CanJPsychiatry 2012;57(2):7884

Original Research

Medication-Adherent First-Episode Psychosis Patients Also Relapse: Why?

Emmanuelle Levy, MD, FRCPC1; Nicole Pawliuk, MA2; Ridha Joober, MD, PhD3; Sherezad Abadi, MSc4; Ashok Malla, MBBS, FRCPC5
1

Assistant Professor, Department of Psychiatry, McGill University, Montreal, Quebec; Staff Psychiatrist, PEPP-Montral, Douglas Mental Health University Institute, Montreal, Quebec. Research Data Manager, PEPP-Montral, Douglas Mental Health University Institute, Montreal, Quebec. Associate Professor, Department of Psychiatry, McGill University, Montreal, Quebec; Associate Director, PEPP-Montral, Douglas Mental Health University Institute, Montreal, Quebec. Research Coordinator and Agent de planification et evaluation, PEPP-Montral, Douglas Mental Health University Institute, Montreal, Quebec. Professor, Department of Psychiatry, McGill University, Montreal, Quebec; Director, PEPP-Montral, Douglas Mental Health University Institute, Montreal, Quebec. Correspondence: Department of Psychiatry, McGill University, Douglas Mental Health University Institute, 6875 Boulevard Lasalle, Montreal, QC H4H 1R3; ashok.malla@douglas.mcgill.ca.

2 3

4 5

Key Words: first-episode psychosis, medication adherence, relapse Received May 2011, revised, and accepted July 2011.

Objective: Poor adherence to medication is a major determinant of relapse following treatment of first-episode psychosis (FEP). However, medication-adherent patients also relapse. We examined what factors influence the risk of relapse after controlling for adherence. Method: We selected a sample of fully adherent patients (n = 65) who had achieved remission at one point. We then compared patients who relapsed, using 2 different definitions of relapse, to those who did not relapse by 12 months on age, sex, premorbid adjustment, duration of untreated psychosis, length of prodrome, and substance abuse. Results: Among the 65 medication-adherent patients in remission, 9 (14%) relapsed according to criteria for relapse requiring a change in medication. These patients differed from those who remained in remission only in the pattern of premorbid adjustment (greater proportion with deteriorating pattern), although this was not independent of other variables. No differences were found on any other variable. Using a more commonly used metric for relapse, based on symptom ratings alone, an additional 14 (21.5%) patients relapsed. Substance abuse significantly predicted relapse, with substance abusers having more than 25 times the odds of relapsing by 12 months (OR 25.6; 95% CI 2.4 to 278.1, P = 0.008). Conclusion: Using a more conservative definition of relapse in this adherent-tomedication population, we find a very low rate of relapse associated, at least partially, with poor premorbid adjustment. As substance abuse was a significant predictor of symptomatic relapse, this would suggest that there should be a greater emphasis on interventions focused on reducing substance abuse in FEP. WWW Objectif : La mauvaise observance des mdicaments est un dterminant majeur de la rechute la suite du traitement dun premier pisode de psychose (PEP). Cependant, les patients qui sont fidles aux mdicaments rechutent eux aussi. Nous avons examin les facteurs qui influencent le risque de rechute aprs un contrle de lobservance. Mthode : Nous avons slectionn un chantillon de patients lobservance complte (n = 65) qui avaient atteint la rmission un moment donn. Nous avons ensuite compar les patients qui rechutaient, laide de 2 dfinitions diffrentes de la rechute, avec ceux qui ne rechutaient pas 12 mois, selon lge, le sexe, lajustement prmorbide, la dure de la psychose non traite, la dure du prodrome, et labus de substances. Rsultats : Sur les 65 patients fidles aux mdicaments en rmission, 9 (14 %) ont rechut daprs des critres de la rechute indiquant un changement de mdicaments. Ces patients diffraient de ceux qui sont demeurs en rmission uniquement par le modle dajustement prmorbide (une proportion plus grande par le modle de dtrioration), mme si ce ntait pas indpendant des autres variables. Aucune diffrence na t observe sur toute autre variable. laide dune mesure de la rechute plus frquemment utilise, selon lestimation des symptmes uniquement, 14 (21,5 %) autres patients ont

78 W La Revue canadienne de psychiatrie, vol 57, no 2, fvrier 2012

www.LaRCP.ca

Medication-Adherent First-Episode Psychosis Patients Also Relapse: Why?

rechut. Labus de substances prdisait significativement la rechute, et les toxicomanes avaient plus de 25 fois de probabilits de rechuter 12 mois (RC 25,6; IC 95 % 2,4 278,1; P = 0,008). Conclusion : En utilisant une dfinition plus conventionnelle de la rechute dans cette population fidle aux mdicaments, nous constatons un taux trs faible de rechute associ, du moins partiellement, un mdiocre ajustement prmorbide. Comme labus de substances tait un prdicteur significatif dune rechute symptomatique, cela peut suggrer de mettre davantage laccent sur les interventions axes sur la rduction de labus de substances dans le PEP.

n routine care, most patients relapse within the first 5 years following treatment for FEP.1 Relapse rates after 12 months of treatment for FEP range between 16% to 44%,17 with the lowest rates found in SEI services,3,7 where high levels of adherence to antipsychotics are generally reported. Poor adherence to medications is a major determinant of relapse.8 In a recent study3 of patients treated in an SEI service with high rates (87%) of adherence to medication, relatively low relapse rates were reported for 1 and 2 years (21% and 29%, respectively). These relapse rates, although lower than in routine care, are not inconsequential. Adherence apparently does not fully prevent patients from ever relapsing.

Clinical Implications
Despite a higher level of adherence to antipsychotics in specialized early intervention services, patients do relapse, albeit less often. For adherent patients, relapse requiring a change in treatment may be influenced by relatively nonmalleable factors, such as a deteriorating premorbid functioning, although this is not an independent effect and likely reflects an inherently unstable pattern of illness course. In the presence of full adherence to medication, patients with substance abuse have a significantly higher risk of relapse, based on symptom change alone. Few patients in this study relapsed. A larger sample of patients with FEP would provide greater statistical power and could potentially reveal other more malleable predictors. Substance abuse was assessed only at baseline. Other time points in treatment were not examined. Presence or absence of substance abuse was assessed with the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (for DSM-IV-TR) Patient Edition, but quantities of substances consumed were not systematically recorded.

Limitations

Other risk factors associated with relapse include substance abuse3,9 and poor premorbid adjustment.1 In trying to determine other factors associated with relapse, some studies have attempted to control statistically for medication adherence.1 What predicts relapse in patients treated for FEP other than adherence to medication requires controlling for adherence experimentally, and not merely statistically, by conducting a study of relapse in patients with FEP who are adherent to medication. To the best of our knowledge, no previous study has examined relapse in medication-adherent patients. The objective of our study is to determine the rate of relapse in patients otherwise considered to be adherent to medication within the first 12 months of treatment and to examine what other factors might influence the risk of relapse once adherence is controlled for.

Material and Methods

Context and Sample
Patients aged 14 to 30 years, received treatment for FEP in the PEPP-Montreal clinic at the Douglas Mental Health University Institute in Montreal, Quebec. All patients met criteria for primary diagnosis of psychotic disorder on admission, using the SCID-I/P (for DSM-IV-TR) Patient Edition,10 assessed by trained interviewers, and agreed on through consensus between 2 senior research psychiatrists. All raters were independent of any involvement in treatment. Our study was part of a larger longitudinal outcome evaluation of treatment of FEP in an SEI service. A protocol for our study, including symptom evaluations and informed consent, had been approved by the Research Ethics Board of the Douglas Mental Health University Institute. All participants in our study had signed the informed consent. Patients were not previously treated with antipsychotics
www.TheCJP.ca

for more than 30 days. Patients were excluded if they were unable to speak French or English, had an IQ of less than 70, or had an organic brain syndrome.

Measures

Selection Criterion: Adherence. All patients were assessed for medication adherence 7 times over the first year of treatment (at baseline and at months: 1, 2, 3, 6, 9, and 12), through direct interview followed by additional corroboration with case managers and families. In an independent investigation of adherence to treatment, this method of assessing adherence was shown to be highly correlated with pill counting (r = 0.90, P < 0.001; n = 42).11,12 The level of patient adherence was determined using a 5-point scale, ranging from not taking medication at all to being adherent more than 75% of the time (0 [0%], 1 [1% to 25%], 2 [26% to 50%], 3 [51% to 75%], and 4 [76% to 100%]). A participant was considered adherent to medication if he or she took it more than 75% of the time, at
The Canadian Journal of Psychiatry, Vol 57, No 2, February 2012 W 79

Original Research

Table 1 Demographics of patients who relapse and those who did not
First 12 months
Demographic Age, years Sex, male, % Education, %

Relapsed n=9 23.1 67 67 11 22 78 0 22

Stayed in remission n = 56 22.5 66 52 19 29

Analysis t = 0.41, df = 63 Fisher exact test c2 = 0.74, df = 2

P 0.68 0.64 0.69

High school College University

Marital status, %

c2 = 14.1, df = 2 82 18 0 0.001

Single Married Separated or divorced

6 out of the 7 evaluations. Only patients who were adherent to medication, thus defined, were included in our study. Assessment of Symptoms. Positive and negative symptoms of psychosis were assessed by trained interviewers using the SAPS13 and the SANS14 at entry to the program for treatment, monthly for 3 months, and thereafter every 3 months until month 12. Clinical notes were used to determine the presence or absence of psychotic symptoms between evaluations. Assessments of interrater reliability on the symptom scales revealed good agreement (kappas for SAPS = 0.76; SANS = 0.77). Remission. We considered patients to be in remission of positive symptoms if they showed minimal or no psychotic symptoms (for example, delusions, hallucinations, thought disorder, and bizarre behaviour) for a period of at least 4 weeks established through symptom ratings of 2 or less on each of the global subscales on the SAPS. We used a definition of remission consisting only of positive symptoms because of our focus on relapse of psychosis. Relapse. Several studies1,4,5,7,15,16 define relapse as the recurrence of psychotic symptoms, established only through changes in scores on various rating scales, while others have used hospitalization or a combination of several

behavioural changes not necessarily confined to psychotic symptoms.17 Definition of relapse based solely on crossing a threshold on ratings of symptoms does not provide any information on clinical relevance, such as whether it was accompanied by need to change treatment or, alternatively, the symptom change corrected itself within a relatively brief time. We define relapse as the recurrence of psychotic symptoms after the patient had been in remission (see criterion above) when SAPS global severity scores are 3 or more on any of the subscales (hallucinations, delusions, bizarre behaviour, and formal thought disorder) lasting for at least 1 week and when such change resulted in a change in antipsychotics (including adjustment of dosage). However, we also conducted a secondary analysis using a more liberal definition based on symptom change alone. Premorbid Adjustment. Premorbid adjustment was assessed with van Mastrigt and Addingtons18 modification of the PAS19 to evaluate adjustment during childhood and early adolescence on both social and educational dimensions. We excluded the late adolescent and adult periods, as these would overlap with the onset of prodromal as well as psychotic symptoms. Total PAS score is calculated by adding scores on all items and then dividing them by the total possible score. The final score is the proportion bound between 0 (best possible) or 1 (worst possible). The PAS change over time was defined using 3 premorbid functioning PAS subtypes: deteriorating, stable good, and stable poor, following the criteria provided by Haas and Sweeney20 and Larsen et al.2,21,22 Following these criteria, deterioration was calculated as a 2-point change from the previous age period. The rest of the participants were classified as stable good if their PAS score was above the median, or stable poor if the PAS score was below the median. Duration of Untreated Psychosis. Early signs and symptoms were determined using the Circumstances of Onset and Relapse Schedule.23,24 This instrument provides information regarding lifetime history of the illness prior to the onset of the presenting psychotic episode and allows the estimation
www.LaRCP.ca

Abbreviations
DUP FEP PAS PEPP SANS SAPS SCID SEI duration of untreated psychosis first-episode of psychosis Premorbid Adjustment Scale Prevention and Early Intervention Program for Psychoses Scale for Assessment of Negative Symptoms Scale for Assessment of Positive Symptoms Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition specialized early intervention

80 W La Revue canadienne de psychiatrie, vol 57, no 2, fvrier 2012

Medication-Adherent First-Episode Psychosis Patients Also Relapse: Why?

Table 2 Clinical characteristics of patients who relapsed and those who did not
First 12 months
Characteristic Diagnosis type, (%)

Relapsed n=9 6 (67) 3 (33) 25.1

Stayed in remission n = 56 45 (80) 11 (20)

Analysis Fisher exact test

P 0.39

Schizophrenia spectrum disorders Affective psychosis

Baseline symptoms

0.37 29.8 t = 0.9, df = 63 c2 = 3.5, df = 5 5 (56) 13 mg/day 3 (33) 1.25 mg/day 1 (11) 225 mg/day 0 0 2 (25) 3 (38) 4 (50) 1 (12) 30 (54) 12.5 mg/day 17 (30) 1.6 mg/day 1 (1.8) 300 mg/day 2 (3.6) 2 (3.6) 3 (7) 7 (13) 14 (30) 11 (24) 21 (46) Fisher exact test Fisher exact test 0.31 0.03 0.48

SAPS total, average

Antipsychotic at baseline, n (%) average dosage

Olanzapine Risperidone Quetiapine Olanzapine + Conventional Risperidone + Conventional No antipsychotics

Substance abuse or dependence, (%) Premorbid adjustment course, (%)

Stable poor Deteriorating Stable good

of the DUP. Specifically, DUP was calculated as the period between the time of onset of psychotic symptoms to the time of adequate antipsychotic treatment, defined as taking antipsychotics for 1 month.25 Prodromal Period. Length of prodrome was defined as the time from onset of the nonpsychotic symptoms (for example, depression, anxiety, and social withdrawal) prior to and contiguous with onset of psychosis. Substance Abuse. Presence or absence of substance abuse was defined by the baseline SCID-I/P assessment of current substance abuse or dependence.10

Results
Sample Characteristics and Adherence to Medication

Statistical Analysis

Following the calculation of rate of relapse for the defined sample, we compared patients who relapsed with those who remained in remission on demographic variables, premorbid adjustment, DUP, length of prodrome, and substance abuse. Chi-square analyses or t tests were conducted with SPSS version 15.0,26 depending on the nature of the variables. Nonnormal distributions were normalized through a log-linear transformation (for example, DUP and prodromal period). Univariate analyses were followed by logistic regression, using only those independent variables as covariates that showed a significant association in the univariate analysis at the 0.05 level. Cox regression analysis was used to determine the influence of any covariates on time to relapse.
www.TheCJP.ca

Among the 203 patients accepted in the PEPP clinic between January 2003 and June 2007, 163 were followed up actively through a 12-month period. Among the 40 patients who did not continue to 12 months of treatment, 29 refused follow-up, 8 had moved, and 3 were later determined to be ineligible for the program for diagnostic reasons. The 40 patients who discontinued treatment were similar to the 163 who continued, on demographic and diagnostic variables, except for older age at program entry (24.3 years, compared with 22.7 years, P = 0.02) and ethnicity, with 68% of patients who discontinued treatment being non-White, compared with 33% of those who continued (P < 0.001). Remission data at 12 months were available for 140 patients. There were no differences between patients with remission data and those without. Fourteen patients of the 140 did not remit by month 12. Nonremitting patients did not differ from patients who remitted in demographic and diagnostic variables, except for older age at program entry (24.9 years, compared with 22.6 years, P = 0.05). Medication-adherence data, at all time points over the first 12 months, were available for 116 patients who had remitted. Sixty-five (56%) of these patients were
The Canadian Journal of Psychiatry, Vol 57, No 2, February 2012 W 81

Original Research

fully adherent to medications; 51 (44%) were defined as nonadherent at varying levels. Adherent patients did not differ from nonadherent patients in demographic and diagnostic variables, except for living status at baseline (91% of adherent compared with 76% of nonadherent patients lived with others, P = 0.04).

Relapse as Defined by Symptom Change and Need to Change or Increase Medication

Rate and Clinical Correlates. Among the 65 consistently adherent patients, 9 (14%) relapsed in the first year of treatment while 56 (86%) remained in remission. Time to remission did not differ significantly between patients who relapsed (mean time = 8.6 weeks) and patients who stayed in remission (mean time = 10.8 weeks). The mean time to relapse was 22.7 weeks (SD 17.1 weeks). Using a Cox regression analysis, time to relapse was not associated with any variable of interest (DUP, length of prodrome, substance abuse, sex, premorbid adjustment). Table 1 and 2, respectively, present a comparison of demographic and clinical differences between adherent patients who relapsed and those who did not. Patients who relapsed did not differ from those who remained in remission with respect to age of onset, sex, education, ethnicity, living circumstances, IQ, symptoms, time to remission, DUP, substance abuse, or diagnosis type. Among 56 patients who remained in remission, 10 (18%) were married and none were separated or divorced. In contrast, among the 9 patients who relapsed, none were married and 2 (22%) were separated or divorced (chi-square P = 0.001). A higher proportion of patients who relapsed showed a deteriorating course of premorbid adjustment (50%, compared with 24%; Fisher exact test P = 0.03) (Table 1). When entered stepwise into a logistic regression, neither marital status nor PAS course significantly predicted relapse over the 1-year period.

remission, DUP, or diagnosis type. However, a higher proportion of patients who relapsed were diagnosed with current substance use (36%) compared with those who remained in remission (5%) (Fisher exact test P = 0.008). Moreover, all patients with substance abuse issues were cannabis abusers, with 2 of them having a cannabis dependence. SAPS-relapsed patients also had a significantly longer prodrome before psychosis (mean = 213 weeks) compared with those who stayed in remission (mean = 48 weeks) (P = 0.02). A greater proportion of SAPS-relapsed patients showed a stable poor course of premorbid adjustment (56%, compared with 24%; Fisher exact test P = 0.01). Finally, SAPS-relapsing patients had a higher total global SAPS score (9.4) than those staying in remission (1.6). No other symptom comparisons showed differences between groups. The meaningful variables that significantly differed between the 2 groups were entered stepwise into a logistic regression to predict SAPS relapse by month 12. Substance abuse was the only variable found to significantly predict relapse. Patients suffering from substance abuse at entry had more than 25 times the odds of SAPS relapsing by 12 months (OR 25.6; 95% CI 2.4 to 278.1; P = 0.008).

Discussion

We examined what factors contribute to relapse in fully adherent patients. During the first 12 months of treatment, using a more conservative definition of relapse, we found a relatively low rate of relapse (14%) as expected in a sample of fully adherent patients. This low rate is also similar to the rates of relapse in the first year of follow-up of other studies of relapse in FEP carried out in SEI services.1,3 Using this definition of relapse, we found a deteriorating course of premorbid adjustment to be associated with risk of relapse. Poor premorbid adjustment had been previously described to influence poor clinical outcome and risk of relapse.1,2,27 Our findings suggest that the risk of relapse is associated with a particular pattern of premorbid adjustment, which may render patients vulnerable to relapse because of an inherently unstable illness course. Our finding also suggests that this association between relapse and poor premorbid adjustment is independent of adherence to medication. While single marital status also appeared to be associated with risk of relapse, the small number of patients who were married in the sample makes it difficult to attach much importance to this finding. An examination of correlations between marital status and premorbid adjustment did not reveal any significant association ( = 0.23, P = 0.1). Using a logistic regression analysis, neither the premorbid adjustment nor marital status had an independent effect on risk of relapse. Conversely, when using only the rating scale definition, relapse was associated with poor premorbid adjustment that had been stable from childhood. It is difficult to speculate why the 2 definitions correlate with different aspects of premorbid adjustment. The definition requiring change in treatment
www.LaRCP.ca

Using Symptom Rating-Based Relapse Criteria

We conducted additional analyses using only changes in SAPS scores (severity scores are 3 or more: hallucinations, delusions, bizarre behaviour, and formal thought disorder) to establish relapse without the criterion of change in medication. Hereafter, we refer to this group as SAPS relapse. This is in keeping with the definition of relapse used in several previous studies.1,2,46 Rate of and Time to Relapse. Among our adherent population of 65, 14 additional patients (21.5%) met the definition of SAPS relapse. The mean time to SAPS relapse was 15.6 weeks (SD 10.6 weeks). None of the following variables were associated with time to relapse: age, DUP, substance abuse, sex, and premorbid adjustment (Cox regression analysis was nonsignificant). A comparison of adherent patients who met this second definition (SAPS relapse) with those who did not, showed no significant differences in age of onset, sex, education, ethnicity, living circumstances, IQ, symptoms, time to
82 W La Revue canadienne de psychiatrie, vol 57, no 2, fvrier 2012

Medication-Adherent First-Episode Psychosis Patients Also Relapse: Why?

may identify patients whose relapse is more prolonged and (or) severe, and a deteriorating premorbid adjust ment may be a precursor to a relapsing course irrespective of medication. In either case, the lack of an independent effect of premorbid adjustment on risk of relapse would suggest that the impact of premorbid adjustment is influenced by other unknown factors. Definitions of relapse vary across the literature, with the most commonly used definitions relying exclusively on changes in ratings of positives symptoms.1,2,46 We find substance abuse to be strongly associated with a risk of relapse when using a definition based only on changes in ratings of positive symptoms. Note that we do not find this association with substance abuse with our first definition of relapse. This implies that clinicians may not consider patients to have relapsed if such relapse follows substance abuse because symptoms may not be sustained when patients reduce their substance use. Hence clinicians may wait for the effect of reduction in substance use by the patient before prescribing changes in medication. That factors influencing relapse depend implicitly on the definitions used for relapse suggests that the heterogeneity of findings in the literature may be due to the heterogeneity in definitions of relapse. Hence it would be useful to find a common empirical definition of relapse so that findings can be compared across studies. We suggest that rating scales with concomitant medication change in a definition more closely tied to clinicians evaluation of relapse. Among substances, cannabis use is very common in patients with psychotic disorders.28 Cannabis abuse had been associated with a 4-fold increased risk of relapse, independent of sex, expressed emotion, and age of onset of psychosis.29,30 Cannabis use has also been found to predict relapse after statistically controlling for medication adherence, subjective life stress, family environment, and DUP.31 In a comprehensive study,9 relapse rates were 3 times greater in FEP patients with substance abuse even after controlling for medication adherence, age, sex, and DUP. Cannabis abusers also had a shorter time to relapse into positive symptoms, and heavier abusers had an increased relapse risk.9 Note that these studies had used rating scales without concomitant medication change to establish relapse. To our knowledge, our study is the first to report the association of cannabis abuse on relapse in medication-adherent FEP patients. Our failure to find any effect of substance abuse on time to relapse using the Cox proportional hazards regression analysis may be related to the small number of patients in the relapse group. Our failure to find significant finding in both Cox regression analyses suggests that our study design would need to be replicated in a larger population of adherent patients. The strength of our report is that all results are based on well-characterized, previously untreated patients (patients had not received antipsychotics prior to entry to the treatment program for more than 30 days), treated in the
www.TheCJP.ca

only SEI service for all FEP available within a catchment area, and hence likely to be representative of patients with FEP in the community. Limitations of our study include relying on the presence of substance abuse (as per SCID-I/P interview) only at baseline. Therefore, no information is presented on substance use around the time of relapse. Further, information concerning the quantities of substances consumed was not collected systematically and the effect of the amount of cannabis consumption on the rate or risk of relapse could not be determined. In summary, in our medication-adherent population, we find a very low rate of relapse (14%) as defined by changes in SAPS scores, with concomitant medication changes. For this subsample of patients with FEP who were adherent to medication, this low rate should not come as a surprise. Premorbid adjustment was found to be a significant correlate of full clinical relapse, suggesting that patients with a deteriorating course are more likely to relapse, independent of their adherence to medication. As a result, more psychosocial interventions, with the aim of social reintegration, should be used to alleviate some of the deficits in patients with poor premorbid adjustment. Patients who showed only symptombased relapse, without any change in medication, were more frequently substance abusers, with an estimated 25-fold increase in risk of such relapse. This result suggests that more treatment needs to be directed toward reduction of substance abuse to reduce relapse rates.

Acknowledgements

Our study was part of a larger outcome study in FEP funded from several sources, including grants to Dr Malla from the Canadian Institute of Health Research and the Douglas Institute. Dr Malla is also funded through the Canada Research Chairs Program.

References

1. Robinson D, Woerner M, Alvir J, et al. Predictors of relapse following response from a first episode of schizophrenia or schizoaffective disorder. Arch Gen Psychiatry. 1999;56:241247. 2. Larsen TK, Moe LC, Vibe-Hansen L, et al. Premorbid functioning versus duration of untreated psychosis in 1 year outcome in first-episode psychosis. Schizophr Res. 2000;45:19. 3. Malla A, Norman R, Bechard-Evans L, et al. Factors influencing relapse during a 2-year follow-up of first-episode psychosis in a specialized early intervention service. Psychol Med. 2008;38:15851593. 4. Rabiner CJ, Wegner JT, Kane JM. Outcome study of 1st-episode psychosis. 1. Relapse rates after 1 year. Am J Psychiatry. 1986;143:11551158. 5. Ucok A, Polat A, Cakir S, et al. One year outcome in first episode schizophreniapredictors of relapse. Eur Arch Psy Clin N. 2006;256:3743. 6. Wiersma D, Nienhuis FJ, Slooff CJ, et al. Natural course of schizophrenic disorders: a 15-year followup of a Dutch incidence cohort. Schizophr Bull. 1998;24:7585. 7. Petersen L, Nordentoft M, Jeppesen P, et al. Improving 1-year outcome in first-episode psychosis: OPUS trial. Br J Psychiatry Suppl. 2005;48:S98S103. The Canadian Journal of Psychiatry, Vol 57, No 2, February 2012 W 83

Original Research
8. Leucht S, Barnes T, Kissling W, et al. Relapse prevention in schizophrenia with new-generation antipsychotics: a systematic review and exploratory meta-analysis of randomized, controlled trials. Am J Psychiatry. 2003;160:12091222. 9. Wade D, Harrigan S, Edwards J, et al. Substance misuse in firstepisode psychosis: 15-month prospective follow-up study. Br J Psychiatry. 2006;189:229234. 10. First M, Spitzer R, Gibbon M, et al. Stuctured Clinical Interview for DSM-IV-TR axis 1 disorders, research version, patient edition. New York (NY): Biometrics Research, New York State Psychiatric Institute; 2002. 11. Cassidy C, Rabinovitch M, Joober R, et al. 117Is any one method of measuring medication adherence in first-episode psychosis better than others? Schizophr Res. 2008;98(Suppl 1):81. 12. Cassidy C, Rabinovitch M, Schmitz N, et al. A comparison study of multiple measures of adherence to antipsychotic medications in first episode psychosis. J Clin Psychopharmacol. 2010;30:6467. 13. Andreasen N. The Scale for the Assessment of Positive Symptoms (SAPS). Iowa City (IA): University of Iowa; 1984. 14. Andreasen N. The Scale for the Assessment of Negative Symptoms (SANS): conceptual and theoretical foundations. Br J Psychiatry Suppl. 1989;7:S49S58. 15. Craig TK, Garety P, Power P, et al. The Lambeth Early Onset (LEO) Team: randomised controlled trial of the effectiveness of specialised care for early psychosis. BMJ. 2004;329(7474):1067. 16. Linszen DH, Dingemans PM, Lenior ME, et al. Relapse criteria in schizophrenic disorders: different perspectives. Psychiatry Res. 1994;54:273281. 17. Csernansky J, Mahmoud R, Brenner R. A comparison of risperidone and haloperidol for the prevention of relapse in patients with schizophrenia. N Engl J Med. 2002;346:1622. 18. van Mastrigt S, Addington J. Assessment of premorbid function in first-episode schizophrenia: modifications to the Premorbid Adjustment Scale. J Psychiatry Neurosci. 2002;27:92101. 19. Cannon-Spoor H, Potkin S, Wyatt R. Measurement of premorbid adjustment in chronic schizophrenia. Schizophr Bull. 1982;8:470484. 20. Haas GL, Sweeney JA. Premorbid and onset features of first-episode schizophrenia. Schizophr Bull. 1992;18:373386. 21. Larsen T, McGlashan T, Moe L. First-episode schizophrenia: I. Early course parameters. Schizophr Bull. 1996;22:241256. 22. Larsen T, Friis S, Haahr U, et al. Premorbid adjustment in first-episode non-affective psychosis: distinct patterns of pre-onset course. Br J Psychiatry. 2004;185:108115. 23. Malla A, Norman R, Schmitz N, et al. Predictors of rate and time to remission in first-episode psychosis: a two-year outcome study. Psychol Med. 2006;36:649658. 24. Norman R, Malla A, Verdi M, et al. Understanding delay in treatment for first-episode psychosis. Psychol Med. 2004;34:255266. 25. Malla AK, Norman RMG, Manchanda R, et al. One year outcome in first episode psychosis: influence of DUP and other predictors. Schizophr Res. 2002;54:231242. 26. SPSS Inc. SPSS, version 15.0 [software]. Chicago (IL): SPSS Inc; 2006. 27. Malla A, Payne J. First-episode psychosis: psychopathology, quality of life, and functional outcome. Schizophr Bull. 2005;31:650671. 28. Regier D, Farmer M, Rae D, 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;264:25112518. 29. Linszen D, Dingemans P, Nugter M, et al. Patient attributes and expressed emotion as risk factors for psychotic relapse. Schizophr Bull. 1997;23:119130. 30. Linszen DH, Dingemans PM, Lenior ME. Cannabis abuse and the course of recent-onset schizophrenic disorders. Arch Gen Psychiatry. 1994;51:273279. 31. Hides L, Dawe S, Kavanagh D, et al. Psychotic symptom and cannabis relapse in recent-onset psychosis. Br J Psychiatry. 2006;189:137143.

84 W La Revue canadienne de psychiatrie, vol 57, no 2, fvrier 2012

www.LaRCP.ca