Relatives’ Experiences of Mental Health Care, Family Burden and Family Stigma: Does Participation in Patient-Appointed Resource Group Assertive Community Treatment (RACT) Make a Difference?

Abstract

The aim of this exploratory cross-sectional study was to investigate the experiences of relatives of individuals with severe mental illness with and without participation in patient-appointed Resource Group Assertive Community Treatment (RACT). A total of 139 relatives (79 with and 60 without RACT) completed the Family Involvement and Alienation Questionnaire, the Burden Inventory for Relatives of Persons with Psychotic Disturbances, and the family version of the Inventory of Stigmatizing Experiences. We found that relatives participating in RACT experienced a more positive approach from the healthcare professionals, as well as a lower degree of alienation from the provision of care. Relatives who did not participate in RACT were more afraid that their ill next of kin would hurt someone. No other differences in family burden were found. Experiences of family stigmatization were similar in both groups. In conclusion, participating in patient-appointed RACT may contribute to a higher level of satisfaction for relatives in their encounter with healthcare professionals and a more positive alliance. Implementation of RACT in new settings would require adaptation to local conditions to facilitate cooperation between healthcare staff and other relevant services. Staff training focuses on the case manager function and needs assessment, as well as how to create an alliance with the patient and his/her relatives.

Introduction

Developing a severe mental illness (SMI) is a process that can be associated with considerable suffering, both for the ill person and for that person’s relatives. At the same time, the process can stimulate feelings of love and empathy within the family (Ewertzon et al., 2012; Johansson et al., 2010, 2012; Weimand et al., 2011, 2013). Persons affected by SMI often value their relatives’ practical and emotional support, which also can be of importance in the recovery process (Schön, 2009; Topor et al., 2006). To support the affected person, relatives often expressed a need to be involved in the professional care (Goodwin & Happell, 2007; Weimand et al., 2011), and such involvement is encouraged in the Swedish national guidelines for support to persons with schizophrenia and similar conditions (Swedish National Board of Health and Welfare, 2018). Despite good intensions on the part of service providers, relatives report feeling alienated from the professional care provided for their next of kin with SMI (Ewertzon et al., 2010; Johansson et al., 2015). In recent decades, the involvement of relatives in professional care has received increasing attention. An example of a model involving relatives is Resource Group Assertive Community Treatment (RACT). The present study focuses on the experiences of relatives with and without experiences of RACT.

Background

Research has shown that the daily life demands of supporting a person with SMI can increase the psychological and social burden for relatives (Flyckt et al., 2013; Miller, 2017; Östman et al., 2005; Rowe, 2012; Weimand et al., 2010). Relatives have also described a feeling of being alone in their situation (Ewertzon et al., 2012; Johansson et al., 2010; Weimand et al., 2011). Not having someone to share the situation with can lead to an increased burden and this can affect relatives’ health (Weimand et al., 2010). For example, Johansson et al. (2015) demonstrated that parents of adult children suffering from a long-term mental disorder had a lower level of health-related quality of life compared to national age-related norms. Results were predominantly related to ratings of family burden, as well as symptoms of anxiety and depression. The situation of relatives can be adversely affected by stigma, which might lead to avoidance of contact with others, or, inversely, that others would avoid contact with them (Larson & Corrigan, 2008). Stigma impact on quality of life for the individual relative, as well as stigma impact on the whole family, were associated with overall family burden in a Swedish study set at outpatient clinics for persons with psychotic disorders (Allerby et al., 2015)

The situation for relatives may improve with time. Information and sharing of knowledge from the healthcare professionals can be supportive in achieving positive change in the relative’s situation (Walton-Moss et al., 2005). Relatives of patients with SMI may have unique insight and knowledge about their next of kin and often take a pronounced responsibility for his or her wellbeing. Therefore, they should be considered a vital resource by healthcare professionals. However, several reviews have found that in many cases relatives experienced barriers and hindrances in their contact with healthcare services (Doody et al., 2017; Miller, 2017; Rowe, 2012). Relevant issues include not being provided with pertinent information, not being involved in the planning of care, as well as being met with a lack of interest when wishing to convey their own experiences to the care professionals (Johansson et al., 2014; Weimand et al., 2011). A study by Ewertzon et al. (2010) among family members of people with psychosis found that a majority of the participants had experienced an approach from the professionals characterized by a lack of cooperation and confirmation. Furthermore, a significant association was identified between the participants’ experience of approach from the health care professionals and their own feeling of alienation, which indicated that the professionals’ approach had bearing on whether the participants experienced involvement in the care provided. A subsequent study by Ewertzon et al. (2011) showed a low level of agreement between the family members’ actual experience of the healthcare professionals’ approach and what they considered to be important. A discrepancy between relatives’ expectations and their experience of perceived support from staff has also been shown in the inpatient setting (Miller, 2017). Good relations and communication were considered by relatives as being important aspects of the quality of the psychiatric services (Schröder et al., 2007) and important for genuinely being involved in the professional care (Goodwin & Happell, 2007). Such involvement will enable relatives to support their mentally ill family member (Syrén, 2010) and to manage their own life situation (Piippo & Aaltonen, 2009).

In recent decades, cooperation with relatives has received increasing attention within psychiatric services. The introduction of practice standards regarding information sharing is one such example, one that increased the relatives’ opportunity to become involved in the provision of psychiatric care (Lakeman, 2008). Furthermore, treatment models for people with SMI have been developed, in which the relatives’ experiences of the model are followed up. Weimand et al. (2018) presented an example in a Norwegian study among family members whose next of kin were cared for in accordance with the Assertive Community Treatment (ACT) model. ACT is an outreach team model for case management that was developed in the United States during the 1970s to coordinate treatment for people with SMI. ACT teams are multidisciplinary and include the patient, a psychiatrist, a nurse, a social worker, a physiotherapist and social services staff, depending on the needs to be met. Actions performed are primarily in the everyday environment and are based on treatment plans, for which the patient, together with the team, develops goals based on his or her individual needs.

A further development of ACT, introduced by Fallon and Fadden (1993) is RACT. The RACT approach is described in detail in the meta-analysis of Nordén, Malm, et al. (2012). Briefly, this is a structured care model that emphasizes the mentally ill individual’s own resources, with a focus on shared decision-making. The goal is to increase empowerment and self-determination, rather than to reduce psychopathology per se. The resource group (RG) is central to RACT, and the addition of the RG distinguishes RACT from ACT. Members of the RG are appointed by the patient, and include the case manager, the psychiatrist, and significant others from the mentally ill person’s informal network (e.g., family members and/or friends) and from the formal network (social workers and, when applicable, residential support staff). The RG becomes the platform for facilitating a structured collaboration with care professionals. During regular meetings, the patient’s own personal development goals and wishes are discussed and plans for recovery are jointly made. The RACT approach also emphasizes the importance of psychoeducational strategies, directed at both patients and their relatives (Berglund, 2019). These strategies are prioritized by the Swedish National Board of Health and Welfare (2018) based on the existing body of evidence. Relevant and objective information about the illness, delivered in a pedagogical and non-judgmental manner, has been shown to reduce relatives’ feelings of guilt, and to increase hope and reduce family burden (Hjärthag et al., 2017). The aims of psychoeducational interventions are to increase knowledge about the illness, to increase relatives’ understanding of the disability, recognition of early warning signs, as well as steps to take in case of deterioration (Berglund, 2019). The results of a meta-analysis revealed that care according to the RACT model was associated with reduced disease symptoms, improved functional level and increased well-being in patients with psychosis (Nordén, Malm, et al., 2012). It has also been observed that for people with schizophrenia, receiving care according to RACT improved functional level and satisfaction with care at a five-year follow-up (Malm et al., 2014). In a qualitative study, Nordén, Eriksson, et al. (2012) analyzed case managers’ project reports after a training course in RACT and found that the professionals gave the relatives more attention and that the relatives were positive to participation in RGs. However, findings in that study were from the perspective of the case managers, and there is limited knowledge from the point of view of the relatives themselves. In order to further develop RG work with a person-centered approach, in which the whole family’s need for support is taken into account, it is important to examine relatives’ own experiences of the model. New knowledge could aid the development of an evidence-based practice for the whole family.

The overall aim of this study was to investigate the experiences of relatives with and without participation in RACT, in relation to their experience of encounters with psychiatric services. We were interested in their experience of staff approach, as well as feelings of alienation. Further, we wanted to investigate family burden and family stigma. We hypothesized that relatives with participation in an RG would experience (a) a more positive encounter with psychiatric services, (b) less family burden, and (c) less family stigma.

Methods

This study was conducted as an explorative non-randomized observational cross-sectional study.

Sample and recruitment procedures

Study participants were relatives of adults who were in contact with mental health services due to SMI. For the purpose of this study, the term relative was broadly defined as a person who was either part of the immediate family, other close relatives or other important persons, such as a close friend. SMI was in this study defined as a psychiatric disorder resulting in serious functional impairment, which substantially influences or limits daily life.

For inclusion, a person had to be the relative of an adult (>18 years) with SMI, who had contact with mental health services. Swedish language capability had to be sufficient to understand the study information and questionnaire.

Data were consecutively collected during the period October 1, 2017 to December 31, 2018. Recruitment was carried out at psychiatric services in western Sweden (Sahlgrenska University Hospital’s nine psychosis outpatient units and Skaraborg Hospital’s four outpatient units and one psychosis inpatient unit), as well as two NGOs (the Swedish Schizophrenia Fellowship and the Association for Psychiatric Cooperation in Skaraborg). Information was provided by case managers at the psychiatric units and by the research team at the NGOs. Relatives interested in participating in the study signed a written consent form and were given a self-report questionnaire, along with a pre-addressed envelope. Those who did not return the questionnaire were sent one postal reminder.

The postal questionnaire

The survey gathered sociodemographic data, including the participants’ gender, age, the country in which they mainly grew up, their education level, marital status and occupational status. Further, they were asked about their relation to the patient, whether they lived together with the patient, the amount of time spent together with the patient, and whether or not they had participated in RG meetings (i.e., whether they had taken part in a planned and prepared meeting that is part of structured care according to RACT, involving the case manager, doctor, patient, and others invited by the patient).

Data were collected using three instruments: The Family Involvement and Alienation Questionnaire—Revised (FIAQ-R) (Ewertzon et al., 2018), the Inventory of Stigmatizing Experiences (ISE, family version) (Stuart et al., 2008), and the Burden Inventory for Relatives of Persons with Psychotic Disturbances (BIRP) (Hjärthag et al., 2008).

The FIAQ-R measures family members’ experiences of the healthcare professionals’ approach, as well as feelings of alienation from the provision of professional care (Ewertzon et al., 2018). In an evaluation set in Sweden, the FIAQ-R was found to have satisfactory psychometric properties in the contexts of the care of older people, geriatric care, psychiatric care, palliative care and diabetes care (Ewertzon et al., 2018). The FIAQ-R contains 29 items that are divided into two scales: (1) Experience of Approach (21 items), covering the concepts of openness, confirmation, cooperation and continuity, and (2) Feeling of Alienation (8 items), covering the concepts of powerlessness and social isolation. Each item in the Experience of Approach scale is evaluated in two ways; (A) by experience and (B) by subjective importance. The first of these (A) focuses on the relative’s actual experience of the professionals’ approach toward them and the second (B) considers the importance that the relative ascribes to this. The four response alternatives regarding experience of approach (A) range from “Completely disagree” (1) to “Completely agree” (4). A high score indicates a positive response. Likewise, there are four alternatives regarding subjective importance (B), ranging from “Of no importance” (1) to “Of the very highest importance” (4). A high score indicates that the professionals’ approach toward them is important.

The response categories of the Feeling of Alienation scale range from “Completely disagree” (1) to “Completely agree” (4). A low score indicates that the participants experience themselves as not being alienated from the professional care and a high score indicates a feeling of alienation.

BIRP, which was developed in Sweden to assess the burden of relatives of persons with psychotic disturbances, has been found to have good validity and reliability (Hjärthag et al., 2008). The instrument consists of 10 items that focus on practical and emotional burden as well as on own health. For the purpose of this study, we used the following emotional burden items: “He/she is sometimes like an adult, sometimes like a child,” “He/she rejects me,” “I am afraid that he/she will hurt someone,” “I am scared that he/she will commit suicide.” There are four response options (1 = never, 2 = sometimes, 3 = often and 4 = always). Responses were dichotomized as “yes” (sometimes, often, always) vs “no” (never).

For the purpose of this study, selected items from the family version of the ISE (Stuart et al., 2008) were used to rate participants’ experiences of stigma. The ISE has been found to have good reliability in relatives of persons with various psychiatric diagnoses. A Swedish version was developed in collaboration with the developer and applied in a study of family members’ experiences of stigma and family burden (Allerby et al., 2015). Items employed in the current study included “Has stigma affected the family’s ability to get and keep friends?,” “Has stigma affected your ability to socialize with other relatives?,” “Has stigma affected quality of life in the family?,” “Do you try to avoid situations that may be stigmatizing for the family?,” and “Have you ever tried to reduce stigma by explaining and telling friends or relatives about mental illnesses?” Participants were asked to respond to each question from (a) their own individual experience and (b) the experience of the entire family. Responses were dichotomized as “yes” vs “no or uncertain.”

Statistical analysis

The t-test was used to compare the means of the continuous variables. Chi square and Fisher’s exact test were used to analyze differences in proportions. As in previous studies using the FIAQ (Ewertzon et al., 2010, 2011; Johansson et al., 2019; Weimand et al., 2018), data were analyzed using rank-based, non-parametric statistical methods (Altman, 1991; Siegel & Castellan Jnr, 1988)

The median level (Md) and quartiles (Q₁, Q₃) were used to describe the distributions of the response profiles. A global score was calculated for each scale by the median scoring technique for multi-items (Svensson, 2001). The Mann-Whitney test was used to investigate potential differences between relatives with and without RG experience.

A power calculation according to Altman (1991) (80% power) showed that 120 participants would be needed to capture a 25% group difference in proportions (significance 0.05) to test categorical data. The significance level was set at p < 0.05. Data was analyzed using SPSS Statistics 22.1 (IBM Corp., Armonk, NY, USA).

Ethical approval

The participants were given oral and written information about the study. After receipt of the information, written consent was obtained from all participants. The study was approved by the Regional Research Ethics Committee in Gothenburg, No 151-17.

Results

A total of 188 relatives initially agreed to participate in the study, and 141 of these returned questionnaires. Responses from two relatives had to be excluded due to missing data, yielding a total of 139 participants (response rate 74%).

Sociodemographic characteristics

Over two-thirds of the participants were female (n = 97, 70%). There were 41 males (29%) and one further person identified as being of other sex. The mean age was 63 years (SD 12.4, range 28–88). Almost three quarters were married (n = 101, 73%). Fifteen (11%) were divorced, 12 (9%) were a widow/widower, and 10 (7%) were single. Most (n = 128, 92%), had grown up in Sweden. Eighty-six (61%) had 12 years of schooling or more, and 76 (55%) were currently working.

Most were parents (n = 95, 68%), while other participants were siblings (n = 19, 14%), partners (n = 13, 9%) or had another relation (n = 8, 6%), such as being a son or daughter. One fourth of the participants (n = 34, 24%) lived together with the patient. Most of the participants (n = 123, 88%) spent time together with the patient, with an average of 20.4 h/week (SD 30, range 1–166 h, median 10 h).

Seventy-nine of the relatives (57%) had participated in RG. No evidence of differences was found regarding sociodemographic characteristics in those with and without RG participation (Table 1).

Table 1. Characteristics of relatives to persons with severe mental illness, with and without involvement in Resource Group Assertive Community Treatment (n = 139).

	Resource group
	Yes (n = 79)		No (n = 60)		p
Age (mean, SD)	64, 11.7		62, 13.3		0.31 ^a
	n	%	n	%
Female	53	67	44	73	0.54 ^b
Marital status
Married/partnership	54	68	46	77	0.40 ^b
Divorced	9	11	6	10	0.75 ^b
Widow/widower	9	11	3	5	0.19 ^b
Single	5	6	4	7	0.97 ^c
Swedish origin	75	95	53	88	0.21 ^c
Education > 12 years	50	63	36	60	0.69 ^c
Currently employed Relationship to the ill person	43	54	33	55	0.84 ^b
Parent	58	73	37	62	0.08 ^b
Sibling	8	10	11	18	0.18 ^b
Partner	5	6	8	13	0.18 ^b
Other relation	5	6	3	5	0.71 ^c
Lives together with the ill person	16	20	18	30	0.18
Spent time together with the ill person/week (mean, SD)	19, 27.3		22, 32.6		0.54 ^a

a t-test.

b Chi square test.

c Fisher’s exact test.

Experiences of the healthcare professionals’ approach and its importance

Table 2 shows the distribution of the responses on the “experience of approach” scale, i.e. the participants’ actual experience of the professionals’ approach, and on the “subjective importance” scale, i.e. the importance they ascribed to the professional’s approach. For the group as a whole, the median agreement level to the statements of “experience of approach” was “partly agree” (3), which indicated that the participants on average experienced a rather positive actual approach from the healthcare professionals. The median level of “subjective importance” was “of the very highest importance” (4), which indicated that the participants on average considered the healthcare professionals’ approach as being extremely important, and no difference was observed between relatives with and without RG.

Table 2. The distribution of relatives’ response profile in the total group and subgroups with and without participation in resource group regarding experience of approach and subjective importance.

		Resource group
	The total group	Yes	No
Scales	Md (Q1, Q3)	(Md (Q1, Q3)	Md (Q1, Q3)		p-value^a
Experience of approach^b	n = 132 3 (3, 4)	n = 79 3 (3, 4)		n = 53 3 (2, 3)	0.002
Subjective importance ^c	n = 130 4 (3, 4)	n = 78 4 (3, 4)		n = 52 4 (3, 4)	N.S.

^aMann–Whitney test.

^bResponse categories: 1 (completely disagree), 2 (partly disagree), 3 (partly agree), and 4 (completely agree).

^cResponse categories: 1 (of no importance), 2 (of little importance), 3 (of great importance), and 4 (of the very highest importance).

Table 2 shows further that relatives who had participated in an RG agreed to a significantly (p < 0.002) greater extent with the statements in “experience of approach” than did relatives who had not participated in an RG. This indicates that relatives who had participated in an RG experienced a more positive actual approach from the healthcare professionals than their counterparts without RG experience

Feeling of alienation from the provision of professional care

Table 3 shows the distribution of the responses on the “feeling of alienation” scale. For the group as a whole, the median agreement level for feeling of alienation was “partly disagree” (2), which indicates that, on average, the participants felt only partially alienated from provision of professional care.

Table 3. The distribution of the response profile for the total group and subgroups with and without resource group experiences regarding feeling of alienation.

		Resource group
Scale	The total group	Yes	No	p-value^a
	Md (Q1, Q3)	Md (Q1, Q3)	Md (Q1, Q3)
Feeling of alienation^b	n = 126 2 (1, 2)	n = 75 2 (1, 2)	n = 51 2 (2, 3)	0.008

^aMann–Whitney test.

^bResponse categories: 1 (completely disagree), 2 (partly disagree), 3 (partly agree), and 4 (completely agree).

Furthermore, the relatives who had participated in an RG agreed to a significantly (p < 0.008) lower extent with the statements in “Feeling of alienation” compared to their counterparts who had not participated in an RG. This indicates that relatives with RG experience felt to a lower extent that they were alienated from the provision of care.

Family burden and stigmatization in relation to RG participation

Table 4 shows the distribution of the emotional burden and family stigma variables. For the group as a whole, the most common emotional burden was strain due to their next of kin being sometimes like an adult and sometimes like a child (60.2%). The least common burden was fear that their next of kin would hurt someone (13.2%). However, proportions with this fear differed significantly between groups. Over one-fifth of the relatives who had not participated in an RG were afraid that their next of kin would hurt someone, which can be compared with 6% among those with RG experience. No other significant differences in proportions were found with regard to family burden (results not shown).

Table 4. Family burden and stigma in relatives with and without resource group participation.

			Resource group
	Total, n = 139		Yes, n = 79		No, n = 60
	n	%	n	%	n	%	Χ ^2
Family burden ^a
He/she is sometimes as like an adult and sometimes like a child, n = 133	80	60.2	50	65.8	30	52.6	0.125
He/she rejects me, n = 138	50	36.2	26	33.3	24	40.0	0.419
I am afraid that he/she will hurt someone, n = 136	18	13.2	5	6.4	13	22.4	0.006
I am scared that that he/she will commit suicide, n = 136	44	32.4	22	28.6	22	37.3	0.282
Family stigma ^b
Has stigma affected the family’s ability to get and keep friends?, n = 121	22	18.2	13	18.1	9	18.4	0.965
Has stigma affected your ability to socialize with other relatives?, n = 128	40	31.3	23	30.3	17	32.7	0.771
Has stigma affected quality of life in the family?, n = 126	45	35.7	31	41.3	14	27.5	0.110
Do you try to avoid situations that may be stigmatizing for the family?, n = 128	46	35.9	27	36.5	19	35.2	0.880
Have you ever tried to reduce stigma by explaining and telling friends or relatives about mental illnesses?, n = 126	92	73.0	57	78.1	35	66.0	0.133

a BIRP: An item was considered to be endorsed if the response was sometimes, often, and always.

^bISE: An item was considered to be endorsed if the response was yes.

Almost three quarters of the relatives responded that they handled stigma by explaining the situation, telling friends and relatives about mental illness. Over a third reported that they tried to avoid situations that might stigmatize the family and a similar proportion reported that stigma had affected the family’s quality of life. Proportions with stigma concerns did not differ in persons with and without RG experience (Table 4).

Discussion

Results

To the best of our knowledge, this is the first study to investigate the situation of relatives with and without participation in RACT in relation to their experiences of encounters with psychiatric services, family burden, and family stigma. Relatives who took part in RGs experienced a more positive approach from the healthcare professionals. Further, they felt less alienated from the provision of care compared to those without the RACT experience. Those who had not participated in RGs experienced a greater fear that their next of kin would hurt someone. Experiences of family stigma did not differ in relatives with and without participation in RACT.

The positive experiences of relatives who took part in RACT may be explained in part by the case manager’s role as a coordinator, providing relatives with an easy access to health care professionals. Another partial explanation may be the fact that the RACT model emphasizes psychoeducational interventions as pivotal to detect early warning signs, as well as mutual agreement regarding actions to be taken if the patient becomes impaired. The patient and the relative are jointly involved in identifying treatment goals and decisions are jointly taken. All of these aspects may influence the relatives’ experience of the healthcare professionals’ approach toward them, reducing the risk of alienation.

The total group of participants in the present study experienced a positive approach from the healthcare professionals. They considered the professionals’ approach to be extremely important, and on average reported not feeling alienated from the provision of care. Our results parallel those of a previous study focusing on relatives’ experiences of care in accordance with the ACT model (Weimand et al., 2018). In that study, a majority of the participants experienced a positive approach from the ACT team. Results for relatives with RG and ACT differ from those reported from previous studies that involved relatives’ interactions with various unidentified healthcare models (Ewertzon et al., 2010; Johansson et al., 2015, 2019). Those previous studies showed that the relatives experienced a less positive approach from the health professionals. Further, they felt alienated from the provision of care to a higher extent. Taken together, it appears that service programs such as ACT and RACT stimulate a more positive health care environment in terms of relatives’ experiences. These programs are in line with the call for the involvement of relatives in integrated services for people with SMI, as stressed by the authors of a recent review (Yesufu-Udechuku et al., 2015).

In our study, ratings of all family burden and stigma variables were higher compared to a previous Swedish study (Allerby et al., 2015). The latter focused on persons on oral antipsychotics, who had a fairly high level of cognitive function (Brain et al., 2014). Further, only one tenth lived together with the family member who participated in the study, which can be compared to 24% in the current study. Relatives who did not live with their mentally ill next of kin have been found to be less affected by stigma (Östman & Kjellin, 2002).

We found an association between the family burden variable “I am afraid that he/she will hurt someone” and lack of RG participation. While this finding may reflect selection bias, as persons with elevated risk of violent behavior will be less likely to participate in an RG, it is also possible that taking part in an RG improves relatives’ sense of security. They learn more about the illness and its treatment, and they can earlier recognize signs of deterioration in their relative. Importantly, they are prepared to act if and when such symptoms occur. Relatives without RG participation might be more apprehensive about the risk of violent acts. It is not uncommon that relatives report that the support they receive from mental health services is insufficient in relation to their needs (Weimand et al., 2011). Families of patients identified as having increased risk of violence are in need of extra support (Katz et al., 2015) as aggressive behaviors affect family safety as well as the safety of others.

Our data do not support the hypothesized associations regarding less family stigma in relatives who took part in RG. One explanation might be that a longer period of exposure to RACT services would be required before an impact on this variable may be observed.

Methodological considerations and limitations

Several methodological considerations need attention. This explorative study investigated a group of relatives across 14 mental healthcare units and members of two NGOs in a large geographic area. This design offers a greater transferability than would be achieved if the study were restricted to a single psychiatric unit or clinic. Three-fourths of the relatives who initially agreed to participate in the study actually returned the questionnaires, which suggest a relatively good response rate considering the low response rates often observed in survey-based studies. The sample size was acceptable in relation to the power calculation. However, there are limitations related to the study design. The optimal manner to test our study questions would be to set up a cluster-randomized trial, in which half of the services were randomized to involve relatives in RACT and half continued with treatment as usual. This was not feasible as services had already begun to implement RACT, and the degree of implementation varied among services. Another important limitation was that we could not determine whether characteristics of the ill relatives differed in study participants with and without RACT, as no patient data were collected in connection with this study. Due to the exploratory nature of the study, we have chosen not to adjust for multiple comparisons. There are also considerations related to sampling procedure. In keeping with our ethical approval, we could collect no data for those who did not accept participation. Thus, we do not know how many were uninterested in participation, nor do we know whether those who filled in the questionnaire were representative of the background population.

Consent of the person with SMI was not required for the relatives to participate in this study. Both ethical and legal aspects were considered here. One important argument for not requesting patient consent was that the questionnaire focuses on the relative’s own experiences. If patient consent were a prerequisite for inclusion, some relatives who would like to share their experiences would not be able to do so. Also, a requirement of patient consent would likely have resulted in a biased selection of participants, as patients with particularly severe symptoms or ongoing family conflict might be less likely to consent to the participation of a family member. In Sweden, there is no legal support allowing a family member to prohibit other adult members of the same family from participating in research studies.

The collected data in the current study were self-rated and based on recall. According to Flyckt et al. (2013), instruments based on recall may underestimate relatives’ experiences of burden and family stigma. However, using an alternative method such as diary might not be suited to quantify mental strain as assessed by BIRP and ISE. All instruments used in the study have been evaluated with satisfactory psychometric properties among relatives of people with psychiatric diagnosis; FIAQ-R (Ewertzon et al., 2018), ISE (Stuart et al., 2008), and BIRP (Hjärthag et al., 2008). Hence, we conclude that the instruments are suitable in the current study context.

Implications and further research

The results in this study show that relatives participating in RACT experience a more respectful approach from mental health providers as well as a lower level of alienation. From the perspective of relatives, our findings suggest that this model could be more widely implemented in clinical practice. However, this is an explorative study with a relatively small sample. Larger randomized controlled studies with broader representative samples are required to compare RACT with other care models for persons with SMI. We need to know more about how RACT works for relatives with diverse sociodemographic and cultural backgrounds and those with various types of relationships with the the ill person. Furthermore, it is important to examine the extent to which case managers actually follow RACT methodology in clinical practice.

Conclusion

The study contributes new knowledge about relatives’ experiences of their encounters with psychiatric services for persons with SMI. The inclusion of a close friend or relative in the RG distinguishes the RACT approach from earlier family intervention approaches. RACT has not been compared to other family involvement interventions, but an advantage of RACT is that it creates a direct channel for relatives’ involvement in mental health care. The RACT approach might prove particularly relevant in low- and middle-income countries, where the burden on family caregivers may be great. For implementation outside the Swedish context, we stress the need for culturally sensitive modifications. The RACT model will require adaptation to local conditions to facilitate cooperation among health care professionals, patients, their relatives, and other relevant services. Implementing the model requires staff education and training in the core RACT elements, including the case manager function, how to create an alliance with the patient and his/her relatives, and how to map the patient’s needs.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Research Council for Health, Working Life and Welfare, grant number 2014-4510 and the Swedish state under an agreement between the Swedish government and the county councils concerning economic support of research (ALF).