 |
 |
 |
 |
 |
 |
 |
|
|
|
|||||||||
|
|||||||||||
Director, Clinical Research and Resource Centre, Waitemata District Health Board; Senior Clinical Lecturer, School of Medical and Health Sciences, University of Auckland, Auckland, NZ
Reprints: Ms. Wheeler, Mental Health and Addiction Services, Snelgar Building, Waitakere Hospital, Waitemata District Health Board, 55-75 Lincoln Rd., Waitakere, Auckland, New Zealand, fax 09 8381883, amanda.wheeler{at}waitematadhb.govt.nz
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionConclusionsReferences |
|---|
METHODS: A retrospective chart review was conducted for adult outpatients receiving care from community mental health services on October 31, 2004. Data collected for all patients prescribed an antipsychotic included demographics (sex, age, ethnicity); principal diagnosis (Diagnostic and Statistical Manual of Mental Disorders, 4th edition); comorbid conditions; duration of mental illness; psychiatric admissions; and treatment information (psychotropic medications, with dose and route of administration). If clozapine had been started after the introduction of full government prescription subsidy (February 1999), additional data, including year of initiation and prior antipsychotic history, were collected. Analysis included all outpatients with a diagnosis of schizophrenia (including schizoaffective disorder).
RESULTS: Antipsychotics were prescribed for 2796 schizophrenia patients; 32.8% were prescribed clozapine, with a mean dose of 372 mg/day and an average duration of illness of 9.7 years before starting clozapine. Patients who had started treatment after clozapine was funded by the government (59.3%) had received a median of 3 antipsychotic drugs prior to starting clozapine; most of the treatment regimens included 1 second-generation antipsychotic (91.2%). Clozapine patients were less likely to be coprescribed another antipsychotic compared with non-clozapine patients (11.7% vs 17.6%; p < 0.001). Both the clozapine and non-clozapine groups had a low total number of psychotropic medications prescribed (median 2); for clozapine patients, the second drug was most likely to be for treatment of hypersalivation.
CONCLUSIONS: Outpatients with treatment-resistant schizophrenia were prescribed clozapine at expected rates; however, treatment was delayed longer than recommended. There is some evidence that access to clozapine for treatment-resistant schizophrenia has improved, possibly as the result of the introduction of government subsidy, guideline dissemination, or increasing experience of clinicians with use of clozapine. In this real-world environment, the number of concomitant psychotropic medications for outpatients with schizophrenia was found to be low; when used concomitantly with clozapine, they were most commonly used to manage adverse effects.
Key Words: audit, clozapine, prescribing, schizophrenia
Published Online, May 13, 2008. www.theannals.com, DOI 10.1345/aph.1K662
In addition to its superior efficacy in treatment-resistant schizophrenia, clozapine offers clinical advantages over other antipsychotics. These include a low risk of extrapyramidal symptoms6 (reflected in its approval in most countries for use in people intolerant of other antipsychotics, particularly tardive dyskinesia); reduction of suicide in schizophrenia7 (approved for this indication in the US); and reduction in the frequency and duration of hospital admissions.8,9 This research evidence has led current clinical practice guidelines to recommend that clozapine be considered at the earliest opportunity for people with treatment-resistant schizophrenia. This is defined as a lack of satisfactory clinical response to trials (with recommended treatment duration 6–8 wk) of at least 2 antipsychotics (at least 1 of which should be a non-clozapine second-generation antipsychotic such as risperidone, olanzapine, or quetiapine).10-12
However, clozapine is associated with other problematic effects such as sedation, hypersalivation, weight gain, constipation, insulin resistance, and diabetes,13,14 and, more rarely, risks of cardiac adverse events such as myocarditis and cardiomyopathy.15,16 It is the significant hematologic risks (severe neutropenia and agranulocytosis) that have necessitated the need for regular blood testing and registration with a national monitoring database in most countries that use clozapine.14 In New Zealand, weekly white blood cell counts are required for the first 18 weeks of therapy, followed by monthly testing thereafter.17,18 Guidance for regular monitoring of metabolic and cardiac adverse events for all antipsychotic medications has also been outlined by the New Zealand Mental Health Metabolic Working Group Initiative.19
Estimates of the number of people with treatment-resistant schizophrenia who are expected to benefit from clozapine range from 25% to 50%. Previous research indicated that clozapine is underused.20,21 Similarly, other research suggested that treatment is delayed for longer than is desirable to improve clinical, social, and functional outcomes; that is, clozapine is not prescribed as soon as 2 other antipsychotics have failed.22 Factors affecting the use of clozapine may include issues relating to patient consent, patient adherence to monitoring, adverse effects, the increasing availability of other second-generation antipsychotics,21,23,24 and cost of this medication.
Although clozapine was registered for use in New Zealand in early 1993, funding was restricted to the discretionary use of a hospital's budget intended for inpatient treatment only. Difficulties for patients in gaining access to clozapine treatment were highlighted in the mid-1990s25 and, in February 1999, clozapine became a fully funded medicine for all patients in the community or hospital setting. National underprescribing of clozapine continued to be reported,26 and steps were taken to monitor its use regionally. At that time, national data extracts of dispensing data from electronic pharmacy databases did not provide information about the number of individuals taking a particular medication or about the individual's diagnosis. This study was therefore set up with the aim of using clinical audit and feedback27 as a tool to compare everyday best practice with recommended practice for adults engaged with a community mental health center (CMHC) in Auckland (the largest city in New Zealand) and Northland (a mixed urban and rural region north of Auckland).
This paper reports first on the treatment pathway and pattern of clozapine prescribing in schizophrenia, and second on the extent to which other psychotropic drugs were coprescribed with clozapine. The coprescription of psychotropic drugs with clozapine in patients with schizophrenia is compared with schizophrenia patients prescribed alternative antipsychotic treatment.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
DATA COLLECTION
Study data were extracted by a trained senior psychiatric research nurse
from the paper and/or electronic clinical file used at each CMHC and from each
DHB's patient information management system. Data were entered by a clerical
assistant into a custom-designed Microsoft Access database with predetermined
response options to ensure data quality and facilitate analysis. Data
collected included demographics (sex, age, ethnicity), principal diagnosis,
comorbid conditions, duration of mental illness (time, in years, from first
contact with a treatment provider until 2004), psychiatric admission history,
and treatment data (psychotropic medication, with total daily dose and route
of administration). The principal Axis I diagnosis (Diagnostic and
Statistical Manual of Mental Disorders, 4th
ed.32)
documented at the psychiatric assessment closest to October 31, 2004, was
recorded as the major mental disorder (these assessments are usually conducted
every 3 months to coincide with prescription requirements). Axis I and II
comorbidities were recorded if documented at the psychiatric assessment
closest to the study review point. Documented physical health comorbidities
were recorded, as well as medications prescribed concurrently (database
allowed recording of 2 additional Axis I, 2 Axis II, and up to 8 Axis III
comorbidities). Other information was extracted from clinical summaries, case
reviews, discharge summaries, and the multidisciplinary progress notes
recorded for the 3 months on either side of the review point. Psychiatric
admission history was verified with the National Health Events database
maintained by the National Health Information Service.
All psychotropic data (including antipsychotics, antidepressants, mood stabilizers, sedative–hypnotics, and drugs used for management of neurologic adverse effects of antipsychotics) were recorded from the medication chart and verified in the body of the clinical notes or with duplicate copies of the prescription wherever possible. For patients prescribed clozapine, the year of initiation was also recorded. When clozapine initiation was from 1999 onward (the year community funding became available), information about the type and number of antipsychotics prescribed prior to clozapine was collected. The number of antipsychotic treatments was defined as the number of different antipsychotics prescribed regularly at a therapeutic antipsychotic dose for a minimum of 6 weeks. If the drug was used more than once, it was counted as a single antipsychotic treatment. Where there was a clearly documented break in clozapine treatment (>12 wk, and alternative antipsychotic treatment was prescribed), the clozapine restart date was recorded. The national clozapine registration database was used to verify the treatment start/restart date(s) and discontinuation date and reason for discontinuation where relevant. For the purposes of comparison, antipsychotic doses were converted into chlorpromazine equivalents (CPZe). The methodology used for this is described elsewhere.33
The study was approved by the Auckland Regional Ethics Committee (AKL2000/065).
DATA ANALYSIS
Data analyses were conducted using SPSS version 15 (SPSS, Chicago, IL).
Continuous variables were compared between groups using independent
t-tests or analysis of variance. Discrete variables were analyzed
using 
2 tests. When sample distributions did not
satisfy assumptions of normality (eg, age, duration of illness, psychiatric
admissions), nonparametric tests were used (Mann-Whitney U or
Kruskal-Wallis) or the variable was log transformed for analysis (eg,
antipsychotic dose). Logistic regression was used to investigate variables
associated with 2 outcome measures: pre-clozapine treatment duration (year of
first contact with a treatment provider to year clozapine started) and
antipsychotic coprescription with clozapine. To minimize the type I error
associated with multiple comparisons, the threshold required to demonstrate
statistical significance was set at p less than 0.01
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
Comparisons between the clozapine and non-clozapine treatment groups are
shown in Table 1. There were
more females in the non-clozapine group. The ethnicity profile also differed,
with fewer European and Asian and more M
ori and Pacific people treated
with clozapine. Patients prescribed clozapine had been unwell for longer and
were less likely to have comorbid mental disorder(s), primarily seen with
lower rates of movement disorders diagnosed as Axis I neuroleptic-induced
disorders (Table 1). There was
no significant difference between groups in the proportion with a comorbid
Axis II disorder. Physical comorbidity was grouped (0, 1, 2–4, and 
5
conditions) for analysis. Again, there was no significant difference between
groups, although patients treated with clozapine were found to have higher
rates of gastrointestinal conditions.
|
The average daily dose of clozapine was 372 mg (median 350, SD 152, range 50–900). Average daily doses did not vary significantly across the 4 DHBs (clozapine dose log transformed F3,913 = 0.77; p = 0.29).
PATHWAY TO CLOZAPINE TREATMENT
Almost the entire clozapine group (99.6%) had data that allowed calculation
of the duration between year of first contact with a treatment provider and
the clozapine starting year. Thirty-seven percent had started clozapine within
5 years of first contact with a provider (18.3% in 
2 y, 18.6% in 3–5
y), 24.8% within 6–10 years, and 38.3% more than 10 years after first
contact. The mean pre-clozapine treatment duration was 9.7 years (median 8; SD
7.8; range 0–43). Longer pre-clozapine treatment duration was found in
older patients (aged 50 y, mean 20.4 y; aged 30–49 y, 10.1 y; aged
15–29 y, 3.2 y; p < 0.001). It was also longer in European patients
(mean 11.3 y; Mori, 8.7 y; Pacific, 7.8 y; Asian, 6.7 y; p < 0.001),
and for patients who started clozapine treatment before it was funded in 1999
(mean 10.5 vs. 9.2 y; p < 0.001). There was no significant difference in
the pre-clozapine treatment duration between sexes (p = 0.285) or DHBs (p =
0.107).
Full government prescription subsidy for clozapine started in February 1999; 59.3% (n = 544) started clozapine after this date. A history of antipsychotic treatment was available for 533 of these patients (98.0%). Figure 1 shows the distribution of patients by the number of antipsychotic treatments prior to starting clozapine. The average number of antipsychotics prescribed was 3.5 (median 3.0, SD 1.7, range 1–10). The majority had been treated with an oral second-generation antipsychotic (91.2%), 61.9% were prescribed at least one oral first-generation antipsychotic, and 46.7% were prescribed a depot antipsychotic before the switch to clozapine. Only 39 patients had been treated with all 3 oral second-generation antipsychotics available in New Zealand at the time of the audit (risperidone, olanzapine, quetiapine) before starting clozapine.
|
|
The rates of antipsychotic polytherapy varied across the 4 health services,
with a trend for higher rates in DHB-D patients (21.6% vs DHB-A, 7.3%; DHB-B,
13.2%; DHB-C, 13.6%; p = 0.011). There was also some evidence to support
differences based on pre-clozapine treatment duration, although these
differences did not reach statistical significance (duration >10 y, 14.2%
polytherapy; duration 2 y, 8.5% polytherapy; p = 0.058). Combination
antipsychotic prescription was not associated with sex (p = 0.559), ethnicity
(p = 0.856), or age group (p = 0.139).
When rates of antipsychotic polytherapy were compared, there was a reduced
likelihood of coprescription of another antipsychotic in the clozapine versus
non-clozapine group (11.7% vs 17.6%; 2 16.5; df = 1; p
< 0.001).
Other Coprescribed Medication
Overall, the clozapine group was prescribed an average of 1.83 psychotropic
medications including the antipsychotic (SD 0.95; median 2; range 1–6).
Comparison with the non-clozapine group showed no significant difference (mean
1.93; SD 1.07; median 2; range 1–7) (Mann-Whitney U p =
0.125).
Table 3 shows the other
psychotropics coprescribed with clozapine. When rates of coprescription of
other psychotropics were compared with the non-clozapine treatment group,
there was an increased likelihood of being prescribed a medication to manage
adverse effects with clozapine (25.1%) versus non-clozapine treatment (19.0%)
(2 13.95; df = 1; p < 0.001). This difference is
predominantly due to the use of terazosin (an
1-antagonist), which is prescribed specifically for
clozapine-induced
hypersalivation.34
Overall, there was no significant difference in rates of anticholinergics
prescribed between the clozapine and non-clozapine groups (18.4% vs 17.0%,
respectively; 2 0.822; df = 1; p = 0.364). The rate of
coprescribed sedative–hypnotic treatment was, however, significantly
lower for the clozapine compared with the non-clozapine group (10.7% vs 16.7%;
2 17.84; df = 1; p < 0.001). Although there was no
difference in terms of rates of coprescribed antidepressants (clozapine, 17.4%
vs non-clozapine 16.6%; 2 0.35; df = 1; p = 0.556),
there was a trend toward lower rates of mood stabilizer coprescription in the
clozapine group (13.7% vs non-clozapine, 16.7%; 2 4.12;
df =1; p = 0.042).
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
The clozapine-treated group had experienced longer illness duration but
overall was not experiencing greater mental or physical health morbidity than
was the comparison non-clozapine–treated group. There are 3 findings of
particular note. The first of these findings is the lower rates of comorbid
neuroleptic-induced movement disorders associated with clozapine treatment.
This is consistent with lower rates of extrapyramidal symptoms reported with
clozapine, especially when compared with the first-generation antipsychotics
and
risperidone.1,14,35
Second, while it is well documented that clozapine is associated with
disorders such as obesity, diabetes, hypertension, and
hyperlipidemia,13
there also are reports of significant excess mortality and morbidity from
cardiovascular disease in schizophrenia patients versus the general
population.36-39
In this study similar rates of cardiac and endocrine comorbidity were found in
both the clozapine and non-clozapine groups (
8% and 10%,
respectively). This reflects age-matched cohort studies, which have reported
that increased risks may not solely be due to the direct effect of
antipsychotics and that these risks are seen with both first- and
second-generation antipsychotic
treatments.36,39
Finally, rates of gastrointestinal comorbidity were higher with clozapine
treatment. Constipation is a common problem
(14%)14
and is likely due to the drug's potent antimuscarinic activity. Other
gastrointestinal adverse effects reported with clozapine include heartburn,
nausea/vomiting, and reduced gut
motility.14,40
The average clozapine dose prescribed in this study (372 mg/day) was within the recommended therapeutic dose range,17,18 comparable to that shown in Australian outpatient surveys,41,42 and consistent with clinical trial data,43 but lower than doses reported in outpatient studies from the US.44
On average, it took almost 10 years from first contact with a health provider to starting clozapine treatment. Patients who had a longer delay in treatment were more likely to be older, of European ethnicity, and to have started clozapine when funding was restricted (1993–1998). Almost two-thirds of the patients were started on clozapine after government funding was initiated. On average, they had been prescribed 3 different antipsychotics prior to starting clozapine; for most patients, this included at least one second-generation antipsychotic (91%). The funding change appears to have been associated with a positive increase in the number of outpatients able to access clozapine for treatment-resistant schizophrenia.
Reassuringly, most outpatients were not sequentially given trials of every second-generation antipsychotic available in New Zealand before being started on clozapine; rather, only 7% had a treatment history that included all 3 second-generation antipsychotics. This contrasts with reports from the US and China, where clozapine use has decreased as the market share increased with non-clozapine second-generation antipsychotics.21,23,24
The duration before clozapine was initiated appears to be significantly delayed. This is concerning in light of evidence-based guidelines recommending early use, the UK long-term effectiveness study CUtLASS 2 (Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study),4 and emerging research supporting early clozapine treatment in first-episode patients.45 The CUtLASS study showed that to achieve a good response at one year in patients who had already failed 2 antipsychotic trials, it was better to commence clozapine than to switch to another second-generation antipsychotic.4 In a recently published US study of first-episode patients who had failed to respond to 2 antipsychotics, introducing clozapine (as early as 25 wk into start of treatment) resulted in a better response than switching. The authors suggested that clozapine use should be considered within the first 6 months of schizophrenia treatment and that it should be repositioned as a second-rather than third-line treatment.45 The issue of earlier use of clozapine was also examined by Kerwin.46 He reviewed arguments for and against earlier clozapine use than its current license allows and concluded that it should be used sooner rather than later and after a single unsuccessful trial of one second-generation antipsychotic.
In a small review of 112 schizophrenia inpatients in the UK, clozapine treatment was found to be delayed an average of up to 5 years.22 Five years is shorter than the New Zealand average of 10 years; however, comparing the 2 studies is not possible because of differences in the calculation of treatment delay. Treatment delay in the UK study was calculated from the sixth week of treatment with a second antipsychotic to first use of clozapine and excluded the time before clozapine was available in the UK (pre-1990). If this definition had been applied to the New Zealand data, it would have shortened the calculation, but the degree to which it would have changed the overall result is unknown. However, in line with the findings of this study, the UK study also found that older patients and those diagnosed before the introduction of clozapine had a longer delay in starting clozapine.
Antipsychotic coprescription with clozapine in this study was found to be
low (12%). Patients with longer duration before clozapine was initiated
were more likely to be coprescribed another antipsychotic, suggesting the
existence of a group with a longer history of poor response who were
potentially more treatment resistant. When rates of antipsychotic
coprescription with clozapine were compared with the rates in the
non-clozapine group, they were found to be lower (12% and 18%,
respectively).
Reported rates of antipsychotic coprescribing with clozapine vary. The cross-sectional rate in this study is comparable to that shown in a Chinese outpatient survey24 (13%) and lower than rates reported in the US (22% at inpatient discharge,47 36.2% polytherapy for >60 days,48 21.0% for polytherapy >90 days44).
There is little evidence to support antipsychotic combinations, but in severe treatment-resistant schizophrenia there are mixed reports supporting clozapine augmentation with another antipsychotic, in particular sulpiride49 (not available in New Zealand) and risperidone.50-53 Outside of this situation, antipsychotic polytherapy is typically recommended for only brief periods of switching, acute relapse,12,54 or perhaps for treatment-resistant schizophrenia when clozapine is contraindicated or when a patient refuses.55 In this study, just over half of the clozapine coprescribing was with risperidone. This suggests that a number of prescribers may have been augmenting a partial response to clozapine; however, it was beyond the scope of this study to examine clinical indications for coprescribing. Some practices observed have no evidence to support coprescribing; multiple (>2) antipsychotics and combinations including a depot antipsychotic should be avoided because of the risk of prolonged bone marrow suppression.17,18
This study also found that schizophrenia patients, irrespective of whether they were treated with clozapine or alternative antipsychotic(s), were prescribed an average of only 2 psychotropics, including the antipsychotic. This is possibly due to low rates of comorbid mental illness reported in both treatment groups (20% and 25%, respectively). A slightly higher number of psychopharmacologic medications (average of 3) were reported in a US study of schizophrenia patients in routine practice; 41% of this group had comorbid mental illness.56
Most commonly coprescribed psychotropics in this review were for the
management of adverse effects, 25% of the clozapine group compared with 19% in
the non-clozapine group. Terazosin and benztropine have been reported to be
effective in the management of clozapine-induced
hypersalivation,57
which occurs in approximately 30% of
patients.14
With non-clozapine antipsychotics, anticholinergics are recommended most
commonly for the management of acute extrapyramidal symptoms. About 11% of the
clozapine group was coprescribed a sedative–hypnotic, which was lower
than the rate for the non-clozapine group (17%). A possible explanation is
that sedation is the most commonly reported adverse effect caused by
clozapine's pharmacologic profile
(39%).14
Clonazepam, a long-acting benzodiazepine, was the most frequently coprescribed
sedative–hypnotic. This may be explained by augmentation of clozapine
partial responders (psychotic agitation, irritability,
anxiety),58
rather than use as a sedative–hypnotic. Concomitant antidepressants and
mood stabilizers were prescribed for less than 20% of both treatment groups,
and there was no significant difference in the rates between groups.
LIMITATIONS
The findings of this study represent real-world prescribing for all
schizophrenia patients attending a CMHC in Auckland and Northland. It reflects
the total antipsychotic regimen that the prescriber intended the patient to
take at the time of the review; however, the study did not explore adherence
to or the clinical effects of treatment. The methodology employed
cross-sectional data that were documented in the outpatient clinical records
and collected retrospectively. Therefore, the quality of the study data was
dependent to a large extent on the quality of the clinical recordings made by
the mental health clinicians. It was beyond the scope of this study to examine
whether patients were ever offered clozapine and reasons why prescribers may
or may not have recommended a clozapine trial or why patients may or may not
have agreed to such a trial.
The study findings may not be generalizable to other settings, such as elderly services or primary care; however, the findings reflect antipsychotic prescribing by more than 50 prescribers for approximately one-third of adult outpatients with schizophrenia or schizoaffective disorder living in New Zealand.
|
Conclusions |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
|
Footnotes |
|---|
I appreciate the assistance of the Mental Health Services and staff at the 4 DHBs. I thank Kirsten Norris, Morgan Kelly, and Karen Day for data collection and entry, Dr. Grant Paton-Simpson for database support, Elizabeth Robinson for statistical advice, and Prof. Peter Joyce for support and feedback.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionConclusionsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
 |
 |
 |
 |
