Ms. Fjell, Dr. Friis, Dr. Melle, and Dr. Smeby are affiliated with Ullevål University Hospital, Oslo, Norway. The late Dr. Thorsen, Dr. Johannessen, Dr. Larsen, and Ms. Øxnevad are with Stavanger University Hospital, Stavanger, Norway. Dr. Friis, Dr. Larsen, and Dr. Melle are with the Institute of Psychiatry, University of Oslo. Ms. Lie is with Haugesund Hospital, Haugesund, Norway. Ms. Lyse and Dr. Simonsen are with Mid-Sector Roskilde Psychiatric University Hospital, Fjorden, Roskilde, Denmark. Dr. McFarlane is with Maine Medical Center, Portland. Dr. Vaglum is with the Department of Behavioral Sciences in Medicine, University of Oslo. Dr. McGlashan is with the Department of Psychiatry, Yale University, New Haven, Connecticut. Send at correspondence to Dr. Friis at the Division of Psychiatry, Ullevål University Hospital, N-0407 Oslo, Norway (e-mail: svein.friis@medisin.uio.no). Lisa B. Dixon, M.D., M.P.H., served as guest editor for this column.

Psychoeducational multifamily group treatment decreases relapse and rehospitalization among patients with schizophrenia (1). This treatment modality was shown to be more effective than the single-family version in a sample composed largely of first-episode patients with schizophrenia (2). However, no studies of multifamily group psychoeducation in an exclusively first-episode sample have been reported.

In this column we describe use of psychoeducational multifamily group treatment with first-episode patients and their families in the context of a research project on early detection of psychosis.

The Treatment and Intervention in Psychosis (TIPS) project for early detection and treatment of patients with first-episode psychosis was carried out in four health care sectors, three in Norway (Stavanger, Haugesund, and Ullevål) and one in Denmark (Roskilde) (3,4,5,6). Both the Regional Ethical Committee and the Norwegian Data Inspectorate approved the study. All patients who participated signed informed consent.

All patients entering treatment for a first episode of psychosis in these catchment areas from January 1, 1997, through December 31, 2000, were invited to participate in the TIPS study. The participants were required to live in one of the catchment areas, to be 18 to 65 years old (15 to 65 in Stavanger and Haugesund), and to meet diagnostic criteria for a first-episode, nonorganic, nonaffective psychosis. Inclusion also meant participating in a two-year standard treatment program. The treatment protocol had three elements: antipsychotic medication, weekly supportive psychotherapy, and psychoeducational multifamily group treatment every second week. Families were not offered multifamily group treatment if a key family member had a serious psychiatric or somatic illness, if there was a recent history of sexual abuse within the family, or if the family lived far away. When parents were divorced, both parents were invited, as were siblings or children if they were older than 18 years. Family members were informed about the multifamily group and asked to participate.

Of the 301 patients 246 were invited to participate and 147 agreed. Reasons for nonparticipation by families were as follows: a key family member suffered from a serious psychiatric or somatic illness or there was a recent history of sexual abuse within the family or lack of command of Norwegian or Danish (26 patients, or 9%), no family or the family was too far away (28 patients, or 9%), the patient refused (79 patients, or 33%), or the family refused (20 patients, or 8%). One patient died before multifamily group participation was offered.

Of the participating families, 131 (89%) participated with the patient present. Relatives attended a significantly higher proportion of the meetings than the patients (65% compared with 45%; t=7.11, df=146, p<.005). Forty-four percent of the relatives (65 relatives) attended 75% or more of the family group meetings, whereas only 26% of the patients (38 patients) achieved that high an attendance rate.

The most important predictor of multifamily group participation was age. There was a dramatic decline in the participation rate with increasing age. The mean±SD age of patients with participating families was 23.6±6.4 years, compared with 30.3±10.1 for those who refused and 34.6±10.8 for those who were not invited to participate (F=38.3, df=2 and 298, p<.001). Although group participation was offered as quickly as possible, most families had to wait between six and 12 months before starting in a group because there was not a sufficient number to start a group. A substantial portion (16%) had to wait more than one year.

We based the family work of the TIPS project on the standard elements of a psychoeducational multifamily group program. We engaged the family members as partners in the patient's treatment and rehabilitation by using a structured treatment manual (7). We organized the program in three stages: joining sessions to engage patients and their family members, a multifamily educational workshop, and 90-minute multifamily group meetings every second week, organized around problem solving. All group leaders in the project underwent a training program of more than 60 hours of lectures and role play. They also underwent monthly group supervision during the project (8).

We modified the program to meet the needs of first-episode patients and their families. Once the family members gave informed consent, the group leaders met for at least three single-family sessions with individual families separately from the patient and with individual patients separately from their families; the clinicians focused on developing a treatment alliance and providing information and guidance specific to that family. The alliance meetings with first-episode families were centered on the families' denial and shock and upon reframing expectations for their family member. These meetings provided an opportunity for most family members to talk for the first time with a clinician about their frustration, grief, and hopelessness. Most families had no previous experience with psychotic illness and were willing to learn and to get help in order to develop better coping skills. They could easily recall the early warning signs of illness because the signs had occurred recently. The acknowledgment of their observation skills established the start of the growing partnership between clinicians and families. The alliance meetings with the patients were explicitly focused on their experience of the need for support and help.

After the alliance meetings, all families met every second week with group leaders on a single-family basis until the start of the multifamily group. Each group included five patients and up to three members of each family, with two group leaders. We arranged separate educational workshops (9) for patients and family members before the groups started. In each workshop the most experienced clinicians in the project gave lectures focusing on the stress-vulnerability model of symptom exacerbation. The lectures in the family workshop highlighted the process of personal development and the diagnostic ambiguity of first-episode syndromes. The workshop for patients focused on developing an understanding of how symptoms exacerbate with stress and the need for rest and time-outs. The workshop was repeated for all group members after one year of group meetings. Patients were asked to invite additional persons from their network to participate in the second workshop.

The main lesson learned is that psychoeducational multifamily group treatment can be implemented with patients experiencing a first episode of psychosis and with their families and that it was well received. We also learned about the difficulties associated with involving patients with their families in this critical period of the illness. Patients' reluctance was the most prominent reason for nonparticipation and may well have stemmed from a lack of experience with family work on the clinical wards. Many patients needed time to consider the offer to participate and adjust to the idea of participating in a large group with their family. It appears important to train therapists in motivational enhancement to help patients deal with the challenges and stimulation in this family work. Patients' nonparticipation seemed to increase with increasing age—with age, patients' reluctance increased and fewer patients had contact with their families. Some chose not to participate because of anticipated stress resulting from engagement with their families. Others had delusions about their families or denied their own illness. With increasing age, there also seemed to be an increasing wish for independence from families.

Another lesson learned is that patients have different needs according to age and life circumstance. Different types of groups may be optimal—for example, multifamily couples groups for married patients or educational groups for single adult patients that focus on how to develop social relations and friendships.

We also learned of the need to engage the family rapidly. It was time-consuming to gather the necessary number of families to start a new group, partly because of the low incidence of first-episode psychosis. Our strategy to address this delay was to offer interim individual family counseling. However, we consider such delays a barrier to optimal clinical results. For most family members, this was their first experience with psychosis of a loved one. In times of crisis both patients and their families have a critical need for support and information. We recommend that new patients and their families enter existing multifamily groups, which would greatly reduce the stress on families at the start of a major mental disorder of a loved one. One way to reduce time to engagement would be to create multifamily group centers across regions and municipalities that can serve several health care districts simultaneously. A larger population base will generate sufficient numbers of new patients so that groups can be created within weeks rather than months.

Since the end of the TIPS project, substantial expansion of multifamily groups in the project sites and in other Danish and Norwegian psychiatric services has been noted. This expansion of evidence-based family work is the result of the training and supervision of group leaders during the project period and ongoing dissemination of information about the significance of family work in the treatment of severe psychiatric disorders. Patients with other diagnoses and a longer duration of illness are now being offered family work. These developments create positive expectations for greater inclusion of the family perspective in treatment programs.

Multifamily group treatment as currently constituted can be usefully provided to patients experiencing a first episode of psychosis. To counteract the tendency of older patients to refuse participation, group programs should be tailored to the needs of older patients. In addition, efforts should be made to engage patients and their families quickly.

The study was supported by grants 133897/320 and 154642/320 from the Norwegian National Research Council, by the Norwegian Department of Health and Social Affairs, by grants 1997/41 and 2002/306 from the Norwegian National Council for Mental Health/Health and Rehabilitation, and by Rogaland County and Oslo County, Norway (Dr. Vaglum, Dr. Johannessen, Dr. Friis, Dr. Larsen, and Dr. Melle). Support was also received from the Theodore and Vada Stanley Foundation; the Regional Health Research Foundation for Eastern Region, Denmark; Roskilde County, Denmark; Helsefonden Lundbeck Pharma; Eli Lilly and Janssen-Cilag Pharmaceuticals (Dr. Simonsen); a Distinguished Investigator Award from the National Alliance for Research on Schizophrenia and Depression (NARSAD); grant MH-01654 from the National Institute of Mental Health (Dr. McGlashan); and a NARSAD Young Investigator Award (Dr. Larsen).

The authors report no competing interests.