Schizophrenia Bulletin vol. 39 no. 6 pp.

1242–1251, 2013
doi:10.1093/schbul/sbt138
Advance Access publication September 26, 2013

The Effect of Motivational Interviewing on Medication Adherence and
Hospitalization Rates in Nonadherent Patients with Multi-Episode Schizophrenia

Emile Barkhof*,1, Carin J. Meijer1, Leo M. J. de Sonneville2, Don H. Linszen3, and Lieuwe de Haan1

Downloaded from http://schizophreniabulletin.oxfordjournals.org/ at UNAM Direccion General de Bibliotecas on April 30, 2015
1
Department of Psychiatry, Academic Medical Centre, Amsterdam, The Netherlands; 2Department of Clinical Child and Adolescent
Studies, Leiden University, Leiden, The Netherlands; 3Department of Psychiatry, Academic Hospital Maastricht, Maastricht,
The Netherlands
*To whom correspondence should be addressed; Department of Psychiatry, Academic Medical Centre, Meibergdreef 5, Amsterdam 1105
AZ, The Netherlands; tel: 31-20-8913500, fax: 31-20-8913702, e-mail: e.barkhof@amc.uva.nl

Background: Medication nonadherence in patients with Introduction
schizophrenia presents a serious clinical problem.
Nonadherence to antipsychotic medication is highly
Research on interventions incorporating motivational
prevalent in patients with schizophrenia1 and is associated
interviewing (MI) to improve adherence have shown
with sharply increased readmission rates, more aggressive
mixed results. Aims: Primary aim is to determine the
incidents, more suicides, significant emotional and social
effectiveness of a MI intervention on adherence and hos-
burden for patients and their families, and higher finan-
pitalization rates in patients, with multi-episode schizo-
cial costs.2–4 In the last decades, several interventions have
phrenia or schizoaffective disorder, who have experienced
been developed to improve adherence rates.5 Recent treat-
a psychotic relapse following medication nonadherence.
ment recommendations promote focusing on specific tar-
Secondary aim is to evaluate whether MI is more effec-
gets that may contribute to nonadherence.6 Motivational
tive in specific subgroups. Methods: We performed a
interviewing (MI) is a client-centered, directive method
randomized controlled study including 114 patients who
for enhancing intrinsic motivation to change by exploring
experienced a psychotic relapse due to medication nonad-
and resolving ambivalence.7 MI was originally designed
herence in the past year. Participants received an adapted
as a therapeutic approach to treat abuse of alcohol and
form of MI or an active control intervention, health edu-
other substances, for which it proved to be very effec-
cation (HE). Both interventions consisted of 5–8 ses-
tive.8,9 Following positive results in other health care
sions, which patients received in adjunction to the care
domains with regard to behavior change,10–12 effectiveness
as usual. Patients were assessed at baseline and at 6 and
of (adapted) MI for improving medication adherence
12 months follow-up. Results: Our results show that MI
has also been studied in patients with psychotic disor-
did not improve medication adherence in previously non-
ders.5,13 Compliance therapy, an intervention based on
adherent patients who experienced a psychotic relapse.
MI and other cognitive approaches, showed substantial
Neither were there significant differences in hospitaliza-
improvements in medication adherence in patients with
tion rates at follow-up between MI and HE (27% vs 40%,
schizophrenia,14,15 although this could not be replicated
P = .187). However, MI resulted in reduced hospitaliza-
in another trial.16 Studies focusing on a similar interven-
tion rates for female patients (9% vs 63%, P = .041), non-
tion (“adherence therapy”), also containing MI, yielded
cannabis users (20% vs 53%, P = .041), younger patients
mixed results.17,18
(14% vs 50%, P = .012), and patients with shorter illness
More recently, an individually tailored approach incor-
duration (14% vs 42%, P = .040). Conclusions: Targeted
porating MI proved to be effective in prompting service
use of MI may be of benefit for improving medication
engagement and medication adherence.19
adherence in certain groups of patients, although this
This means that to date evidence concerning the effec-
needs further examination.
tiveness of MI as a means to improve medication adher-
ence in patients with schizophrenia is still inconclusive.
Key words:  RCT/intervention study/medication The present study was designed to investigate the
adherence/motivational interviewing/schizophrenia effect of MI on adherence to antipsychotic medication

© The Author 2013. Published by Oxford University Press on behalf of the Maryland Psychiatric Research Center. All rights reserved.
For permissions, please email: journals.permissions@oup.com

1242

) comprised 4 phases. 2015 to investigate whether specific subgroups of patients may active stance of the therapist.  MI is a client-centered. cerning medication) as well as cognitive deficits such as sible explanation for the unequivocal results on the effect attention problems.20 The cur. and these tapes were used in monthly Interventions supervised 4-hour sessions for all therapists. mize skills and minimize contamination effects between lectual dysfunction (intelligence quotient <70). As there sion length. greater flexibility in ses- benefit more from this intervention than others.” In the next phase. and community adequate mastery of the Dutch language and be able and mental health nurses. cies between the present behavior and the patient’s own cation nonadherence. exploring the patient’s own personal goals. Therapy sessions were audiotaped if the patient con- sented to this. both following Procedure nonadherence to the antipsychotic treatment. who were trained by a professional prepared to give written informed consent. The overall goal is to increase patients generally have poor social functioning and an the patient’s intrinsic motivation for change.). intervention–MI–or the control group–HE–by means of antipsychotic treatment had to be resumed with at least a computerized cluster randomization program. a manual was their medication adherence is urgently needed. the of MI on adherence so far.1 we cational information. with a multi-episode schizophrenia spectrum selectively reinforcing “change talk. and route (available upon request from E. incorporating some adaptations directed at rent study explicitly aimed to include this severely dis. to ensure treat- directive method. food. ongoing interventions were discussed. HE comprised indi- rioration with nonadherence as a probable cause. or mini. The phases involved introduction and engagement. along with gender and age. Next. physical exercise etc. treatment of psychotic disorders. illness duration. by the patient. ing a block of codes for every 6 consecutive inclusions. Participants had experienced a recent (<1 y) psychotic relapse and/or a clinical deterioration. The last phase was committed to evaluation selected from inpatient and outpatient facilities for the and consolidation of the motivation to change. and a more active provision of psychoedu- are several factors associated with nonadherence. who are often either excluded symptoms. much improved. Therapists criteria: an age of 18–65  years.oxfordjournals. These so-called “revolving door” future goals are amplified. a second aim of the study was adapted form therefore encompasses a somewhat more Downloaded from http://schizophreniabulletin. interventions according to the treatment manuals and on It explores personal ideas and ambivalences. trainer on MI (eight 4-h sessions) as well as on HE (two teria were an organic disease with a possible etiological 4-h sessions) according to a structured manual.21 was avoided. consecutive order. a psychiatrist. Effects of Motivational Interviewing on Adherence in patients. specific positive symptoms (delusions con- from studies or are unwilling to participate. These sessions were focused on delivering strategically directed conversations about their problems.7 unfavorable prognosis. eliciting and preventing contamination of MI to HE. though not in a strict may mediate effectiveness of MI. produc- some symptomatic improvement. 1243 . Participants were reinforced. and Methods the “readiness for change.B. vidual lectures on general health topics (such as healthy potential participants were invited to receive informa. or 3 (very much improved.  The control group was provided patients that showed a psychotic relapse or a clinical dete. in which the Motivational Interviewing. who were hospitalized or unstable due to medi. and therefore improvement of Based on the basic principles of MI. severity of illness.” by which discrepan- disorder. interventions. Exclusion cri. These were used as a framework. which were one by one revealed by the coordinating mal improved) on the Clinical Global Impression Scale researcher.22 Participants were required to have an tions were psychologists. exploring attitudes and beliefs toward treat- ment. which favorable attitudes and beliefs toward change were tions in the greater Amsterdam area. designed.org/ at UNAM Direccion General de Bibliotecas on April 30. Participants were asked to tion on the study and to verify the following inclusion choose one of these topics for each session. Health Education (HE) sessions. informa- Participants tion was provided and ambivalences were amplified along The study was conducted in 3 mental health care institu. through which patients are engaged in ment fidelity. a clinical diagnosis of delivering HE used an active and didactic attitude with schizophrenia or a schizoaffective disorder confirmed by specific scripts to ensure that discussing treatment issues the Structured Clinical Interview for DSM disorders. such as negative turbed group of patients. 2. defined as score of 1.22 Subsequently. to maxi- relation to the psychotic disorder and/or a severe intel. MI according to the intervention manual herence such as cannabis use. Clinicians of the par- ticipating facilities regularly reviewed their caseloads for Health Education. specific problems of the study group. of medication administration. In contrast to the original MI. The therapists who performed both interven- for improvement. To find a pos. resulting in hospitalization and/or a change on the Clinical Global Participants were allocated to either the experimental Impression Scale. though only after explicit consent aimed to explore whether known risk factors for nonad.

medication administration route. Demographics. of admissions. With an estimated assessments were stored separately. higher sample t tests were performed with the dichotomous scores indicating more belief in the personal benefit of subgroups of the variables gender. Next to the interventions. Therapists were not otherwise Psychopathology  Severity of psychopathology was involved in the treatment of participants. Furthermore. For this purpose.5 with a SD of 1. and illness duration as additional grouping Sociodemographic and Service Receipt Inventory. mixed between-within subjects The LCS yields reliable ratings of the long-term course of ANOVA were performed with PANSS scores as depen- schizophrenia. Amsterdam. depending on the attention span of the par. in both the MI and the the medication. age.27 Assessors were trained on the PANSS Downloaded from http://schizophreniabulletin. medication adminis- prescribed medication were assessed with the Client tration route. The study was approved by the Statistical Analysis Medical Ethical Committee of the Academic Medical To evaluate the effect of randomization on baseline Centre. a sample size of appointed in the same facility. Data on interventions and 50 patients in each group was needed. The session duration varied between 20 and patients had been hospitalized.0. higher scores indicating higher levels of adherence. with a score range of 0–10. Cannabis Abuse  Concomitant cannabis misuse was ment (T0) was performed. Patients were measured with the Positive and Negative Syndrome told they would be allocated to 1 of 2 active interventions Scale (PANSS). Before starting the intervention. Hospitalization  Hospitalization practical barriers. chi-square tests were performed. attrition rate of 20%.24 covering the using the baseline values as covariates to adjust for pre- participant’s report on medication adherence. To assess the influence of age. The DAI showed a good internal consis- HE group.05.5 and the assessments were never performed by researchers with a 2-sided significance level of . with a mini. To achieve a power of 80% of to different researchers. gen- range from 0 to 4. was assessed by the LCS. To assess differences Assessments in hospitalization rates. and levels of adherence. with higher scores indicating higher der. All assessments were per- formed by trained psychologists and psychiatrists. scores intervention scores. these variables cation adherence as judged by caregivers and treating were separately entered as a second fixed factor in 2-way physicians were measured by the 5-point adherence item between-groups ANOVA. discussed. The DAI is a self-report on severity of psychopathology. age. adherence was assessed with the medication adherence The main effects of interventions on adherence mea- questionnaire (MAQ). When the therapist judged there were problems to keep patients engaged in the interventions or in case of Secondary Outcomes. separate independent 10-item questionnaire. participants received care between assessors.23 variables. which was further facilitated by reg- as usual. again after the intervention was completed (T1) and after 6 months follow-up (T2). fewer sessions were given. Less than 5 sessions was counted as dichotomous variable was constructed indicating whether a dropout. regardless of the number 45 minutes.org/ at UNAM Direccion General de Bibliotecas on April 30. demographic and disease-specific parameters. a mum of 5 sessions. 1244 . the continuous variables age and illness duration were transformed to a dichotomous Primary Outcome. using original training videos to maximize concordance tion. of the life chart schedule (LCS).25 Based on the LCS data. with scores ranging from To evaluate the effect of the interventions on the sever- 1 to 5. cannabis use. consisting of functional assertive community ular supervised meetings in which videotaped assess- treatment for patients treated on an outpatient basis and ments were scored and inconsistencies between assessors routine clinical care for hospitalized patients. t tests were performed for the continuous variables and chi-square tests for the categorical variables. Attitudes toward medication were assessed with the To detect possible differences between subgroups Drug Attitude Inventory (DAI). Medication Adherence  Medication variable using the median value.26 or HE. Participants were interviewed assessed by means of urine analysis at baseline. illness duration on adherence outcomes. Participants received a travel expenses compensation of 5 euro for each intervention session or assessment in which they participated.oxfordjournals. levels of internal validity and reliability.  Information on demographic data and including gender. ticipant during a session. ity of psychopathology. cannabis use. a baseline assess. patients were offered 8 sessions of either MI tency and validity. a 4-item questionnaire with good sures were assessed using 1-way between-groups ANOVA. 2015 and were not told which was the experimental condi. We aimed to find a difference on the adherence measures a coordinator assigned the interventions and assessments of 0. Barkhof et al Within a period of 26 weeks.E.25 dant variables. cannabis use. who Sample Size were masked to which condition a patient was allocated. we aimed to include 120 patients. data concerning medi. ensuring that the interventions detecting such difference with a medium effect size of 0.

The toward medication (see table 2). Of these. P = .7) than those who completed interventions (M = 36.35. which was not a significant difference (P = . we were unable to perform an intention to treat analysis and decided to perform a per-protocol Interventions analysis instead. depending on the analysis. The mean number of sessions in the Results MI group was 6.2) vs 6. group that dropped out was younger (M  =  30. Nevertheless.6 (SD = 0. Four hundred and three patients were referred for par- ticipation in the study. 10 in measures.  Consort status.8. φ  =  0. 1245 . Of the 186 remaining patients.072.8. with regard to the effect of MI and HE on adherence. df = 112. Cohen’s d = 0. 2015 Not all participants received the full 8 sessions of the interventions. Downloaded from http://schizophreniabulletin. lost to follow-up. Cohen’s d.oxfordjournals. 72 refused Adherence to participate. we examined the differences between subgroups 11. and medication administration route SD = 9. differences between MI and HE on the two adherence During the intervention. 226 did not meet inclu- sion criteria. 18 patients dropped out. SD = Next. there were no differences in attitudes the MI group and 8 in the HE group (see figure 1).9) in Participants the HE group.44) as independent variables and PANSS scores as dependent and showed a higher use of cannabis use at trend level variables.020. or phi coefficient.2% vs 10. Likewise. t = −2.22). 1. Demographic characteristics and disease-specific Because the majority of patients who dropped out of parameters at baseline did not significantly differ between the intervention-refused follow-up assessments or were conditions (see table 1). Effect sizes of all significant results are expressed these characteristics were equally distributed over both in (partial) eta squared. Patients referred N = 403 Patients meeting criteria N = 186 Refused to participate N = 72 Informed consent + randomised N = 114 Experimental Intervention Control Intervention (Motivational Interviewing) (Health Education) N = 55 N = 59 MI MI HE HE completed drop-out completed drop-out N = 45 N = 10 N = 51 N=8 Fig.4%. One hundred and fourteen patients were At both follow-up assessments. conditions. Effects of Motivational Interviewing on Adherence duration of illness. there were no significant randomized and allocated to the two treatment arms. (22.60).org/ at UNAM Direccion General de Bibliotecas on April 30.7. P  =  .4 (SD = 1.

8 (6. there were no significant interactions between on the MAQ and the LCS when they received MI. n (%) 91 (80 %) 43 (78%) 48 (81%) . b MAQ.03 2.2 (15.3) Downloaded from http://schizophreniabulletin.13 1.34 0. between the DAI score and age group: F(1.36 0.00 1.32 0.14  Oral 85 (75%) 44 (80%) 41 (70%)  Depot 29 (25%) 11 (20%) 18 (30%) PANSS total.97 1. 2015   Schizoaffective disorder.87 Duration of illness. mean (SD) 35.71 admissions.9) 71.2) 3. suggesting that patients younger than cation administration: F(1.9) .24   T1 (posttreatment) 3. P = .  Effects of Interventions on Adherence Rates Motivational Interviewing (n = 30) Health Education (n = 32) Mean SD Mean SD Pa MAQb   T0 (baseline) 3. At T2.4) 45 (76.36.93.64 2.8 (5. n (%) 53 (47%) 26 (47%) 27 (46%) .9 (10.13 1. score ranges 1–5. there pared with HE. tion at 6 months follow-up when they received MI.12 . score ranges 0–10.7 (1.67 2.67 Ethnicity: white European.38 1.oxfordjournals.  Baseline Characteristics Total (n = 114) Motivational Interviewing (n = 55) Health Education (n = 59) Pa Age.17 0. 35 years showed more favorable attitudes toward medica- These results suggest that patients using depot medica.32 Sex: male.18 6.34   T2 (6-mo follow-up) 2.14. At  T1.89 2.50 6. b Available N = 59.42 3.49)  =  3. com- any of the covariates and intervention type. η2 = 0. n (%) 27 (23.11 .99 3.4) . n (%) .64 Cannabis-positive urine sample.39 6. Drug Attitude Inventory.52 .07. n (%) 14b (24%) 5 (16%) 9 (33%) .4) 6.53. P = .6 (5.0 (3.009.3) 37 (1.83 Note: aUnivariate analysis with baseline value as covariate. com- tion show higher adherence rates at 6 months follow-up pared with HE.21 DAIc   T0 (baseline) 6.0) .8) .71 .6) 8. Positive and Negative Syndrome Scale. a Chi-square tests for dichotomous variables.8 (6. t tests for continuous variables. mean (SD) Diagnosis .48   First-generation antipsychotic 38 (33%) 17 (31%) 21 (36%)   Second-generation antipsychotic 70 (62%) 36 (65%) 34 (58%) Other (mood stabilizer) 6 (5%) 2 (4%) 4 (6%) Medication administration route. life chart schedule.99  Schizophrenia.3) 4. mean (SD) 71.8 (18.45 0. 1246 .9) 70.34 3.11 Note: PANSS.7) Medication at baseline.59) = 4.86 2. η2 = 0. medication adherence questionnaire.05.81 .31 0.70 LCS adherence score by physician and/or caregiverd   T0 (baseline) 4.6) 14 (23. Table 2.76 4.9 (20. η2 = 0.30   T1 (posttreatment) 6.07. n (%) . and the LCS score and route of medi. d LCS.org/ at UNAM Direccion General de Bibliotecas on April 30. P = . n (%) 87 (76.4) 34.82 4. years (SD) 7.3) 42 (76.7) 13 (23.E.45) = 7.33 0.11 Number of psychiatric 3. c DAI. was a significant interaction between the MAQ score Furthermore.96   T1 (posttreatment) 4. score ranges 0–4.84 4.72   T2 (6-mo follow-up) 6.74   T2 (6-mo follow-up) 4. Barkhof et al Table 1.98 .8 (4.037. there was a trend level interaction and route of medication administration: F(1.38 1.

041. t (11) = −2.48).20 . route there were no significant differences between MI and HE of medication administration) may benefit more from MI on hospitalization rates in the group of patients on oral than others.48).49 .963 T1: posttreatment (n = 94) 17/45 (38%) 19/49 (39%) 0. we did not find significant differences in Among patients with an illness duration shorter than adherence over time between MI and HE.99).27) nor in the group on depot Besides a small difference in adherence rates favoring medication (n = 25.org/ at UNAM Direccion General de Bibliotecas on April 30. df  =  1. For the other (χ2  =  6.002 . 6 years (n = 41).273 Note: HE.83). in both groups. df  =  1. In the Focusing on potential treatment effects in terms of group >35 years (n = 50). 14% of patients were hospitalized in the As previous studies incorporating MI in their inter- MI condition vs 42% in the control condition (χ2 = 4. might have been of influence. no significant multi-episode schizophrenia that were unstable due to differences were found (P = . φ  =  −0.21. but there were no differences (P = . In the group of patients younger than 35  years on PANSS subscales for positive symptoms (P  =  .484 • Age ≤35 (n = 43) 3/21 (14%) 11/22 (50%) 6. pitalization rates.06 .187 • Females (n = 19) 1/11 (9%) 5/8 (63%) 6. φ  =  −0. we performed analyses concerning specific sub. medication (n = 68. 27% of patients in the MI tion in severity of psychopathology (F(2.12 .012. In the group with a longer that the heterogeneity of the investigated patient groups illness duration (n = 52).921 T2: 6-mo follow-up (n = 93) 12/45 (27%) 19/48 (40%) 1. 2015 follow-up period in the MI group vs 50% in the control (P = .52). P = . age.01 .09]). 1247 . there were no significant differ.4).38).  Effects of Interventions on Hospitalization Rates MI (Ratio Hospitalized) HE (Ratio Hospitalized) χ2 P T0: baseline (n = 114) 24/55 (44%) 26/59 (44%) 0.012 • Age >35 (n = 50) 9/24 (38%) 8/26 (31%) 0. P  =  .475 • Cannabis negative (n = 40) 5/25 (20%) 8/15 (53%) 4. fering on cannabis use.14–19 we hypothesized df = 1.21 .041. or general symptoms Downloaded from http://schizophreniabulletin.62).12.040 • Duration of illness >6 y (n = 52) 9/23 (39%) 11/29 (38%) 0. group (Δ 7.35).041 • Males (n = 74) 11/34 (32%) 14/40 (35%) 0.75. health education. φ = −0. There was was a nonsignificant difference between the two groups.84 . with both groups showing a reduc- At T2 (6 months follow-up). MI. we therefore ences between conditions (P = . P  =  . Regarding adherence and hos- patients (n  =  43) that refused urine analysis (P  =  .041 • Urine analysis refused (n = 43) 7/17 (41%) 8/26 (31%) 0.380 • Oral antipsychotic (n = 68) 10/35 (29%) 12/33 (36%) 1. In the group with a negative urine analysis on cannabis Discussion use at baseline (n = 40).93). female patients showed ences between conditions (P = . there were no significant differ.01 .33). There were also no differences between interventions groups.810 • Cannabis positive (n = 10) 0/3 (0%) 3/7 (43%) 1. motivational interviewing.38). SD  =  4.005 [partial η2 = 0. vention have shown mixed results. just as in the group of medication nonadherence. φ  =  −0.24.4.oxfordjournals. between the two interventions (P = .110)  =  5. P  =  . P = . 44% of patients were Total PANSS scores showed no significant interaction hospitalized at baseline. a large effect for time.59. gender. group were hospitalized vs 40% in the control group P = .035. illness duration.75 . group (χ2  =  6.930 • Depot antipsychotic (n = 25) 2/10 (20%) 7/15 (47%) 1. there between intervention type and time (P = . Effects of Motivational Interviewing on Adherence Hospitalization Rates Psychopathology As shown in table 3. investigated whether specific subgroups of patients (dif- Regarding the route of administration of medication. 20% of patients were hospital- ized in the MI condition vs 53% in the control condition This study aimed to investigate the effect of an adapted (χ2  =  4. df  =  1.40.9. Among those form of MI on medication adherence in patients with with a positive test on cannabis (n = 10). symptoms for specific subgroups. P = . between conditions (P = .616 • Duration of illness ≤6 y (n = 41) 3/22 (14%) 8/19 (42%) 4. Among male variables.74 .32 .87) (see table 4).040.0) compared with the HE group tion was hospitalized vs 63% in the control condition (Δ 3.81). P = . (n  =  43). a larger decrease in general PANSS symptoms in the MI Nine percent of female patients in the MI condi. 14% of patients were hospitalized during the negative symptoms (P  =  .24 . In addition.25 .19). there were no significant effects on changes in patients (n  =  74).57). Next. At T1 (postintervention). SD = 2. there were no significant differences PANSS scores.68). MI over HE in the group who used depot medication on Table 3.

is younger group nonadherence was more prevalent.0 6.4 12. found to be lower for patients younger than the median As hospitalization rates do not reveal the total time age of 35  years and patients with shorter illness dura. including pitalization rates cannot be explained by cannabis as a cannabis abuse.9   T2 (6-mo follow-up) 64. In most observational studies.2 7.4 6.0 30. d General subscale score ranges 16–112.30 confounder. spent in hospital. The fact that but that measures used in this study are not reliable or there were gender-specific treatment effects of MI result. In our study. b.33 32. 2015   T2 (6-mo follow-up) 15.2 16.5 16. and that for patients using can. MI abuse in female patients with schizophrenia in general.3 A possible explanation for this finding may be a suitable tool for improving adherence for younger be that there actually was an improvement in adherence.7 5. no Also.74 35.53   T2 (6-mo follow-up) 16. in our study used cannabis and there is less cannabis nabis specific additional interventions are warranted.2 5. A  younger age and a shorter duration of illness tion of hospitalization between interventions.1 showed virtually the same differences between the sub- Probably related to this.89 Note: Abbreviations are explained in the first footnote to tables 1 and 3.9 72.5   T1 (posttreatment) 65.0 63.83 16.84 15.2 was less prevalent. This may hospital interferes social functioning and hospitalization imply that although nonadherence in the older group is associated with increased costs. sensitive enough.28 in our study.80 19.8 8.6 22. hospitalization rates in the MI condition were differences between interventions were found.29.1.0 5.2 6. MI produced significant tive and residual symptoms. no other mediators of treatment surprising.32 PANSS-negative symptomsc   T0 (baseline) 18. which are strong risk factors for medication nonadherence.98 32. because the time spent in equally distributed over the 2 interventions. a Total score ranges 30–210.28   T1 (posttreatment) 35. adherence and attitudes ing in reduced hospitalization rates in female patients is toward medication were measured using the MAQ and 1248 .31 Female However.0 17. While cannabis use is able course of illness.0 17.66 PANSS general symptomsd   T0 (baseline) 37. while in the hospitalization rates.87 17.c Positive and negative subscales score ranges 7–49. we compared our results with the dura- tion.2 7.  Psychopathology Rates Across Interventions MI (n = 30) HE (n = 32) Mean SD Mean SD PANSS total scorea   T0 (baseline) 72.org/ at UNAM Direccion General de Bibliotecas on April 30. probably due to the fact that there strongly associated with nonadherence in patients with is a later onset of illness and they show less severe nega- schizophrenia.31 17.1 6. it may have been more persistent and The fact that we found differences in subgroups on less susceptible for MI.3 6.2 6. we did find favorable effects of MI with patients with schizophrenia generally show a more favor- regard to hospitalization rates. we found that the group who groups of patients (see online supplementary material). but not in adherence measures. we have lower hospitalization rates for the group who does not found no correlations with severity of psychopathology use cannabis.oxfordjournals.11.E. found to be a risk factor for nonadherence.33 itself may be of value in this respect. Although none of the female patients who do not use cannabis. Barkhof et al Table 4. patients who can be engaged in treatment. On the other hand.3 66.56   T1 (posttreatment) 16.69   T1 (posttreatment) 15.05 Downloaded from http://schizophreniabulletin. the MAQ and the LCS.1. but this difference was This is an important outcome.9 6. This may imply that and illness duration between the female responders and MI is more suitable for improving adherence in patients nonresponders.7 8.29 15.7 PANSS-positive symptomsb   T0 (baseline) 16.07   T2 (6-mo follow-up) 32. it remarkable because adherence is strongly associated with may also have been more amendable.32 Nevertheless. because in the male non-cannabis users.7 9. because it has shown the gender difference in the effectiveness of MI on hos- efficacy in the treatment of substance abuse.1 14. suggesting MI to relapse rates. compared with HE.2 7. gender is not effects in terms of medication adherence were found. dropped out were aged younger.

36 thought to be reliable. this study shows that an adapted form Strengths and Limitations of MI does not produce a significant effect on medica- Limitations of the present study are that subgroups tion adherence or hospitalization. even when subgroups were ana. Other measures that may provide Fourthly. Effects of Motivational Interviewing on Adherence the DAI. and the LCS. because of the did not find changes in psychopathology even though chronicity of their illness or concomitant drug abuse. it can. Also there were ing. our finding that patients on depot effect sizes in addition to significance levels may provide medication show higher adherence rates as assessed with sufficient information for a correct interpretation of the the LCS when they received MI compared with HE. preventing rehospitalization. influenced by the amount of attention that was given in judgemental attitude of caregiver toward the beliefs and the intervention condition. who had recently been nonadherent. so we decided that extra barriers Finally. resulting in diminished reliable and valid outcome measure to assess medication power. ment for additional support of the needs for functioning Moreover. consisting of unstable patients with often ness of patients to participate in research. 1249 . so this will not likely have influenced the not be ruled out that the observed reduction in hospital. Another problem with in the MI group and 14% in the control group) was con- the adherence measures in our study is that a ceiling siderable. This used both subjective (adherence scores) and objective may result in a more open and trustful relation between (hospitalization) outcome measures. the level participate is high (39% of the identified sample). almost the same as in comparable studies. istics of patients for whom MI may be a suitable interven- tion to improve adherence. These focus of our study was to include the most troublesome factors may result in a considerable reluctance to cooper- patients we realized that the willingness to participate ate with a research trial. results.oxfordjournals.19 This high ing devices. method is applied. findings concerning schizophrenia. have been left undetected. the latter represent- the patient and the caregiver.17.6. Nevertheless. hospitalization may be regarded as a more some patients lost to follow-up. between groups. female patients. the proportion of patients that refused to more reliable information on adherence are. This may especially apply to In our study. Furthermore. the disorder in the included patients. but is of medication in blood samples or medication-monitor.15. younger patients. the control group received an active inter- tion rates were found. Although often the Bonferroni patients on depot medication can be regarded as an excep.34 this may have influenced the results. in our analyses no corrections were applied fessional caregiver in the LCS data in the subgroup of for multiple comparisons.19. of which we used information on pists who were not otherwise involved in the treatment adherence by the caregiver and/or a relative. both MI and HE were performed by thera- patients. 2015 tion to this. improvements in medication adherence or hospitaliza. the attrition rates did not differ adherence in this population. which participate37 or excluded from studies. Nevertheless. allowing for a better judge. population. ing a reliable and valid measure of outcome. Because the poor insight. clinical relevant effects may adherence alone. but there were almost no differences of the MI condition (27% vs 40%). although not unexpected given the severity of effect of the LCS and the MAQ might have been operat. thereby ensuring that the effects are not these observations might be that MI promotes a non. would be problematic. In conclusion. the results provide indications subgroups are preliminary and require confirmation in a that MI may yet be suitable for improving adherence in future randomized controlled trial. the latter is expensive and the use refusal rate (partly) reflects the challenging nature of the of invasive measures may negatively influence the willing. On the other hand. This choice for external therapists renders uncer- only rely on the patient as the sole information source. Therefore. Therefore. to ensure treatment integrity of the interven- that we have used multiple instruments and did not tions. Therefore. psychopathology rates improved for both the considerable difference in hospitalization rate in favor groups over time. several authors suggest that all subjective the treating caregiver(s) might have conducted other reports concerning medication adherence are considered interventions or therapeutic styles and to which extend to be relatively unreliable for determining adherence.org/ at UNAM Direccion General de Bibliotecas on April 30. This means of patients. Strength of this study is that it was designed to include lyzed.35 A  possible explanation for vention (HE). between MI and HE. others have argued that the report of Downloaded from http://schizophreniabulletin. Another strength is that we actions of a patient with regard to medication use. compared with HE were relatively small and patients were not randomized in a group of nonadherent patients with multi-episode on these characteristics. is results. which both rely on subjective information of Secondly. non-cannabis users. However. the information obtained from the pro. because of the limited power due to a ization rates is caused by other factors than medication relatively small sample size. However. However. ence. this is consis. the dropout rate during the interventions (18% to inclusion were to be avoided. tainty whether in the meantime during the study period. eg. Although this may appear incongruent with the those patients most in need for improvement of adher- differences found in hospitalization rates. we were able to identify possible character- in society. However. These troublesome patients are often unwilling to tent with other studies on adherence interventions. Thirdly.

van Wijngaarden B. Kirov G. J Clin Nurs. Mojtabai R. Awad AG.189:508–514. Gray R. New York. NY: Guilford Press. Green LW. Velligan DI. and psychosocial treatment of nonadherence in schizophrenia. 13. Gibbon M. Kay SR.63:892–909. Centre (Amsterdam) and at the mental health institutions 17. Eur Psychiatry. Leese M. 2015 Supplementary material is available at http://schizophre motivational interviewing: a practice-friendly review of four niabulletin. Maneesakorn S. 18-month Acknowledgments follow-up. 1997. illness. The expert consensus guideline series: adher. Knapp M. European multicentre randomised controlled trial. 2006.172:413–419. Jeste DV. 1986.E.70(suppl 4):1–46. Weiden PJ.197:448–455. course of schizophrenia: development of a discrete event sim. people with schizophrenia. 16. Susser E. 2009. Anja Lek.16:1302–1312.13:177–183. Leckband SG. Psychol Med. phrenia or schizoaffective disorder. Miller WR.oxfordjournals. Structured 1. Moyers TB. Addiction. The authors have declared that there are no adherence therapy in people with psychotic disorders: ran- conflicts of interest in relation to the subject of this study. Epidemiology. Thailand. Tonigan JS. vidualized approach for each patient of this challeng. Robinson D. David A. ulation model. medication in patients with schizophrenia–a review of the 25. David A. Arch Gen Psychiatry.71:843–861. Spitzer RL. Cost of relapse in schizophrenia. Dick Mai. Williams JBW.56:241–247.67(suppl 5):3–8. Haarlem) and Arkin (Amsterdam). Rollnick S.24:67–74. 23. Client Socio-Demographic and Service 1999.2:207–213. J Subst Misuse. 2007. Br J Psychiatry. Heeg B. Panel on Adherence Problems in Serious and Persistent 26. et  al.319:828–832. Chinn D. 2006. Furthermore. Interventions to improve adherence to antipsychotic adherence. et  al. Hogan TP. et al.13:261–276. et al. 2012. Therefore. de Haan L. de Sonneville LM. Brief interventions for targeted use of MI may be of benefit for improving alcohol problems: a review. Burke BL. ECDEU Assessment Manual for Psychopharmacology. 27. Noonan WC. Adherence therapy for InGeest (Amsterdam. 10. 20. clinical consequences.23(suppl 1):17–33. J References Clin Psychiatry.European Version: development of an 4. 1995. 11. Applewhaite G. Prevalence of and risk factors for medication nonadherence NY: Biometrics Research Department. 1996.oxfordjournals. People for Change. relapse following response from a first episode of schizo. Frank van der Horst.org/ at UNAM Direccion General de Bibliotecas on April 30. 1993. Everitt B. although this needs further examination. Funding 14.org. Linszen DH.327:834. Staring AB. Leucht S. Hayward P.2:8–16. The efficacy of moti- vational interviewing: a meta-analysis of controlled clinical trials. Donohoe G. Motivational interviewing: a review. 1250 . 1998. Gray R. Chisholm D. Curr Psychiatr Rev. J Clin Psychol. 15. Carolien Zeylmaker. ing group.42:67–77. 1996. Lundahl B. Barkhof et al and those with shorter illness duration.65:1232–1245. 21. Gournay K. 1987. An RCT of health nurses who contributed to the interventions and adherence therapy for people with schizophrenia in Chiang assessments: Rinske Schepers. Robson D. et  al. Farr L. Welfare Publication (ADM). Clinical Interview for DSM-IV Axis I Disorders.312:345–349. 2002. 1983. 2000. recent literature. 2006. 22. Reliability of the past decade. MD: US Department of Health. Burke BL. Howel our findings underscore the need to focus on specific D. Eastwood R. The positive and negative 7. Meijer CJ. Morisky DE. trial of methods to promote physical activity in primary care. Schizophr Bull. dictive of drug compliance in schizophrenics: reliability and ence problems in patients with serious and persistent mental discriminative validity. 2000. et al. 2. Rockville. Knapp MR. Kemp R. Opler LA. Bloemzaad. Finnerty M. Harland J.21:419–429. Hayward P. This work was supported by an investigator-initiated and Compliance therapy in psychotic patients: randomised con- nonrestricted grant by the Dr. 2009. Woerner MG. Everitt B. Motivational interviewing in psychotic disorders. Receipt Inventory . Education and Schizophr Bull. Randomised controlled trial of compliance therapy. Buskens E. 2003. Guy W.27:9–18. et  al. 9. Koopmans GT. Levine DM. The Newcastle exercise project: a randomised controlled targets that lead to nonadherence and to apply an indi. Fiszbein A. Meijer CJ. de ­predictive validity of a self-reported measure of medication Haan L. Sajatovic M. Drinkwater C. meta-analyses. Arkowitz H. Weiden PJ. 2010. Modelling the treated instrument for international research. J Clin Psychiatry. Barkhof E.88:315–335. medication adherence in certain groups of patients. J Clin Psychiatry. Dolder CR. Compliance ther- We thank all the participating patients and staff at the apy: a randomised controlled trial in schizophrenia. 2002. Amaddeo F. Barkhof E. Miller WR. 2005. Van der Gaag M. A self-report scale pre- Mental Illness. Treatment van Gool. Br J Psychiatry. 24. Concurrent and 5. Supplementary Material 12. and Ronald 19. Lacro JP. National Institute of Mental 3. Bindman J. Bien TH. Alvir JM. BMJ. life chart schedule for assessment of the long-term course of 6. Br J Psychiatry Suppl. and especially the psychologists and community mental 18. New York. Predictors of Health. Schizophr Res. 8. Sharkey L. O’Donnell C. Dunn LB. et al. BMJ. Expert Consensus schizophrenia. BMJ. The effectiveness and applicability of Downloaded from http://schizophreniabulletin. 1999. Paul Janssen Foundation. Knudsen HC. Department of Psychiatry at the Academic Medical 2003. White M. Olfson M.177:s28–s33. First MB. Br J Psychiatry. Kemp R. trolled trial. 1976:76–338. New York State in patients with schizophrenia: a comprehensive review of Psychiatric Institute. Menchola M. Gaite L. quiz 47. J Consult Clin Psychol. domised controlled trial. Motivational Interviewing: Preparing syndrome scale (PANSS) for schizophrenia. Med Care. Heres S. Pharmacoeconomics.

Determinants of med. and hospi. Cannabis use and the Serv. among patients with serious mental illness. Downloaded from http://schizophreniabulletin. Perneger TV. van Zwieten B. Bromet EJ.196:274–281. Barbui C. 2011. de Haan L. Blyler CR. Guey LT.37:727–736. J Nerv Ment Dis. 2010. Kotov R. 32. Lancet. BMJ. Miller WR.167:987–993. Smeerdijk M.23:637–651. Annu Rev Clin Psychol. et al.316:1236–1238. Lee T. 2015 ication compliance in schizophrenia: empirical and clinical 36. 2002. et  al. 1997. Koeter MW. 2005. Valenstein M. Keet R. Fenton WS. Motivational cation adherence in patients with schizophrenia: the Achilles interviewing and interaction skills training for parents to heel of adherence research. Kikkert MJ. 2008. talization patterns of patients with schizophrenia. Schizophr Res Treatment. Rodell LR.42:195–197. Storosum J. Usall J. What’s wrong with Bonferroni adjustments. Ochoa S. Psychiatr 33.359:83. Family characteristics. Am J Psychiatry.2012:916198. Informed consent differences in schizophrenia and first-episode psychosis: a from behaviourally disturbed patients. 34. 31. substance abuse. Using a pharmacy- schizophrenia: a randomized controlled trial. et al. Beck CM.org/ at UNAM Direccion General de Bibliotecas on April 30. Gender 37. course of schizophrenia: 10-year follow-up after first hospi- 29. based intervention to improve antipsychotic adherence 2012. Rodell DE. Effects of Motivational Interviewing on Adherence 28. Foti DJ. 1998. Cobo J.oxfordjournals. Dekker N.42:1627–1636. comprehensive literature review. Assessment of medi- 30. findings. Hettema J. Heinssen RK. Schizophr Bull. Kulkarni J. Motivational interviewing. Kavanagh J. Muller K. Psychol Med. Labad X. change cannabis use in young adults with recent-onset 35.1:91–111. 1991. 2012. Steele J. Schizophr Bull. Rader LE. talization. Kashner TM. 1251 .