0
Master thesis
The Effects of Social Factors on Hospitalization of
Patients with Severe Mental Illness
Name Terry Kriz
E-Mail Terry.kriz1@gmail.com
Student number 6290787/10002713
University University of Amsterdam (UvA)
Course Clinical Psychology
Supervisor of internship Lindy-Lou Boyette
Institution of internship Arkin Klaprozenweg and Sporenburg Supervisor of institution Martijn Kikkert and Mariken de Koning Period of internship April – October2016
1 INDEX 1. Summary 2 2. Introduction 3 3. Methods 6 Participants 6 Procedure 6 Operationalization 6 Materials 7 4. Data-analysis 9 5. Results 10 Demographic features 10
Social factors on hospitalization 11
MATE and MAQ scores 11
5. Discussion 13
2 SUMMARY
Patients with severe mental illness (SMI) are by definition in care for a long time and have not achieved full recovery, often with emotional distress and dysfunction as a result. It is still unclear what causes hospitalization for these patients. The objective of the study is to examine whether social factors predict the risk and duration of hospitalization of SMI patients. These relations were also controlled for substance abuse and medication adherence. Methods: A group of 162 participants was interviewed using structured interviews to elicit data on demographic characteristics, clinical characteristics, the social support they received, characteristics of their social networks, the discrimination they
experienced, substance abuse and their adherence to medication. Data on hospitalization of participants over a 5-year period were collected from case files. Results: Sixty-two of the participants were
hospitalized during the study period. There were no correlations found between social support, social network, discrimination and hospitalization. Also there was no difference found after controlling for substance abuse and medication adherence. There was a correlation found between substance abuse and social support and hospitalization. Conclusions: Social factors did not influence the risk of
hospitalization. When controlled for substance abuse and medication adherence there was no difference on social factors between hospitalized and non-hospitalized participants. Participants who met criteria for substance abuse received less social support and were also more frequently hospitalized. An important limitation to this study is that a period of 5 years might be too long to find the effects of baseline assessed social factors as they are prone to change. Alternatives for further research are discussed.
3 INTRODUCTION
Severe mental illness (SMI) patients are defined as patients who have been using mental health care continuously for at least two years, have a DSM diagnosis of which the primary diagnosis is not dementia, a mental handicap or addiction, show residual signs of their illness in the form of disorders in for example cognitive functioning, sleep, and attention, limited social functioning and a chronic course of illness (Schreurs, 1990). The majority of these patients fit the criteria for a psychotic disorder (Swarts et al., 1998). Research by Peen et al. (2011) shows that the number of patients with SMI in Amsterdam increased with 50% to 4600 patients between 2000 and 2005. Of this population 10% is hospitalized annually.
This increase in SMI patients is quite problematic for many reasons. Aside from the strain on resources and the higher costs of admitting patients with prior hospitalization (Ascher-Svanum et al. 2010) it also means that there are a lot of people that are not functioning in a stable enough way to live independently. This causes a lot of distress for both the patients as their friends and family. There is still a lot unknown about what causes this increase in SMI patients and what causes hospitalization for these patients.
Previous research found that medication noncompliance, more severe psychotic and depressive symptoms and substance use are correlated with hospitalization (Ascher-Svanum et al., 2010). These factors are predictors for relapse for both patients that have been admitted for their first psychosis (Alvarez-Jimenez et al., 2012) as well as SMI patients (Haywood et al., 1995, Botha et al., 2010). Apart from risk factors such as substance use and medication noncompliance there are also protective factors; having a job, a partner and a sheltered living situation are correlated with fewer hospitalizations (Frick et al., 2013). These protective factors all have a social component, which could mean that social influences also affect hospitalization rates.
Previous research found that social support correlates with less positive symptoms and fewer hospitalizations at follow-up in a study with patients experiencing their first psychotic episode (Norman et al., 2005). This is also supported from the patients’ perspective; patients reported social support to be the core category assisting their recovery (Schon et al., 2009). In another study patients reported that positive relations with professionals and family members helped in recovery by making them feel special and worth doing something for (Topor et al., 2006). These findings could be prone to bias as patients might not want to admit that regular intake of medication or limiting substance use are factors that contributed to their wellbeing, as those factors might not feel as rewarding as positive social interactions.
According to the social-cognitive perspective (Cohen & Lakey, 2009) perceived social support increases positive self-esteem which leads to better mental health. This theory is based on Beck’s cognitive models of emotional disorders (Beck et al., 1979). In this perspective perceiving social support will eventually lead to stable beliefs about the supportiveness of others which will increase the
4
support increases positive self-evaluation (Barrera & Li, 1996), social support and positive self-esteem then increase positive emotion which positively influences mental health (Cohen & Lakey, 2000).
On the other hand there is reason to believe that negative social interactions such as
discrimination lead to negative thoughts about the self which can cause emotional distress which may increase the risk of hospitalization. This could be one of the reasons why being part of an ethnic minority group increases the risk for psychotic illnesses (Fearon et al., 2006) and explain how internalized stigma leads to low self-esteem and negative outcomes regarding recovery in mental patients (Yanos et al., 2008). It is found that patients who often experience discrimination are hospitalized more often (Hansson et al., 2014). Also expressed emotion and criticism by the key relative is seen as an
independent predictor for symptomatic relapse suggesting that negative social experiences are in fact an influential factor (Vaughn et al., 1984). These findings confirm that negative social factors might play an important role in mental health.
So far previous research found several relations between social influences and hospitalization, however it is not clear whether social factors predict hospitalization or vice versa. Most of the research mentioned has been cross-sectional and does not mention the possibility that more severely ill patients could have more negative and less positive social experiences as a result of their illness being more severe. It is probable that this relation goes both ways which makes it very useful to find out whether social factors can actually predict hospitalization. If social factors do predict hospitalization a focus on positive social relations and stimuli might prove beneficial for recovery and the prevention of
hospitalization.
On top of that, it seems social factors are also correlated with substance abuse and medication adherence which could imply a more complicated role of social factors. Psychiatric patients who also abuse alcohol are stigmatized more than patients who do not abuse substances (Corrigan et al., 2005). This stigma on substance abuse seems to be enduring even a year after treatment (Link et al., 1997). Similar results are found for the association between self-stigma with treatment adherence (Fung et al., 2008). These relations could be of influence on the relation between social factors and hospitalization.
In this study SMI patients are interviewed about the social support they receive, their social network and the level of discrimination they experience which is compared to their hospitalization data and later controlled for substance abuse and medication adherence. The aim of this study is to answer the following two questions:
1. Do social factors (received social support, social network and the level of experienced discrimination) predict the number and duration of hospitalizations?
2. Are social factors still predictive of hospitalization when controlled for substance abuse and medication noncompliance?
Because the study covers a long time-period the number of hospitalizations only provides partial information as it does for example not distinguish between being hospitalized once for a month and
5
being hospitalized once for three years. To give complete information hospitalization is measured as the number of hospitalizations as well as total hospitalization days.
Social network is added as a social factor next to social support and discrimination because it is possible that any social experiences already have a more positive effect than living in relative social isolation. So even though these factors might partially overlap, there is still enough difference to test out both variables.
It was expected that higher social support and a better social network predict a lower number and a shorter duration of hospitalizations. It is expected that a higher level of experienced discrimination predicts a lower number and a shorter duration of hospitalization.
Lastly it was expected that when controlled for substance abuse, received social support, social network and the level of experienced discrimination would have a similar but smaller predictive value on the number and duration of hospitalizations.
6 METHODS
Participants
Data from the study ‘Vermaatschappelijking van de chronische patiënt in de grote stad’ (Theunissen et al., 2008) was used. The subjects were 876 patients participating in long term programs at GGZ clinics in Amsterdam between 2000 and 2005. They were randomly selected from registration data. Inclusion criteria were the SMI criteria; being in care for at least two years, a DSM diagnosis that isn’t addiction, dementia or social handicap, residual symptoms, limited social functioning and a chronic course of illness (Scheurs, 1990). Other inclusion criteria were being in care at one of Amsterdam’s mental health clinic and being between 18 and 65 years of age. Patients were measured in 2005 and 2011. For this report only the data from the first measurement of the participants who were not admitted in a psychiatric hospital at the time of first measurement were included. In addition
registration data about the number of hospitalizations and total hospitalization days of the participants was used.
At the time of the first measurement in 2005, 323 participants were included and interviewed. By the time of the second measurement in 2011, 307 participants could be traced of which 225 people were included. The loss of participants was mainly caused by refusal to participate for a second time (47 participants) or by death (35 participants). 63 participants were excluded from analysis because they were admitted in a psychiatric hospital at the time of first measurement, which left 162 participants. Procedure
The participants were randomly selected from GGZ registration data after which their
practitioners were asked for permission to contact the participants to ask for participation in the study. The participants were interviewed at their home or their care facility. Before starting the interview participants were asked to sign an informed consent form. Several questionnaires were administered as an interview which took on average one and a half hour. The participants were rewarded 15 euros for their participation. The participants' psychiatrists and nursing staff were asked for additional
information using various other questionnaires which were conducted at the care facility. Operationalization
Table 1 shows the operationalization and questionnaire used for each construct. The only construct that is not operationalized by a questionnaire is hospitalization. The hospitalization data is taken from participants’ case files.
Table 1 Operationalization and Questionnaires Used for Each Construct
Construct Operationalization Questionnaire
Social support Social network Social Network Questionnaire Social support Social Support Scale
7
Substance abuse Substance abuse MATE
Medication compliance Medication compliance MAQ Hospitalization Total hospitalization days -
Number of hospitalizations -
Materials
The Social Network Questionnaire (SNQ) is developed to measure the size and quality of a patient's social network (Van Wijngaarden, 1987). The SNQ consists of four items that ask about the monthly frequency of social contact and the number of social contacts within the primary and secondary network. The scores of these items are added together in a total score. The minimum is 0 and there is no maximum score. The higher the total score, the larger and more frequent the social network. The questionnaire has been used and adjusted in earlier research; we used the same adjusted version that was used by Duurkoop (1995). This adjusted version of the questionnaire was found to be a reliable (alpha .82) and valid instrument for measuring a patients' social network (Duurkoop, 1995).
Social support was assessed by the Social Support Scale (SSS) (Van Dam-Baggen et al., 1988), an instrument developed to measure experienced social support. This list consists of two parts, part A and B. In part A participants are asked how many acquaintances and close friends they have. Part B assesses the subjective level of support that is received. Because part A overlaps with what is asked in the SNQ only part B was used. Part B consists of 11 yes or no questions such as "Others come to me for support and advice" and "I talk to others in confidentiality". The minimum score is 0 and the maximum score 11, a higher score means more subjective social support. The validity of this scale is good and is found to measure a different construct than social network (Van Dam-Baggen et al., 1988)
The Discrimination Scale (DS) is the daily discrimination section of the discrimination part of the National Survey of Midlife Development in the United States (Kessler, Michelson & Williams, 1999). It measures the extent to which the participants feel discriminated on seven different types of
discrimination. It has been translated to Dutch and consists of seven items which are rated on a four-point Likert-scale; 1 often, 2 sometimes, 3 rarely and 4 never. The minimum score is 7 and the maximum score is 28, the higher the score the less discrimination experienced. The internal consistency of the discrimination scale is high with an alpha of .97.
Measurements in the Addictions for Triage and Evaluation (MATE) was used to measure substance use, physiological dependency and abuse according to DSM-IV-criteria. The MATE is a scale based on the Europeans Addiction Severity Index but focuses more on the criteria for substance abuse and uses more structured questions which makes it less susceptible to subjective interpretation. In this research some extra questions were added in order to use EuropASI composition scores (McLellan ea., 1980). These are scores that indicate whether the participant meets the DSM-IV-criteria for substance abuse where 0 indicates no substance abuse and 1 indicates substance abuse. According to Schippers et al. (2010) the MATE is a reliable and valid instrument to assess substance abuse and dependency according to DSM-IV-criteria.
8
Medication Adherence Questionnaire (MAQ) (Thompson ea. 2000) is an instrument designed to assess medication compliance. The MAQ consists of 7 yes or no questions and one 5 point Likert-scale question. Participants are for example asked whether they took their medication the day before and how often they have difficulty in taking all their medication. The maximum score is 11 and the minimum score is 0, a higher score indicates lower compliance. Both the reliability and the test-retest reliability after a two week interval are .76. For this research the scores on the MAQ are rescored into compliant and noncompliant. Participants score noncompliant if they answer yes on one of the 7 questions.
9
IBM SPSS Statistics for Windows Version 22.0 was used to conduct the analyses (IBM Corp, Armonk, NY, 2013).
The explorative analyses frequencies and descriptives were used to explore the demographic data (age, gender, ethnicity, education, diagnosis and housing). This was also done for the tested variables (hospitalization, social network, social support, discrimination, medication adherence, substance abuse). These data were also used to explore whether there are any differences between hospitalized and non-hospitalized patients in combination with a chi-square test or a Mann-Whitney test.
It was planned to do a linear regression analysis to analyze the predictive value of social factors on hospitalization. However the assumptions of normality of residuals and homoscedasticity were violated and neither a root square, a log10 nor a log linear data transformation were sufficient to meet the assumptions. Therefore the data was analyzed using a non-parametric test, the Spearman rank correlation in which all dependent and independent factors including substance abuse and medication adherence were analyzed. The substance abuse and medication adherence were analyzed for
explorative reasons in this analysis. This analysis does not analyze the predictability but it does analyze correlations between the studied factors.
Because there is no non-parametric solution to control the social factors for substance abuse and medication adherence the analysis was changed to a binary logistic regression. This type of analysis is less sensitive to violations of the assumptions than a linear regression. To conduct this analysis, hospitalization was changed to a binary variable with hospitalized and non-hospitalized as 1 and 0. Substance abuse and medication adherence were put in box 1 and social support, social network and discrimination were put in box 2.
10 Demographic features
The study included 162 participants, 93 male and 69 female with a mean age of 45.9 (SD=10.1). An overview of the demographic and clinical features and results can be seen in Table 2. The results in Table 2 are shown for all participants (n=162), participants that were hospitalized at least once (n=62, 38.3%) and participants that were not hospitalized (n=100, 61.7%).
Of the 162 participants 103 (63.6%) lived in independent housing whilst the rest lived in
sheltered housing (36.4%). The average number of days hospitalized is 111.4 (SD=303.5). The majority of the participants had schizophrenia as their main diagnosis (67.3%) followed by mood disorders (9.2%), substance use disorders (6.1%), anxiety- or panic disorders (1.9%) and other disorders (12.3%). These findings are similar to previous research on SMI patients where 68% had a diagnosis for a psychotic disorder (Swarts et al., 1998).
A Chi-square test or a Mann-Whitney test were done to find out whether there were differences between hospitalized and non-hospitalized patients. The only feature that significantly differed between hospitalized and non-hospitalized participants was the diagnosis (X²=11.853, p<.05). This might mean that patients with a certain diagnosis are more likely to be hospitalized than patients with a different diagnosis.
Table 2. Demographic and Clinical Features of the Sample Characteristics of sample (n=162) Hospitalization
(n=62) No hospitalization (n=100) Statistics X² or U df P Age, mean (SD) 45.9 (10.1) 46.3 (10.5) 45.7 (9.9) 3074.5 .931 Gender, n (%) 0.038 1 .846 Male 93 (57.4) 35 (56.5) 58 (58) Female 69 (42.6) 27 (43.5) 42 (42) Ethnicity, n (%) .984 1 .321 Western 102 (63) 42 (67.7) 60 (60) Non-Western 60 (37) 20 (32.3) 40 (40) Education, n (%) 1.172 2 .557 Primary/secondary education 122 (75.4) 54 (72.6) 77 (77) Higher education 33 (20.4) 13 (21) 20 (20) Unknown 7 (4.3) 4 (6.5) 3 (3) Diagnosis, n (%) 11.854 5 <.05 Schizophrenia 109 (67.3) 37 (59.7) 81 (81) Mood disorder 15 (9.2) 7 (11.3) 8 (8) Anxiety- or panic disorder 3 (1.9) 1 (1.6) 2 (2)
Substance use disorder 16 (6.1) 11 (17.7) 6 (6)
11 Comorbid substance use
disorder, n (%) 48 (29.6) 21 (33.9) 27 (27) .867 1 .352
Housing, n (%) 6.844 4 .144
Independent housing 103
(63.6) 40 (64.5) 63 (63) Sheltered housing 59 (36.4) 22 (35.5) 37 (37) Total hospitalization days,
mean (SD) 111.4 (303.5) 291.1 (435.8) - Number of
hospitalizations, mean (SD) 0.9 (1.6) 2.3 (1.6) - Monthly substance usage
Alcohol, mean (SD) 24 (58.3) 24.7 (69.5) 23.6 (50.5) 2736 .150 Joints, mean (SD) 9 (28.7) 12.0 (37.6) 7.2 (21.5) 2983 .576 Hard drugs, mean (SD) 6.9 (23.2) 6.6 (18.4) 7.0 (25.9) 3033 .729 Social network, mean
(SD) 125.9 (62.1) 124.5 (64.5) 126.4 (60.1) 2984 .691
Social support, mean (SD) 25.2 (7.5) 24.5 (7.3) 25.7 (7.6) 2817 .330 Discrimination*, mean
(SD) 15.4 (6.4) 15.1 (6.0) 15.7 (6.6) 2917 .596
Non adherent (MAQ) **,
n (%) 66 (40.7) 25 (40.3) 41 (41) .004 1 .947
Substance abuse (MATE)
***, n (%) 24 (14.8) 13 (21.0) 11 (11) 2.571 1 .109
* N=161 due to missing data ** N=149 due to missing data *** N=131 due to missing data Social factors on hospitalization
A Spearman rank correlation with a significance level of p=.05 was used to analyze the correlations between the scores on social support, social network, discrimination and total hospitalization days and number of hospitalizations. There were no significant correlations found between any of the dependent and independent variables as can be seen in Table 3. This means that social network, social support and discrimination do not correlate with total hospitalization days or with number of hospitalizations. The results are unexpected as a negative significant correlation between social support, social network and both total hospitalization days as well as number of hospitalizations were expected. Also a positive correlation between perceived discrimination and both hospitalization measures was expected but we found no correlation between those factors either.
Table 3 Correlations between Social Factors and Hospitalization Total hospitalization
days, correlation, (p) Number of hospitalizations correlation, (p) Social Support Scale (SSS) -.102 (.198) -.089 (.261)
Social Network Questionnaire (SNQ) -.037 (.643) -.021 (.791) Discrimination Scale (DS)* -.021 (.397) -.051 (.262)
12 * N = 161 instead of 162 due to missing data
Substance abuse and medication adherence
After being controlled for substance abuse and medication adherence, there was still no relationship between social factors and hospitalization. There was no significant difference between hospitalized and non-hospitalized participants on social support (p=.336), social network (p=.778), discrimination (p=.460).
Substance abuse and medication adherence were analyzed in the Spearman rank correlation to explore the correlation with total hospitalization days and number of hospitalizations. The medication adherence did not significantly correlate with any of the variables. Substance abuse was positively correlated with the number of hospitalizations (ρ= .18, n = 131, p<.05) indicating that participants that met the criteria for substance abuse were admitted more frequently than participants that did not meet those criteria. Another finding is that substance abuse is negatively correlated with social support (ρ = -.31, n=131, p<0.01) which means that participants who met the criteria for substance abuse received less social support than participants who did not.
13
In this study the predictive value of social factors on hospitalization were examined. However, perceived social support, social network and the level of experienced discrimination did not predict the number and duration of hospitalization. Also, when controlled for substance abuse and medication adherence there was no difference on social support, social network and discrimination between hospitalized and non-hospitalized participants. Therefore none of the hypotheses could be confirmed.
The explorative analysis found that patients who met the criteria for substance abuse also received less social support and were hospitalized more frequently than patients who did not meet these criteria. There was no relation between substance abuse and the other social and hospitalization factors and there were no relationships between noncompliance and any of the factors.
There are a couple of reasons that could explain the lack of relationships between the studied factors. As it could be that there are no actual relationships between the researched factors, even factors whose relationship has been supported by previous research were often found not to correlate within this study. This could mean that there is a problem with the way the research has been set up and how the variables were measured.
Firstly by measuring the influence of baseline assessed social factors over a 5 year period it is assumed that there is a long-term effect of social factors on hospitalization. The social network might be somewhat unstable for psychiatric patients as one of the criteria for any disorder is social or professional dysfunction (DSM-IV-TR, 2000). Stressful social events such as discrimination or a lack of social support might trigger symptoms and decrease self-esteem which could in turn make hospitalization needed. These changes can occur quickly and the significance of these factors is missed when there is only one measurement over a long time period. This could be solved by doing a repeated measures design where the participants are measured for example every six months for a couple of years. Then the connection can be made between the most recent results and the occurring hospitalization rates a short period thereafter.
Previous research however, showed correlation between the number of admissions and social support over a three-year period (Norman et al, 2005). This study was done with patients who had a first episode so it is unsure whether the results were different because of the period between measurements and results or because patients with first-episode psychosis might rely more on social support than SMI patients because they are still unfamiliar with being mentally ill.
The same argument can be made for medication adherence. It is probable that medication adherence has a short-term effect on functioning and therefore also on hospitalization as medication non-adherence could trigger the return of symptoms. However this is not the only issue with medication in this study. When asked in an interview participants might feel inclined to give socially acceptable answers which could bias the data towards being more adherent to medication than in reality. On top of that, the cut-off score of answering one question against being adherent already scored the participant as non-adherent. There is a difference between forgetting the medication a couple of times a month and stopping the medication deliberately for a longer period of time. The latter probably comes with a higher risk of relapse and hospitalization so it might be useful to assess the difference between
14
occasional non-adherence and intentional or long-term non-adherence and to find a more reliable way to measure adherence. Other research that found a relation between hospitalization and medication adherence considered their participants non-adherent when they reported to take “at least half” or less of their medication which might be a better measure of adherence (Ascher-Svanum et al., 2010).
Previous research uses a less strict measurement in the use of substances (Botha et al., 2008, Ascher-Svanum et al., 2008). In these studies recent substance use is measured rather than substance abuse, this way of measuring might have made the difference as participants who did show problematic substance use but did not meet criteria for substance abuse might have similar hospitalization rates but are rated differently. This might explain why substance abuse correlates with only some of the expected variables.
In short, patients who experienced more social support, had a better social network and experienced less discrimination were not found to be hospitalized less often or for a shorter amount of time and this relation remained unchanged after controlling for substance abuse and medication adherence. These findings are not in accordance with the social-cognitive perspective (Cohen & Lakey, 2000). However, the results of this study are limited by the used research methods. Taking these limitations into consideration it would be beneficial to replicate this kind of research as there is the possibility that the results will be different. On top of that it is important to not give up on providing mental patients with social support as lower hospitalization rates are not the only possible benefits positive social influences could provide.
15
Alvarez-Jimenez, M. et al. (2012). Risk factors for relapse following for first episode psychosis:
A systematic review and meta-analysis of longitudinal studies. Schizophrenia Research, 139, 116-128.
American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (4th ed.,
text rev.). Washington, DC: Author.
Ascher-Svanum, H. et al. (2010). The cost of relapse and the predictors of relapse in the treatment of schizophrenia. BMC Psychiatry, 10. DOI: 10.1186/1471-244X-10-2.
Barrera, M. & Li, S. A. (1996). The relation of family support to adolescents’psychological distress and behavior problems. Handbook of Social Support and the Family. New York: Plenum
Beck, A. T., Rush, A. J., Shaw, B. F. & Emery, G. (1979). Cognitive therapy of depression. New York: Guilford
Botha, U. A. et al. (2008). The revolving door phenomenon in psychiatry: comparing low-frequency and high-frequency users of psychiatric inpatient services in a developing country. Social Psychiatric
Epidemiology, 45, 461-468.
Corrigan et al. (2005). How adolescents perceive the stigma of mental illness and alcohol abuse.
Published Online: http://dx.doi.org/10.1176/appi.ps.56.5.544.
Duurkoop, P. (1995). Terug naar Amsterdam, Longitudinaal onderzoek naar het functioneren van
chronische patiënten in nieuwe woonsituaties. Amsterdam: Academisch proefschrift Vrije
Universiteit.
Fearon, P et al. (2006). Incidence of schizophrenia and other psychoses in ethnic minority groups: results from the MRC AESOP Study. Psychological Medicine, 36, 1541-1550.
Frick, U. et al. (2013). The Revolving Door Phenomenon Revisited: Time to Readmission in 17’415 Patients with 37’697 Hospitalisations at a German Psychiatric Hospital. Readmission after
Psychiatric Hospitalisation, 8, 1-9.
Fung, K. M., Tsang, H. W. H. & Corrigan, P. W. (2008). Self-stigma of people with schizophrenia as predictor of their adherence to psychosocial treatment. Psychiatric Rehabilitation Journal, 32, 95-104.
Hansson, L., Stjernswärd, S. & Svensson, B. (2014). Perceived and anticipated discrimination in people with mental illness – An interview study. Nordic Journal of Psychiatry, 68, 100-106.
Haywood, T. W. et al. (1995). Predicting the “revolving door” phenomenon among patients with schizophrenic, schizoaffective, and affective disorders. American Journal of Psychiatry, 152,856-861.
IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.
Kessler, R. C., Mickelson, K.D. & Williams, D. R. (1999). The prevalence, distribution, and mental health correlates of perceived discrimination in the United States. Journal of Health and Social Behavior, 208-230.
16
Lakey, B. & Cohen, S. (2000). Social support theory and measurement. Social Support Measurement and
Intervention: A Guide for Health and Social Measurements. Oxford University Press, New York
Link, B. G., Struening, E. L., Rahav, M., Phelan, J. C. & Nuttbrock, L. (1997). On stigma and its
consequences: Evidence from a longitudinal study of men with dual diagnoses of mental illness and substance abuse, Journal of Health and Social Behavior, 38, 177-190.
McLellan, A. T., Luborsky, L., Woody, G. E. & O’Brien, C. P. (1980). An improved diagnostic evaluation instrument for substance abuse patients. The Addiction Severity Index. Journal of Nervous and
Mental Disease, 168, 26-33.
Norman, R. M. et al. (2005). Social support and three-year symptom and admission outcome for first episode psychosis. Schizophrenia Research, 2, 227-234.
Peen, J., Theunissen, J., Duurkoop, P., Kikkert, M. en Dekker, J. (2011). Na de extramuralisering; een retrospectief onderzoek naar omvang en zorggebruik van de groep chronische patiënten in de Amsterdamse ggz. Tijdschrift voor Psychiatrie,53, 509-517.
Schippers, G., Broekman, T. G. & Buchholz, A. (2011). Manual and protocol for assessment, scoring and
use of the MATE 2.1. Nijmegen, The Netherlands: Bureau Beta
Schippers, G. ea. (2010). Measurements in the Addictions for Triage and Evaluation (MATE): an instrument based on the World Health Organization family of international classifications.
Society for the Study of addiction, 105, 862-871.
Schon, U.K., Denhov, A. & Topor, A. (2009). Social relationships as a decisive factor in recovering from severe mental illness. The International Journal of Social Psychiatry, 55, 336-347.
Schreurs, M. (1990), Over chronische psychiatrie in Midden-Twente. Een epidemiologisch onderzoek in Midden-Twente naar het functioneren en de zorgbehoefte van mensen met chronische psychiatrische problematiek. Projectgroep Multifunctionele Eenheid, Hengelo.
Swart, M. S. et al. (1998). Violence and Severe Mental Illness: The effects of substance abuse and nonadherence to medication. American Journal of Psychiatry, 155, 226-231.
Theunissen, J. R. et al. (2008) Vermaatschappelijking van de chronische patiënt in de grote stad. Thompson, K., Kulkarni, J. & Sergejew, A.A. (2000). Reliability and validity of a new Medication
Adherence Rating Scale (MARS) for the psychoses. Schizophrenia Research, 42(3), 241-247. Topor et al. (2006). Others: The role of family, friends, and professionals in the recovery process.
American Journal of Psychiatric Rehabilitation, 9, 17-37.
Van Dam-Baggen, C. M. J., Huiskes, C. J. A. E., Kraaimaat F. W. & Schreurs, P. J. G. (1988). Inventarisatie
Sociale Steun. Utrecht: Rijksuniversiteit Utrecht.
Van Wijngaarden, B. (1987). Sociaal Netwerk. Utrecht: Interne publicatie afdeling psychiatrie, Academisch Ziekenhuis
Vaughn, C. E. et al. (1984). Family factors in schizophrenic relapse: Replication in California of British research on expressed emotion. Archives of General Psychiatry, 41, 1169-1177.
17
Yanos, P. T., Roe, D., Markus, K. & Lysaker, P.H. (2008). Pathways between internalized stigma and outcomes related to recovery in schizophrenia spectrum disorders. Psychiatric Services, 59, 1437–1442.
