[PubMed Citation] [Related Articles in PubMed]
TABLE OF CONTENTS
[Method] [Results] [Discussion] [References] [Tables] [Figures]
American Journal on Mental Retardation: Vol. 106, No. 4, pp. 344–358.
Evaluation of the Dissemination of Active Support Training in Staffed Community Residences
Edwin Jones, David Felce, Kathy Lowe, and Clare Bowley
University of Wales College of Medicine
Opportunity Housing Trust (Cardiff, Wales)
Drive Pontyclun (Cardiff, Wales)
Cartrefi Cymru (Cardiff, Wales)
Our aim was to replicate Active Support (Jones et al., 1999 ), a staff training intervention designed to increase the assistance given to adults with severe mental retardation living in community residences in order to increase their participation in activities. Training was conducted in 38 residences, involving 303 staff members and 106 persons with mental retardation. Active Support resulted in significant increases in assistance and engagement in activity. Significant correlations between participant abilities, receipt of assistance, and levels of engagement were found. Active Support was found to be of greater benefit to people with lower adaptive behavior and to have a diminishing value for people with higher adaptive behavior, implying that its implementation should be matched to the support needs of residents.
(Received 2/22/00, accepted 12/1/00.)
Deinstitutionalization in Wales, a country with a population of about 2.9 million, has reflected trends throughout the developed world (Hatton, Emerson, & Kiernan, 1995 ; Lakin, Prouty, Anderson, & Sandlin, 1997 ; Mansell & Ericsson, 1996 ). Welsh policy emphasized that the community accommodation provided to replace the long-stay hospitals should be small-scale and use ordinary housing stock (Welsh Office, 1983 ), and the resulting provision has a mean group size of 2.8 residents (Perry, Beyer, Felce, & Todd, 1998 ). One of the most conspicuous failings of institutional services was not providing individuals with sufficient opportunity to be occupied constructively (e.g., Felce, de Kock, & Repp, 1986 ; Oswin, 1971 ). Small-scale community reprovision carried the expectation that access to more normative settings and opportunities would be reflected in greater resident engagement in the range of activities that characterize the lifestyle of people without disabilities. However, there is evidence that inadequate support from staff and low activity, at least among people with severe mental retardation, are enduring problems, even in decent home-like environments with high staffing levels and apparent adherence to contemporary service philosophies (Emerson & Hatton, 1996 ; Emerson et al., 1999 ; Felce & Perry, 1995 ).
Measuring the level of engagement in typical daily living activities has been one way of evaluating whether community services provide an improved environment and resident quality of life (Emerson & Hatton, 1996 ; Felce & Emerson, 2000 ). Emerson et al. (1999) , Felce and Perry (1995) , and Felce et al. (1998) have shown that the extent of resident engagement in activity is related to resident abilities as measured by a standardized adaptive behavior scale. People with higher assessed adaptive behavior have been found to participate in typical social, domestic, personal, and leisure activities for the majority of the time (Felce & Perry, 1995 ). People with lower assessed adaptive behavior have been found to be unoccupied for the majority of the time (Emerson et al., 1999 ; Felce & Perry, 1995 ). One of the consequences of the restricted skill development associated with severe and profound mental retardation is the need for greater support to engage in the activities of everyday living.
The problem of inadequate staff support may no longer be extremely low rates of attention from staff per se but, rather, the quality of attention. Hewson and Walker (1992) , Felce and Perry (1995) , and Felce et al. (1999) found that the great majority of staff attention given to residents in community homes was in the form of conversation, which made little contribution to enabling residents to participate in an activity. Although people with more severe or profound mental retardation require much greater support from staff, they do not typically receive it (Felce et al., 1998, 1999 ). These findings have led to the advocacy of an approach termed Active Support (Emerson & Hatton, 1996 ; Felce, 1996 ; Mansell, 1998 ). Broadly based on demonstrations within the applied behavior analysis literature that staff and resident responding alike can be affected by attention to antecedent, concurrent, and consequent conditions, Active Support was first developed and evaluated in an early British community-housing demonstration project for adults with severe or profound mental retardation (Felce, 1989, 1991 ). It combines five elements that have been shown separately to be related either to the extent of attention residents receive from staff or to their participation in the activities of daily living.
First, staff members plan opportunities for resident activity proactively. Second, they plan their own division of responsibility for supporting planned resident activity. Clear allocation of staff members to duties and arranging for them to work alone with small resident groups, rather than with larger groups and with other members of staff, have been shown to result in increased staff attention to residents (Felce, Repp, Thomas, Ager, & Blunden, 1991 ; Harris, Veit, Allen, & Chinsky, 1974 ; Mansell, Felce, Jenkins, & de Kock, 1982 ). Third, when supporting participation, staff supplement verbal instruction with gestural or physical prompting, demonstration, or physical guidance as necessary, increasing the help provided until the person is able to participate in the activity successfully. Repp, Barton, and Brulle (1981) found that nonverbal instruction either with or without physical assistance was the form of instruction most likely to help residents respond correctly. Fourth, staff members give the majority of their attention to residents when residents are constructively occupied. Provision of attention contingent upon engagement has been shown experimentally to increase activity levels (Porterfield, Blunden, & Blewitt, 1980 ; Mansell, Felce, de Kock, & Jenkins, 1982 ). Fifth, staff monitor the opportunities provided to individuals each day. Self-monitoring and feedback have been shown to increase on-task staff performance (e.g., Richman, Riordan, Reiss, Pyles, & Bailey, 1988 ).
An experimental evaluation of Active Support in five staffed community residences, which involved direct observation using categories similar to those defined later in this paper, showed that it resulted in residents receiving significantly more assistance from staff and being significantly more engaged in activity (Jones et al., 1999 ). Implementation of Active Support increased the mean proportion of time each resident (a) received assistance from staff (from 5.9% to 23.3%); (b) engaged in domestic activities (from 12.8% to 32.2%); and (c) engaged in social, domestic, personal, leisure, or educational activity (from 33.1% to 53.4%). Across individuals, increases in assistance and engagement in activity were significantly and positively correlated. Both were significantly inversely related to resident adaptive behavior, as measured by an adaptive behavior scale. In baseline, staff members gave more attention and assistance to people with higher adaptive behavior scale scores. After the introduction of Active Support, receipt of attention was unrelated to adaptive behavior, but those with lower adaptive behavior scale scores received more assistance. The disparity in activity between those with higher and lower adaptive behavior was reduced. The effectiveness of assistance in promoting engagement, as measured by the conditional probability of engagement following the provision of assistance, significantly increased (Felce et al., 2000 ). Gains were maintained in the majority of houses 6 months after the last postbaseline data point, 8 to 12 months after intervention.
These changes were achieved without the addition of staff members. Active Support training, therefore, resulted in a more effective use of staff resources. In cost–benefit terms, Active Support was outcome effective, to the extent that resident participation was considered as a key objective, and almost cost neutral, in that training costs were minor compared to those bound up in staff salaries. Replication of these results in similar services could increase their cost effectiveness. Our purpose in the current study was to conduct a widespread replication of Active Support training in similar community residences run by three Welsh not-for-profit agencies, one of which had participated in the earlier study and another two with similar service philosophies.
Method Return to TOC
Measurement of Participant and Setting Characteristics
In addition to age and gender, we collected information on each participant's adaptive behavior, social impairment, and challenging behavior by interviewing staff members who knew the person well. We used (a) the Adaptive Behavior Scale (ABS) Part One–Community and Residential Version (Nihira, Leland, & Lambert, 1993 ); (b) the Aberrant Behavior Checklist (Aman & Singh, 1986 ); (c) the Psychopathology Instrument for Mentally Retarded Adults (Matson, 1988 ); and (d) items relating to participants' language and means of communication, the quality of their social interaction, and the extent to which they engaged in symbolic activities, stereotypic behavior, and echolalia as contained within the Disability Assessment Schedule (Holmes, Shah, & Wing, 1982 ). The latter set of items were employed to assess