International literature has explored the role of family in the outcome of chronic schizophrenia, but only few studies are available on the relatives of people experiencing their first episode of psychosis (FEP). This is a period when most of the changes in family dynamics are observed (Addington et al., 2003; Koutra et al., 2014; Yesufu-Udechuku et al., 2015).

Schizophrenia impacts not only on the patients but also to those who care for them. The term caregiving burden since the 80s has been used to explore these consequences (Hatfield and Lefley, 1987) and contributed to highlight. Family members of persons with a severe mental disorder experience negative effects linked to these burdensome situations (Budd et al., 1998; Magliano et al., 2005).

With the shift from hospital-based to community-based care which occurred in most western countries, the relatives of mentally ill people have become an integral part of the care system. Thus, taking into account the carers' perspectives on mental health provision and on overall families' needs and support has become of crucial importance.

Research has shown that about one third of the relatives of schizophrenic patients suffer from emotional or behavioral distress and that patterns of familiar interpersonal interaction can worsen the course of schizophrenia (Magliano et al., 2000; Moller-Leimkuhler, 2006). Several studies have explored carers' needs and perceptions, coping styles, mental health and satisfaction. Follow-up studies on the burden on relatives of persons affected by psychosis treated in community-based routine mental health services are uncommon. Even less studies on the burden that develops in the early stages of psychosis, and on the impact of the interventions provided (Brown and Birtwistle, 1998; Magliano et al., 2000; Moller-Leimkuhler, 2006; Breitborde et al., 2011) have been conducted. These follow-up studies have not provided clear information on the patterns in family burden and thus, there is lack of consensus on burden in the medium long term and also on which specific factors could impact on it, and at which stages.

The GET UP (Genetics, Endophenotypes, Treatment: Understanding early Psychosis) PIANO (Psychosis: early Intervention and Assessment of Needs and Outcome) Trial (Ruggeri et al., 2012, 2015) was conducted in a large epidemiologically based cohort from Italian community mental health centers (CMHCs). It aimed to assess the feasibility and effectiveness of a multi-element psychosocial intervention for FEP patients and their families, compared with treatment as usual (TAU). Experimental treatment was administered as an adjunct to TAU, and included: (1) cognitive behavioral therapy for psychosis (CBTp) to patients; (2) psychosis-focused family intervention (FIp) to families; and (3) case management (CM) to both patients and relatives. The intervention was provided by CMHC staff, trained in the previous 6 months and supervised by experts.

It was hypothesized that in the patients add-on multi-component intervention could favor: (1) greater improvements in symptoms, as measured by the PANSS; and (2) reduction in days of hospital admission over the 9-month follow-up. Secondary hypotheses were that the intervention could improve subjective burden of psychotic symptoms (auditory hallucinations and delusions), social functioning and emotional well-being and produce lower service disengagement rates (Ruggeri et al., 2015).

The hypotheses regarding the relatives were that the multi-element experimental intervention at the 9 month follow-up could: (1) reduce burden of the key relative, as measured using the IEQ; and (2) improve service satisfaction, as measured using the VSSS Relatives scale (Ruggeri and Dall'Agnola, 1993), at follow-up.

In the present paper we aim to provide comprehensive descriptives on the effect on relatives of the 9 month multi-element psychosocial intervention for FEP patients and their families in routine MH services versus treatment as usual (TAU). Specifically we address: (a) changes in the relatives' burden of care and emotional distress and (b) differences in the levels of service satisfaction reached at the end of the follow-up. Testing these hypotheses seems to be of particular relevance in a cultural context, such as Italy, where the majority of the patients continue to live with their families even for long time after the mental illness onset, with potential severe consequences on the everyday life of those caring families.

In the experimental arm out of the 272 patients enrolled, 16 did not have a relative; in 6 cases patients did not give the permission to contact the relative; in 7 cases the relatives did not want to undergo into Family Intervention and in 13 cases the patient refused to enter in the CBT intervention, and thus the relative was also excluded. In the TAU arm, out of the 172 patients enrolled, 10 did not have a relative; in 12 cases patients did not give the permission to contact the relative. Thus, it has been possible to ask the permission to be assessed to 230 relatives in the experimental arm and to 150 relatives in the TAU arm. Of these, 185 subjects in the experimental arm (out of 230 available; 80%) and 75 in the TAU arm (out of 150 available; 50%) accepted to complete the trial's baseline assessments (Figure 1).

Relatives in the two groups did not differ in any socio-demographic characteristics. No other between-group differences were observed at baseline in any variable concerning cohabitation with the patient (see Table 1). It is noteworthy that the majority of the relatives were parents of the affected patient, and that the largest majority lived with the patient and were used to stay with the patient all days and for several hours.

Engagement of the relatives in the Family intervention was good: as shown in Table 2, in the experimental group 170 out of 185 relatives assessed at BL (91.9%) participated in at least 1 family session. Most of them (n = 154; 90.6%) received ≥5 FI sessions; of these 112 (72.7%) received ≥10 sessions.

At follow-up 75 relatives could not be traced and thus could not be assessed: 60 subjects belonged to the experimental arm (32.4% of the cohort) and 15 to the TAU arm (20.0% of the cohort). There were no significant differences in demographics and outcome variables at baseline between completers and non-completers. Both groups had similar baseline levels of burden of care. At follow-up both groups experienced a general trend to improvement in most IEQ items.

However, as shown in Table 3, caregivers belonging to the experimental arm at the 9 month follow-up had significantly greater improvement in burden caused by a series of aspects of everyday life with the patient: “Encouraged to eat,” “Ensured medicine,” “Guarded dangerous acts,” “Guarded taking drugs,” “Disturbed sleep,” “Atmosphere strained?,” “Caused quarrel,” “Annoyed Behavior,” “Others annoyed,” and “Burden” (see Table 3). The majority of these aspects pertains to the “Tension” dimension. The effectiveness of the treatment reached significance only for “Disturbed sleep” and “Atmosphere strained?.”

Emotional distress at baseline, as measured by GHQ, was similar in both groups. As shown in Table 4, at follow-up both groups experienced a general trend for improvement in emotional distress. However, caregivers belonging to the experimental arm had a significantly greater improvement in the majority of GHQ items, with the only exception of “Feeling reasonably happy” (Table 4) where improvement was equally significant in the two groups. The effectiveness of the treatment reached significance only for “Under strain.”

Caregivers in the experimental arm experienced significantly greater levels of service satisfaction after the 9 month treatment in “Ability of psychiatrists to listen and understand patients,” “Recommendations to the closest relative,” “Relatives' knowledge to problems,” “How information is given on diagnosis and prognosis,” “Ability of psychiatrists to listen and understand relatives,” “Thoroughness of nurses,” “Helping relatives,” and “Continuity of care” (see Table 5). All these aspects pertain to two dimensions (“Relatives' Involvement” and “Professionals' Skills and Behavior”), with the only exception of “How information is given on diagnosis and prognosis.”

Relatives of persons affected by psychosis play a crucial role in their pathways to care, being the first to request for help in the majority of the cases (Del Vecchio et al., 2015; Jansen et al., 2015). Several studies have also shown that relatives often feel despair, fears, concerns regarding professional support (Tennakoon et al., 2000; Faridi et al., 2009; Boydell et al., 2014; Lavis et al., 2015; Hickman et al., 2016; Koutra et al., 2014). This paper contributes to shed some light on the potentiality that interventions specific for the FEP have to reduce burden of care in an Italian context that is characterized for the majority of psychotic patients by a close link with they families, who continue to take care of them also for long time after the onset.

Data provided by the GET UP Trial allow to draw interesting conclusions on the role of relatives in the care of their loved ones and on the potentiality that specific forms of individual Family interventions can have already in the early phases of psychosis. The multi-element experimental intervention tested in the GET UP Trial, consisting of CBTp for the patients, Family intervention and Case management is effective in decreasing relatives' burden in several specific aspects of burden mainly related to the “Tension” dimension of the IEQ. The study has also shown that emotional distress is present in the relatives already in the first months after psychosis onset, and proved that the experimental multi-element intervention tested is effective in decreasing that emotional distress. Due to the multi-element treatment design of the study, it is not possible to disentangle the effectiveness that might be attributed to the family intervention. However, in the GET UP Trial, acceptability and engagement in family intervention was very good included a large number of families: in the experimental group 92% of relatives participated in at least 1 family session with the majority that could attend up to 10 sessions (defined as optimal standard for the 9 month treatment).

Caregivers in the experimental arm showed significantly greater satisfaction in 8 items, all pertaining to the dimensions “Relatives' Involvement” and “Professionals' Skills and Behavior,” with the only exception of “How information is given on diagnosis and prognosis,” related to the domain “Information given by the professionals.” This finding is of interest also in the light of literature findings showing that relatives often feel that their role and contribution is undervalued by the clinicians (Iyer et al., 2011; McCann et al., 2012).

Overall, the Family Intervention for psychosis delivered in the GET UP PIANO trial reduced family burden of illness and improved emotional distress and satisfaction with services. It should however be said that this improvement is uneven and that not all aspects of everyday interaction with the patient and with the service have been more positively affected in the experimental group.

Results presented in this paper have some methodological limitations. The relatively high number of relatives that did not give consent to the baseline assessment and the relatively high proportion of relatives that were non completers at follow-up is a limitation of these data even if the lack of significant differences in the sociodemographic characteristics between the two groups hampers this limitation. Moreover, the majority of the differences between changes within groups did not result significant due to the low sample size. Finally, financial constraints have compelled the GET UP trial researchers to plan a follow-up period of only 9 months: Thus, the short duration of the intervention did not allow to fully exploit the effects of treatment in improving burden and emotional distress in the relatives.

Among the strengths of this paper we first enlist the fact that 90% of CMHCs asked to participate accepted and completed the study, thus favoring high representativeness of the services and subjects included in the study. A further strength is that relatives belonging to experimental and TAU arm had similar baseline socio-demographic characteristics. Moreover, as most study's relatives live with the patients and stay with them for several hours/day proves that the trial results provide a realistic picture of problems and potentiality to benefit of treatments regarding those relatives which have the most relevant need of help.

In conclusion, data provided in this paper prove the higher effectiveness over TAU of a specific evidence-based multi-element treatment including family intervention in improving family burden and emotional distress. These data further strengthen previous literature on this issue (Jeppesen et al., 2005; Bird et al., 2010; Gleeson et al., 2010; Onwumere et al., 2011; Lee et al., 2014; Nordentoft et al., 2015) and encourage to promote specific forms of family interventions on caregivers of first-episode psychosis patients, and implement further studies that target their needs in the different stages of psychosis.

This study is conducted according to globally accepted standards of good clinical practice, in agreement with the Declaration of Helsinki and in keeping with local regulations.

GET UP PIANO investigators ensure that all professionals involved in the trial are adequately qualified and informed about the protocol, the study interventions, and their trial-related duties and functions. The coordinating center maintains a list of all appropriately qualified persons involved in the study.

MR conceived the project, coordinated all its phases and wrote the paper. AL collaborated in conceiving the project, coordinated the data collection, and had an important role in writing the paper. PS, MM, SS, STor, AF coordinated the data collection and revised the final draft. CB planned and realized the methodological protocol of the project, coordinated the realization of the on-line system for data collection, planned, and executed statistical analyses and contributed in writing the manuscript. DC realized the on-line system for data collection, merged data from all centers, checked data quality and prepared the dataset for statistical analyses. FP, EL, STos, KDS, CCo, STom, CCr, GP collaborated in data collection and revised the final draft.

Ministry of Health, Italy—Ricerca Sanitaria Finalizzata, Code H61J08000200001.