Volume 100, Issue 1, Pages 39-52 (March 2008)

5 of 42

ABSTRACT

FULL TEXT

FULL-TEXT PDF (801 KB)

The effectiveness of antipsychotic medications in patients who use or avoid illicit substances: Results from the CATIE study

Received 22 October 2007; received in revised form 15 November 2007; accepted 20 November 2007. published online 15 January 2008.

Abstract

Objective

This double-blind study compared a second generation (atypical) antipsychotic drugs compared to a representative older agent for patients with schizophrenia who use or avoid illicit substances.

Methods

Schizophrenic subjects were recruited at 57 U.S. sites and randomly assigned to olanzapine, perphenazine, quetiapine, risperidone or ziprasidone for up to 18 months. The primary aim of this analysis was to delineate differences between the overall effectiveness of these five treatments among patients who used or did not use illicit substances.

Results

There were no significant differences between treatment groups in time to all-cause treatment discontinuation among patients who use illicit drugs (median 3.3 to 6.8 months). Among non-users time to treatment discontinuation was significantly longer for patients treated with olanzapine (median 13.0 months) than perphenazine ( 5.9 months), risperidone (5.6 months), or quetiapine (5.0 months); time to discontinuation for ziprasidone (4.3 months) was even shorter, although the latter difference was not significant. The difference between risperidone and quetiapine, although small, was significant. All remaining differences were non-significant. Similar results were found for discontinuation due to inefficacy. There were no differences between illicit users and non-users in symptom reduction and global improvement, after adjustment for differential duration of treatment. Differences in discontinuation results were attenuated by non-compliance, but the trends persisted after controlling for treatment compliance.

Conclusions

Among patients with chronic schizophrenia who avoid use of illicit drugs, olanzapine was more effective than other antipsychotics as reflected by longer time to all-cause discontinuation, but illicit substance abuse attenuated this advantage, reinforcing the need for concurrent substance abuse treatment.

Keywords: Schizophrenia, Antipsychotics, Substance abuse, Dual disorders

Article Outline

• Abstract

• 1. Introduction

• 2. Experimental design and methods

• 2.1. Study design and measures

• 2.2. Participants

• 2.3. Interventions

• 2.4. Objectives and outcomes

• 2.5. Statistical methods

• 3. Results

• 3.1. Patient characteristics and dispositions

• 3.2. Treatment discontinuation

• 3.3. PANSS and CGI efficacy measures

• 3.4. Adverse events

• 3.5. Potential mediating factors

• 4. Discussion

• Role of funding source

• Contributors

• Conflict of interest

• Acknowledgment

• References

• Copyright

1. Introduction

Substance abuse co-morbidity in schizophrenia is a powerful risk factor for a wide range of poor outcomes (Abram and Teplin, 1991, Cournos et al., 1991, Goodman et al., 2001, Haywood et al., 1995, Rosenberg et al., 2001, Swanson et al., 1997, Swartz et al., 1998, Swofford et al., 1996). Estimates of the prevalence of substance use co-morbidity range from a low of 10%–70% depending on diagnostic assessment methods (Blanchard et al., 2000, Cantor-Graae et al., 2001, Dixon, 1999, Goswami et al., 2004, Kavanagh et al., 2002, Kessler et al., 1997, McCreadie, 2002, Mueser et al., 1998, Regier et al., 1990, Salyers and Mueser, 2001, Swartz et al., 2006a, Swartz et al., 2006b). Clinicians often assume that any level of substance abuse is likely to lead to adverse outcomes (Mueser et al., 1998), but receive little guidance about preferred antipsychotic treatment in this population. Several small-scale studies suggest second generation antipsychotics may reduce illicit substance use (Buckley, 1999, Drake et al., 2000, Green et al., 1999, Green, 2005, Littrell et al., 2001, Petrakis et al., 2006, Smelson et al., 2002, Tsuang et al., 2002), but there are no large representative community-based trials comparing treatment outcomes among users and non-users.

This study employs data from the NIMH Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) project to compare the effects of olanzapine, perphenazine, quetiapine, risperidone and ziprasidone among users and non-users on time to treatment discontinuation (Davis et al., 2003, Keefe et al., 2003, Keefe et al., 2006, Keefe et al., 2007, Lieberman et al., 2005, McEvoy et al., 2006, Rosenheck et al., 2006, Stroup et al., 2003, Stroup et al., 2006, Swartz et al., 2003). Reports on the outcomes from CATIE to date, have found that patients in the initial randomization (Phase 1) on olanzapine had longer time to treatment discontinuation than some of the other drugs, but had greater side effect burden related to weight gain and related metabolic sequelae (Lieberman et al., 2005). None of the other three second generation drugs showed greater efficacy or tolerability than the first generation drug, perphenazine. All of the antipsychotic treatment groups had small but significant improvements in neurocognition and psychosocial functioning, but there was no difference among them (Keefe et al., 2007, Swartz et al., 2007).

The primary objective of the current study was to examine the effect of illicit substance use on time to treatment discontinuation across treatment groups. Secondary objectives included examining subcategories of treatment discontinuation and psychopathology measures.

2. Experimental design and methods

2.1. Study design and measures

The CATIE study was initiated by the National Institute of Mental Health (NIMH) to determine the comparative effectiveness of antipsychotic drugs. Its rationale, design and methods were previously described (Davis et al., 2003, Keefe et al., 2003, Stroup et al., 2003, Swartz et al., 2003) and treatment effects on discontinuation rates and symptoms have been reported (Lieberman et al., 2005, McEvoy et al., 2006, Stroup et al., 2006). The study was conducted between 1/01 and 12/04 at 57 U.S. clinical sites (16 university clinics, 10 state mental health agencies, 7 Veteran's Affairs Medical Centers, 6 private non-profit agencies, 4 private practice sites, and 14 mixed system sites). Patients were randomized to receive olanzapine, perphenazine, quetiapine, risperidone, or ziprasidone under double-blind conditions and followed for up to 18 months or until treatment was discontinued for any reason (Phase I). Patients whose assigned treatment was discontinued could receive other treatments in Phases Ib, II and III (McEvoy et al., 2006, Stroup et al., 2003, Stroup et al., 2006).

2.2. Participants

Eligible patients were 18 to 65 years of age; had received a diagnosis of schizophrenia, as determined on the basis of the Structured Clinical Interview of the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (First et al., 1996); and were able to take oral antipsychotic medication as determined by the study doctor. Patients were excluded if they had diagnosis of schizoaffective disorder, mental retardation or other cognitive disorders; had a history of serious adverse reactions to the proposed treatments; had had only one schizophrenic episode; had a history of treatment resistance, defined by persistence of severe symptoms despite adequate trials of one of the proposed treatments or prior treatment with clozapine; were pregnant or were breast-feeding; or had a serious and unstable medical condition.

The study was approved by the institutional review board at each site, and written informed consent was obtained from the patients or their legal guardians.

2.3. Interventions

Identical-appearing capsules contained olanzapine (Zyprexa, Eli Lilly) (7.5 mg), quetiapine (Seroquel, AstraZeneca) (200 mg), risperidone (Risperdal, Janssen Pharmaceutica) (1.5 mg), or perphenazine (Trilafon, Schering-Plough) (8 mg) or (after January 2002) ziprasidone (Geodon, Pfizer) (40 mg). Patients with current tardive dyskinesia (TD) could enroll but a stratified randomization scheme prevented their assignment to treatment with perphenazine. The dose of the medications was flexible, ranging from one to four capsules daily, based upon the study doctor's judgment. Concomitant medications were permitted throughout the trial, except for additional antipsychotic agents. Patients had monthly visits with study doctors.

2.4. Objectives and outcomes

Preliminary inspection of time to discontinuation for any reason indicated that the time to discontinuation was foreshortened among patients using illicit substances and that the previously reported advantage of olanzapine on this measure (Lieberman et al., 2005) was attenuated among patients using illicit substances. In addition, relatively few patients were alcohol abusers only without using illicit substances. Patients who used or abused alcohol in many ways were more similar to non-substance users. On that basis, we sought to examine the effects of illicit substance use, excluding those solely using or abusing alcohol, on time to discontinuation across treatment groups. Secondary outcomes included three subcategories of time to discontinuation: discontinuation for inadequate therapeutic effects, intolerable side effects, or patient decision, as well as psychopathology outcomes assessed on the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987), the clinical global impression of severity (CGI-S) (Guy, 1976) and episodes of hospitalization.

Illicit drug use (including use or abuse/dependence) was assessed using multiple sources of information collected at the baseline visit and coded into two categories: 1) non-users — subjects with no evidence of any current illicit drug use and 2) current illicit drug use (with or without serious impairment). A small group of patients who used only alcohol with no evidence of illicit substance use were included as illicit substance non-users. Several measures were combined to make these determinations: diagnostic assessment using the SCID (First et al., 1996); positive hair assay (RIAH) or positive urine screen for illicit drug use, self-reported drug use, clinician ratings of use on the initial screening inventory, clinician rating from the Alcohol or Drug Use Scale (Drake et al., 1990), and family report of a problem with alcohol or illicit drugs (Swartz et al., 2006a, Swartz et al., 2006b). The most common illicit drug used was marijuana, followed by cocaine and opiates. Polysubstance use was common. Additional measures include depressive symptoms (Addington et al., 1992) and insight or awareness of illness (McEvoy et al., 1989). Compliance with medication was computed by pill as the percent of the prescribed study drug taken at each visit; missing counts were imputed by averaging the lowest previous ratings.

2.5. Statistical methods

The study sample included the intent-to-treat population of randomized patients who received at least one dose of study medication (N=1432). Analyses were conducted as previously described (Lieberman et al., 2005). Median time to discontinuation was estimated using Kaplan–Meier survival curves: Treatment group comparisons for time to discontinuation were carried out using Cox regression adjusting for baseline clinical exacerbation status, TD (where applicable) and treatment site. Sites with fewer than 15 respondents were combined according to the type health care system they represented. Due to the stratified study design, treatments were compared in a four-tiered approach, as follows: Two hundred thirty-one patients with tardive dyskinesia were excluded from random assignment to perphenzine, and ziprasidone was added after approximately 40% of the sample had been enrolled. Therefore, comparisons involving perphenazine were restricted to patients without TD, while ziprasidone comparisons were restricted to subjects enrolled after the addition of ziprasidone to the study. In the primary analysis step, comparisons among olanzapine, quetiapine, risperidone, and perphenazine were evaluated for non-TD patients using a 3 degree of freedom (df) test. Following a significant result, pair-wise comparisons among the three atypicals in an analysis which included TD patients were evaluated using closed (step-down) testing requiring a significance level of 0.05 or less. Pair-wise comparisons of each of the three atypicals against perphenzine were evaluated using a Hochberg adjustment for multiple comparisons. Ziprasidone was separately compared with the three atypicals and with perphenzine (with TD patients excluded) in a series of pair-wise comparisons, again using Hochberg adjustments (Hochberg, 1988) in analyses of patients enrolled after ziprasidone was added. To further explore the results of the primary analyses, a number of non-tiered exploratory analyses were restricted to bivariate comparisons contrasting olanzapine and the four remaining treatment groups controlling for a diagnosis of tardive dyskinesia (TD) and post-ziprasidone randomization.

The effect of illicit substance use was tested for its association with time to discontinuation as a main effect, and for its interaction with the treatment groups. For the latter we investigated possible interactions between illicit substance use and the antipsychotic treatment groups using both a standard interaction model crossing treatment group with illicit status and a stratified approach based on two subgroups divided on the basis of use or non-use of illicit substances. The stratification approach avoided difficult to interpret multivariable interaction models.

Change from baseline in PANSS and CGI status at end of phase visit were analyzed using analysis of covariance (ANCOVA) procedure using cross-sectional regression analyses for all intent-to-treat patients controlling for time-at-risk (duration of phase 1 treatment), baseline history of clinical exacerbation in the previous 3 months, baseline level of the outcome measure, a diagnosis of TD, and post-ziprasidone randomization. An interaction term crossing treatment group (olanzapine versus other antipsychotic medications) with use of illicit substances was estimated to test for a possible differential effect of illicit substance use on clinical improvement among subjects randomized to olanzapine.

3. Results

3.1. Patient characteristics and dispositions

For the current study 1432 patients were available for analysis. The baseline sociodemographic and clinical characteristics of the illicit drug user (N=643) and non-user (N=789) samples in are shown in Table 1 and demonstrate significant differences on a number of demographic and clinical factors. Patients using illicit drugs were younger, more likely to be male and non-white (primarily African-American), had lower levels of education, and were more than twice as likely to have experienced a recent episode of homelessness. Among the illicit user group there were a few differences between illicit users and abusers: abusers were somewhat younger, more likely to be male, had slightly higher PANSS positive symptom scores, but there were no significant differences in compliance ratings or racial composition.

Table 1.

Sample characteristics for patients with and without illicit substance use


	Statistic	No illicit substance use	Illicit substance use	Probability
	Statistic	(n=789)	(n=643)	Probability
Demographics
Age (years)	Mean±SD	42.6±10.9	38.1±10.8	<0.0001
Sex:				<0.0001
Male	n (%)	541 (69)	521 (81)
Female	n (%)	248 (31)	122 (19)
Race:				<0.0001
White	n (%)	451 (57)	271 (42)
Non-white	n (%)	338 (43)	372 (58)
Hispanic	n (%)	103 (13)	62 (10)	0.0443
Married or living with other	n (%)	156 (20)	118 (18)
Education (years)	Mean±SD)	12.3±2.4	11.8±2.0	0.0001
Homeless — yes	n (%)	16 (2)	35 (5)	0.0005

Clinical history
Years treatment	Median (25%–75%)	17 (8–26)	14 (5–23)	<0.0001
Years since onset	Median (25%–75%)	14 (6–24)	11 (4–22)	0.0003
Childhood conduct symptoms	Median (25%–75%)	0 (0–1)	1 (0–3)	<0.0001

Other diagnosis: past 5 years
Major depression	n (%)	191 (24)	206 (32)	0.001
Alcohol dependence or abuse	n (%)	82 (10)	268 (42)	<0.0001
Drug dependence or abuse	n (%)	0 (0)	409 (64)	<0.0001
Any drug/alcohol diagnosis	n (%)	82 (10)	447 (70)	<0.0001

Clinical status:
CGI severity	Mean±SD	3.9±1.0	4.0±1.0	0.0297
Calgary depression	Median (25%–75%)	3 (1–7)	4 (1–8)	0.0011
PANSS:
Total	Mean±SD	75.0±17.6	76.4±17.6
General psychopathology	Mean±SD	36.7±9.2	37.4±9.4
Negative symptoms	Mean±SD	20.4±6.4	19.9±6.5
Positive symptoms	Mean±SD	18.0±5.6	19.1±5.6	0.0002
Recent exacerbation <3 months:	n (%)	181 (23)	208 (32)	<0.0001
Recent victimization:	n (%)	105 (13)	166 (26)	<0.0001
Insight into illness (ITAQ)	Median (25%–75%)	21 (15–22)	21 (15–22)
Compliance rating (pill count)	Median (25%–75%)	94.5 (80.8–98.8)	89.3 (69.1–97.3)	<0.0001

Study status
Completed 18 mo in Phase 1	n (%)	231 (29)	140 (22)	0.0013

Clinical histories reflected significantly less time in treatment among substance-using subjects, consistent with lower mean age, significantly more childhood antisocial symptoms, and slightly higher odds of having been diagnosed with major depression within the prior 5 years.

Baseline clinical status also indicated significantly more depressive symptoms as measured by the Calgary Depression Rating Scale (Addington et al., 1992) and a significantly greater proportion had a recent clinical exacerbation. The two groups did not differ on the PANSS total scale score, general psychopathology scale, or negative symptom scale, but substance-using patients did exhibit significantly higher PANSS positive symptoms scores.

3.2. Treatment discontinuation

When examined as a main effect in Cox regression models, illicit substance use was not a significant predictor of time to discontinuation (χ21=0.578; p=0.447), however, when an interaction term was added for treatment group and illicit use, they jointly predicted time to discontinuation (χ23=10.762; p=0.013). Visual inspection of median time to discontinuation suggested a stratified approach comparing illicit substance users versus non-users across treatment groups would be more interpretable and informative.

Fig. 1 depicts Kaplan–Meier survival curves for treatment groups stratified by illicit substance non-use versus use and demonstrates differences in time to discontinuation among non-users but little differences among non-users. As seen in Table 2A, Table 2B from the stratified analyses, among non-users discontinuation rate was lower and time to all-cause discontinuation was significantly longer for olanzapine (discontinuation rate 56%, median time 13.02 months) compared to quetiapine (discontinuation rate 81%, median time 5.02 months), risperidone (discontinuation rate 69%, median time 5.57 months), perphenazine (discontinuation rate 74%, median time 5.89 months) but not ziprasidone (discontinuation rate 77%, median time 4.34 months). Among non-users there were significant differences in odds of discontinuation between olanzapine and quetiapine (hazard ratio (HR=0.52, CI 0.40, 0.67, p<0.001), risperidone (HR=0.70, CI 0.53, 0.92, p=0.01) and perphenazine (HR=0.59, CI 0.56, 1.08, p=<0.001), but not ziprasidone (HR=0.78, CI 0.56, 1.08, p=0.13). Odds of discontinuation were also significantly higher for subjects randomized to quetiapine relative to risperidone (HR=1.35; CI 1.05, 1.73, p=0.021). Rates of medication compliance were higher for non-users, but were not different across treatment groups.

View Large Image
Download to PowerPoint

Fig. 1. Kaplan–Meier plots of time to treatment discontinuation for treatment.

Table 2A.

Time to all-cause treatment discontinuation


Condition: no illicit substance use
Assessment	Statistic	OLAN	QUET	RISP	PERF	ZPR
Assessment	Statistic	(n=188)	(n=192)	(n=176)	(n=133)	(n=100)
Modal dose (mg)/no. patients	Mean	20.2/183	562.8/183	4.0/165	21.2/128	112.4/90
Number discontinuing	n(%)	105 (56)	156 (81)	121 (69)	99 (74)	77 (77)
Kaplan–Meier	Median	13.02	5.02	5.57	5.89	4.34
Time to discontinuation (month)	[(%95%CI]	[8.14,a]	[3.79, 6.21]	[3.93, 7.57]	[3.32, 6.50]	[3.11, 7.04]
Compliance percent (SE)		87.3(21.0)	82.0(27.5)	81.1(28.1)	83.0(26.7)	79.1(29.9)

Condition: illicit substance use
Assessment	Statistic	OLAN	QUET	RISP	PERF	ZPR
Assessment	Statistic	(n=142)	(n=137)	(n=157)	(n=124)	(n=83)
Modal dose (mg)	Mean	20.0/129	515.1/126	3.8/140	20.4/117	113.3/75
Number discontinuing	n(%)	105 (74)	113 (82)	124 (79)	93 (75)	68 (82)
Kaplan–Meier	Median	6.75	4.36	4.61	5.25	3.29
Time to discontinuation (month)	[(%95%CI]	[4.57, 9.14]	[3.54, 5.54]	[3.36, 6.36]	[4.00, 6.50]	[2.75, 5.36]
Compliance percent (SE)		78.7(28.7)	71.9(33.3)	75.9(29.2)	78.1(28.0)	75.7(31.4)

Upper bound confidence interval not estimable.

Table 2B.

Time to all-cause treatment discontinuation


Condition: no illicit substance use
Assessment	Statistic	OLAN	QUET	RISP	PERF	ZPR	p-valuea
Assessment	Statistic	(n=188)	(n=192)	(n=176)	(n=133)	(n=100)	p-valuea
Cox models:	HRb		0.52	0.7	0.59	0.78	<0.001
OLAN	[(%95%CI]		[0.40, 0.67]	[0.53, 0.92]	[0.43, 0.81]	[0.56, 1.08]
	p-value		<0.001	0.010	0.001	0.134
	HR			1.35	1.16	1.12
QUET	[(%95%CI]			[1.05, 1.73]	[0.88, 1.54]	[0.82, 1.53]
	p-value			0.021	0.288	0.464
	HR				0.87	1.02
RISP	[(%95%CI]				[0.64, 1.17]	[0.75, 1.40]
	p-value				0.354	0.877
	HR					1.05
PERF	[(%95%CI]					[0.74, 1.49]
PERF	p-value					0.79

Condition: illicit substance use
Assessment	Statistic	OLAN	QUET	RISP	PERF	ZPR	p-valuea
Assessment	Statistic	(n=142)	(n=137)	(n=157)	(n=124)	(n=83)	p-valuea
Cox models:	HR		0.9	0.93	1.16	0.75	0.827
OLAN	[(%95%CI]		[0.67, 0.1.20]	[0.70, 1.24]	[0.75, 1.60]	[0.53, 1.07]
	p-value		0.471	0.627	0.357	0.11
	HR			1.04	1.1	0.89
QUET	[(%95%CI]			[0.79, 1.37]	[0.80, 1.51]	[0.63, 1.24]
	p-value			0.793	0.55	0.486
	HR				1.11	0.76
RISP	[(%95%CI]				[0.81, 1.52]	[0.0.53, 1.09]
	p-value				0.513	0.132
	HR					0.76
PERF	[(%95%CI]					[0.53, 1.10]
PERF	p-value					0.15

p-value column is the overall 3 degree of freedom test comparing olanzapine, quetiapine, risperidone and perphenazine for patients without TD.

Hazard ratios are obtained from analysis of four cohorts defined by the stratified randomization for TD patients and the addition of ziprasidone.

However, among illicit substance users, discontinuation rate and time to all-cause discontinuation were not different for olanzapine (discontinuation rate 74%, median time 6.75 months) compared to quetiapine (discontinuation rate 82% median time 4.36 months), risperidone (discontinuation rate 79% median time 4.61 months), perphenazine (discontinuation rate 75%, median time 5.25 months), and ziprasidone (discontinuation rate 82%, median time 3.29 months). Hazard rates for olanzapine relative to the other four antipsychotics drugs were: quetiapine (HR=0.90, CI 0.67, 1.20, p=0.47); risperidone (HR=0.93, CI 0.70, 1.24, p=0.63); perphenazine (HR=1.16, CI 0.84, 1.60, p=0.36); and ziprasidone (HR=0.75, CI 0.53, 1.07, p=0.11). The contrast between quetiapine and risperidone also was not significant (HR=1.04, CI 0.78, 1.37, p=0.79).

In order to descriptively evaluate the odds of discontinuation due to inefficacy, tolerability or patient decision, odds of discontinuation for the olanzapine treatment group versus the remaining combined treatment groups were evaluated. Among non-users of illicit substances treated with olanzapine, hazard ratios were significantly reduced for discontinuation due to inefficacy (HR 0.34, CI 0.22, 0.53, p<0.0001), and patient decision (HR 0.54, CI 0.37, 0.78, p<0.001), but not for intolerability (HR 1.17, CI 0.79, 1.73, p=0.439). Among users of illicit substances, the hazard ratio for discontinuation due to inefficacy, although almost twice the rate for non-users, was nonetheless, significantly reduced (HR 0.60, CI 0.39, 0.94, p<0.0252). Hazard ratios for discontinuation due to intolerability and patient decision did not differ significantly by treatment group among users.

Differences in discontinuation rates for olanzapine versus other treatment groups combined are contrasted graphically among substance non-users and users in Fig. 2. Lower all-cause discontinuation rates are evident for olanzapine in the non-user group. In addition it appears that patient decision to discontinue treatment is higher among all substance use treatment groups and may also account for much of the increase in discontinuation rates among olanzapine substance users compared to non-users.

View Large Image
Download to PowerPoint

Fig. 2. Discontinuation rates by cause of discontinuation.

3.3. PANSS and CGI efficacy measures

In order to evaluate whether differences in time to discontinuation were mirrored by PANSS and CGI efficacy measures, Table 3 presents estimated changes in PANSS (and its subscales) and CGI ratings from baseline to the end of Phase 1, adjusted to duration of Phase 1 treatment. Many studies consider a 20% reduction in PANSS total score to be clinically meaningful. The total PANSS score decreased modestly (approximately 7 units, or roughly 10%), from baseline to the end of Phase 1 for subjects randomized to olanzapine relative to the other treatment groups (generally a 0–3 unit decrease). Overall there were no differences between olanzapine substance users and non-users or between other treatment groups comparing substance users and non-user subgroups for duration-adjusted change from baseline. In particular, an ANCOVA test of interactions between treatment group and illicit substance use was not significant (data not shown). Similarly olanzapine illicit use and non-use treatment groups made greater improvements relative to other user and non-user treatment groups on PANSS subscales. CGI severity measurements decreased significantly from baseline to the end of Phase 1 among subjects randomized to olanzapine relative to other treatment groups, but as above, the decrease was not affected by illicit substance use (see Table 3).

Table 3.

Efficacy measures at end of Phase 1—LS mean estimates of last observation carried forward differences from baseline, adjusted for average treatment duration


	Olan	Quet	Risp	Perf	Zip
	Units (SE)	Units (SE)	Units (SE)	Units (SE)	Units (SE)
Positive and Negative Syndrome Scale (PANSS) Total Scale:a
No use	7.44(1.19)	0.12(1.32)	2.46(1.57)	3.32(1.62)	0.34(1.85)
Illicit substance use	7.56(1.53)	3.14(1.49)	3.03(1.51)	2.45(1.50)	1.07(1.95)

PANSS Negative Symptom Scale:
No use	−1.57(.45)	−0.41(.44)	−0.51(.49)	−2.27(.49)	−0.46(.62)
Illicit substance use	−1.76(.59)	−0.67(.53)	−0.17(.54)	−0.09(.49)	0.01(.57)

PANSS Positive Symptom Scale:
No use	−2.28(.38)	0.34(.43)	−0.96(.44)	−1.16(.55)	−0.06(.65)
Illicit substance use	−2.79(.43)	−0.70(.53)	−0.98(.51)	−0.88(.53)	−0.31(.72)

PANSS General Psychopathology Scale:
No use	−3.58(.63)	0.19(.70)	−0.99(.86)	−0.89(.88)	0.74(.91)
Illicit substance use	−3.02(.79)	−1.78(.79)	−1.88(.76)	−1.47(.81)	0.77(1.04)

Clinical Global Impression-Severity Scale:
No use	−0.26(.08)	0.16(.08)	−0.11(.09)	−0.16(.10)	0.01(.12)
Illicit substance use	−0.53(.09)	−0.12(.09)	−0.03(.09)	−0.10(.10)	−0.04(.13)

Final Phase 1 measurement less baseline measurement.

Together these findings suggest that psychopathology measures improved more for olanzapine treatment groups, but comparably for substance users and non-users, and that treatment discontinuation may have not been related to clinician-rated efficacy measures but rather factors such as patient decision to discontinue treatment.

3.4. Adverse events

In order to evaluate whether side effects and adverse events might account for differences in rates of treatment discontinuation among substance users and non-users within treatment groups, we examined these reports across treatment groups contrasting any moderate or severe adverse events during treatment (data not shown). We compared treatment groups according to the presence or absence of substance use, using Poisson regression procedures. Patients with illicit substance use with few exceptions reported more side effects, but these increases did not significantly differ across treatment groups. In addition, neurological symptoms were not significantly affected by either illicit use, treatment group or by their interaction.

Illicit substance use had no effect on several measures of weight gain, but as expected from previous reports olanzapine-treated subjects–whether using substances or not–had greater weight gain; including 28.8–31.5% of olanzapine-treated patients gaining greater than 7% of body weight or 6–7 lb during treatment (Lieberman et al., 2005, Stroup et al., 2006).

Two hundred thirty subjects experienced at least one psychiatric hospitalization during Phase 1. Illicit substance users were more likely to be hospitalized (19.3% versus 13.4%, χ2=8.99; p=0.003) differences did not vary significantly across treatment groups and, as previously reported, olanzapine illicit substance users and non-users reported fewer hospitalizations (11% versus 15–20%) (Lieberman et al., 2005).

3.5. Potential mediating factors

As previously seen in Table 1, several sociodemographic and clinical characteristics distinguished illicit substance users from non-users and these differences could account for differences in time to discontinuation among patients randomized to olanzapine or other treatment groups. There were no differences in the distribution of illicit substance users across treatment groups that might account for the larger differences in treatment discontinuation between olanzapine and other treatment groups across the substance use and non-use groups. To test the possibility that other sociodemographic or clinical factors might account for differences in discontinuation rates across illicit substance user and non-user groups, a series of multivariable interaction models were estimated to evaluate possible confounding and/or mediating effects of these sociodemographic or clinical factors. Time to discontinuation for all patients was regressed using Cox regression models examining the interaction between treatment groups and illicit substance use; controlling for clinical exacerbation and TD at baseline and ziprasidone cohort. Subjects randomized to olanzapine and using illicit substances were 1.40 times more likely to discontinue (HR=1.400; 95%CI: 1.054–1.859; χ12=5.0402; p=0.0201) relative to olanzapine-treated non-users. Each potential sociodemographic or clinical mediator was then tested for its putative mediating effect on the olanzapine/illicit substance use interaction. Attenuation of the significance of the olanzapine/substance use interaction term – a possible indicator of mediation – occurred only after the addition medication compliance ratings (HR=1.154; 95% CI: 0.86–1.55; χ12=0.9064; p=0.3411). The interaction was not substantively affected by the addition of any other covariate.

4. Discussion

Clinicians often need to attempt to optimize choice of an antipsychotic for a patient with a history of illicit substance use. Previous CATIE Phase 1 reports demonstrated that olanzapine had a longer time to all-cause treatment discontinuation than quetiapine or risperidone, but not perphenazine or ziprasidone. It has not been clear whether illicit drug use might affect these discontinuation rates and the current study offers evidence of a substantial effect. Somewhat surprisingly among illicit substance users there are no differences in time to all-cause discontinuation across treatment groups. Olanzapine, followed by perphenazine has the longest TTD (median of 6.7 and 5.3 months respectively) but neither is significantly longer than others (quetiapine 4.4, risperidone 4.6, and ziprasidone 3.3 months). Among non-users, olanzapine's superiority in time to discontinuation is strengthened substantially and significantly for all but ziprasidone (olanzapine median 13.0 versus perphenazine 5.9, risperidone 5.6, quetiapine 5.0 and ziprasidone 4.3 months respectively). The difference in medians between non-users and illicit drug users in the olanzapine group is 6.3 months, while for other treatment groups the difference is a month or less. In results not shown we omitted illicit users without abuse and restricted the comparisons to illicit substance abusers versus non-users and the results were essentially unchanged.

Results from PANSS and CGI efficacy measures adjusting for duration of treatment show that patients were making comparable gains whether they used illicit substances or not, suggesting that illicit substance-using patients were generally improving comparably to their non-using counterparts, but were less treatment adherent and discontinued treatment for idiosyncratic reasons. For example, among olanzapine illicit substance users versus non-users patient decision was a predominant cause of discontinuation among users compared to non-users. These were patients who were rated to have discontinued treatment for idiosyncratic reasons unrelated to inefficacy or side effects in the opinion of the clinician, and these decisions may have also been related to the lower medication compliance seen in this group. This is also consistent with multivariable analyses that suggested that compliance at least partially mediated differences between illicit substance use and non-users in time to discontinuation for olanzapine. In addition, as a group substance users reported more side effects but this was not reflected in causes of discontinuation related to intolerability. As expected, olanzapine substance users and non-users experienced more weight gain than other treatment groups—suggesting that increased time to discontinuation with olanzapine still came at the cost of weight gain and was the lead cause of treatment discontinuation among the olanzapine treatment group overall (Lieberman et al., 2005).

Why did illicit substance use attenuate the apparent superiority of olanzapine over the other treatments? Our analyses suggest that illicit drug users differ at baseline in many ways from non-users that may directly and indirectly affect treatment discontinuation: users were sicker at baseline, had less illness course stability, lived in more stressful social environments and during the course of the study were less treatment adherent. Lower treatment adherence appears to explain a significant proportion of the shorter time to discontinuation among illicit drug users, but these other characteristics no doubt contributed to shorter time to discontinuation. It could well be that these other attributes of users (living in stressful neighborhoods, interacting in criminogenic social networks, or other stressors) could also attenuate the benefits of medications. Alcohol use and abuse in the absence of illicit substance use had little effect on time to discontinuation, likely because these patients in the aggregate were older, more stable and more treatment compliant. Taken as a whole these results underscore the need for concurrent substance abuse treatment in order for patients to benefit from antipsychotic treatment.

The differential impact of illicit use and non-use across treatment groups is more puzzling; intuitively, we expected to find uniform decreases in time to discontinuation among users compared to non-users, but the relative attenuation in time to discontinuation for olanzapine in users is far more substantial than for other treatments. An imbalance in illicit drug users across treatment groups could have been a partial explanation, but, in analyses not shown, users were evenly distributed across treatment groups. It is reasonable to conclude that among non-users, olanzapine achieves a more acceptable patient outcome for a longer period of time, as judged by time to all-cause discontinuation than the comparators at the dose ranges studied, but that illicit substance use, including non-compliance attenuated these differences. These results also suggest that the many clinical trials that exclude substance users may offer an incomplete picture of clinical effectiveness in real world populations.

There are several limitations to these findings. We did not quantify levels of illicit substance use at baseline or follow-up so cannot assess the impact of levels of use. Nor do we account for involvement in substance abuse treatment over time, although too few patients in the trial were actively engaged in substance abuse treatment for this to have affected the results. It is also possible that these outcomes were not directly attributable to illicit substance use because substance use could serve as a proxy for a host of negative predictors of clinical course and outcome. A more definitive study of the effect of substance use on treatment outcomes would require a well powered randomized trial. Additional analyses examining whether any of the CATIE medications reduce substance use over time is forthcoming.

Role of funding source

This article was based on results from the Clinical Antipsychotic Trials of Intervention Effectiveness project, supported by the National Institute of Mental Health (NO1 MH90001). The NIMH was involved in the study design, data collection, analysis and interpretation of the data, in writing of the manuscript and decision to submit the paper for publication.

Contributors

Drs Swartz, Stroup, McEvoy, Lieberman, Hsiao and Davis designed the study and wrote the protocol.

Drs. Swartz, Wagner and Swanson conducted the literature searches and analyses.

Drs. Wagner, Swanson and Davis undertook the statistical analyses.

Dr. Swartz wrote the first draft of the manuscript.

All authors have contributed to and have approved the final manuscript.

Conflict of interest

Disclosures: Dr. Swartz reports having received research funding from Eli Lilly and Co., and consulting and educational fees from AstraZeneca Pharmaceuticals LP, Bristol-Myers Squibb, Eli Lilly and Co., and Pfizer Inc. Dr. Wagner has no disclosures. Dr. Swanson reports receiving research funding from Eli Lilly and Co and consultant fees form CNS Neurosciences. Dr. Stroup reports having received research funding from Eli Lilly and Co.; and consulting fees from Janssen Pharmaceutica Products, GlaxoSmithKline, and Bristol-Myers Squibb. Dr. Sonia M. Davis is an employee of Quintiles Inc.; she reports no additional funding. Dr. Rosenheck reports having received research funding from AstraZeneca Pharmaceuticals LP, Bristol-Myers Squibb, and Eli Lilly and Co.; and consulting fees from Bristol-Myers Squibb, Eli Lilly and Co., and Janssen Pharmaceutica Products. Dr. Reimherr and McGee report no additional funding. Dr. McEvoy reports having received research funding from AstraZeneca, Forest Research Institute, Eli Lilly and Co., Janssen Pharmaeutica, and Pfizer Inc.; consulting or advisory board fees from Pfizer Inc. and Bristol-Myers Squibb; and lecture fees from Janssen Pharmaceutica, and Bristol-Myers Squibb. Dr. Lieberman reports having received research funding from AstraZeneca Pharmaceuticals LP, Bristol-Myers Squibb, GlaxoSmithKline, Janssen Pharmaceutica Products, and Pfizer Inc.; and consulting and educational fees from AstraZeneca Pharmaceuticals LP, Bristol-Myers Squibb, Eli Lilly and Co., Forest Pharmaceutical Company, GlaxoSmithKline, Janssen Pharmaceutica Products, Novartis, Pfizer Inc., and Solvay.

Acknowledgments

This article was based on the results from the Clinical Antipsychotic Trials of Intervention Effectiveness project, supported by the National Institute of Mental Health (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.; the program staff of the Division of Interventions and Services Research of the NIMH; and investigators from 56 sites in the United States (CATIE Study Investigators Group). 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. The Foundation of Hope of Raleigh, NC also supported this work.

CATIE Study Investigators Group includes: Lawrence Adler, M.D., Clinical Insights, Glen Burnie, MD; Mohammed Bari, M.D., Synergy Clinical Research, Chula Vista, CA; Irving Belz, M.D., Tri-County/MHMR, Conroe, TX; Raymond Bland, M.D., Southern Illinois University School of Medicine, Springfield, IL; Thomas Blocher, M.D., MHMRA of Harris County, Houston, TX; Brent Bolyard, M.D., Cox North Hospital, Springfield, MO; Alan Buffenstein, M.D., The Queen's Medical Center, Honolulu, HI; John Burruss, M.D., Baylor College of Medicine, Houston, TX; Matthew Byerly, M.D., University of Texas Southwestern Medical Center at Dallas, Dallas, TX; Jose Canive, M.D., Albuquerque VA Medical Center, Albuquerque, NM; Stanley Caroff, M.D., Behavioral Health Service, Philadelphia, PA; Charles Casat, M.D., Behavioral Health Center, Charlotte, NC; Eugenio Chavez-Rice, M.D., El Paso Community MHMR Center, El Paso, TX; John Csernansky, M.D., Washington University School of Medicine, St. Louis, MO; Pedro Delgado, M.D., University Hospitals of Cleveland, Cleveland, OH; Richard Douyon, M.D., VA Medical Center, Miami, FL; Cyril D'Souza, M.D., Connecticut Mental Health Center, New Haven, CT; Ira Glick, M.D., Stanford University School of Medicine, Stanford, CA; Donald Goff, M.D., Massachusetts General Hospital, Boston, MA; Silvia Gratz, M.D., Eastern Pennsylvania Psychiatric Institute, Philadelphia, PA; George T. Grossberg, M.D., St. Louis University School of Medicine—Wohl Institute, St. Louis, MO; Mahlon Hale, M.D., New Britain General Hospital, New Britain, CT; Mark Hamner, M.D., Medical University of South Carolina and Veterans Affairs Medical Center, Charleston, SC; Richard Jaffe, M.D., Belmont Center for Comprehensive Treatment, Philadelphia, PA; Dilip Jeste, M.D., University of California—San Diego, VA Medical Center, San Diego, CA; Anita Kablinger, M.D., Louisiana State University Health Sciences Center, Shreveport, LA; Ahsan Khan, M.D., University of Kansas School of Medicine, Wichita, KS; Steven Lamberti, M.D., University of Rochester Medical Center, Rochester, NY; Michael T. Levy, M.D., PC, Staten Island University Hospital, Staten Island, NY; Jeffrey Lieberman, M.D., University of North Carolina School of Medicine, Chapel Hill, NC; Gerald Maguire, M.D., University of California Irvine, Orange, CA; Theo Manschreck, M.D., Corrigan Mental Health Center, Fall River, MA; Joseph McEvoy, M.D., Duke University Medical Center, Durham, NC; Mark McGee, M.D., Appalachian Behavioral Healthcare, Athens, OH; Herbert Meltzer, M.D., Vanderbilt University Medical Center, Nashville, TN; Alexander Miller, M.D., University of Texas Health Science Center at San Antonio, San Antonio, TX; Del D. Miller, M.D., University of Iowa, Iowa City, IA; Henry Nasrallah, M.D., University of Cincinnati Medical Center, Cincinnati, OH; Charles Nemeroff, M.D., Ph.D., Emory University School of Medicine, Atlanta, GA; Stephen Olson, M.D., University of Minnesota Medical School, Minneapolis, MN; Gregory F. Oxenkrug, M.D., St. Elizabeth's Medical Center, Boston, MA; Jayendra Patel, M.D., University of Mass Health Care, Worcester, MA; Frederick Reimher, M.D., University of Utah Medical Center, Salt Lake City, UT; Silvana Riggio, M.D., Mount Sinai Medical Center—Bronx VA Medical Center, Bronx, NY; Samuel Risch, M.D., University of California—San Francisco, San Francisco, CA; Bruce Saltz, M.D., Henderson Mental Health Center, Boca Raton, FL; Thomas Simpatico, M.D., Northwestern University, Chicago, IL; George Simpson, M.D., University of Southern California Medical Center, Los Angeles, CA; Michael Smith, M.D., Harbor—UCLA Medical Center, Torrance, CA; Roger Sommi, Pharm.D., University of Missouri, Kansas City, MO; Richard M. Steinbook, M.D., University of Miami School of Medicine, Miami, FL; Michael Stevens, M.D., Valley Mental Health, Salt Lake City, UT; Andre Tapp, M.D., VA Puget Sound Health Care System, Tacoma, WA; Rafael Torres, M.D., University of Mississippi, Jackson, MS; Peter Weiden, M.D., SUNY Downstate Medical Center, Brooklyn, NY; James Wolberg, M.D., Mount Sinai Medical Center, New York, NY.

References

Abram and Teplin, 1991. 1.Abram K, Teplin L. Co-occurring disorders among mentally ill jail detainees: implications for public policy. Am. Psychol. 1991;46:1036–1044. CrossRef

Addington et al., 1992. 2.Addington D, Addington J, Maticka-Tyndale E, Joyce J. Reliability and validity of a depression rating scale for schizophrenics. Schizophr. Res. 1992;6:201–208. MEDLINE | CrossRef

Blanchard et al., 2000. 3.Blanchard JJ, Brown SA, Horan WP, Sherwood AR. Substance use disorders in schizophrenia: review, integration, and a proposed model. Clin. Psychol. Rev. 2000;20:207–234. MEDLINE | CrossRef

Buckley, 1999. 4.Buckley PF. New antipsychotic agents: emerging clinical profiles. J. Clin. Psychiatry. 1999;60:12–17.

Cantor-Graae et al., 2001. 5.Cantor-Graae E, Nordstrom LG, McNeil TF. Substance abuse in schizophrenia: a review of the literature and a study of correlates in Sweden. Schizophr. Res. 2001;48:69–82. Abstract | Full Text | Full-Text PDF (126 KB) | CrossRef

Cournos et al., 1991. 6.Cournos F, Empfield M, Horwath E, McKinnon K, Meyer I, Schrage H, et al. HIV prevalence among patients admitted to two psychiatric hospitals. Am. J. Psychiatry. 1991;148:1225–1229.

Davis et al., 2003. 7.Davis SM, Koch GG, Davis CE, LaVange LM. Statistical approaches to effectiveness measurement and outcome-driven re-randomizations in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) studies. Schizophr. Bull. 2003;29:73–80. MEDLINE

Dixon, 1999. 8.Dixon L. Dual diagnosis of substance abuse in schizophrenia: prevalence and impact on outcomes. Schizophr. Res. 1999;35:S93–S100. Abstract | Full Text | Full-Text PDF (93 KB) | CrossRef

Drake et al., 1990. 9.Drake R, Osher FC, Noordsy DL, Hurlbut SC, Teague GB, Beaudett MS. Diagnosis of alcohol use disorders in schizophrenia. Schizophr. Bull. 1990;16:57–67. MEDLINE

Drake et al., 2000. 10.Drake RE, Xie H, McHugo GJ, Green AI. The effects of clozapine on alcohol and drug use disorders among patients with schizophrenia. Schizophr. Bull. 2000;26:441–449. MEDLINE

First et al., 1996. 11.First MB, Spitzer RL, Gibbon M, Williams JB. Structured Clinical Interview for Axes I and II DSM-IV Disorders-Patient Edition (SCID-I/P). New York: Biometrics Research Department, New York State Psychiatric Institute; 1996;.

Goodman et al., 2001. 12.Goodman LA, Salyers MP, Mueser KT, Rosenberg SD, Swartz M, Essock SM, et al. Recent victimization in women and men with severe mental illness: prevalence and correlates. J. Trauma. Stress. 2001;14:615–632. MEDLINE | CrossRef

Goswami et al., 2004. 13.Goswami S, Mattoo SJ, Basu D, Singh G. Substance-abusing schizophrenics: do they self-medicate?. Am. J. Addict. 2004;13:139–150. MEDLINE | CrossRef

Green, 2005. 14.Green AI. Schizophrenia and comorbid substance use disorder: effects of antipsychotics. J. Clin. Psychiatry. 2005;66:21–26.

Green et al., 1999. 15.Green AI, Strous RD, Zimmet SV, Schildkraut JJ. Clozapine for comorbid substance use disorder and schizophrenia: do patients with schizophrenia have a reward-deficiency syndrome that can be ameliorated by clozapine?. Harv. Rev. Psychiatr. 1999;6:287–296.

Guy, 1976. 16.Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology—Revised (DHEW Publ No. ADM 76-338). U.S. Department of Health, Education, and Welfare. Public Health Service, Alcohol, Drug Abuse, and Mental Health Administration, NIMH Psychopharmacology Research Branch, Division of Extramural Research Programs, Rockville, MD, pp. 218–222.

Haywood et al., 1995. 17.Haywood TW, Kravitz HM, Grossman LS, Cavanaugh JL, Davis JM, Lewis DA. Predicting the ‘revolving door’ phenomenon among patient with schizophrenic, schizoaffective, and affective disorders. Am. J. Psychiatry. 1995;151:856–861.

Hochberg, 1988. 18.Hochberg Y. A sharper Bonferroni procedure for multiple tests of significance. Biometrika. 1988;75:800–802. CrossRef

Kavanagh et al., 2002. 19.Kavanagh DJ, McGrath J, Saunders JB, Dore G, Clark D. Substance misuse in patients with schizophrenia: epidemiology and management. Drugs. 2002;62:743–755. MEDLINE | CrossRef

Kay et al., 1987. 20.Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr. Bull. 1987;13:261–276. MEDLINE

Keefe et al., 2003. 21.Keefe RS, Mohs RC, Bilder RM, Harvey PD, Green MF, Meltzer HY, et al. Neurocognitive assessment in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) project schizophrenia trial: development, methodology, and rationale. Schizophr. Bull. 2003;29:45–55. MEDLINE

Keefe et al., 2006. 22.Keefe RSE, Bilder RM, Harvey PD, Davis S, Palmer BW, Gold JM, et al. Baseline neurocognitive deficits in the CATIE Schizophrenia Trial. Neuropsychopharmacology. 2006;4:1–14.

Keefe et al., 2007. 23.Keefe RSE, Bilder RM, Harvey PD, Davis SM, Palmer B, Gold JM, et al. Neurocognitive effects of antipsychotic medications in patients with chronic schizophrenia. Arch. Gen. Psychiatry. 2007;64:633–647. CrossRef

Kessler et al., 1997. 24.Kessler RC, Crum RM, Warner LA, Nelson CB, Schulenberg J, Anthony JC. Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Survey. Arch. Gen. Psychiatry. 1997;54:313–321.

Lieberman et al., 2005. 25.Lieberman JA, Stroup TS, McEvoy JP, Swartz MS, Rosenheck RA, Perkins DO, et al. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. New Engl. J. Med. 2005;353:1209–1223. CrossRef

Littrell et al., 2001. 26.Littrell KH, Petty RG, Hilligoss NM, Peabody CD, Johnson CG. Olanzapine treatment for patients with schizophrenia and substance abuse. J. Subst. Abuse Treat. 2001;21:217–221. Abstract | Full Text | Full-Text PDF (82 KB) | CrossRef

McCreadie, 2002. 27.McCreadie RG. Use of drugs, alcohol and tobacco by people with schizophrenia: case-control study. Br. J. Psychiatry. 2002;181:321–325. MEDLINE | CrossRef

McEvoy et al., 1989. 28.McEvoy JP, Apperson LJ, Appelbaum PS, Ortlip P, Brecosky J, Hammill K, et al. Insight in schizophrenia: its relationship to acute psychopathology. J. Nerv. Ment. Dis. 1989;177:43–47. MEDLINE | CrossRef

McEvoy et al., 2006. 29.McEvoy JP, Lieberman JA, Stroup TS, Davis SM, Rosenheck RA, Swartz MS, et al. Effectiveness of clozapine, quetiapine and risperidone in patients with chronic schizophrenia who failed prior atypical antipsychotic treatment. Am. J. Psychiatry. 2006;153:600–610.

Mueser et al., 1998. 30.Mueser KT, Drake RE, Wallach MA. Dual diagnosis: a review of etiological theories. Addict. Behav. 1998;23:717–734. MEDLINE | CrossRef

Petrakis et al., 2006. 31.Petrakis IL, Leslie D, Finney JW, Rosenheck R. Atypical antipsychotic medication and substance use-related outcomes in the treatment of schizophrenia. Am. J. Addict. 2006;15:44–49. MEDLINE | CrossRef

Regier et al., 1990. 32.Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, et al. Comorbidity of mental disorders with alcohol and other drug abuse. Results from the Epidemiologic Catchment Area (ECA) Study. J. Am. Med. Assoc. 1990;264:2511–2518.

Rosenberg et al., 2001. 33.Rosenberg SD, Goodman LA, Osher FC, Swartz MS, Essock SM, Butterfield MI, et al. Prevalence of HIV, Hepatitis B and Hepatitis C in people with severe mental illness. Am. J. Public Health. 2001;91:31–37. MEDLINE

Rosenheck et al., 2006. 34.Rosenheck R, Leslie D, Sindelar J, Miller EA, Stroup S, McEvoy J, et al. Cost-effectiveness of first and second generation antipsychotics in the treatment of chronic schizophrenia. Am. J. Psychiatry. 2006;163(12):2080–2089. CrossRef

Salyers and Mueser, 2001. 35.Salyers MP, Mueser KT. Social functioning, psychopathology, and medication side effects in relation to substance use and abuse in schizophrenia. Schizophr. Res. 2001;48:109–123. Abstract | Full Text | Full-Text PDF (139 KB) | CrossRef

Smelson et al., 2002. 36.Smelson DA, Losonczy MF, Davis CW, Kaune M, Williams J, Ziedonis D. Risperidone decreases craving and relapses in individuals with schizophrenia and cocaine dependence. Can. J. Psychiatry. 2002;47:671–675. MEDLINE

Stroup et al., 2003. 37.Stroup TS, McEvoy JP, Swartz MS, Byerly MJ, Glick ID, Canive JM, et al. The National Institute of Mental Health Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) project: schizophrenia trial design and protocol development. Schizophr. Bull. 2003;29:15–31. MEDLINE

Stroup et al., 2006. 38.Stroup TS, Lieberman JA, McEvoy JP, Swartz MS, Davis SM, Rosenheck RA, et al. Effectiveness of olanzapine, quetiapine, risperidone, and ziprasidone in patients with chronic schizophrenia after discontinuation of a previous atypical antipsychotic. Am. J. Psychiatry. 2006;163:611–622. CrossRef

Swanson et al., 1997. 39.Swanson J, Estroff S, Swartz M, Borum R, Lachicotte W, Zimmer C, et al. Violence and severe mental disorder in clinical and community populations: the effects of psychotic symptoms, comorbidity, and lack of treatment. Psychiatry. 1997;60:1–22. MEDLINE

Swartz et al., 1998. 40.Swartz MS, Swanson JW, Hiday VA, Borum R, Wagner HR, Burns BJ. Violence and severe mental illness: the effects of substance abuse and nonadherence to medication. Am. J. Psychiatry. 1998;155:226–231.

Swartz et al., 2003. 41.Swartz MS, Perkins DO, Stroup TS, McEvoy JP, Nieri JM, Haak DC. Assessing clinical and functional outcomes in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Schizophrenia Trial. Schizophr. Bull. 2003;29:33–43. MEDLINE

Swartz et al., 2006a. 42.Swartz MS, Wagner HR, Swanson JW, Stroup TS, McEvoy JP, Canive JM, et al. Substance use in persons with schizophrenia: baseline prevalence and correlates from the NIMH CATIE study. J. Nerv. Ment. Dis. 2006;194:164–172. MEDLINE | CrossRef

Swartz et al., 2006b. 43.Swartz MS, Wagner HR, Swanson JW, Stroup TS, McEvoy JP, McGee M, et al. Substance use and psychosocial functioning in schizophrenia among new enrollees in the NIMH CATIE Study. Psychiatr. Serv. 2006;57:1110–1116. MEDLINE | CrossRef

Swartz et al., 2007. 44.Swartz MS, Perkins DO, Stroup TS, Davis SM, Capuano G, Rosenheck RA, et al. Effects of antipsychotic medications on psychosocial functioning in patients with chronic schizophrenia: findings from the NIMH CATIE Study. Am. J. Psychiatry. 2007;164(3):428–436. CrossRef

Swofford et al., 1996. 45.Swofford CD, Kasckow JW, Scheller-Gilkey G, Inderbitzin LB. Substance use: a powerful predictor of relapse in schizophrenia. Schizophr. Res. 1996;20:145–151. Abstract | Full-Text PDF (536 KB) | CrossRef

Tsuang et al., 2002. 46.Tsuang JW, Eckman T, Marder S, Tucker D. Can risperidone reduce cocaine use in substance abusing schizophrenic patients?. J. Clin. Psychopharmacol. 2002;22:629–630. MEDLINE | CrossRef

a Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham NC, United States

b Department of Psychiatry, School of Medicine, University of North Carolina at Chapel Hill, NC, United States

c Department of Psychiatry, University of Utah Medical Center, Salt Lake City, UT, United States

d Appalachian Behavioral Healthcare System, Athens OH, United States

e Quintiles Inc., Research Triangle Park, NC, United States

f Department of Psychiatry, University of Iowa School of Medicine, Iowa City, IA, United States

g Psychiatric Research Institute, Wichita KS, United States

h Department of Psychiatry, Albuquerque Veteran’s Administration Medical Center and the University of New Mexico, Albequerque, NM, United States

i Division of Services and Intervention Research, National Institute of Mental Health, National Institutes of Health, Bethesda, United States

j Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York State Psychiatric Institute, New York, United States

Corresponding author. Department of Psychiatry and Behavioral Sciences, Duke University School of Medicine, DUMC Box 3173, Durham, NC 27710, United States. Tel.: +1 919 684 8676; fax: +1 919 681 7504.

1 The CATIE Investigators are listed in the Acknowledgements.

PII: S0920-9964(07)00542-7

doi:10.1016/j.schres.2007.11.034

5 of 42