Introduction

Numerous authors have already studied stigma experiences among persons receiving professional mental health care and the negative effects on their objective and subjective quality of life. Starting from the original1 and modified2,3 labeling perspective on mental illness, they showed how experiences of social rejection or the fear of devaluation and discrimination among officially labeled persons have detrimental outcomes on job4 and housing5 opportunities, life satisfaction,6 self-esteem, and self-efficacy.7,8

Furthermore, it is often suggested that entering a psychiatric hospital is one of the most stigmatizing experiences one can have. Falk,9 (p. 51) for instance, states that “the most devastating stigmatization of mental illness is imposed on those who are patients in a mental hospital.” Moreover, the labeling perspective on mental illness, which forms the basis of most sociological studies on the consequences of stigmatization of persons with psychological problems, was first formulated decades ago, when mental hospitals were the predominant type of care for persons with mental health problems. It was the same time that these organizations were intensively studied and described as total institutions with several specific internal processes that were believed to have a stigmatizing impact. Patients were described as undergoing a moral career, which involved the loss of all previous roles because of the way mental health care was structured, and which finally lead to a so-called spoiled identity, the result being that reintegration in society was very difficult, if not impossible.10,11 This kind of sociological study has undoubtedly contributed to the already existing negative image of mental hospitals in that time. Ideological calls for change have joined with economic policy, resulting in an enormous change in the provision of mental health care, with deinstitutionalization as a key element.12,13 Therefore, the assumption that deinstitutionalization would imply destigmatization can be considered one of the several factors that have contributed to the change in mental health care delivery.14 Many mental hospitals have been closed since and alternative forms of care have arisen or have come to play a more important role. This movement has resulted in the current situation in which persons with mental health problems receive help from different types of organizations.

Considering this new context, one could ask whether the studies leaning on the labeling perspective on mental illness are still current. Several answers can be formulated on this question. First, many authors have already revealed that public opinion is not changed dramatically: negative attitudes about persons with mental health problems persist among the public.15,16 Furthermore, several recent studies have shown that persons receiving professional mental help in so-called alternative services, such as therapeutic communities, psychiatric wards of general hospitals, assertive community treatment, and clubhouse programs, also experience stigmatization and its negative consequences, and that the original labeling perspective and its several additions or modifications still find empirical validation in the current context.8,17–19

Studies explicitly comparing stigma experiences of clients of different organizations offering professional mental help remain very scarce. However, this subject is not unimportant in light of the continuing existence of stigmatization and its negative consequences for labeled persons, and in the search for destigmatizing factors. Disregard of the context of mental health care itself can be designated as a shortcoming in the current literature on this topic. For instance, at their presentation of the modified labeling perspective on mental illness, Link et al.3 (p. 421) already argued that “we know little about whether variations in the circumstances under which official labels are applied make a difference. Does the patient’s response to the labeling experience vary according to the location and the social atmosphere of the treatment setting?” This study tries to address this shortcoming by discussing some reasons why mental health services could differ concerning the extent to which they label their clients as mentally ill in the current context of Belgian mental health care. The focus lies on the comparison of stigma experiences of clients from a traditional setting—psychiatric hospitals—with those of an alternative setting—psychiatric wards of general hospitals.20 This focus is taken for three reasons. First, apart from foster care in Gheel—which existed long before the rise of psychiatric hospitals—deinstitutionalization has started relatively late in Belgium and psychiatric care in psychiatric hospitals is still common—Belgium still has one of the largest number of psychiatric hospital beds per 100,000 in Europe. Furthermore, despite the deinstitutionalization movement, psychiatric hospitals still play not a negligible role in the general mental health care system in other countries too, such as the United States.21 Second, psychiatric wards in general hospitals are the oldest and the most widespread organizations offering inpatient and outpatient mental health treatment. As with most countries in the European Union, it is the most common therapeutic alternative for psychiatric hospitals.22 Third, although both settings are at first sight rather similar as both offer residential, semiresidential, and outpatient psychiatric care in a clinical context, several theoretical arguments lead to expectations of differences in stigmatization.

Before discussing them, the concept “stigmatization” is clarified. In this study, three dimensions are paid attention to, namely stigma expectations, social rejection experiences, and self-rejection experiences. Stigma expectations are considered as the perceptions of the existence of negative attitudes about, and behaviors against, persons with psychological problems in society at large, as measured by the devaluation/discrimination scale of Link. Rather than measuring concrete stigma experiences, it can be considered as an indicator of an estimation of the tolerance in the public. Link23 reasoned that these expectations do not differ between clients and the public, as everyone has learned them during socialization. Following this reasoning, stigma expectations are not determined by the context, and, hence, they cannot depend on treatment modalities. Therefore, no differences in stigma expectations between the two kinds of hospitals should be found. The second dimension of stigmatization is social rejection, which is conceptualized as experiences of rejection by the environment as a direct consequence of entering a mental health care organization. This concept can be distinguished from stigma expectations as it measures perceptions of actions or behavior of persons in the concrete environment rather than an estimation of general attitudes in the society. The third dimension is self-rejection, which is defined as feelings of shame and inferiority as a direct result of entering the current mental health care center. As social rejection and self-rejection are both concrete dimensions of stigmatization, which are conceptualized as directly resulting from the current involvement in mental health care, it could be assumed that they depend on the direct treatment context. For both, the hypotheses resulting from the reasoning below will be followed.

Several theoretical reasons lead to expectations that clients of psychiatric wards of general hospitals experience less stigmatization in comparison with those of mental hospitals. First, public opinion could vary according to the kind of organization. An early study of Phillips24 indicated that individuals with identical behavior were increasingly rejected as they were described as utilizing no help, a clergyman, a physician, a psychiatrist, or a mental hospital. As psychiatric hospitals could still have the stereotypical image of being “asylums” or meaning “the end of the line,” persons receiving professional help there could experience more stigmatization.25 Furthermore, clients of psychiatric hospitals could be thought of as more ill in comparison with clients of alternative, short-term settings.25 Second, entering an organization that offers multiple health services such as general hospitals provides an opportunity for concealing the mental health problem.26 As concealability or (in)visibility can be regarded as one of the key features of a stigma,11,27,28 the obviousness of the reason of seeking care in a specific organization reveals much. Whereas it is clear that patients of psychiatric hospitals have mental health problems, this is not the case for clients of general hospitals, if they are able to hide having received care in a specialized psychiatric ward. This principle accords with the discussion about the stigmatizing consequences of the selectivity of services, as selectivity or differentiation creates an underclass and “acts, in itself, as a form of labeling.”29 (p. 179) Third, because making distinctions between people can be considered a key element in stigmatization,30 the integration of mental health services in a general hospital could lead to perceptions that persons with mental and physical health problems are equally in need of medicines, leading to less stigmatization of persons with mental health problems, as Angermeyer et al.31 have already mentioned. This reasoning implies that attitudes are more positive when psychological problems are associated with biological explanations and that medicalization stimulates destigmatization. Fourth, the organization of psychiatric care in general local hospitals was supposed to be literally reducing the distance between clients and their environment, in this way avoiding their disintegration or contributing to their reintegration.32 This belief is in accordance with the contact hypothesis, which states that negative attitudes could be changed through contact with persons with psychological problems.33–35

Despite these four arguments in favor of psychiatric wards in general hospitals, however, several counterarguments lead to the alternative hypothesis that clients of these wards could experience stigmatization at least equally, if not more. First, some recent studies (e.g., Nordt et al.36) show that public opinion does not have more negative attitudes toward “ex-psychiatric patients from a mental hospital” in comparison with other persons described as being or having been mentally ill. Second, the locality of general hospitals could impede anonymity because of the possible presence of—or visits from—acquaintances from the neighborhood. This possibility could be less the case in more remote, specialized psychiatric hospitals.31,37 Third, concerning the hypothesis about the benefits of the integration of mental and physical health care, there are several indications that attributions of mental illness to biological factors do not bring about more positive attitudes among the public.38–40 Fourth, a counterhypothesis to the contact hypothesis suggests that persons with mental health problems who stay nearer to the community could experience more social rejection as they could be more exposed to negative reactions than clients of a more remote psychiatric hospital.7 Fifth, different social comparisons in the two kinds of facilities could also play a role, especially concerning self-stigma or internalized feelings of shame among labeled persons. As Chee et al.25 argue, it is possible that clients of general hospitals use the majority of all clients of the entire hospital—who are in the first place physically ill—as a reference group, whereas patients of psychiatric hospitals mainly encounter other clients who also have mental health problems. Thus, patients of psychiatric hospitals could have less feelings of self-rejection because they perceive more peers with similar problems, leading to the belief that they are not the only ones with these problems.

The empirical evidence on the research question is very scarce. As far as we know, only two studies have investigated this topic up to now. The first—a German study of nearly 25 years old—found that clients with schizophrenia from a mental hospital report less stigmatization than their counterparts in the general hospital.31 The second—a recent Singaporean study—revealed that outpatients with schizophrenia in a state mental hospital experienced less stigmatization in comparison with outpatients with the same diagnoses of a general hospital, but that for other mental illnesses, the reverse was true.24 The current study focuses on differences in stigma expectations, social rejection, and self-rejection between clients of general and psychiatric hospitals with a large variety of clinical and background characteristics to examine whether these studies can be replicated by use of recent data in a Western country—Belgium.

When comparing stigmatization between two types of settings, it is not appropriate to ignore the possible differences concerning their client populations, as each type of organization could attract a specific target group. Therefore, it is necessary to account for a range of client characteristics to control for such selection effects. The following client characteristics will be included to examine whether stigma experiences differ, while accounting for some features of the target group that could vary between the two kinds of organizations. First, their diagnosis is introduced, as several studies have already revealed a link between the kind of diagnosis and negative public attitudes on the one hand,41 and clients’ own stigma experiences on the other.42 Second, a measure of severity of symptoms is included. Some opponents of the labeling perspective on mental illness argue that experiences of stigma are rather subjective and attributable to distorted perceptions because of the illness itself, or that “when stigma is a problem, it is more directly related to the person’s current psychiatric status or general ineffectiveness, than it is to having been in a mental hospital.”43 (p. 881) Furthermore, recent research also reveals that severity of illness can be an important predictor of stigma experiences.44 Third, indicators of length of stay and intensity of treatment are taken into account. Clients receiving care in a more intensive way or for a longer period could experience more stigmatization, as there is more chance that being a psychiatric patient becomes a master status. Fourth, number of years since first treatment is included as an indicator of the number of years one is labeled, as there are indications that “stigma is a powerful and persistent force in the lives of long-term mental patients.”7 (p. 80) Fifth, education and income are used as indicators of socioeconomic status to account for selection effects, as there are indications that psychiatric hospitals count relatively more clients from lower socioeconomic groups.45 Finally, some background variables are included as controls: gender, age, and marital status.

Methods

Sample

Data were gathered in 2005, in the context of a larger study on stigma and well-being in five types of professional mental health organizations in Flanders, Belgium: psychiatric hospitals, psychiatric wards in general hospitals, psychiatric rehabilitation centers, day activity centers, and community mental health centers. This article concentrates on the two first mentioned settings for the reasons mentioned above. In Belgium, mental health care is mainly privately provided but publicly funded, according to the social health insurance model. This implies that 99% of the population is covered by obligatory health insurance, which pays back approximately 89% of their health care spending.46 Clients have a free choice of hospital and no system of catchment areas or obligatory referral exists.

Whereas the total population in Flanders consists of 41 psychiatric hospitals and 36 general hospitals with one or more psychiatric wards, 10 centers from each type are randomly selected. As some organizations refused to participate, the final sample consists of 8 psychiatric hospitals and 7 general hospitals. In each center, all wards fitting the following research criteria were selected. Concerning age, the research was limited to clients in the adult group, excluding wards consisting exclusively of persons younger than 18 or older than 60 or 65. Furthermore, wards for clients with cognitive disorders or mental retardation were excluded. Within the selected wards, clients were excluded who had cognitive disorders or mental retardation, those who were in a too acute stage of illness to be able to participate (determined by the staff), and those who did not have enough knowledge of Dutch to understand the questionnaire. All clients who fit the criteria and who were present on an agreed-upon-beforehand date with the supervisor were asked to participate. Informed consent was obtained after an introduction by the researcher. Of the 783 eligible clients, 555 (71%) agreed to fill in a written structured questionnaire. The completion, which occurred at the presence of a researcher who accompanied the clients where necessary, took up on average 45 min. As the results section will reveal, not all surveys were complete, as some clients dropped out or refused to answer some questions.

The client characteristics are described in Table 1. The sample consists of 291 women and 264 men. Their age ranges from 16 to 73, with an average of 39. Concerning the educational level, 3.8% finished only primary education (until 12 years), 21.8% the first 3 years of secondary education (until 15 years), 50.8% the last 3 years of secondary education (until 18 years), and 23.6% finished college (until 21 or 22 years). One fourth of the participants are married or cohabiting. The mean length of current treatment is 13 months, whereas the mean number of years since first treatment was nearly 10 years. The average intensity of treatment is 106 h/week. Concerning the main diagnosis, which was obtained from the staff, information is missing on nearly 6% of the participants. Of those clients whose main diagnosis was obtained, 29% has a mood disorder, 22% a psychotic disorder, and 35% a substance-related disorder. The similarities and differences between the two kinds of settings will be discussed in the results section.

Table 1 Client characteristics of total sample, psychiatric hospitals, and psychiatric wards of general hospitals
Full size table

Measures

Dependent variables

Stigma expectations are measured by means of Link’s devaluation–discrimination scale,23 which is translated by the authors of the present study and slightly adapted by replacing references to “mental hospital” and “ex-...” with “persons who receive(d) psychological help.” This scale consists of 12 items, with 4 answer categories from “absolutely disagree” to “absolutely agree,” ranging from 1 to 4. A total score is obtained by averaging the 12 scores, with higher scores revealing more stigma expectations (mean = 2.68, SD = 0.43, alpha = 0.81). The internal consistency of this scale is comparable to other studies using the instrument.47,48 To measure social rejection, five items are adapted from the social rejection scale of Fife and Wright.49 These items were introduced by the sentence “because I come to this center...,” to explicitly designate current rather than potential past experiences. An exemplary item is “Since I come to this center, some people treat me with less respect.” The items are coded from 1 (absolutely disagree) to 5 (absolutely agree), and averaged to obtain a main score, with higher scores meaning more social rejection (mean = 3.15, SD = 1.21, alpha = 0.91). The scale of self-rejection comprises five items referring to feelings of shame and inferiority because of receiving mental health care in the current center, with the same introductory sentence as for social rejection. The items are partially inspired by the social isolation subscale of Fife and Wright.49 An example is “Since I come to this center, I have come to feel inferior.” The scoring procedure is the same as for social rejection, with higher scores indicating more self-rejection (mean = 2.89, SD = 1.27, alpha = 0.91).

Independent variables

Hospital type is a dichotomous variable with a score of 1 for general hospitals and 0 for psychiatric hospitals. Symptoms are measured by the Brief Symptom Inventory-1850, using the Dutch translation of the items of the SCL-90-R.51 The mean score on the 18 items that are coded from 0 to 4 is computed to obtain a total score, with higher levels indicating more symptoms (alpha = 0.94). The length of the current treatment period is measured in months.Footnote 1 The number of years since first treatment is computed as the difference between current age and the age at which one first received professional mental health care. The intensity of current treatment is computed as the number of hours a week one spends in the current treatment setting. Concerning the diagnosis, diagnostic information is obtained from their psychiatrist or psychiatric nurse in charge. In the analyses, three main diagnostic categories are used as dichotomous variables (1 = present, 0 = absent): mood disorders, psychotic disorders, and substance-related disorders. Education is measured using 4 categories (primary degree = 1, college degree = 4). Income is measured through a proxy variable that measures how easy it is for one to get by with the money one receives, on a continuum from 1 (very difficult) to 6 (very easy). Finally, some background variables are included: gender (men = 1, women = 0), age (in years), and marital status (married or cohabiting = 1; single, divorced or widowed = 0).

Results

Comparison of the characteristics of the client populations

Before discussing the main research question, the client characteristics of the two settings are compared. Table 1 reveals several differences between the target groups of both kinds of centers. Psychiatric wards of general hospitals count a higher percentage of women, and the mean age is slightly higher. Furthermore, more than a third of their clients are married or cohabiting, in comparison with only a fifth of the clients in psychiatric hospitals. Another difference is that the respondents from the general hospitals report more symptoms. The largest difference, however, is the mean length of current treatment: in psychiatric hospitals this length is nearly 16 months, in comparison with only one and a half months for clients of the psychiatric wards in general hospitals. Furthermore, the total number of years since first treatment is on average 2 1/2 years longer for clients of psychiatric hospitals. Finally, the main diagnoses differ: psychiatric wards of general hospitals count more clients with a mood disorder and less with a psychotic disorder. The patients of the two settings do not differ, however, regarding their socioeconomic status or the intensity of the current treatment. To conclude, the comparison of both settings reveals several differences concerning their client population. Therefore, it is essential to control whether potential variation in stigmatization cannot be attributed to this diversity.

Comparison of stigma experiences

Multiple regression analyses are used to test the link between the organization type and several treatment and background characteristics of clients on the one hand, and the measures of stigmatization on the other hand (Table 2). At first sight, hospital type is only linked with social rejection, as shown by the baseline model. However, when introducing the background variables, organization type is linked with two of the stigma measures: stigma expectations and social rejection experiences. The dichotomous variable measuring hospital type, where 1 stands for general hospitals, has a negative and significant coefficient both measures. This means that, when controlling for the background characteristics, clients attending a psychiatric ward of a general hospital report less fear of devaluation and discrimination, and that they experience less social rejection in their environment since they go to the center, in comparison with clients from psychiatric hospitals. Concerning feelings of shame and embarrassment because of the attendance to the mental health care center, as measured by the scale of self-rejection, no link can be found with the institutional setting. It is important to mention that these three results are found controlling for several client characteristics that are shown to differ between the two kinds of settings, as mentioned in Table 2.

Table 2 The link between hospital type and stigmatization: results of multiple regression analyses (N = 494)
Full size table

Regarding the other variables in the analysis, current symptoms are generally the strongest predictor of stigma experiences. For the three measures of stigmatization, the coefficients were positive and significant on the 0.000 level, meaning that clients with more symptoms have more fear of devaluation and discrimination and report more experiences of social rejection and self-rejection since they came to the current mental health care setting. This can be interpreted in two ways. First, the result could be because of a perception bias, as clients reporting more symptoms could take a black view of everything, including reactions of others to their situation. Second, clients with more symptoms could have more actual stigma experiences as they (are known to) differ more from persons without psychological problems. With regard to the diagnosis, the results of the analysis reveal a link with only the concrete rejection experiences. The significant positive coefficient for psychotic disorders reveals that clients with this diagnosis report more social rejection from their environment in comparison with consumers without that diagnosis. This finding is consistent with Holzinger et al.,42 who found that stigma expectations do not differ between clients with different diagnoses, whereas actual stigma experiences were higher for clients with schizophrenia. In addition, the significant negative coefficients for mood and substance-related disorders indicate that clients with these diagnoses report less self-rejection. Furthermore, the results reveal that clients with more financial resources experience less social rejection. No other relationships between the client or treatment characteristics on the one hand, and stigma expectations or social rejection on the other, were found. However, several features are positively linked with self-rejection. Older people and clients who receive help for more hours a week are more inclined to have feelings of shame and inferiority. In addition, consumers who received their first treatment earlier will report less self-rejection, whereas those with longer current treatment duration report more.

Discussion

The study of labeling and stigmatization of persons receiving professional mental health care and its consequences is not new. Partially because of several studies that pointed at stigmatizing features of psychiatric hospitals themselves and the beliefs that deinstitutionalization would imply destigmatization, the landscape of mental health care has undergone major changes since. However, many recent studies reveal that stigmatization persists. This article went back to earlier research linking stigma experiences with the direct treatment environment by exploring whether differences in stigmatization can be linked with type of care. In particular, stigma experiences of clients of a traditional setting—psychiatric hospitals—were compared with those of their counterparts in psychiatric units in general hospitals, the most common alternative in Europe. Theoretical arguments give no decisive answer on this question and the empirical evidence on this topic is very scarce. The current study leans most on Angermeyer et al.,31 but it is elaborated on by including more organizations, more different diagnoses of the clients, more different stigma indicators, and more background variables. As the analyses in the current study revealed a range of differences between the target groups, it is crucial to bear in mind that these characteristics are controlled for in the main analysis, as potential differences in stigmatization could be because of variation in the client population.

What did the results reveal? First, clients of psychiatric hospitals experience more social rejection in comparison with those of psychiatric wards in general hospitals. This result could be explained in different ways. First, public attitudes about psychiatric hospitals are possibly still rather negative, as they are still viewed as asylums. Furthermore, clients from general hospitals could have more opportunities to hide the fact that they have mental health problems and/or to associate their problems with physical illness. Finally, the more remote location of psychiatric hospitals could contribute to more disintegration of their clients. Despite these arguments about the organizational differences in social rejection, the same variation was not found for self-rejection. Psychiatric hospitals could be more associated with asylums among the public than among clients. Another possibility is that clients from the general hospitals themselves know about the mental health problem and thus feel ashamed and embarrassed, whereas their acquaintances may think they reside in the hospital for physiological health reasons. Furthermore, negative social comparison mechanisms could play a role in general hospitals, as Chee et al.25 argued. A third result revealed that, when controlling for background characteristics, clients from general hospitals have lower stigma expectations, contrary to the expectations on the basis of the reasoning of Link.23 However, as most clients in the current study already have received treatment for a certain time, past experiences could affect their answers to questions on stigma experiences. This reasoning is in accordance with Wright et al.7 who argued on the basis of their study on long-term psychiatric patients that stigma-related experiences can drive clients’ stigma-related attitudes. Persons who experience more social rejection could also be more pessimistic about the negative attitudes in society at large. As social rejection can be linked with type of treatment, so can these general stigma expectations. To summarize the findings, the study reveals that clients from general hospitals and psychiatric hospitals do have different stigma experiences, which cannot be attributed to a large range of differences concerning their background characteristics.

When comparing the current results with both former studies on this topic, several differences can be found. Angermeyer et al.31 found that clients with schizophrenia in general hospitals reported more stigma expectations, whereas Chee et al.25 revealed more stigma experiences in outpatients with schizophrenia in the general hospital, but the reverse for other mental illnesses. This leads to the hypothesis that the target group—especially their diagnosis—plays a significant role. The fact that clients with psychotic disorders form only a minority of the present sample could explain the different findings. However, the data do not allow replicating their analysis separately for each diagnostic group, as few clients in the database have a psychotic disorder, especially in the general hospitals. On the other hand, this study confirms that differential findings can be found according to the dimension of stigmatization. In accordance with Chee et al.25, no organizational variation for feelings of shame and inferiority was found. Therefore, it remains interesting to differentiate between several domains of stigmatization.

Before concluding, some shortcomings of the study should be pointed out. First, concerning the general hospitals, only specialized psychiatric wards were included. However, general hospitals can also offer mental health care in other wards. It is quite possible that these clients still experience less social rejection, especially as their care is still more integrated in the general health care. Second, the number of clients from general hospitals was rather small, which did not allow us to replicate the analyses for different diagnostic categories of Chee et al.25 A third shortcoming is that the study is cross-sectional, which restricts making definite conclusions about selection effects or the direction of causality on the basis of the available data. However, several arguments can be formulated to treat the stigma-related variables as dependent rather than independent variables. First, when clients would systematically actively choose the type of treatment on the basis of the possible stigmatizing consequences, clients with higher stigma expectations could be supposed to prefer the general hospital, which is the opposite of what is found. Second, with regard to the concrete stigma experiences, the rejection that clients experienced after they entered the current treatment setting is focused on by explicitly formulating the questions in that way.

Another possible concern could be the degree of difference between the clients of both settings, as much dissimilarity concerning their background emerged. Therefore, it was essential to take these background variables as controls. However, the analyses revealed that few of the differentiating variables are linked with the outcome and that the difference in outcome cannot be attributed to it. However, certain other selection mechanisms cannot be completely ruled out, as clients from both settings could still differ in other ways. This is especially true as no obligatory referral or catchment area systems exist, and as clients have a free choice of hospital. For instance, general hospitals could attract clients with a higher socioeconomic status who could have less fear of (the consequences of) stigmatization. Although educational level and a proxy variable for income, which do not differ between the two settings, are controlled for, no information concerning job status or exact income is available. Whereas many relevant differences between the populations are already taken into account, it must be admitted that not all potential confounders could be controlled for.

Finally, the inclusion of more than one hospital of each type contributes to the generalizability of this study. However, the analyses were limited to a general comparison between the two types, neglecting the possible variation within each type. It is beyond the scope of this study to extend the discussion to also studying differences in stigma expectations and experiences within a variety of centers belonging to one kind of treatment setting, or across different types of settings. Furthermore, it could be interesting to extend this research to other alternative settings. This is especially true as this study focused on a traditional setting and the most common alternative therapeutic setting in Europe, which implies that it is relatively “conservative.” As the general (international) tendency in mental health care is toward community care, it would be very interesting to study potential differences in stigmatization between these latter services. Furthermore, variation in the client population should be accounted for as was the case in the current study. Special attention could be paid to socioeconomic differences, as these could be more important in other (welfare) states than in Belgium with its universal social insurance based health care system.

Implications for Behavioral Health

Despite the enormous changes in the landscape of mental health care since the first appearance of studies about stigmatization of its clients, current studies on the public opinion and on clients reveal that stigmatization still exists. These studies imply that deinstitutionalization did not bring about destigmatization in general. However, when studying the so-called alternative settings, one could ask whether their clients are as equally stigmatized as their counterparts in more traditional services. This research question is but seldom explicitly studied. This study revealed that, when accounting for clients’ background characteristics, clients from psychiatric wards of general hospitals report on average less stigma expectations and social rejection experiences when compared with clients from psychiatric hospitals, whereas no differences were found for self-rejection. Several explanations could be offered for this result, such as the differential possibilities of secrecy, the location of the setting, the composition of the client population, and a different public opinion according to the kind of setting. However, this empirical study did not allow us to decide which of these alternative explanations holds because no information concerning different public opinion was available and because stigmatization of clients was not linked with specific organizational features. Therefore, it would be interesting to link studies of public opinion and clients’ experiences and to include the organizational dimension into discussions about stigmatization and destigmatization of mental health care. Then, whether variation in stigmatization between several treatment settings could be linked to particular features of the organization of care could be investigated. If specific characteristics that are associated with less stigmatization could be pointed out, eventually a first step could be taken toward attempts to destigmatization on the level of the mental health service itself.