California Institute for Mental Health Page 1
IMPLEMENTING SAMHSA EVIDENCE-BASED PRACTICE TOOLKITS
Family Education and Support[1]
Brief description of the practice
Target group:
Families of persons with severe and persistent mental illness, particularly schizophrenia although there have been studies with other diagnostic groups. ”Family” refers to anyone who cares about patient, not necessarily actual relatives.
Practice components:
Family education and support is a “method of working with families in partnership with families to impart current information about the illness and help them to develop coping skills for handling problems posed by mental illness in one member of the family.” (Toolkit)
The family education and support approach arose in the late 1970s. It grew from the recognition that (a) conventional family therapy not only ineffective but potentially damaging, (b) families need information about mental illness and (c) families need to learn coping skills which may be counterintuitive and (d) families can make a difference (either positive or negative) in a client’s recovery.
The approach was conceived and developed by three different investigators, Falloon, Anderson, and McFarlane, who each had a somewhat different emphasis and format for the intervention, so that the term has become broader and covers multiple cognitive, behavioral, and supportive interventions delivered in multiple formats.
The suggested minimum time for the intervention is 9 months. The Toolkit suggests that additional time for the intervention up to two years will improve outcomes.
The Toolkit is designed to be used by professionals but has been used successfully by paraprofessionals. The Toolkit can be used in a multifamily or a single family format.
The core components of the model as presented in the Toolkit include the following:
q Joining—developing a respectful, trusting, helpful relationship with family
q Education about the nature of mental illness
q Problem-solving
q Structural change in treatment—focus on family strengths, as well as stress/burden
q Multifamily contact in order to develop support. This can be accomplished through the use of a multifamily intervention or if used with a single family will encourage the family to join NAMI.
Evidence supporting family education and support practices
A. Narrative reviews
The three review articles in this category include as co-authors the two co-authors of the Toolkit: William McFarlane and Lisa Dixon. The reviews present an overview of the basic elements of the intervention along with the supporting evidence for the practice. While the reviews present a large number of studies they do not purport to be thorough analyses of the literature. There is minimal description of the methodology of the studies which are cited in the reviews but the summaries of findings refer almost exclusively to randomized control studies.
McFarlane, Dixon, et.al. (2003)[2] summarizes early studies as follows: “Overall, the relapse rate for patients provided family education and support has hovered around 15% per compared to a consistent 30-40% for individual therapy and medication or medication alone.” A table of 11 “major outcome” studies shows lower relapse rates in seven studies which compared family education and support with standard treatment. Differences in relapse rates in these studies ranged from 17% vs. 83% to 44% vs. 64%.The authors explain the four studies with nonconforming results as not having adequate interventions (either too short or not culturally sensitive) or as resulting from comparison against an enriched standard care. “The nonconforming studies tend to validate the effectiveness of the studies in which an effect was found—by suggesting that the core elements in fact make a difference.” The authors conclude: “Family education and support has a solid research base and a consensus among the leaders of the field regarding its marked efficiency, essential components and techniques; it should continue to be recommended for application in routine clinical practice. However, there continue to be important gaps in the knowledge needed to make comprehensive evidence-based practice recommendations and to implement them with a variety of families.”
B. Systematic reviews (with meta-analysis)
There are two systematic reviews focused on family interventions which use meta-analysis. Study selection criteria limited the review to higher quality studies.
Pitschel-Walz, et.al.(2001)[3] selected 25 studies from 1977 to 1997 which contained 40 different comparisons. All studies in the review had consumer rehospitalization or relapse as the major outcome. Only RCTs were included. All but four of the studies were limited to clients with schizophrenia and schizo-affective diagnoses.
1. Family intervention vs. standard care. There were 12 studies with 14 different comparisons. “Effect size” is a standardized measure of difference between experimental and control groups. The effect size in these studies was what is conventionally viewed as small. The authors conclude: “A treatment which includes family intervention is clearly superior to the usual care of schizophrenia patients.”
q Stability of impact. There appeared to be no decrement in effect as there was no difference between the results at different follow-up periods up to 24 months.
q Length of intervention. There were 7 studies of short term interventions (<3 months) and 7 with long-term interventions (>9 months). Both groups of studies were more effective than standard care, but long-term was more effective than short-term.
q Type of family intervention. The interventions were classified into those with more of a education and support focus and those with more of a therapeutic orientation. Both types of interventions were more effective than standard care. There was no significant difference between the two types of interventions.
2. Combination of family and individual intervention vs. individual intervention. This comparison included two studies in which all the clients received a comprehensive psychosocial program with some receiving in addition a family intervention. These showed no difference between the groups suggesting “no additional effect of a family intervention could be found, on condition that the patients received a comprehensive treatment.”
The other high quality review is Pharoach, et al (2003).[4] The authors utilized the Cochrane Collaboration Guidelines for assessing quality of studies. All the family interventions included an educational component; other components included in some but not all studies were cognitive behavioral interventions, role playing, relaxation, motivational interviewing, and home exercises. A total of 28 studies were included in the review. Studies were conducted in Australia, Canada, China, and the U. S. The major comparisons were between family intervention and standard treatment.
q Hospital use. Results from 7 studies comparing intervention to standard care are equivocal. The authors conclude “The reviewers do recognize the enormous difficulty of conducting randomized trials in this area, but, nevertheless if family intervention is to be widely used, it could be expected that its implementation should be based on more convincing data than these.”
q Relapse. Family intervention reduced the rate of the variety of 'relapse events' compared to standard care in 14 studies at 12 months and in 6 studies at 24 months. Results remain statistically significant at 12 months when heterogeneity is resolved, but authors caution there may be a small study bias.
q Mental status. “The overall impression [regarding mental status] is mixed with findings both favorable for family intervention and equivocal.”
q Treatment compliance. “About 15% of participants left the study before completion of one year. Compared with other trials of the care of people with schizophrenia, this level of follow up is excellent. Family intervention did not seem to promote or hinder this attrition. The experimental intervention did, however, promote compliance with medication.”[5]
q Social functioning. “There is an impression that family intervention does improve general functioning in this domain” but there was no impact on more objective outcomes such as employment, independent living, prison use. “This may be an example of rating scales being sensitive to slight changes that may not have repercussions for routine life.”
q Cost savings. Three studies indicated significant savings in direct or indirect costs.
C. Individual Studies Since Reviews
q Randomized studies since the publication of the two major review articles add information to a few of the items of interested including comparisons of different modalities, application to different cultural groups, and applicability of the models to community settings.
q Dyck et.al. (2002)[6] compared service usage of a group of outpatients with multi-family psycho-education vs. standard care. Clients were diagnosed with schizo or schizo-affective disorders; most clients had chronic conditions with mean illness duration of 10 years. The intervention consisted of weekly sessions for two years conducted by two clinicians using a standardized protocol based on McFarlane. They were trained and supervised to strict adherence to the manual. Results showed a significant reduction in hospitalization at one year (9% vs. 22%) controlling for prior year usage.
q Chien et.al (2004)[7] compared two types of family intervention in Hong Kong—mutual support and psychoeducation - to standard care in 96 families with a schizophrenia family member. Special attention was paid to cultural aspects such as importance of extended family and sense of filial responsibility. The mutual support intervention which was led by peers who were well trained and supervised outperformed the other two interventions on specific measures of functioning and re-hospitalizations with the family education and support outcomes in the middle.
q Bradley, et.al (2006)[8] studied multiple family psycho-education in Australia for a newly arrived non-English speaking (Vietnamese) and an English-speaking group. Relapse rates were lower for the family intervention group at the end of the one year of intervention (12% vs. 36%) and after 18 months (25% vs. 63%).
q A 2006 German multicenter study found rehospitalization rates were reduced from 58% to 41% and hospital days from 78 to 39 days using an 8 session family and client intervention.[9]
D. Consensus Panel Recommendations Regarding Integrated Treatment of Co-Occurring Disorders
Family education and support is included in a number of consensus panel recommendations and practice guidelines.
q The SAMHSA ACT Toolkit is a result of the consensus panel convened by Robert Wood Johnson.[10]
q The Schizophrenia Patient Outcomes Research Team (PORT) project recommends that all families in contact with their consumer/relative be offered a family education and support intervention of at least nine months duration which includes education about mental illness, family support, crisis intervention, and problem solving skills. (As reported in Dixon et al, 2003[11])
q The World Schizophrenia Fellowship produced a set of principles which include a number of items related to the provision of services to families including helping to improve communication within the family, providing training for family in structured- problem solving, encourage family to expand its social network including NAMI, being flexible in meeting needs of the family, listening to families, treating families as partners in planning and delivering Tx, exploring the family’s expectations for Tx and for the consumer, assessing strengths and limitation of family’s ability to support consumer, and helping resolve family conflict by responding sensitively to emotional distress. (As reported in Dixon et al, 2003)
q The American Psychiatric Association Guidelines say “On the basis of the evidence, persons with schizophrenia and their families should be offered a family intervention, the key elements of which include a duration of at least 9 months, illness education, crisis intervention, emotional support, and training in how to cope with illness symptoms and related problems.”
E. Evidence regarding adaptability to cultural and other subpopulations
q There is a sizable body of literature about the applicability of the family education and support intervention to different cultural groups and in different cultural environments. All of the studies which have dealt with the issue have stressed the importance of adapting the intervention to account for cultural variations in the understanding of mental illness, the role of the family, and the ways in which the mental health services are and are not conducive to the culture in question (e.g. help seeking behavior, role o authority, medical model), variations in the meaning of socializing, and variations in communication styles..
q The Cochrane Review notes the diversity of countries within which the research has been conducted and concludes “The provision of health care for mentally ill people in the countries in which the trials were undertaken is diverse but the relative consistency of results suggests that their outcomes may be generalized to other health service traditions.”
q The Toolkit cites the use of the intervention in multiple settings including African-American clients in Los Angeles and New York, a Mexican-American population in Los Angeles. It also cites the use of the intervention in numerous European countries, New Zealand and Australia, Japan and China.
q The one negative finding is from a study by Telles et al which compared 40 Hispanic-American non-English speaking families randomly assigned to family intervention or control (meds and case management). There was no difference between the groups for those families who were judged to be more acculturated and the family intervention resulted in exacerbated symptoms for those who were less acculturated. The authors suggest that some of the messages in the family intervention may be dystonic with culture of Spanish speaking immigrant families.
Capsule Summary of Evidence: Effective, Efficacious, Promising, or Emerging, Not Effective, or Harmful
Family education and support can be considered to be Effective as regards the single outcome of relapse prevention. It meets the criteria of having no substantial harm, having manuals, being replicated in at least two rigorous randomized controlled trials, having positive impacts for at least one year, having a reliable outcome measure, and having the overall weight of the evidence in its favor.
There are, however, a number of limitations and cautions to this rating which will be critical in counties’ decisions about whether and/or how extensively to implement this practice. These are briefly discussed below.
1. The rating of an effective practice is limited to the one outcome of relapse. The results for other outcomes are far less certain. What seemed like early positive findings on reductions in hospital use have not been supported by more recent studies done by other than the intervention’s innovators. The findings on improvements in medication compliance are equivocal. Findings on mental health status and a variety of social functioning outcomes remain equivocal. In a mental health system moving increasingly toward s a recovery orientation practices may need to demonstrate effectiveness against a broad range of outcomes.