 10.0

18 months Sample size Mean (SD) Median

1665 10.7 (4.8) 10.0

111 213.3 (130.7) 200.0

469 4.1 (2.2) 4.0

60 346.6 (190.1.7) 300.0

55 13.9 (8.7) 10.0

24 months Sample size Mean (SD) Median

1478 10.6 (4.9) 10.0

102 215.0 (129.1) 200.0

381 4.0 (2.1) 4.0

52 369.3 (181.8) 400.0

41 12.7 (8.8) 10.0

Figure 1 Proportion of patients, with prescription pattern data available, who maintained their initially prescribed antipsychotic monotherapy for 24 months (with odds ratio and NNT referenced to olanzapine). Cp ≤ .001 compared with clozapine, Hp ≤ .001 compared with haloperidol, Qp ≤ .001 compared with quetiapine, Rp ≤ .001 different compared with risperidone, q.001 b p b .05 compared with quetiapine. CI = Confidence interval; N/A= not applicable; NNT = number needed to treat.

Antipsychotic prescription patterns in outpatient settings

175

Figure 2 Reasons for antipsychotic treatment modification in patients who were prescribed monotherapy at study entry and who switched antipsychotics, added antipsychotics, removed and then resumed their treatment or changed the dose of their treatment during the 24 months. Cp ≤ .001 compared with clozapine, Hp ≤ .001 compared with haloperidol, Qp ≤ .001 compared with quetiapine, R p ≤ .001 different compared with risperidone, c.001 b p b .05 compared with clozapine, h.001 b p b .05 compared with haloperidol, q .001 b p b .05 compared with quetiapine, r.001 b p b .05 compared with risperidone. Differences in statistical significance were based on the odds ratios of modifying antipsychotic treatment for a given reason.

risperidone (olanzapine vs. risperidone, OR: 0.6; 95% CI: 0.5–0.8; p ≤ .001, NNT: −9.2), quetiapine (olanzapine vs. quetiapine, OR: 0.5; 95% CI: 0.3–0.7; p ≤ .001, NNT: − 5.6) or haloperidol (olanzapine vs. haloperidol, OR: 0.5; 95% CI: 0.3–0.7; p ≤ .001, NNT: − 5.3).

3.5. Predictive factors associated with maintaining antipsychotic monotherapy A number of factors were associated with the odds of remaining on the originally prescribed monotherapy for the first 24 months of treatment (Table 4). Overall, the baseline factors significantly related to higher odds of remaining on the monotherapy prescribed at study entry were; prescription of olanzapine monotherapy, presence of EPS, initiation (rather than switching) antipsychotic treatment, no previous use of antipsychotic agents for the treatment of schizophrenia, lower overall CGI-S score at study entry, and enrolment from a region other than Asia. Analyses revealed regional variations in the Table 4

odds of remaining on the monotherapy prescribed at study entry, with patients from Asia experiencing significantly lower odds of remaining on their original monotherapy, compared to patients from other regions. Further, patients in Latin America had significantly higher odds of remaining on monotherapy compared to patients in Central and Eastern Europe (Latin America vs. Central and Eastern Europe, OR: 1.4; 95% CI: 1.1– 1.6; p ≤ .001).

3.6. Concomitant medications Concomitant psychotropic medication use generally decreased following 24 months of treatment (Table 5). The odds of taking at least one concomitant psychotropic medication at 24 months were significantly lower for olanzapine and clozapine patients, compared to risperidone or haloperidol recipients (p ≤ .001), and for patients treated with risperidone compared to haloperidol (risperidone vs. haloperidol, OR: 0.34; 95% CI: 0.18–0.65; p = .001, NNT: 4.2). When compared to haloperidol-

Predictive factors for remaining on monotherapy during the 24-month study period

Baseline characteristic Treatment Risperidone Quetiapine Clozapine Haloperidol CGI-S Overall score Extrapyramidal symptoms Switched treatment First time antipsychotic use Region Africa and Middle East Central and Eastern Europe Latin America

Reference category

Adjusted odds ratio

95% confidence interval

(Olanzapine) (Olanzapine) (Olanzapine) (Olanzapine) (per unit difference in baseline score) (Presence) (Initiated treatment) (Previously untreated)

0.49 0.25 0.55 0.20 0.88 0.67 0.70 0.70

(0.41, (0.17, (0.41, (0.14, (0.82, (0.57, (0.56, (0.52,

(Asia) (Asia) (Asia)

2.00 1.75 2.36

(1.53, 2.62) (1.37, 2.23) (1.85, 3.02)

p value b 0.001

0.58) 0.36) 0.77) 0.27) 0.95) 0.79) 0.84) 0.94)

b 0.001 b 0.001 b 0.001 0.019 b 0.001

176

I. Bitter et al.

treated patients, the proportion of patients taking anticholinergics at 24 months were lower for patients prescribed any of the atypical antipsychotics, but further testing was prevented by insufficient patient numbers in the clozapine and quetiapine treatment groups. The proportion of patients prescribed antidepressants at study entry ranged from 11% (clozapine and haloperidol monotherapy groups) to 23% (quetiapine), with no considerable reduction in antidepressant use at 24 months. The odds of using anxiolytics or hypnotics were also significantly lower for patients treated with olanzapine, compared to patients treated with haloperidol (p ≤ .001). The use of mood stabilizers remained relatively constant over the 24-month period.

4. Discussion This report examined prescription patterns and factors influencing treatment change for patients with schizophrenia treated in clinical practice settings over a period of 24 months. Data were collected in regions where insight into treatment patterns had previously been sparse. The flexible design of the IC-SOHO study allowed the treating physician to change or add antipsychotic medication throughout the treatment period. Further, patients were free to discontinue at any time. In effect, patient and clinician judgment of efficacy and tolerability of the prescribed antipsychotic were reflected in their willingness to maintain the baseline

Table 5 Concomitant medication use at study entry and at 24 months for patients remaining on their initially prescribed antipsychotic monotherapy 24 months % (n)

Odds ratio⁎ (95% confidence interval)

Number needed to treat (NNT)

At least one concomitant medication Olanzapine 50.7 (1633) Clozapine 45.3 (107) Risperidone 67.3 (752) Quetiapine 52.9 (100) Haloperidol 82.0 (210)

34.7 34.0 52.7 47.2 76.3

(529) (35) (205) (25) (45)

1R,H (N/A) 0.95R,H (0.62, 1.47) 2.11H (1.68, 2.66) 1.71h (0.98, 2.98) 6.15 (3.32, 11.38)

– −141.3 5.6 8.0 2.4

Anticholinergics Olanzapine Clozapine Risperidone Quetiapine Haloperidol

12.4 (398) 10.6 (25) 30.4 (340) 11.1 (21) 61.7 (158)

5.1 3.9 27.8 3.8 59.3

(78) (4) (108) (2) (35)

No statistical analyses conducted due to insufficient patient numbers

– −81.2 4.4 −74.6 1.8

Antidepressants Olanzapine Clozapine Risperidone Quetiapine Haloperidol

14.7 (474) 11.4 (27) 16.9 (189) 23.3 (44) 11.3 (29)

12.9 10.7 13.6 20.8 18.6

(197) (11) (53) (11) (11)

1 (N/A) 0.82 (0.43, 1.05 (0.75, 1.68 (0.84, 1.68 (0.85,

– −44.7 141.5 12.8 17.5

Anxiolytics/hypnotics Olanzapine Clozapine Risperidone Quetiapine Haloperidol

32.4 (1044) 29.2 (69) 37.2 (416) 31.7 (60) 46.5 (119)

20.2 22.3 27.0 30.2 42.4

(308) (23) (105) (16) (25)

1H,r (N/A) 1.15h (0.70, 1.89) 1.42h (1.09, 1.84) 1.72 (0.93, 3.18) 2.89 (1.68, 4.97)

– 46.9 14.7 10.0 4.5

Mood stabilizers Olanzapine Clozapine Risperidone Quetiapine Haloperidol

8.9 (287) 9.3 (22) 9.3 (104) 7.4 (14) 10.2 (26)

7.6 7.8 6.9 7.5 13.6

(116) (8) (27) (4) (8)

1 (N/A) 0.99 (0.47, 0.91 (0.58, 1.03 (0.36, 1.78 (0.82,

– 623.3 −150.2 −1683.9 16.8

Concomitant medication

Study entry % (n)

1.59) 1.46) 3.34) 3.32)

2.12) 1.41) 2.93) 3.89)

⁎ Odds ratio compared to olanzapine at 24 months, adjusted for baseline variables. p ≤ .001 compared with haloperidol. R p ≤ .001 compared with risperidone. h .001 b p b .05 compared with haloperidol. q .001 b p b .05 compared with quetiapine. r .001 b p b .05 compared with risperidone. N/A= Not applicable. H

Antipsychotic prescription patterns in outpatient settings treatment without modification or augmentation (Lieberman et al., 2005). This information is clinically relevant as it facilitates treatment decisions, particularly if there are deviations from published treatment guidelines. This study confirms previous reports that antipsychotic treatment change is prevalent in clinical practice (Edlinger et al., 2005a; Faries et al., 2005; Ganguly et al., 2004) and the likelihood of remaining on antipsychotic monotherapy is mainly dependent upon the antipsychotic medication prescribed (Faries et al., 2005; Lieberman et al., 2005). Monotherapy is recognized as the preferred mode of therapy (Weiden and Casey, 1999), and the identification of factors related to the maintenance of antipsychotic monotherapy during long-term, naturalistic treatment should be encouraged. Recognition of these factors may facilitate the development of management strategies designed to enhance adherence to long-term antipsychotic monotherapy in clinical practice. Overall, the results from the IC-SOHO study support previous publications that have compared patient adherence to monotherapy antipsychotic treatments. Consistent with our findings, research from US trials associated olanzapine treatment over 12 and 18 months with a greater likelihood of remaining on monotherapy, compared to treatment with risperidone or quetiapine (Faries et al., 2005; Lieberman et al., 2005). Although data from the ICSOHO were obtained from 27 countries over a period of 24 months, results confirmed a greater adherence to olanzapine monotherapy, compared to monotherapy treatment with clozapine, risperidone, quetiapine and haloperidol. Results from the CATIE study (Lieberman et al., 2005) indicate that the lowest discontinuation rate was reported for patients prescribed olanzapine (64%), and the highest for patients who received quetiapine (82%). When compared to the CATIE results, the proportion of patients remaining on monotherapy was higher in the IC-SOHO study, but this difference may have been inflated by differences in statistical methodology. Patients in the IC-SOHO study who discontinued or had missing data were excluded from the analysis whereas similar patients in the CATIE study were assumed to have changed treatment and were, therefore, considered not adherent to the originally prescribed monotherapy (Lieberman et al., 2005). Results of this study are also generally consistent with a study conducted two to three years after quetiapine was approved by the US Food and Drug Administration in which quetiapine-treated patients were shown to have the highest rate of switching (Leslie and Rosenheck, 2002). Although the IC-SOHO study was conducted several years later, the proportion of patients to switch to another antipsychotic treatment was still highest in the quetiapine treatment group. This may reflect the need to further optimize quetiapine therapy and/or a lack of confidence in quetiapine therapy among treating clinicians. In support of the former explanation, our study has shown a consistent increase in the quetiapine dose throughout the 24-month study period. Statistical modeling indicated that IC-SOHO patients were more likely to remain on their originally prescribed antipsychotic monotherapy if, at study entry, they had been experiencing EPS, had initiated rather than switched medications, had not previously received antipsychotics for

177 the treatment of schizophrenia, had a lower overall CGI-S score, or were enrolled in a region other than Asia. It may initially be difficult to account for the influence of EPS symptoms at study entry on the odds of remaining on the originally prescribed antipsychotic monotherapy, but one potential explanation may be that the presence of distressing motor dysfunction at the commencement of the study could lead to improved adherence if the new therapy did not cause continuation or exacerbation of these symptoms. An alternate scenario could be that EPS at study entry and the odds of remaining on monotherapy could both be correlated to a third (unmeasured) variable. As a further explanation for some of these findings, when compared to chronic, multi-episode patients, first episode patients are known to be more responsive to treatment with many achieving full symptom remission and level of recovery (Lieberman, 2006). It is possible that improved clinical outcomes in first episode patients may be associated with continued adherence to the originally prescribed antipsychotic. Geographic region was identified as a factor related to remaining on the originally prescribed antipsychotic monotherapy, with patients treated in Asia more likely to discontinue treatment. Differences in clinical environments and prescribing practices could be expected between the regionally diverse countries that participated in the IC-SOHO study, and further investigation into this issue is certainly warranted. In the IC-SOHO study, the odds of modifying treatment due to lack of effectiveness were significantly lower for olanzapine patients compared to recipients of clozapine or quetiapine monotherapy. Further, the odds of treatment modification due to issues of intolerability were significantly lower with olanzapine treatment when compared to risperidone or haloperidol. This finding is inconsistent with existing data (Lieberman et al., 2005), but olanzapine was also associated with significantly greater odds of modifying treatment due to patient's request, when compared to risperidone, quetiapine or haloperidol. It is possible that the higher rate of discontinuation due to patient's request in the present study may actually reflect issues of intolerability. This may be at least partially related to the greater treatment-related weight gain observed with olanzapine treatment (Haddad, 2005) as weight gain observed during treatment with psychotropic medication has been reported to be a common reason for nonadherence to prescribed treatment (Sachs and Guille, 1999). A number of comments can be made about the dose of antipsychotics prescribed in the IC-SOHO study. The doses of olanzapine and quetiapine were lower than was reported in the CATIE study (Lieberman et al., 2005). Patients in the CATIE study were randomly assigned antipsychotic treatments whereas in the IC-SOHO study, the choice of treatment was at the discretion of the clinician. Therefore, differences may have existed in the appropriateness of treatments received, which in turn may have influenced the range of doses required. Clozapine recipients in the IC-SOHO study were treated with mean doses (230.4 mg/day at 24 months) that were substantially lower than what were commonly used in American studies (N 400 mg/day; Fleischhacker et al., 1994) or in Austrian clinical practice (N340 mg/day; Edlinger et al., 2005b). Nevertheless, several European studies have shown that clozapine doses around 200 mg/day can be

178 effective (Bitter et al., 2004; Fleischhacker et al., 1994). Although not statistically assessed, there was a trend during the study period for a gradual increase in the mean quetiapine dose, when compared to other antipsychotics evaluated in this report. The relatively low dose for quetiapine reported in IC-SOHO patients may be a consequence of uncertainty associated with the label recommendations and the available literature at the time this study was conducted regarding the most appropriate quetiapine dose required for the treatment of patients with schizophrenia (Citrome et al., 2005). The gradual increase in antipsychotic dose evident in the IC-SOHO study would indicate that clinicians exercised best clinical practice by starting with a low dose, which was gradually increased until the lowest effective dose had been prescribed. The reduction in the mean dose of haloperidol during the study period may have been due to the emergence of adverse effects, causing dose reduction or treatment discontinuation. Finally, the higher odds of remaining on a constant dose with olanzapine, compared to risperidone, suggest that the starting dose of olanzapine may be closer to the therapeutic dose. When compared to data obtained from clinical records at one hospital during the time the IC-SOHO study was being conducted (Edlinger et al., 2005b), the mean dose for quetiapine was approximately 50% lower in IC-SOHO patients. Despite potential concern regarding the appropriateness of quetiapine dose reported in this study, the observational design of the IC-SOHO study allowed clinicians to use their own discretion and clinical judgment to determine the optimal treatment for each patient. For this reason, the doses for all medications provided in this paper should be accepted as reflecting actual treatment decisions. Approximately one-fifth of IC-SOHO patients were prescribed more than one antipsychotic agent at study entry. However, it was not possible to determine if this was due to long term maintenance polypharmacy, or a temporary situation due to cross tapering associated with the initiation of a new antipsychotic. At the 24-month visit, the proportion of IC-SOHO patients who were prescribed more than one antipsychotic agent was consistent with literature reporting 10% of outpatients maintained long-term antipsychotic polypharmacy (Covell et al., 2002), but was substantially lower than other research citing 40% to 58% of patients received antipsychotic polypharmacy (Faries et al., 2005; Ganguly et al., 2004; Sim et al., 2004). Treatment guidelines regarding antipsychotic polypharmacy (Gaebel et al., 2005) may differ widely across regions, and the rate of antipsychotic polypharmacy in a study sample may vary depending upon the participating countries. The regional variations in the odds of antipsychotic treatment change observed in this study were likely to be influenced by the composition of countries for a given region. Countries differ in terms of treatment guidelines (Gaebel et al., 2005), formulary practices, national health care policies (including psychiatric training), and availability and affordability of medications (Kohn et al., 2004), all of which may influence prescription patterns. In addition, the dynamics of patient–doctor relationships may also vary between cultures. Studies exploring regional variations in treatment change are lacking and additional research is

I. Bitter et al. required to determine more specific reasons for any differences (Ito et al., 2005). Concomitant psychotropic medications, such as antidepressants and mood stabilizers, are frequently prescribed to patients with schizophrenia to augment antipsychotic response (Casey et al., 2003; Citrome, 2002) or to address affective symptoms (Glick et al., 2006). The use of these medications by IC-SOHO patients was relatively constant throughout the 24-month treatment period. However, the benefit of long-term concomitant prescription of antidepressants and mood stabilizers has been questioned in the management of stabilized, chronic patients. In fact, careful withdrawal of these adjunctive medications in stabilized patients has been found to have no influence on clinical outcomes (Glick et al., 2006). In terms of the study design, IC-SOHO has several strengths. This study incorporated a pragmatic design to closely reflect normal clinical practice with treating clinicians able to choose the most appropriate antipsychotic treatment and dose, and concomitant medications as clinically indicated. Therefore, the findings of this report closely reflect prescription patterns in clinical practice. Evidence from the IC-SOHO study is particularly relevant as insight into prescription practices in non-Western countries is relatively scarce, despite the value of this information to benchmark against published guidelines. Furthermore, the study was comparative and prospective in design, was of long-term duration and included a large sample size with broad geographical coverage. The study design, however, also has several limitations. As previously discussed (Dossenbach et al., 2004), some of these limitations are intrinsically associated with observational studies. First, the non-randomized design and lack of blinding opened the potential for bias, but this was reduced in this study by adjusting comparisons using baseline covariates that were recognized to have clinical significance. Nevertheless, it is acknowledged that these baseline corrections may not completely address the potential biases resulting from the design of this study. Second, as a consequence of the study design, which stipulated recruitment of approximately equal numbers of patients in the olanzapine and non olanzapine groups, there were a higher number of patients in the olanzapine group, compared to other treatment groups. This treatment allocation method differs from clinical practice and does not reflect clinical prescription patterns. However, the relatively large size of the olanzapine group enhanced the precision of the estimates for the outcomes associated with olanzapine treatment. Third, attrition bias may have influenced the results, as many of the enrolled patients were not available for this analysis once they discontinued the originally prescribed monotherapy. Fourth, due to the large number of tests undertaken and the absence of formal adjustments for multiple comparisons, results should be interpreted with caution. Fifth, it is not possible to comment on the actual number of days that each patient received treatment, as data were recorded only at study visits and only brief summaries were available regarding treatment modifications made between scheduled visits. Sixth, clinical assessments were made by clinicians of different professional and cultural backgrounds and, therefore, the diagnosis of schizophrenia and the application of the CGI-S scale may

Antipsychotic prescription patterns in outpatient settings have differed between countries or regions. Seventh, it is not possible to determine the extent of influence that corporate sponsorship of the trial may have on the treatment decisions made by the clinician. Finally, psychosocial influences (e.g. psychoeducation, supportive psychotherapy and cognitive behavioral therapy) on treatment adherence and effectiveness were not measured in this study. Psychosocial interventions are an important part of schizophrenia management (Lehman et al., 2003) and this may have varied considerably between centers. In conclusion, this study provides insight into the prescription patterns of antipsychotic medications in reallife clinical practice and identifies and highlights differences between atypical and typical agents. Further, factors that may influence treatment change in outpatients with schizophrenia were identified. In this study, patients were more likely to remain on monotherapy if, at study entry, they had been experiencing EPS, had initiated rather than switched medications, had not previously received antipsychotic agents for the treatment of schizophrenia, had a lower overall CGI-S score at study entry, or were enrolled from a region other than Asia. Further, switching antipsychotic medication was more common than adding another antipsychotic agent to the existing medication. Antipsychotic prescribing patterns are affected by a myriad of non-clinical factors that are beyond the scope of this report. However, the pragmatic design and naturalistic setting of this large, intercontinental study underscore the clinical relevance of these results.

Role of the funding source This research was funded by Eli Lilly and Company, Indianapolis, USA. Employees of Eli Lilly and Company were involved in the study design, in the collection, analysis and interpretation of data, in the writing of the manuscript, and in the decision to submit the manuscript for publication.

Contributors Drs Bitter, Treuer, Martenyi, Ungvari and McBride were involved in the interpretation of the data and the development of this manuscript. Ms Dyachkova undertook the statistical analysis, assisted in the interpretation of the data and contributed to the manuscript. All authors approved the final version for submission.

Conflict of interest Tamas Treuer, Yulia Dyachkova, Ferenc Martenyi, and Margaret McBride are employees of Eli Lilly and Company. Istvan Bitter was an employee of Eli Lilly from November 2000 until August 2003 at which time the study was ongoing. Dr Bitter is an advisory board member/ consultant/lecturer for Bristol-Myers Squibb, Eli Lilly, EGIS, Janssen, Lundbeck, Novartis, Pfizer and Richter. Dr Ungvari has received sponsorship from Smith Kline & Beecham, Servier-Synthelabo and Eli Lilly.

Acknowledgements The authors acknowledge the medical writing services provided by ProScribe Medical Communications in the preparation of this manuscript.

179

References American Psychiatric Association, 1994. DSM-IV. Diagnostic and Statistical Manual of Mental Disorders, 4th ed. American Psychiatric Association, Washington. Bitter, I., Dossenbach, M.R., Brook, S., Feldman, P.D., Metcalfe, S., Gagiano, C.A., Furedi, J., Bartko, G., Janka, Z., Banki, C.M., Kovacs, G., Breier, A., 2004. Olanzapine versus clozapine in treatment-resistant or treatment-intolerant schizophrenia. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 28, 173–180. Casey, D.E., Daniel, D.G., Wassef, A.A., Tracy, K.A., Wozniak, P., Sommerville, K.W., 2003. Effect of divalproex combined with olanzapine or risperidone in patients with an acute exacerbation of schizophrenia. Neuropsychopharmacology 28, 182–192. Citrome, L., Jaffe, A., Levine, J., Allingham, B., 2002. Use of mood stabilizers among patients with schizophrenia, 1994–2001. Psychiatr. Serv. 53, 1212. Citrome, L., Jaffe, A., Levine, J., Lindenmayer, J.P., 2005. Dosing of quetiapine in schizophrenia: how clinical practice differs from registration studies. J. Clin. Psychiatry 66, 1512–1516. Covell, N.H., Jackson, C.T., Evans, A.C., Essock, S.M., 2002. Antipsychotic prescribing practices in Connecticut's public mental health system: rates of changing medications and prescribing styles. Schizophr. Bull. 28, 17–29. Davis, J.M., Marder, S.R., Tamminga, C.A., 2006. Switch or stay? Am. J. Psychiatry 163, 2032–2033. Dossenbach, M., Erol, A., el Mahfoud, K.M., Shaheen, M.O., Sunbol, M.M., Boland, J., Hodge, A., O'Halloran, R.A., Bitter, I., 2004. Effectiveness of antipsychotic treatments for schizophrenia: interim 6-month analysis from a prospective observational study (IC-SOHO) comparing olanzapine, quetiapine, risperidone, and haloperidol. J. Clin. Psychiatry 65, 312–321. Edlinger, M., Baumgartner, S., Eltanaihi-Furtmuller, N., Hummer, M., Fleischhacker, W.W., 2005a. Switching between second-generation antipsychotics: why and how? CNS Drugs 19, 27–42. Edlinger, M., Hausmann, A., Kemmler, G., Kurz, M., Kurzthaler, I., Walch, T., Walpoth, M., Fleischhacker, W.W., 2005b. Trends in the pharmacological treatment of patients with schizophrenia over a 12 year observation period. Schizophr. Res. 77, 25–34. Essock, S.M., Covell, N.H., Davis, S.M., Stroup, T.S., Rosenheck, R.A., Lieberman, J.A., 2006. Effectiveness of switching antipsychotic medications. Am. J. Psychiatry 163, 2090–2095. Faries, D., Ascher-Svanum, H., Zhu, B., Correll, C., Kane, J., 2005. Antipsychotic monotherapy and polypharmacy in the naturalistic treatment of schizophrenia with atypical antipsychotics. BMC Psychiatry 5, 26. Fleischhacker, W.W., Hummer, M., Kurz, M., Kurzthaler, I., Lieberman, J.A., Pollack, S., Safferman, A.Z., Kane, J.M., 1994. Clozapine dose in the United States and Europe: implications for therapeutic and adverse effects. J. Clin. Psychiatry 55 (Suppl B), 78–81. Gaebel, W., Weinmann, S., Sartorius, N., Rutz, W., McIntyre, J.S., 2005. Schizophrenia practice guidelines: international survey and comparison. Br. J. Psychiatry 187, 248–255. 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, 1377–1388. Glick, I.D., Pham, D., Davis, J.M., 2006. Concomitant medications may not improve outcome of antipsychotic monotherapy for stabilized patients with nonacute schizophrenia. J. Clin. Psychiatry 67, 1261–1265. Haddad, P., 2005. Weight change with atypical antipsychotics in the treatment of schizophrenia. J. Psychopharmacol., Suppl. 19, 16–27. Haro, J.M., Edgell, E.T., Jones, P.B., Alonso, J., Gavart, S., Gregor, K.J., Wright, P., Knapp, M., 2003a. The European Schizophrenia

180 Outpatient Health Outcome (SOHO) study: rationale, methods and recruitment. Acta Psychiatr. Scand. 107, 222–232. Haro, J.M., Kamath, S.A., Ochoa, S., Novick, D., Rele, K., Fargas, A., Rodriguez, M.J., Rele, R., Orta, J., Kharbeng, A., Araya, S., Gervin, M., Alonso, J., Mavreas, V., Lavrentzou, E., Liontos, N., Gregor, K., Jones, P.B., 2003b. The Clinical Global ImpressionSchizophrenia scale: a simple instrument to measure the diversity of symptoms present in schizophrenia. Acta Psychiatr. Scand., Suppl. 416, 16–23. Ito, H., Koyama, A., Higuchi, T., 2005. Polypharmacy and excessive dosing: psychiatrists' perceptions of antipsychotic drug prescription. Br. J. Psychiatry 187, 243–247. Kohn, R., Saxena, S., Levav, I., Saraceno, B., 2004. The treatment gap in mental health care. Bull. World Health Organ. 82, 858–866. Lehman, A.F., Buchanan, R.W., Dickerson, F.B., Dixon, L.B., Goldberg, R., Green-Paden, L., Kreyenbuhl, J., 2003. Evidence-based treatment for schizophrenia. Psychiatr. Clin. North Am. 26, 939–954. Leslie, D.L., Rosenheck, R.A., 2002. From conventional to atypical antipsychotics and back: dynamic processes in the diffusion of new medications. Am. J. Psychiatry 159, 1534–1540. Leucht, S., Barnes, T.R., Kissling, W., Engel, R.R., Correll, C., Kane, J.M., 2003. Relapse prevention in schizophrenia with newgeneration antipsychotics: a systematic review and exploratory meta-analysis of randomized, controlled trials. Am. J. Psychiatry 160, 1209–1222. Lieberman, J.A., 2006. Neurobiology and the natural history of schizophrenia. J. Clin. Psychiatry 67, 14. 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.,

I. Bitter et al. Lebowitz, B.D., Severe, J., Hsiao, J.K., 2005. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N. Engl. J. Med. 353, 1209–1223. National Institute of Clinical Excellence, 2003. Schizophrenia: Full National Clinical Guideline on Core Interventions in Primary and Secondary Care. Gaskell and the British Psychological Society, London. National Institute of Mental Health, 1976. Clinical Global Impressions (CGI). In: Guy, W. (Ed.), ECDEU Assessment Manual for Psychopharmacology. US Department of Health Education, and Welfare, Bethesda, pp. 217–222. Sachs, G.S., Guille, C., 1999. Weight gain associated with use of psychotropic medications. J. Clin. Psychiatry Suppl. 60, 16–19. Schooler, N.R., 2003. Relapse and rehospitalization: Comparing oral and depot antipsychotics. J. Clin. Psychiatry 64 (Suppl 16), 14–17. Sim, K., Su, A., Fujii, S., Yang, S.Y., Chong, M.Y., Ungvari, G.S., Si, T., Chung, E.K., Tsang, H.Y., Chan, Y.H., Heckers, S., Shinfuku, N., Tan, C.H., 2004. Antipsychotic polypharmacy in patients with schizophrenia: a multicentre comparative study in East Asia. Br. J. Clin. Pharmacol. 58, 178–183. U.S. Food and Drug Administration, 2003. Report to Congress-Reports on Postmarketing Studies. Available at: http://www.fda.gov/cber/ fdama/pstmrktfdama130.htm. Accessed September 20, 2003. Weiden, P.J., Casey, D.E., 1999. Polypharmacy: combining antipsychotic medications in the treatment of schizophrenia. J. Pract. Psychiatry Behav. Health 7, 229–233. World Health Organisation, 1992. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. World Health Organisation, Geneva.