The National Institute of Mental Health

Clinical Antipsychotic Trials of Intervention

Effectiveness (CATIE) Project: Schizophrenia
Trial Design and Protocol Development

phase 3, in which clinicians help subjects select an

open-label treatment based on individuals' experiences
in phases 1 and 2.
Keywords: Randomized clinical trial, schizophrenia, antipsychotic drugs, effectiveness, longitudinal.
Schizophrenia Bulletin, 29(1):15-31,2003.

Abstract
The National Institute of Mental Health initiated the
Clinical Antipsychotic Trials of Intervention
Effectiveness (CATIE) program to evaluate the effectiveness of antipsychotic drugs in typical settings and
populations so that the study results will be maximally
useful in routine clinical situations. The CATIE schizophrenia trial blends features of efficacy studies and
large, simple trials to create a pragmatic trial that will
provide extensive information about antipsychotic
drug effectiveness over at least 18 months.
The protocol allows for subjects who receive a
study drug that is not effective to receive subsequent
treatments within the context of the study. Medication
dosages are adjusted within a defined range according
to clinical judgment The primary outcome is all-cause
treatment discontinuation because it represents an
important clinical endpoint that reflects both clinician
and patient judgments about efficacy and tolerability.
Secondary outcomes include symptoms, side effects,
neurocognitive functioning, and cost-effectiveness.
Approximately 50 clinical sites across the United
States are seeking to enroll a total of 1,500 persons
with schizophrenia. Phase 1 is a double-blinded randomized clinical trial comparing treatment with the
second generation antipsychotics olanzapine, quetiapine, risperidone, and ziprasidone to perphenazine, a
midpotency first generation antipsychotic If the initially assigned medication is not effective, subjects may
choose one of the following phase 2 trials: (1) randomization to open-label clozapine or a double-blinded
second generation drug that was available but not
assigned hi phase 1; or (2) double-blinded randomization to ziprasidone or another second generation drug
that was available but not assigned in phase 1. If the
phase 2 study drug is discontinued, subjects may enter

The advent of the newer atypical or second generation

antipsychotic medications developed in the wake of clozapine (olanzapine, quetiapine, risperidone, and ziprasidone)
has raised considerable hope that treatments for schizophrenia will be more effective and better tolerated than in
the past. The use of atypical medications has steadily
grown, and this trend is expected to continue. However,
there is no definitive evidence that these expensive medications can reduce morbidity and hospital use and
improve community functioning, as compared to the conventional or first generation antipsychotic drugs.
The existing evidence regarding the relative effectiveness of first and second generation antipsychotics for
schizophrenia is predominantly based on short-term studies and does not adequately address long-term effectiveness and cost issues. The studies of second generation
antipsychotics to date, which have for the most part been
sponsored by pharmaceutical companies and designed to
achieve regulatory approval based on evidence of efficacy
and safety, typically have lasted 4 to 12 weeks, have
involved initially hospitalized patients with no medical or
psychiatric comorbidities, and have permitted few concomitant medications. The main outcome measures of
these studies have mainly been the core psychopathology
of schizophrenia and well-known side effects (e.g.,

Send reprint requests to Dr. T. S. Stroup, CB# 7160, 10626

Neurosciences Hospital, University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599-7160; e-mail: [email protected].

15

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

by T. Scott Stroup, Joseph P . McEvoy, Marvin S. Swartz, Matthew J. Byerly,

Ira D. Qlick, Jose M. Canive, Mark F. McQee, Qeorge M. Simpson,
Michael C. Stevens, and Jeffrey A. Lieberman*

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

agitation and psychosis.) Soon after the contract period

began in October 1999, the original group of investigators
at UNC, Duke, and Yale expanded the Protocol Development Committee by including other schizophrenia
researchers who were not part of the original team. The
trial designs submitted in the contract application process
were a starting point for the protocol development process.
An initial day-long meeting of the committee took place in
Chapel Hill, NC (November 1999), with subsequent
weekly meetings held by teleconference. In this way, the
protocol was developed by the CATTE team of investigators through an iterative, systematic process. Additional
input was sought from various "stakeholder" groups,
including consumer representatives, mental health care
administrators, and policy makers.
In January 2000, a draft study design was presented to
consumer representatives at an NIMH-convened workshop
entitled Clinical Trials Recruitment: What Motivates People to Participate? Feedback from this meeting led to substantial protocol modifications. In March 2000, a revised
draft protocol was submitted to the NIMH for review by
its Scientific Advisory Committee for the CATDE program.
This committee made additional recommendations that
were incorporated into the final protocol.

Rationale for Trial Design

The CAllfc schizophrenia trial addresses the aims of the
CATTE RFP. It seeks to fill some of the existing gaps in
translating efficacy to effectiveness. To do this, the study
has adapted many of the characteristics of "pragmatic"
randomized controlled trials (RCTs). Pragmatic RCTs aim
to answer "real-life" questions of practical significance to
clinicians and patients in real-world clinical settings
(Hotopf et al. 1999). Such settings have varied levels of
ancillary services and serve patients with comorbid psychiatric illnesses, general medical problems, and substance
use disorders. Medication adherence is not ensured, as is
the case in routine practice. However, such pragmatic
studies randomize interventions and endeavor to ensure
that treatments and the assessment of outcomes are
blinded where possible.
The primary aim of the CATIE schizophrenia trial is
to determine the comparative effectiveness of a representative conventional antipsychotic and the different atypical
antipsychotic medications for a representative sample of
patients seeking treatment for chronic schizophrenia as
measured by all-cause treatment discontinuation rates and
associated measures of effectiveness and safety. Potential
causes for discontinuation include a clinical determination
of inadequate therapeutic effect (efficacy failure) or unacceptable side effects (tolerability failure), patient inability
or refusal to take the assigned antipsychotic (adherence/

16

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

extrapyramidal side effects [EPS]) that are particularly

problematic for some of the first generation drugs. These
studies have neither definitively demonstrated the "real
world" effects of the newer atypical antipsychotics in the
broad patient populations that receive them nor adequately
examined the broad range of side effects that may occur.
The limited types of assessment measures used and short
study durations have not provided adequate information
about treatment for this highly variable and chronic condition. Moreover, the patient samples involved in these studies and the conditions imposed by the restrictions of the
protocols limit the generalizability of the results. Additional information is needed to inform clinicians about the
appropriate role of atypicals. Because atypical antipsychotic medications represent a major budget item for
health benefit plans, including State Medicaid agencies,
objective information about their cost-effectiveness is
direly needed by policy makers.
Consequently, as part of its public mental health care
treatment initiative, the National Institute of Mental Health
(NIMH) established a request for proposal (RFP) for a
research program to evaluate the comparative effectiveness of antipsychotic drugs in disorders for which this
class of drugs is indicated (RFP 99-DS-0001) (see
Lebowitz et al., this issue). The RFP issued by the NIMH
specified that the protocol should focus on atypical
antipsychotic drugs and should follow a public health
model. In particular, the RFP specified that the trial should
do the following: focus on effectiveness and broader measures of outcome rather than efficacy alone; examine costeffectiveness and the impact of external factors on treatment delivery, adherence, and outcomes; enhance
generalizability by being as inclusive as possible, without
exclusions for comorbid psychiatric disorder, drug abuse,
or medical illness; place a premium on demographic and
geographic diversity; use multiple types of treatment settings that generally represent the systems of care in which
patients with schizophrenia are treated; and generate
results that inform community clinical practice. In addition, the RFP called for the program to develop a network
of sites and investigators able to respond to future needs
for effectiveness research, to serve as a base for ancillary
clinical investigations, and to include a mechanism for
adding new antipsychotic medications in the future. The
Clinical Antipsychotic Trials of Intervention Effectiveness
(CATTE) program was funded through a contract awarded
to the University of North Carolina (UNC).
This article describes the development of the study
designed to assess the effectiveness of antipsychotic drugs
for persons with chronic schizophrenia. (The article by
Schneider et al. in this issue describes the development of
the study designed to assess the effectiveness of antipsychotic drugs for persons with Alzheimer's disease with

The NIMH-CATIE Project

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

compliance), and administrative reasons (e.g., subject

moves out of the study area). Further, if differential discontinuation rates occur among the medications, the study
is designed to identify key causes for discontinuation.
In the protocol development process, study design
decisions were made by the Schizophrenia Protocol
Development Committee with input from various "stakeholder" groups. Some of the key design decisions and their
rationales are reviewed below.

the study that allows random assignment to a conventional

antipsychotic. In phase 1, patients with TD enter phase
1A, which includes randomization to only oral olanzapine, quetiapine, risperidone, or ziprasidone.
Long-acting intramuscular forms of conventional
medications were carefully considered for inclusion in this
study. A depot comparison gained strong support because
of its clinical relevance for persons with a history of medication nonadherence or substance abuse, or with an unstable living situation. Ultimately, however, the following
factors led to the decision not to include a depot comparison group in phase 1:

The expectation that a depot form of an atypical

medication would soon be marketed, limiting the
future significance of our findings
The difficulty of blinding injections and possible
bias against the unblinded (or blinded) treatments
The problem of testing a treatment indicated for
patients with a history of medication nonadherence in a study that requires all patients to agree
to take medications
The expected difficulty of enrolling participants
in a depot study with a conventional medication

Ultimately, we concluded that a comparison of oral to

depot medications would not be a primary goal of this
study and that the effectiveness of depot medications
should be examined in a study primarily focused on that
question.
Double-Blinded Treatment Conditions in Phases 1 and
2. The treatments in phases 1 and 2 are blinded (except
for clozapine) to enable rigorous comparisons of treatment effectiveness between drugs. The rationale for blinding is sufficiently compelling despite the complications it
introduces to the study's practical implementation and its
ecological validity. Because the oral medications used in
phases 1 and 2 have been marketed for several years, clinicians and patients have preconceived notions about the
efficacy and side effect profiles of the medications.
Therefore, oral medications are blinded and physicians,
raters, and patients do not know which medication has
been assigned to an individual. This is intended to
decrease bias in ratings and biased decisions about treatment failure. Clozapine is open label because of the logistical complexities that bunding it would involve given the
need for blood sampling for hematologic monitoring.

Inclusion of Conventional Antipsychotic Medications.

This study offers an important opportunity to determine
whether atypicals are indeed superior to conventional
antipsychotic drugs. Perphenazine was selected as the
conventional medication because it is a midpotency medication with only a moderate incidence of EPS (relative to
high-potency medications and other midpotency medications) and sedation (relative to low-potency medications).
We considered the appropriateness of assigning patients
entering the study on atypical medications to a conventional antipsychotic. Because our aim of determining the
relative real-world effectiveness of conventional and atypical antipsychotic medications is critical to clinicians and
policy makers, we determined that only tardive dyskinesia
(TD) is a contraindication for this treatment decision.
Therefore, patients with TD are excluded from the arm of

Patient and Clinician Involvement in Decision Making.

In response to feedback from consumer and advocacy
groups to an earlier version of this protocol at the NTMHconvened workshop mentioned above, it was decided to
include participant input when determining possible treat-

17

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Patient Sample. To make the results of the trial generalizable and representative of the broad group of chronic
schizophrenia patients, there are few exclusion criteria.
Patients with medical or psychiatric comorbidities and
those who require concomitant other medications are
included in the trial.
The patients who enroll in this study have chronic or
recurrent schizophrenia. First episode patients and wholly
treatment-refractory patients are excluded. First episode
patients are excluded because of their high rates of response
to antipsychotic medications at relatively low doses. Treatment-refractory patients are excluded because their severe
illness may preclude detection of differential effectiveness
that would be apparent in treatment-responsive patients.
A wide spectrum of patients with schizophrenia enroll
in the study, ranging from partially remitted outpatients to
exacerbated inpatients. Partially remitted patients, who
have received benefit from antipsychotic medication but
who remain symptomatic (because of lack of efficacy or
inability to tolerate an efficacious dose) or who suffer significant side effects, commonly consider a change in medications. Patients whose condition is exacerbated also
commonly need a change in medication. Patients who are
seemingly doing well on their current medication but who
wish to consider a change for reasons of greater improvement or better tolerability also enroll.

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

ment options. Consumer advocates suggested that this

feature would make the study more appealing to consumers and their family members and thus would improve
enrollment and retention rates. When participants enter
phase 2, they choose between two studies, one that compares clozapine to the other atypical medications not previously taken in the CATIE study, and another that compares ziprasidone to the other atypical medications not
previously taken. In addition, subjects and their clinicians
make a shared decision about an appropriate open-label
treatment in phase 3. When a participant enters phase 3,
partial information about phase 1 and phase 2 treatment
assignments is given to the study clinicians and patients
so that an informed treatment decision can be made based
on previous experience in the study. The names of the
phase 1 and phase 2 drugs are provided alphabetically
rather than chronologically in order to help preserve the
blind.

a study medication only when guideline-recommended

treatment options have been exhausted and clear clinical
deterioration persists, or for administrative reasons.
Although participant retention is important, the safety and
well-being of individual patients are the primary considerations in all clinical decision making.

Augmentation With a Second Antipsychotk Medication.

Simultaneous treatment with two antipsychotics is available in phase 3 as a result of feedback that the Protocol
Development Committee received from numerous stakeholders, including several mental health care administrators and policy makers, who reported that combining
antipsychotic drugs is a common treatment strategy when
response to a single antipsychotic is suboptimal. We will
obtain descriptive information about this treatment strategy because dual antipsychotic treatment is included as an
option in phase 3. The results may help decide whether to
include a two-antipsychotic arm in future CATTE trials.

Methods
Specific Aims and Hypotheses. The specific aims of the
schizophrenia trial are as follows:

Primary Outcome Variable. Time to all-cause treatment

failure, marked by its discontinuation, was selected as the
primary outcome variable. Although symptoms, side
effects, functioning, costs, and so on are important outcomes, we chose treatment discontinuation as the primary
outcome because it is a distinct measure that reflects both
efficacy and side effects. All-cause treatment discontinuation is a clinically meaningful outcome that reflects the
input of both the patient and the clinician. In the course of
a patient's treatment, the need to change medications
reflects the possibility that the treatment was not sufficiently effective or tolerable, or the belief that another
treatment would be superior.
Guidelines ensure that physician-initiated treatment
changes are made for only logical and justifiable reasons,
so that each treatment to which patients are assigned is
given every chance to be effective before a switch is made.
Guidelines also are provided to help ensure that clinicians
and patients attempt to optimize dosing and use adjunctive
medications before determining that the drug is not useful
or tolerable. The goal is for patients to be withdrawn from

18

To determine the long-term effectiveness and tolerability of the newer atypical antipsychotics, relative to perphenazine. We hypothesize that treatment with the newer atypical antipsychotics is
associated with greater long-term effectiveness
and tolerability than treatment with perphenazine.
To determine the comparative long-term effectiveness and tolerability of the newer atypical
antipsychotics in adults with schizophrenia. We
hypothesize that the newer atypical antipsychotics are similarly effective in treating psychopathology but have different side-effect liabilities.
To determine, among patients who fail treatment
with an initially assigned newer atypical antipsychotic, the long-term effectiveness and tolerability of the newer atypical antipsychotics, relative
to clozapine. Patients with inadequate resolution
of psychopathology, or marked sensitivity to
EPS, are recommended to enter this phase 2 trial.
We hypothesize that treatment widi clozapine is
associated with greater long-term effectiveness
and tolerability than treatment with a newer atypical drug other than the one the patient initially
received.

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Inclusion of Newly Approved Antipsychotics. Because

of our desire that the CATEE trials be maximally informative regarding current clinical practice, we made contingency plans to include, if possible, new antipsychotic
drugs that gained U.S. Food and Drug Administration
(FDA) approval during the study's enrollment period. In
February 2001, ziprasidone was approved by the FDA and
was then added to phases 1, 2, and 3 of the study. In
November 2002 aripiprazole was approved by the FDA
and was then added only to phase 3 as an open-label treatment. Thus we expect information from randomized,
blinded portions of the trial about ziprasidone, but will
obtain only open-label, descriptive data about aripiprazole.

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

The NIMH-CATIE Project

To determine, among patients who fail treatment

with an initially assigned newer antipsychotic,
the long-term effectiveness and tolerability of the
newer atypical antipsychotics, relative to ziprasidone. Patients with weight gain, hyperglycemia,
or hyperlipidemia are recommended to enter this
phase 2 trial. We hypothesize that treatment with
ziprasidone is associated with greater long-term
effectiveness and tolerability than treatment with
a newer atypical drug other than the one the
patient initially received.

Study Design. The CATTE schizophrenia trial is a multiphase RCT of antipsychotic medications involving up to
1,500 persons with schizophrenia followed for 18 months.
Figure 1 contains a schematic diagram of the trial design.
Participants are broadly representative of persons with the
chronic and recurrent forms of schizophrenia. Patients are
to be followed for at least 18 months. The study includes
three treatment phases and a naturalistic followup phase.
Persons who do well on an assigned treatment remain on
that treatment for the duration of the 18-month treatment
period. If an assigned treatment is deemed a failure, the
patient moves to the next phase of the study to receive a
new treatment. All study medications are double-blinded

Figure 1. Schizophrenia trial design. Responders stay on assigned medication for duration of 18month treatment period. Phase 1A = Participants with tardive dyskinesia are not randomized to perphenazine; phase IB = Participants who fail to respond to perphenazine are randomized to an atypical
(oianzapine, quetiapine, or risperidone) before they are eligible for phase 2; R = randomized.
Phase 3

Phase 2

Phase;!
Double-blind, random
treatment assignment

Participants who discontinue

phase 1 choose either the
clozapine or the ziprasidone
randomization pathways

CLOZAPINE

OLANZAPINE

(opn label)

OLANZAPINE,
QUETIAPINE, or
RISPERIDONE

QUETIAPINE

1,500
participants
with
schizophrenia

Participants who discontinue

phase 2 choose one of the
following open-label treatments
ARIP1PRAZOLE
CLOZAPINE
FLUPHENAZINE
DECANOATE
OLANZAPINE
PERPHENAZINE

RISPERIDONE

ZIPRASIDONE
QUETIAPINE
OLANZAPINE,
QUETIAPINE, or
RISPERIDONE

ZIPRASIDONE

RISPERIDONE
ZIPRASIDONE

PERPHENAZINE

No one assigned to same

drug as in Phase 1

It

Two of the above

antipsychotics

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

in treatment phases 1 and 2 except for clozapine, which is

dispensed open label because of the clinical requirement
to monitor maintenance of adequate white blood cell
counts.
Outcome measures for the trial extend beyond efficacy and safety to capture more details about participants'
quality of life, cognitive capacity, and use of general, mental health, substance abuse, and rehabilitative services. The
representativeness of the subjects is ensured by recruiting
into the trial a broad range of "real-world" patients,
including those with comorbid conditions (e.g., substance
use disorders, medical problems) that would exclude them
from most clinical trials. The study is being conducted at
sites that represent a broad array of clinical settings (e.g.,
State mental health, academic, Department of Veterans
Affairs, health maintenance organization, managed care)
in order to generate representative, generalizable, and
practically relevant study findings.
In phase 1 of the trial, patients are randomly assigned
to double-blinded treatment with oral perphenazine,
oianzapine, quetiapine, risperidone, or ziprasidone.
Patients with TD bypass phase 1 (and the risk of assignment to perphenazine) and are assigned to treatment with
oral oianzapine, quetiapine, risperidone, or ziprasidone
(phase 1A).

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

Participants who discontinue treatment with the conventional antipsychotic (oral perphenazine) in phase 1 are
randomly assigned to double-blinded treatment with
olanzapine, quetiapine, or risperidone (phase IB).
In phase 2, participants who discontinue their assigned
treatment with an atypical antipsychotic in phase 1,1A, or IB
are recommended to one of two treatment assignment pathways based on their reason for discontinuation, as follows:

Once and twice daily medication

schedules.

Quetiapine and ziprasidone require twice daily (BED) dosing, whereas perphenazine, olanzapine, and risperidone
can be taken once daily (QD). Because it would deviate
from clinical practice patterns to insist on BID dosing in
all cases, we sought to minimize the number of patients
who are required to take their medication BID, and still
maintain blinding to the extent possible. Throughout the
study, half of the patients randomized to perphenazine,
olanzapine, and risperidone are assigned to BID dosing,
and half are assigned to QD dosing. All quetiapine and
ziprasidone patients receive BID dosing unless they are
determined to need only one capsule per day.
Because quetiapine has a relatively slow recommended initial titration schedule, patients assigned to a
BID dosing schedule begin their treatment using an initial
dose strength and follow the titration schedule described in
table 4. After the initial titration period using a starter pack
of capsules, the total BID dose of all blinded study medications is identical to the QD dosing schedule. It is
expected that there will be roughly equal efficacy and
safety responses for the QD and BID dosing of these medications.
Dosing for open-label medications. As noted in table
5, the recommended dose range for open-label clozapine
in phases 2 and 3 is 200 to 600 mg daily. The recommended dose range for fluphenazine decanoate injections
available in phase 3 is 12.5 to 50 mg intramuscularly (IM)
every 2 weeks for most patients. The recommended dose
range for aripiprazole in phase 3 is 10 to 30 mg daily.
Transition to study medications. Because this study
addresses long-term effectiveness and tolerability, not
acute treatment efficacy, efforts are made to facilitate the
transition of patients onto their study medications and prevent their destabilization by this process. For patients who
are already taking an antipsychotic medication prior to any
phase, tapering the previous medication over 2 weeks is
recommended, but cross-titration is permitted for up to 28
days. Clinicians have substantial leeway in implementing
these transitions according to clinical judgment.
Summary of pharmacological treatments. The methods allow double-blind comparisons among the newer
atypical antipsychotics, and between the newer atypical
antipsychotics and perphenazine. The blinding procedures
allow dosing across commonly prescribed dose ranges of
the study drugs (olanzapine 7.5-30 mg daily, quetiapine
200-800 mg daily, risperidone 1.5-6 mg daily, ziprasidone
40-160 mg daily, and perphenazine 8-32 mg daily) with
equivalent numbers of capsules, that is, one to four daily.

The first pathway in phase 2 offers a 50:50 randomization to ziprasidone versus one of the other atypical antipsychotics (olanzapine, quetiapine, or risperidone) not previously received by the patient in this
study. Patients who discontinued their previous treatment with an atypical antipsychotic because of intolerance to the previous regimen (e.g., weight gain) are
expected to preferentially enter this study.
The second pathway in phase 2 offers a 50:50 randomization to clozapine versus an atypical antipsychotic (olanzapine, quetiapine, or risperidone) not
previously received by the patient in this study.
Patients who discontinued their previous treatment
with an atypical antipsychotic because of inadequate
efficacy are expected to preferentially enter this
study.

Phase 3 is for persons who discontinue the treatment

assigned in phase 2. Research and clinical staff examine reasons for failure of the treatments assigned in phases 1 and 2.
Based on this review, study treatment guidelines are reviewed
and the research participant is followed in phase 3 on an
open-label treatment chosen collaboratively by the clinician
and patient.
The followup phase is for persons who are no longer willing to continue taking study medication or who have discontinued their phase 3 medication before 18 months from the time
of initial randomization has elapsed. Followup phase participants are not provided with study medication but are followed
naturalistically on their treatment of choice. Fbllowup visits to
collect basic outcome data are scheduled quarterly through 18
months from the time of a participant's initial randomization.
Subjects. A sample of 1,500 men and women, 18 to 65
years of age, who currently meet or have met in the past
DSMIV diagnostic criteria for schizophrenia, based upon
the Structured Clinical Interview for DSM-IV (SCID;
First et al. 1994), review of their clinical records, and
input from available informants, are enrolled. Inclusion
and exclusion criteria are listed in tables 1 and 2.
Treatments
Pharmacological treatments.
Dosing of study medications. Standard study capsules
contain olanzapine 7.5 mg, quetiapine 200 mg, risperi-

20

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

done 1.5 mg, ziprasidone 40 mg, or perphenazine 8 mg

(table 3). Clinicians can prescribe one to four capsules
daily, based on individual patients' therapeutic response
and side-effect burden.

The NIMH-CATIE Project

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

Table 1. Inclusion criteria

1. Patients must be 18-65 years of age.
2. Patients must currently meet or have met in the past the DSM-1Vcriteria for schizophrenia.
3. Patients entering the study must, according to their own judgment in consultation with their physician, have a condition
appropriate for treatment with an oral medication.
4. Patients must demonstrate adequate decisional capacity to make a choice about participating in this research study and
must provide informed consent to participate.

Table 2. Exclusion criteria

1. Patients with a DSM-tVdiagnosis of schizoaffective disorder, mental retardation, pervasive developmental disorder,
delirium, dementia, amnesia, or other cognitive disorders are excluded.

3. Patients in their first episode of schizophrenia are excluded. Patients will be considered to be in their first episode if
they first began antipsychotJc drug treatment for psychosis within the previous 12 months and have had psychotic
symptoms for less than 3 years.
4. Patients with well-documented histories of failure to respond to even one of the proposed treatment arms are
excluded. A treatment failure has occurred If the patient continued to demonstrate severe psychopathology in spite of
fully adhering to treatment at an adequate dose of the medication for an appropriate length of time. Specific dose and
duration criteria are as follows:

Olanzapine at dosages 30 mg/day for 6 consecutive weeks

Quetiapine at dosages 800 mg/day for 6 consecutive weeks
Perphenazine at dosages 32 mg/day for 6 consecutive weeks
Risperidone at dosages 6 mg/day for 6 consecutive weeks
Ziprasidone at dosages 160 mg/day for 6 consecutive weeks

5. Patients currently, or in the past, treated with clozapine for treatment resistance are excluded. Patients who have taken
clozapine for reasons other than treatment resistance may be eligible.
6. Patients who are currently stabilized on haloperidol decanoate or fluphenazine decanoate and who require long-acting
injectable medication to maintain treatment adherence are excluded.
7. Women who are pregnant or breastfeeding are excluded. Women of childbearing potential must agree to use appropriate contraception in order to enroll in this study.
8. Patients with tardive dyskinesia are excluded from assignment to conventional antipsychotic treatment arms.
9. Patients with a contraindication to any of the drugs to which they might be assigned are excluded.
10. Patients with a medical condition that is serious and acutely unstable are excluded.
11. Patients with the following cardiac conditions are excluded:

Recent myocardial infarction (<6 months)

QTc prolongation (screening electrocardiogram with QTc > 450 msec for men, QTc > 470 msec for women)
History of congenital QTc prolongation
Sustained cardiac arrhythmia or history of sustained cardiac arrhythmia
Uncompensated congestive heart failure
Complete left bundle branch block
First-degree heart block with PR interval 0.22 seconds

12. Patients on concurrent treatment with dofetilide, sotalol, quinidine, other Class la and III antiarrhythmics, mesoridazlne,
thioridazine, chlorpromazine, droperidol, pimozide, sparfloxacin, gatjfloxacin, moxifloxacin, halofantrine, mefloquine,
pentamidine, arsenic trioxide, levomethadyl acetate, dolasetron mesylate, probucol, or tacrolimus are excluded.
13. Patients who have taken any investigational drug within 30 days of the baseline visit are excluded.

21

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

2. Patients with well-documented, drug-related, serious adverse reactions to even one of the proposed treatment arms
are excluded.

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

Table 3. Standard study dosing

Blinded medications

Dose per standard study capsule

Olanzapine
Perphenazine
Quetiapine
Risperidone
Ziprasidone

Dose range
7.5-30
8-32
200-800
1.5-6
40-160

7.5 mg
8mg
200 mg
1.5 mg
40 mg

mg/day
mg/day
mg/day
mg/day
mg/day

Table 4. Initial twlce-a-day dosing schedule

Blinded
medications

Initial dosage
capsules
7.5
8
100
1.5
40

mg
mg
mg
mg
mg

Day1

Day 2

7.5 QD
8QD
100 QD
1.5 QD
40 QD

7.5 QD
8QD
100 QD
1.5 QD
40 QD

Day 3

Day 4 a.m.

7.5 BID
8 BID
100 BID
1.5 BID
40 BID

7.5
8
100
1.5
40

Day 4 p.m.
7.5
8

200
1.5
40

Note.BID = twice a day; QD = once a day.

Research participants are offered psychosocial interventions directed at improving patient and family understanding of the illness, decreasing the burden of illness on
the family, maximizing treatment adherence, minimizing
relapse, enhancing access to a range of community-based
rehabilitative services, and improving study retention.
Interventions include the following:

Table 5. Recommended dosing for open-label

study medications
Open-label medications

Dose range

Fluphenazine decanoate
Clozapine
Aripiprazole

12.5-50 mg every 2 weeks

200-600 mg/day
10-30 mg/day

This design also allows comparison of QD versus BID

dosing strategies.
Clinicians can adjust the dose of the assigned antipsychotic as clinically indicated within the range of one to four
standard study capsules. Clinicians can prescribe adjunctive
and concomitant medications whenever they believe these
are indicated, but they must record the indications for their
use. Antipsychotics other than the assigned study medicines
are not allowed. The antipsychotic cannot be discontinued or
changed without considering the treatment a failure. Additional antipsychotics may not be prescribed after cross-titration is complete except in emergencies or in phase 3 if
patients elect the dual antipsychotic treatment option. The
ongoing need for an additional antipsychotic is a criterion for
treatment failure.

Patient and family education

Opportunities for family support
Adherence enhancement
Broker/Unking style case management or liaison
with existing case management services

Patient and family education. All participants are

offered an individually tailored educational plan adapted
from the successfully implemented Texas Medication
Algorithm Project (TMAP). The TMAP education plan is
conducted in several phases and invites family participation if desired by the patient. Each phase involves one or
more individual sessions, with patients moving to the next
phase as appropriate. All sessions are presented orally,
with supplementary print materials offered. Videotape
educational presentations are also made available. To
assess the intensity of educational efforts, research staff
log the educational sessions provided to each subject.
The phases of the educational plan include the following:

Psychosocial Interventions. Meta-analytic reviews (e.g.,

Mojtabai et al. 1998) support the additive effectiveness of
psychosocial treatments delivered in combination with
pharmacological treatment. Adding a psychosocial treatment produces greater improvement on measures of outcome than pharmacological treatment alone. The addition
of a psychosocial treatment decreases relapse rates by 20
percent, on average, relative to medication treatment
alone.

22

Individual introductory education. Content

includes diagnosis, medications, and symptom
self-monitoring.
Individual followup. Content includes selfassessment of symptoms and side effects, assessing changes in symptoms, and so forth.

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Olanzapine
Perphenazine
Quetiapine
Risperidone
Ziprasidone

Dally Dose (mg) During BIDTttratlon

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

The NIMH-CATIE Project

Ongoing education. Content includes more extensive information about the illness, and treatment
and reinforcement of earlier content.
Group education/support. This phase offers
opportunities for patients to discuss the illness
with each other and provides referrals to other
support groups.

Family support. Study sites are encouraged to work

with the National Alliance for the Mentally 111 (NAMI) to
facilitate offerings of family support groups at study sites
or in the local community. Where feasible, structured family education programs (Journey or Hope/Family-toFamily program) are also offered.
Adherence enhancement Kemp et al.'s (1996) compliance therapy intervention was adapted for CATTE to
guide and enhance patient and family education on medication adherence. This cognitive behaviorally oriented
approach equips therapists to address common problems
with denial of illness and resistance to treatment. In the
initial phase, the patient's stance toward treatment is
elicited. In the second phase, ambivalence to treatment is
explored, common misgivings are addressed, misconceptions are corrected, and indirect benefits of medications
are emphasized. The third phase of compliance therapy
focuses on treatment maintenance. Research associates
and study clinicians have access to adherence enhancement materials on the CATTE Web site.
Case management. Sites vary in case management
availability. Nonetheless, all participants have access to
locally available community-based treatment and rehabilitative services so as to maximize potential rehabilitation
and functioning. For subjects with active case management, study sites do not duplicate these services but serve
as a liaison to ongoing case management services. Where
subjects do not have case management provided, site personnel serve as broker/linking style case managers,
attempting to gain entry for subjects to community services.
Measures and Assessment of Outcomes. Subjects attend
monthly study visits for clinical examinations and to participate in assessments following the study's Schedule of
Events (table 6). Outcome measures are obtained from a
variety of sources, including patient self-report, clinician
ratings, and ratings by trained study personnel.
Clinical and functional assessments.
Primary outcome. Time to all-cause pharmacological
treatment failure, reflected by a decision to discontinue
and change medications, is the primary outcome in the
study. For time to withdrawal to serve as a meaningful
variable, treatment retention must be maximized. Therefore, except under unusual circumstances, patients are

23

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

withdrawn from the study only when treatment options

have been exhausted and clear clinical deterioration persists. Nonadherence, exacerbation of psychopathology, or
hospitalization do not necessarily require diat a patient be
withdrawn from the treatment that he or she is receiving in
the study (e.g., if a patient experiences a recurrence of
symptoms because he or she went away on a trip for 1
week and forgot to bring the medication and therefore did
not take it). Similarly, only a side effect that produces significant subjective distress or objective dysfunction that
cannot be resolved by dose adjustment or adjunctive medications should lead to treatment discontinuation. However, this focus on patient retention is balanced by the fact
that the safety and well-being of patients are the primary
considerations in all clinical decision making. Under no
circumstances are patients maintained on their current
study medications if there is a clear clinical reason to
change or they prefer to move on to the next phase of the
study or discontinue study medication or participation
altogether.
In addition to this primary outcome measure, the clinical and functional outcome measures listed in table 7 are
used to assess psychopathology, recovery, relapse, and
severity of illness over the course of the study. Descriptions of these measures and the rationales for their selection appear in the article by Swartz et al. in this issue.
Side effects and adverse events measures. In addition to the measures listed in table 7, metabolic effects of
the drugs are monitored regularly. Each participant's
pulse, blood pressure, and weight are assessed regularly.
To evaluate the effect of antipsychotic treatments on
weight gain, glucose and lipid metabolism, waist-hip ratio
(WHR), and body-mass index (BMT) are recorded regularly. In addition, fasting blood glucose level, hemoglobin
A1C, lipid profile (total cholesterol, high-density lipoprotein [HDL] cholesterol, triglycerides), complete blood
count, and serum prolactin level are collected according to
the Schedule of Events (table 6).
Adverse events and side effects. Given the substantial
clinical experience already available with these drugs, we
elected to focus our monitoring on 18 well-known adverse
events and side effects (AE/SE) that have been commonly
reported with one or more of the study antipsychotics. By
asking systematically about these 18 AE/SE, we will
determine the relative frequencies and severities of these
common and important AE/SE across the study antipsychotics. However, we also included a general inquiry section to record responses to questions about general physical health.
Family/caregiver experiences. Family burden is
measured to assess the impact of schizophrenia on the
objective and subjective burden on the family using a
revised version of the Family Experiences Interview

 11

!.!.

X
X

- | itn(

t~ n

. jW >Cm.mO

BX

- - - i l l

ac

Fa onM

PKtkKOt

Thyrwd-atfetiKfrg h m M fTW)

Nil

F M J T > * 3 > *M*m fct i t U i

(TAQ

O^lylUfc{IW M UClTiirir<^fcyWUi
Sofa, MtaM kmm km- L * w Qu*r r UN

KMIfn

r II

knwAMMUl

MmnMOMMM

X
X
X
X

X
X

X
X
X
X
X

X
X

X
X

X
X

X,

X
X

x
X

X.

X
X

X
X

X
X
X

X
X

X
X

X
X

X
X

X
X

X
X

X
f,

1?

X
X

X
X

If

X
X

X
X

X
X

1?

X.

1]

x
X

X
X

1?

X,

11

'9

X
X

X
X

X
X

x
x

X.
X

X
X

X.

X
X

X
X

X,

X
X

X.

X.

X
X,

X
X

C^yQ>
X

nun

tfpplcJ(aB)

X
X

am wtatM tea*, >* eg w .

MMXA, U f a , ftaM M U u M

IWMn , * . fan*

tan

X
X
X

X
X
X

C. . . _. L ' - 4 _ . . U . J i n i i mi i

ttacCAT-CD

Owrt A M ttMM

MUhpM

mdp+t

flid I M * *mm

WMkWboWcM Farm

UVp-yclMrie Uttory

DSUk

Mi

tiinittufc

Schedule of Events

Table 6. Schedule of Events

X
X

X
X

X
X

23

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

C/3

to

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

The NIMH-CATIE Project

i
M

H
N

S
-

<

x x x x X X X X

m-

gf:

Schedule of Events cont'd

i
!

x X

I
I

x X x x

i
i !
siI
I

i! 1
fs
i J
j
i 1 11
!
I

[I
i

<

lr.

WRAT-m Reading Test (Wilkinson 1993)

Controlled Oral Word Association Test (Benton
and Hamscher 1978)
Category Instances (Category Fluency) (Benton
and Hamscher 1978)
Wechsler Intelligence Scale for ChildrenRevised, Mazes (Wechsler 1991)
Hopkins Verbal Learning Test (Brandt 1991)

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Schedule (Tessler and Gamache 1995). Families and

caregivers are asked about instrumental support provided, limit setting provided for disturbing behavior,
time spent directly related to care for the patient, lost
productivity, and financial contributions to the patient's
care.
Medication adherence. Medication adherence is
conceptualized as an intermediate outcome variable in this
study. However, because adherence with treatment is a
prerequisite for determining the effectiveness of the study
medications as indicated by all other outcome measures,
efforts are made to maximize participant adherence to
medication regimens. Although these efforts to enhance
adherence may confound the assessment of adherence as
an outcome measure, relative to usual care, they do so
equally across all treatment groups.
To enhance understanding of the relationship between
patient attitudes toward medications, insight into illness,
and medication adherence, we administer the Drug Attitude Inventory (Awad 1993) and the Insight Into Treatment Attitude Questionnaire (McEvoy et al. 1989) at intervals specified in the Schedule of Events (table 6).
A pill count is conducted monthly. The patient is
asked a standard set of questions about medication adherence at each appointment. Study personnel synthesize the
patient report, pill count, and any other available information into a global judgment of medication adherence.
Substance use. Substance use is examined in multiple ways. Subjects are asked quarterly about drug and
alcohol use. Urine drug screens are performed quarterly.
In addition, hair samples for radioimmunoassay to detect
substances of abuse are collected during patient visits and
shipped to a central laboratory for assessment
Neurocognitive assessments. Because the choice of
tests for a neurocognitive battery is often controversial,
the tests were chosen with consensus approval from the
leaders in schizophrenia and dementia research. A
Neurocognitive Advisory Group was selected in concert
with the NIMH project officer and the External Scientific
Advisory Board. Selection of members was designed to
ensure maximum consensus and integration of personnel
across committees.
The neurocognitive battery is completed according to
the Schedule of Events in table 6. The tests are listed
below.

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

Table 7. Clinical and functional outcome measures

Psychopathology

Adverse events/side effects

Psychosocial

Addlngton et al. 1990.

Barnes 1989; Carpenter et al. 2000.
Simpson and Angus 1970.
4
Heinrichs et al. 1984.
6
Lehman1988.
6
Steadman et al. 1998.
2

Face Emotion Discrimination Task (Kerr and

Neale 1993)
Wechsler Adult Intelligence Scale-Digit Symbol
Test (Wechsler 1974)
Letter-Number Span (Gold et al. 1997)
Grooved Pegboard (Lafayette Instrument Company 1989)
Computerized Continuous Performance Test
(Cornblatt et al. 1988)
Computerized Test of Visuospatial Working
Memory (Hershey et al. 1999)
Computerized Wisconsin Card Sorting Test
(Heaton et al. 1993)

The high cost of atypical antipsychotic medications

can be justified if their use leads to savings in other health
care or non-health care costs or if they yield improvements in the well-being of patients, their families, or the
communities in which they live. Cost analyses will be conducted from the perspective of society and will include the
cost of medications and all other health care costs.
Costs will be measured as the product of the units of
service X the cost per unit, and will be assessed through
the Service Use and Resources Form (SURF; Rosenheck
et al. 1998). The SURF comprehensively documents use
of health services and non-health resource consumption
such as involvement with the criminal justice system and
public support payments. We will conduct a series of sensitivity analyses using different unit-cost estimates. Medication costs will be based on the specific dosing of all
agents to each patient.
Effectiveness will be evaluated in two ways. First, we
will use the utilities generated by the Health Utilities Index
(HUI; Feeny et al. 1996), a health-related quality-of-life
index that measures preference-based utilities and health
states in quality-adjusted life years (QALYs). Second, we
will combine other quality-of-life and symptom measures
and scale them to generate disease-specific QALYs, using
methods developed elsewhere (Rosenheck et al. 1998).
Sensitivity analysis will be conducted using both measures
of effectiveness.
Initially, costs and effectiveness measures will be
examined separately. If the increased medication costs of

These tests and the neurocognitive functions measured

by them are described by Keefe et al. elsewhere in this issue.
Health services and cost-effectiveness assessments.
External factors such as the contacts between the patient
and his or her health care setting and psychosocial support
system (including family interactions, living situation,
environmental safety, and financial resources) may affect
outcomes and also reflect the effectiveness of treatments.
The methods used to examine the organizational context
of care, the subject service utilization, and the basis for
the study's cost-effectiveness and cost-benefit analyses in
CATIE are outlined by Schneider et al. (2001) in an article describing the Alzheimer's disease trial and used by
Rosenheck and colleagues elsewhere (Rosenheck et al.
1997, 1999). These methods are summarized here.

26

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Family experiences

Clinical Global Impressions ScaleSeverity and self-report versions

Positive and Negative Syndrome Scale
Calgary Depression Rating Scale1
Barnes Akathisia Scale2
Simpson-Angus Extrapyramidal Side Effect Scale3
Abnormal Involuntary Movement Rating Scale
Adverse events/side effects form
Heinrichs-Carpenter Quality of Life Scale4
Lehman Quality of Life Interviewselected items5
MacArthur Community Violence Instrument6
Medical Outcomes Study Short Form-12
Insight Into Treatment Attitude Questionnaire
Drug Attitude Inventory
Family Experiences Interview Schedule (adapted)

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

The NIMH-CATIE Project

atypical antipsychotics are offset by reduced services

costs, and outcomes are superior, then the atypical antipsychotics will emerge as dominant choices and further analysis will not be necessary. If both costs and benefits
increase with atypical antipsychotics, further analysis will
be undertaken using cost-effectiveness acceptability
curves (Van Hout et al. 1994). These analyses take into
consideration various estimates of the monetary value of a
QALY and estimate the probability of achieving various
cost-effectiveness ratios.

Safety
Serious adverse events, as defined by the FDA, are
assessed according to FDA guidelines. Treatment-emergent adverse effects are monitored using the AE/SE measure described above.
Clinical Monitoring and Laboratory Tests. The schedule for clinical monitoring and laboratory testing appears
in the Schedule of Events (table 6). Screening evaluations
are intended to identify medical abnormalities and to
establish baseline levels. Parameters to be monitored
include clinical chemistry and electrolytes (sodium, potassium, bicarbonate, chloride, blood urea nitrogen, creatinine, phosphorus, uric acid, calcium, total protein, albumin, creatine kinase, alkaline phosphatase), liver function
tests, thyroid-stimulating hormone, complete blood count,
urinalysis, fasting blood glucose, lipids (total cholesterol,
HDL cholesterol, and triglycerides), urine pregnancy test
(for all women), prolactin, and an electrocardiogram
(ECG). Laboratory assessments are collected and
processed via a central laboratory, while ECGs are
processed through a central ECG laboratory.
Each patient's pulse, blood pressure, and weight are
recorded at screening, month 1, and month 3, and quarterly thereafter. Height is recorded at screening. To evaluate the effect of antipsychotic treatments on weight gain,
we calculate the WHR and the BMI on this same schedule.
In addition to the screening ECG, an ECG is conducted at
the month 18 visit, at each end-of-phase visit, and 1 month
after a participant starts on a new study medication.
An ophthalmoscopic exam to evaluate for the presence of cataracts is conducted as part of the screening
physical exam. Followup ophthalmoscopic exams are conducted every 6 months and at end-of-phase visits.

Statistical Methods and Analytic Plan

Primary OutcomePhase
receive rerandomization to an
they had not previously been
assigned atypical treatment

The statistical methods and analytic plan of the CATIE

schizophrenia trial are described in detail by Davis et al. elsewhere in this issue. They are briefly summarized below.

27

2. Patients are eligible to

atypical treatment (to which
assigned) if their originally
fails. Because the second

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Primary OutcomePhase 1. The primary analysis will

consist of a comparison of treatment discontinuation rates
between the conventional treatment group and the pooled
atypical treatment groups in phase 1 from the beginning
of the trial (olanzapine, quetiapine, risperidone).
Treatment discontinuation will be defined as withdrawal
from randomized treatment or the study itself for any reason, including lack of safety, lack of efficacy, poor medication adherence, or patient decision (after clinical reasoning and discussion with the patient by the clinician
have taken place). The analysis will be conducted on the
intent-to-treat population, with success or failure defined
for all patients randomized. The objective is to determine
if the atypical treatments maintain patients on their original therapy for a longer period than the conventional therapy. To this end, Kaplan-Meier estimates of the time to
treatment discontinuation will be produced. Hypothesis
tests for assessing treatment differences will be based on
the Wilcoxon rank test for differences in survival distributions (Kalbfleisch and Prentice 1980) between the pooled
atypical treatments and the conventional treatment, adjusting for baseline status (exacerbated, partially remitted)
and clinical site. Once daily and BID dose groups will be
pooled for all primary analyses.
In the second step of the primary analysis, the three
atypical treatment groups will be compared. Assuming a
significant difference was found between the pooled atypicals and the conventional in phase 1, comparisons for time
to all-cause treatment discontinuation will be made among
the three atypical treatment groups for phases 1 and 1A
combined to determine if any treatments are substantially
different from the others.
All other analyses of the primary outcome or analyses
of secondary outcomes will be considered descriptive in
nature, and statistical significance will therefore be evaluated relative to p = 0.05 as a descriptive tool.
Secondary analyses of the primary outcome, controlling for baseline covariates, will be performed using proportional hazards regression models (Cox 1972). In particular, an analysis of treatment discontinuation adjusted for
clinical site, baseline status, and whether the patient is randomized to a QD or BID treatment regimen will be conducted, and any interaction between baseline status or regimen and treatment will be assessed. Other covariates of
interest will include baseline demographic variables (e.g.,
gender) and overall disease severity. Interactions between
significant covariates and treatment effects will be
assessed in the context of the proportional hazards models.

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

Persons who are initially assessed not to have the

capacity to make an informed decision but who wish to
participate are offered a brief structured educational program based on the one used by Carpenter et al. (2000). The
program focuses on the prospective participant's difficulties in learning and manipulating information in an effort
to improve the person's level of functioning to the point
where a competent decision can be made about whether to
enter the study. After this program is completed, decisional
capacity is again assessed using the MacCAT-CR. Persons
who demonstrate decisional capacity at that time sign a
consent form and enter the screening phase of the study.
Persons who are determined not to have the capacity
to consent to participate in a research study even after
remediation are not enrolled in the study unless a guardian
is available and guardian consent for research is allowed in
the particular study site's jurisdiction.
To assess decisional capacity longitudinally, and to
examine the effects of the study medications on decisional
capacity, the MacCAT-CR is administered to participants
at screening, month 6, month 18, and each end-of-phase
visit. The screening, month 6, and month 18 administrations coincide with neurocognitive assessments.

Comparison of ziprasidone versus olanzapine,

quetiapine, and risperidone combined in the trial
for patients who discontinued the first randomized treatment because of tolerability issues
Comparison of clozapine versus olanzapine, quetiapine, and risperidone combined in the trial for
patients who discontinued the first randomized
treatment because of lack of efficacy

Analyses will be performed using similar methodologies as those described for the first phase.

Informed Consent and Subject Advocate Procedures.

Consent is obtained at the outset for all phases of the
study except for the followup phase. (Consent for the followup phase is obtained at the time of entry into that
phase.) All participants are asked to appoint a person who
serves as a "subject advocate," who is involved with the
informed consent discussion and assists the subject with
decision making. Optimally, the subject advocate is a person of the patient's choosingusually a family member
or close friend. In the event that such a person is not available, each research site designates a person not otherwise
involved with the research project (e.g., a social worker, a
case manager) to serve in this role. If at any subsequent
point the participant is believed to have severely impaired
capacities to protect his or her own interests, then the subject advocate makes a decision as to whether the patient
should leave the study. That determination is based on
whether there has been such a substantial decrement in
the benefit-risk ratio for this participant that the assumptions on which the participant's original decision to enter
the study were predicated are no longer valid.
Involvement by the subject advocate is ongoing but is
especially important at entry into phases 2 and 3.

Other Outcomes. Initial analyses of all secondary outcome variables will be based on the phase 1 and 1A treatment randomization in an intent-to-treat fashion, regardless of the treatment the patient is receiving at the end of
the study. The goal of this approach is to assess which
treatment is superior as initial therapy regardless of any
subsequent randomizations.

Human Subjects Considerations

Given the heightened public and scientific concern about
the participation of mentally ill subjects in research, a
CATTE Ethics Committee, chaired by Paul Appelbaum,
M.D., was established to review all human subjects considerations. This committee reviewed the protocol and recommended the procedures outlined below.

Assessment of Decisional Capacity. All prospective

research participants are screened for decisional capacity
using the MacArthur Competence Assessment
Tool-Clinical Research (MacCAT-CR) (Appelbaum and
Grisso 1996). Persons who demonstrate adequate decisional capacity to participate in this research study are
allowed to decide whether to enter the study. Those who
choose to participate sign a consent form and enter the
screening phase.

Summary
This article has described the rationale, aims, and design
of the CATTE schizophrenia trial; its procedures; and its

28

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

stage treatment is randomly assigned among atypical

treatment arms, the study design will support treatment
comparisons based on this rerandomization. However,
because patients and investigators will choose between
two different trials in phase 2, data from these two trials
will not be combined for statistical comparisons. Within
each trial, analyses will be conducted with respect to
selected outcomes in order to compare the performance of
the different treatment groups as second lines of treatment. In particular, an analysis of time from re-randomization to all-cause treatment discontinuation will be conducted using survival analysis techniques as described
above. The primary analysis will focus on two main
hypotheses:

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

The NIMH-CATIE Project

proposed method of evaluating the effectiveness of

antipsychotic medications in the broad population of
patients with schizophrenia who are treated clinically with
them. The protocol represents a type of pragmatic trial that
is intended to evaluate the effectiveness of treatments in
the settings where they are used for the purpose of generating results that benefit clinicians, administrators, policy
makers, and patients. In the context of clinical therapeutics
and interventions research, such studies provide an important bridge for the verification and translation of findings
from traditional randomized controlled trials to clinical
practice and public mental health care.

Pharmacoeconomics in Clinical Trials. 2nd ed. Philadelphia, PA: Lippincott-Raven, 1996. pp. 239-252.
First, M.; Spitzer, R.; Gibbon, M.; and Williams, J. Structured Clinical Interview for DSM-FV Axis I Disorders
Administration Booklet. Washington, DC: American Psychiatric Press Inc., 1994.
Gold, J.M.; Carpenter, C; Randolph, C ; Goldberg, T.E.;
and Weinberger, D.R. Auditory working memory and Wisconsin Card Sorting Test performance in schizophrenia.
Archives of General Psychiatry, 54:159-165, 1997.

References

Heaton, R.K.; Chelune, GJ.; Taley, J.L.; Kay, G.G.; and

Curtiss, G. Wisconsin Card Sorting Test Manual: Revised
and Expanded. Odessa, FL: Psychological Assessment
Resources, 1993.

Addington, D.; Addington, J.; Schissel, B. A depression

rating scale for schizophrenics. Schizophrenia Research,
3(4):247-251, 1990.

Heinrichs, D.W.; Hanlon, T.E.; and Carpenter, W.T. The

Quality of Life Scale: An instrument for rating the schizophrenic deficit syndrome. Schizophrenia
Bulletin,
10:388-398, 1984.

Appelbaum, P.S., and Grisso, T. MacArthur Competence

Assessment ToolClinical Research Version. Worcester,
MA: University of Massachusetts, 1996.
Awad, A.D. Subjective response to neuroleptics in schizophrenia. Schizophrenia Bulletin, 19(3):609-618, 1993.
Barnes, T.E. A rating scale for drug-induced akathisia.
British Journal of Psychiatry, 154:672-676, 1989.

Hershey, T.; Selke, G.; Fucetola, R.; and Newcomer, J.W.

Spatial long-term memory but not working memory
decreases over time in schizophrenia. [Abstract]. Society
for Neuroscience Abstracts, 25(Partl):572, 1999.

Benton, A.L., and Hamscher, K. Multilingual Aphasia

Examination Manual (revised). Iowa City, LA: University
of Iowa, 1978.

Hotopf, M.; Churchill, R.; and Lewis, G. Pragmatic randomised controlled trials in psychiatry. British Journal of
Psychiatry, 175:217-223,1999.

Brandt, J. The Hopkins Verbal Learning Test: Development of a new verbal memory test with six equivalent
forms. 77K- Clinical Neuropsychologist, 5:125-142, 1991.

Kalbfleisch, J.D., and Prentice, R.L. The Statistical Analysis of Failure Time Data. New York, NY: John Wiley and
Sons, 1980.

Carpenter, W.T., Jr.; Gold, J.M.; Lahti, A.C.; Queern, C.A.;

Conley, R.R.; Bartko, J.J.; Kovnick, J.; and Appelbaum,
P.S. Decisional capacity for informed consent in schizo-

Kay, S.R.; Fiszbein, A.; and Opler, L.A. The positive and
negative syndrome scale (PANSS) for schizophrenia.
Schizophrenia Bulletin, 13(2):261-276, 1987.

phrenia research. Archives of General Psychiatry,

57(6):533-538, 2000.

Kemp, R.; Hayward, P.; Applewhaite, G.; Everitt, B.; and

David, A. Compliance therapy in psychotic patients: Randomised controlled trial. British Medical Journal,
312:345-349, 1996.

Cornblatt, B.A.; Risch, NJ.; Fans, G.; Friedman, D.; and

Erlenmeyer-Kimling, L. The Continuous Performance
Test, Identical Pairs Version (CPT-IP): I. New findings
about sustained attention in normal families. Psychiatric
Research, 26(2):223-238, 1988.

Kerr, S.L., and Neale, J.M. Emotion perception in schizophrenia: Specific deficit or further evidence of generalized
poor performance? Journal of Abnormal Psychology,
102:312-318, 1993.

Cox, D.R. Regression models and life tables. Journal of

the Royal Statistical Society, B34:187-220, 1972.

Lafayette Instrument Company. Grooved

Davis, S.M.; Koch, G.G.; Davis, C.E.; and LaVange, L.M.

Statistical approaches to effectiveness measurement and
outcome-driven re-randomizations in the Clinical Antipsychotic Trials of Intervention Effectiveness (CAT1K) studies. Schizophrenia Bulletin, 29(l):73-80, 2003.

Pegboard

Instruction Manual, Model 32025. Lafayette, IN:

Lafayette Instrument Company, 1989.
Lebowitz, B.D.; Vitiello, B.; and Norquist, G.S.
Approaches to multisite clinical trials: The National Institute of Mental Health perspective. Schizophrenia Bulletin,
29(1):7-13, 2003.

Feeny, D.H.; Torrance, G.E.W.; and Furlong, W.J. Health

Utilities Index. In: Spiker, B., ed. Quality of Life and

29

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Guy, W. ECDEU Assessment Manual for PsychopharmacologyRevised. Rockville, MD: Department of Health,
Education, and Welfare, 1976.

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

T.S. Stroup et al.

Lehman, A.F. The quality of life interview for the severely

mentally ill. Evaluation and Program Planning, 11:51-62,
1988.
McEvoy, J.P.; Apperson, LJ.; Applebaum, P.S.; Ortlip, P.;
Brecosky, J.; HammilL K.; Geller, J.L.; and Roth, L. Insight
in schizophrenia: Its relationship to acute psychopathology.
Journal of Nervous and Mental Disease, 177:43-41,1989.

chiatric facilities and others in the same neighborhoods.

Archives of General Psychiatry, 55:393-401,1998.

Mojtabai, R.; Nicholson, R.A.; and Carpenter, P.N. Role of

psychosocial treatments in management of schizophrenia:
A meta-analytic review of controlled outcome studies.
Schizophrenia Bulletin, 24(4):569-587, 1998.

Tessler, R., and Gamache, G. Family Experiences Interview Schedule Toolkit. Cambridge, MA: Human Services
Research Institute, 1995.
Van Hout, B.A.; Al, MJ.; Gordon, G.S.; and Rutten, F.T.
Costs, effects, and C/E ratios alongside a clinical trial.
Health Economics, 3:309-319, 1994.
Ware, J. Jr.; Kosinski, M.; and Keller, S.D. A 12-item
short-form health survey: Construction of scales and preliminary tests of reliability and validity. Medical Care,
34<3):220-233, 1996.

Swartz, M.S.; Perkins, D.O.; Stroup, T.S.; McEvoy, J.P.;

Nieri, J.M.; and Haak, D.C. Assessing clinical and functional outcomes in the Clinical Antipsychotic Trials of
Intervention Effectiveness (CAITE) schizophrenia trial.
Schizophrenia Bulletin, 29(1):XXX-XXX, 2003.

Rosenheck, R.; Cramer, J.; Allan, E.; Erdos, J.; Frisman,

L.K.; Xu, W.; Thomas, J.; Henderson, W.; and Charney, D.
Cost-effectiveness of clozapine in patients with high and
low levels of hospital use. Archives of General Psychiatry,
56(6):565-572, 1999.

Wechsler, D. Wechsler Adult Intelligence ScaleRevised.

San Antonio, TX: Psychological Corporation, 1974.
Wechsler, D. Wechsler Intelligence Scale for Children. 3rd
ed. San Antonio, TX: Psychological Corporation, 1991.

Rosenheck, R.A.; Cramer, J.; Xu, W.; Grabowski, J.;

Douyon, R.; Thomas, J.; Henderson, W.; and Charney, D.
Multiple outcome assessment in a study of the cost-effectiveness of clozapine in the treatment of refractory schizophrenia. Health Services Research, 33:1235-1259, 1998.

Wilkinson, G.S. WRAT-3 Administration Manual. Wilmington, DE: Wide Range, 1993.

Acknowledgments

Rosenheck, R.A.; Cramer, J.; Xu, W.; Thomas, J.; Henderson, W.; Frisman, L.K.; Fye, C ; and Charney, D., for the
Department of Veterans Affairs Cooperative Study Group
on Clozapine in Refractory Schizophrenia. A comparison
of clozapine and haloperidol in the treatment of hospitalized patients with refractory schizophrenia. New England
Journal of Medicine, 337:809-815, 1997.

This article was based on results from the Clinical

Antipsychotic Trials of Intervention Effectiveness
(CATTE) project supported with Federal funds from the
National Institute of Mental Health (NIMH) under contract NO1 MH90001. The aim of this project is to examine
the comparative effectiveness of antipsychotic drugs in
conditions for which their use is clinically indicated,
including schizophrenia and Alzheimer's disease. The
project was carried out by principal investigators from the
University of North Carolina, Duke University, the University of Southern California, the University of
Rochester, and Yale University in association with Quintiles, Inc., and the program staff of the Division of Interventions and Services Research of the NIMH and investigators from 84 sites in the United States. AstraZeneca
Pharmaceuticals LP, Bristol-Myers Squibb Company, Forest Pharmaceuticals, Inc., Janssen Pharmaceutica Products, L.P., Eli Lilly and Company, Otsuka Pharmaceutical
Co., Ltd., Pfizer Inc., and Zenith Goldline Pharmaceuticals, Inc., provided medications for the studies.

Schneider, L.S.; Ismail, M.S.; Dagerman, K.; Davis, S.;

Olin, J.; McManus, D.; Pfeiffer, E.; Ryan, J.M.; Sultzer,
D.L.; and Tariot, P.N. Clinical Antipsychotic Trials of
Intervention Effectiveness (CAllh): Alzheimer's disease
trial. Schizophrenia Bulletin, 29(l):57-72, 2003.
Schneider, L.S.; Tariot, P.N.; Lyketsos, C.G.; Dagerman,
K.S.; Davis, K.L.; Davis, S.; Hsiao, J.K.; Jeste, D.V.; Katz,
I.R.; Olin, J.T.; Pollock, B.G.; Rabins, P.V.; Rosenheck,
R.A.; Small, G.W.; Lebowitz, B.; and Lieberman, J.A.
National Institute of Mental Health Clinical Antipsychotic
Trials of Intervention Effectiveness (CATTE): Alzheimer
disease trial methodology. American Journal of Geriatric
Psychiatry, 9(4):346-360, 2001.
Simpson, G.M., and Angus, J.W. A rating scale for
extrapyramidal side effects. Acta Psychiatrica Scandinavica, 212(Supplement): 11-19, 1970.

The Authors

Steadman, H.J.; Mulvey, E.P.; Monahan, J.; Robbins, P C ;

Appelbaum, P.S.; Grisso, T.; Roth, L.H.; and Silver, E.
Violence by people discharged from acute inpatient psy-

T. Scott Stroup, M.D., M.P.H., is Associate Professor,

Department of Psychiatry, University of North Carolina

30

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

Mueser, K.T.; Drake, R.C.; Clark, R.E.; McHugo, GJ.; Mercer-McFadden, C ; and Ackerson, T.H. Toolkit on Evaluating
Substance Abuse in Persons With Severe Mental Illness.
Cambridge, MA: Human Services Research Institute, 1995.

The NIMH-CATIE Project

Schizophrenia Bulletin, Vol. 29, No. 1, 2003

Director, Psychopharmacology Research Unit, Valley

Mental Health, Salt Lake City, UT. Jeffrey A. Lieberman,
M.D., is Thad and Alice Eure Professor of Psychiatry,
Pharmacology, and Radiology and Vice-Chairman of
Psychiatry, University of North Carolina School of Medicine, Chapel Hill, NC.
Drs. Stroup and McEvoy are Co-Principal Investigators (Pis) of the CATIE schizophrenia trial. Dr. Lieberman
is PI of the CATTE project and of the schizophrenia trial.
Dr. Swartz is Co-PI of the CATTE Clinical and Functional
Assessment Unit. Drs. Byerly, Glick, Canive, McGee,
Simpson, and Stevens are the site Pis who have recruited
the largest number of study subjects at the time of publication and have made significant contributions to the study.
For the Schizophrenia Trial Center Members: C.E.
Davis, Ph.D.; Kenneth Davis, M.D.; Sonia Davis, DrPH;
Susan Essock, Ph.D.; Don Goff, M.D.; John Hsiao, M.D.;
Richard Keefe, Ph.D.; Gary Koch, Ph.D.; Barry Lebowitz,
Ph.D.; Jeffrey Lieberman, M.D.; Steve Marder, M.D.;
Joseph McEvoy, M.D.; Herbert Meltzer, M.D.; Alec
Miller, M.D.; Del Miller, M.D.; Diana Perkins, M.D.,
MPH; Robert Rosenheck, M.D.; Nina Schooler, Ph.D.;
Marvin Swartz, M.D.

31

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

School of Medicine, Chapel Hill, NC. Joseph P. McEvoy,

M.D., is Associate Professor, Department of Psychiatry,
Duke University, Durham, NC; and Deputy Clinical
Director, John Umstead Hospital, Butner, NC. Marvin S.
Swartz, M.D., is Professor, Department of Psychiatry and
Behavioral Science, Duke University, Durham, NC.
Matthew Byerly, M.D., is Assistant Professor of Psychiatry, The University of Texas Southwestern Medical Center at Dallas, TX. Ira Glick, M.D., is Professor, Department of Psychiatry and Behavioral Science, Stanford
University School of Medicine, Stanford, CA. Jose
Canive, M.D., is Director, Psychiatry Research, New
Mexico VA Health Care System, and Associate Professor
of Psychiatry, University of New Mexico, Albuquerque,
NM. Mark McGee, M.D., is Chief Clinical Officer of
Appalachian Behavioral Healthcare, and Assistant Professor of Psychiatry, Department of Specialty Medicine,
Ohio University College of Osteopathic Medicine,
Athens, OH. George Simpson, M.D., is Professor of
Research Psychiatry and Interim Chair, Department of
Psychiatry, University of Southern California Keck
School of Medicine, Los Angeles, CA. Michael Stevens,
M.D., is Medical Director, Valley Mental Health, and

The 12th Association of European Psychiatrists Congress will take place in Geneva, Switzerland,
14-18 April 2004.
Deadline for Receipt of Abstracts: 30 October 2003
For further information, please contact
12th EAP Congress
Kenes International
Global Congress Organisers and Association Management Services
17 Rue du Cendrier, P.O. Box 1726
CH-1211 Geneva 1, Switzerland
Tel: 441 22 908 0488; Fax: 441 22 732 2850
e-mail: [email protected]

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ by guest on February 8, 2015

12th AEP Congress