Best Practices
‘To understand weaknesses leads to the awareness of the means to understand what is strong (Stufflebeam, 1983 pp.117 )’
Elements of Successful Interventions
The advocacies of early intervention programs for children with known challenges have evolved as ameliorative in nature, to support the upsurge of children with special needs. Whether the increase is due to legislative support or advancement in detection methods, the proliferation of intervention programs has fueled evaluative research (Guralnick, 1998; Ramey & Ramey, 1992). Radical differences in philosophies, theories and implementation of intervention models vary considerably for subgroups of children having similar developmental disorders (Bailey & Woolery, 1992).
The next section will discuss three categories critical for interventions. These include (a) essential elements, (b) structure, design, methodology, and planning and (c) evaluative methods for analysis. Each category will be reduced to smaller subheadings that briefly discuss each element. These elements or criteria are cumulated from the literature to support the development of successful interventions.
(a) Essential Elements
Early Intervention
Early intervention is viewed as necessary for children with developmental needs. A wealth of literature supports the salience of earliest intervention effects on the plasticity of children’s brain development (Ramey & Ramey, 1998; Shore, 1997). Although the literature suggest age of entry into programs is an important variable, there is lack of consensus as to “earlier is better” (White, Bush, Casto, 1987; Powers, 1992) theory for intervention services. Some studies support the benefits of earliest intervention and have substantiated this claim by a regression in intellectual functioning in the absence of early intervention within the first 5 years of life (Guralnick, 1998; Simeonsson et al., 1987). Studies also support the benefits of intensive therapies that include active parental involvement, number of hours, duration of therapy and individual educational experiences (Ramey & Ramey, 1998).
Developmental Influence
One characteristic for success of intervention programs is a changing trend toward incorporating developmental treatment strategies (Howlin, 1990. Mirenda & Donnellon, 1989, Hauser-Cran; Prizant & Wetherby, 1998; Dawson & Osterling, 1997; Powers, 1992; Marcus & Schoppler, 1989; Ramey & Ramey, 1998). Stanley Greenspan (1998) proposes that the approach must consider how the child develops along multiple lines of development versus unitary lines of development to include physical, cognitive, emotional social and familial functioning.
Others also view a developmental perspective to treatment as fundamental in the nature of autism as a developmental disorder with impairment in cognitive, communication and social learning processes. Viewing autism as a disorder of development versus a psychiatric disturbance (Marcus & Schoppler, 1989; Powers, 1992) requires an understanding and sensitivity to the pervasiveness and chronic nature of the condition. Parents and professional need to expect developmental changes and anticipate associated problems with the child’s developmental progression.
Therefore, understanding the course of typical child development provides educators with an awareness of the discrepancy between where the child’s level of functioning is and where it ought to be. This knowledge serves to provide appropriate age level curricular context (Mirenda & Donnellon, 1989; Powers, 1992), and to properly challenge the child’s abilities yet maintain sensitivity to his or her inabilities (Bricker & Carlson, 1981).
Behavioral-Developmental-Ecological Approach
The literature also supports a behavioral-developmental-ecological perspective in program planning (Olley & Stevenson, 1989; Mirenda & Donnellon, 1989). Incorporating behavioral and developmental approaches in an ecological framework as a basis for instructional teaching bassists the child to function in his/her family and community (Mirrenda & Donnellan, 1989; Howlin, 1987; Salisbury et al. 1980; Bricker & Carelson, 1981; Powers, 1992). This element is critical to the family dynamic and level of familial stress involving the caring for a child with autistic spectrum disorder. Few interventions incorporate these three elements in their programming. It is important for intervention not only to incorporate these elements in their programming but also when doing so, to carefully match skills to the individual’s functional developmental abilities (Freeman, 1997; Mirrenda & Donnellan, 1987, Freeman, 1997)).
The literature on behavioral methods dominates intervention planning. Behavioral programs prevail for several reasons, in that they immediately address presenting behaviors associated with the diagnosis of autistic disorder (Nicholas, 1977; Powers, 1992). Additionally, behavioral programs conduct functional analysis of behavioral challenges to systematically reduce undesirable behaviors.
The efficacy of the behavioral approach, including operant learning principles has been popularized in intervention practice (Salisbury et al., 1980; Guralnick & Bennett, 1987; Lovaas, 1987). Operant principles that include reinforcement and shaping of behaviors are the cornerstone to most interactions with young children in intervention programs since the literature supports the efficacy of this approach. The implementation of operant learning techniques in behavioral styles of intervention, discriminate this approach from other therapies. Evidence has accumulated that early interventions characterized by variables that include, structure, high intensity, high-quality service provision and rigorous research design and planning, tend to produce significant positive effects (Freeman, 1997). Several key variables have been identified in the literature that affects intervention success. Although some studies have operationalized such variables, many have not evaluated the effects of these variables (Bush, White, Casto, 1982).
The literatures on developmental models rely on Piagetian principles. The premise here is child development, which follows an invariant sequence to mental growth. Evaluating the child’s developmental level of functioning is assessed through various standardized measures. The results thus act as a guideline for program planning where curricula are appropriately challenging to the child’s abilities.
An ecological influence aims to enhance the functional ability of the individual with autism. Here children are is taught functional life skills necessary for him or her to live in natural environments, as he or she requires skills. This differs from the developmental approach in that these skills are taught as the child needs these skills according to life situations, rather than relying on developmental criteria to decide when it is appropriate to teach these skills.
Thus, harmonizing these three approaches (behavioral-ecological-developmental) serves as a model that addresses the needs of the child and family. A multi-theoretical approach, diverse in nature, could be synthesized to address the complexities of autistic spectrum disorder (Mirenda & Donnellan, 1989).
Individualized Interventions
There is an abundance of literature that stresses the importance of individualized intervention planning (Salisbury, 1980; Mirrenda & Donnellan, 1987; Prizant, 1999; Howlin, 1987). Few studies however truly individualize their intervention approach to the child’s needs. Developmentally appropriate curricula based on formal standardized assessment results are scarce. Evaluating what level the child is functioning assists interventionists by targeting functional skills that are taught with the goal of moving the child along the developmental continuum (Adelman, 1986). According to Ramey & Ramey (1998), the individualized plan should be developed collaboratively by parents and professionals.
Comprehensive Approach
Comprehensive assessments are essential to programming. Results from the assessments are invaluable road maps for developmental clinicians to plan a course of intervention. Individualized programming based on these assessments ensures that the program accommodates the needs of the child. Thus, the literature supports a comprehensive clinical approach, where assessments are critical to the intervention process at various intervals throughout treatment (Greenspan, 1998; Mirranda & Donnellon, 1989; Prizant, 1999; Bricker & Littman, 1982; Ramey & Ramey, 1998). Comprehensive therapies that emphasize learning to generalize to other developmental domains are supported in the literature as having greater effects versus specific interventions with narrower focus (Ramey & Ramey, 1998; Thomas & Marshall, 1977; Howlin, 1987; Schopler et al., 1982). Research has shown that parents often feel overwhelmed and confused about the array of services offered to them, and often they perceive services as isolated events versus coordinated as a whole in intervention planning. Thus, interdisciplinary approaches provide parents with greater support (Thomas & Marshall, 1977).
Parental & Family Planning
There is a changing trend towards family-based planning (Schaefer & Briesmeister, 1989; Powell, 1990, Schoppler, 1982; Simeonsson et al., 1987; Powers, 1992), where parents are not only involved in program development and in the selection and prioritizing of goals but also play active roles as parent-therapists. Powell (1990) suggests modeling strategies to parents and role-playing to teach parent-child interactions. In doing so, parents will be provided with strategies and immediate feedback in their interactions and communication with their child.
The plan should also address levels of family characteristics (e.g., social supports, financial resources, marital status, family dynamics, parent-child interactions) (Guralnick, 1998) and potential stressors that could affect the development of the child during this intervention period. Coping with the needs of the family as a whole is often ignored in programming. Intervention outcomes have been significantly correlated with familial impact on the child’s development (Mahoney et al., 1998; Feldman, 1993). Feldman (1993) also noted significant increases in emerging language performance of infants and toddlers as a function of quality mother-child interactions. Interventions sensitive to parental and familial needs in planning should focus on increasing family adjustment and understanding the nature of their child’s disorder.
Additionally programs should promote patterns of management through ongoing support, parental counseling, education and guidance (Hewitt, 1977; Mahoney et al., 1998). “Effective parenting is a matter of learning, not instinct (Prickarts, 1971), particularly for the family of a young handicapped child, provision of a continuing parent support is a natural component of preventative and compensatory programs of early childhood education” (Briber, 1969).
High Quality Service Provision
High-quality service provision is key to the successful implementation of intervention services (Luce et al., 1992; Lovaas, 1987; Guralnick et al., 1987). Recruitment of qualified and competent staff, possessing advanced university degrees should control for this according to Luce et al., (1992). Staff should be well grounded in developmental principles in order to assess typical versus atypical development. Staff should also be able to function in an interdisciplinary team and be sensitive to familial functioning. Staff should be knowledgeable of available community resources in order to disseminate valuable information to families (Guralnick et al., 1987). Additionally staff orientation and training workshops and continual evaluation of service provision is essential for quality assurance.
Intensity
Research has exemplified an association between intensity and the children’s intellectual functioning (Guralnick, 1998; Dawson & Osterling, 1997; Simeonsson et al., 1987). Research has shown that children make considerable gains in intensive treatments versus less intensive. Intensity is often operationalized as the number of hours and frequency of intervention, duration of intervention, (Simeonsson et al., 1987; Dawson & Osterling, 1997). Duration is implicated since it suggests that the longer the intervention the comprehensive the programming. Duration affects also the family interaction patterns where the intervention acts as a buffer for family issues that arise over time (Guralnick, 1998). Thus, duration of intervention supports the family from interference of stressors that can negatively effect the development of their child. Interventionists can provide families with strategies and resources to assist in troubleshooting challenges as they arise by delineating undesirable behaviors, and therefore instilling a level of confidence for families in dealing with issues.
Social skills Development
New generation research proclaims the merits of pairing typically developing peers with children with autism (Howlin, 1987; Strain, 1987). In peer-mediated learning, active reciprocal interactions are taught by unobtrusively coaching the typical peer to engage in meaningful play relations with the child with autism (Dawson & Osterling, 1997). Research has shown that therapies often focus on cognitive development versus encouraging normal reciprocity within social interactions and increasing awareness for socio-emotional cues (Howlin, 1987).
Social skill development incorporating elements of spontaneous imaginative play are often not addressed in intervention research. Interventions that involved elements of play often focussed on pre-selected or pre-structured situations and activities (Dawson & Osterling, 1987). Imaginative play and spontaneous child-directed play is often an element not incorporated into program planning. Research stresses the importance of structured intervention (Lovaas, 1987; Rutter, 1973; Koegel et al, 1982). Consequently this contradicts the goal of teaching spontaneous play. Children with autism prefer to function with a level of structure in their routines. Less predictable situations are compounded with inappropriate behaviors and adjustment issues (Howlin, 1987). Therefore, the child often reverts to self-stimulatory perseverative behaviors that are soothing. Downtime, (where the child is not engaged in structured routines) is often associated with perseveration. Downtime periods could serve as opportunities for the child to explore independent unstructured play routines. Incorporating the development of reciprocal social interaction, social communication, social imagination and understanding into programming has important implications (Powers, 1992; Wing, 1988).
Outcome evaluations
Evaluations of maintenance and retention of learned skills and concepts are scarce. The lasting effects of the child’s gains due to the treatment intervention, exemplify the most successful interventions available. This is true however with extended follow up assessments since most studies tend to evaluate gains with follow up assessments soon after the studies were completed. To evaluate retention of treatment gains that would ensure effectiveness of the intervention, longer follow up assessments would reflect true effectiveness. It is generally understood that the therapeutic benefits reflected in the child’s behavior are specific to the environment learned, thus poor generalization. This difference in behavior either elicited or spontaneous is often not considered in outcome results. Therefore, outcome evaluations result in false claims of true effectiveness. It has been suggested in the literature that to observe true benefits of treatment, the outcome measures must span various settings to include spontaneous and elicited behaviors (Howlin, 1987; Guralnick et al., 1987).
Outcome evaluations are affected by the measures used to assess change in the children’s development. The debate regarding the appropriateness of psychometric measures versus observational measures to assess change in handicapped samples endures (Bricker & Lippman, 1982). Many psychometric measures have inadequate normative data on the performance of handicapped samples and are insensitive to specific changes in atypical children’s behavior and to close administrations. It is also argued that existing measures are insensitive to children with severe impairments such as autistic spectrum disorder, in that various testing items are inappropriate (eg. identifying objects that are not unfamiliar to the child) and communicating test items is very challenging (eg. delivering long oral instructions to a child who has poor comprehension) (Bricker & Lippman, 1982). Observation measures have been known to show greater therapeutic benefits. Howlin (1987), recommends using a combination of both psychometric and observational measures to assess therapeutic benefits since both complement one another and act as a comprehensive assessment of the child’s abilities. The literature also recommends the use of parental reports as part of the evaluative process to assess satisfaction regarding the child’s gains and overall impression of the intervention itself (Howlin, 1987).
In terms of follow up results of treatment effectiveness, the literature recommends a longer follow period to truly reflect the impression of treatment gains. Depending on the intervention approach, either a shorter or longer follow-up may result in impressive claims of treatment efficacy. Reports of significant gains in relatively short periods could be biased since behaviors may regress with length of time. Therefore, length of time affects claims made by treatment evaluation studies (Howlin, 1987). Thus, the greater the follow-up period, the more objective the findings. Follow-up assessments should also include individual and group results to measure change scores (Howlin, 1987). Individual scores reflect changes that may explain overall group scores if, for example, group scores were significantly high or low. Individual scores can also identify certain subgroups of children that either makes significant improvements or little gains. Therefore the treatment may benefit specific subgroups of children that share similar characteristics of autism.
(b) Design and Methodology
Many intervention programs have methodological/procedural, statistical and conceptual weakness. The purpose of this next section is to guide development of an efficacious intervention program incorporating better evaluative methods. Although some studies of intervention programs have employed rich empirical methods (Lovaas, 1973), true experimental design is ethically impossible to employ for individuals with developmental disorders including PDD. Many existing programs lack rigorous scientific design because of the ethical concerns and also the lack of agreement for children regarding the etiology, prognosis and best treatment, due to the heterogeneity of PDD. There are several problems with true empirical methods in clinical experiments. (a) The availability of measures may not be sensitive for children with special needs; therefore results become uninterpretable (Bricker & Lippman, 1982). (b) The ethical concerns of appropriate existing analytic designs that involve control groups of children for treatment (Kendall et al., 1999). (c) The difficulty with random assignment for conducting strict scientifically controlled experimental studies and the difficulty with comparative studies including population, variability and instrumentation difficulties (Kendall et al., 1999).
Research design is concerned with the attributing of success and effectiveness of the treatment to the program itself. Increasing internal validity of program design using methods that would seek to rule out alternative plausible causes supports attributing exclusive effectiveness of the treatment to the program. Programs can include controlled design features such as random assignment to treatment groups, including comparison or control groups, or waiting list controls, where a generated list of clients waiting for treatment could provide a control group (Rogers, 1989; Sheikopf & Siegel, 199;Ozonoff & Cathcart, 1998; Vincent & Salisbury, 1980).
More design features include pre-post and interim evaluation using a combination of qualitative and quantitative measurements and using built in multiple baseline or Meta evaluation. A good understanding of the value and effect of the treatment by using baseline or multiple baseline behaviors implicates the effects of treatment attributable to the therapy (Rutman, 1984). Data analysis relies on complete reliable research design. Incomplete or ambiguous information makes results difficult to interpret. Although some precautions can be taken to maximize internal validity, several specific threats may interrupt research evaluation methods, including attrition (participants leaving the program), maturation, selective-maturation (when different subjects are maturing at different rates), instrumentation, and statistical regression (unreliable pretreatment measures) (Rutman, 1984; Bricker, 1981).
Research design is also concerned with generalizability. The program should include a broad representation of clients (perhaps of different severity), locations and situations (teaching environments), and multiple teachers. Programming for generalization is critical for therapy maintenance to ensure learned skills are demonstrated to be consistent in various environments (Salisbury et al., 1980; Bricker, 1981). There should also be a focus on using natural materials in teaching and natural reinforcers for transportability. Learning is most meaningful when teaching is diverse and teaching involves relevant behaviors in its natural context (Powers, 1992). Thus the likelihood for spontaneous generalization of learned skills are more likely.
Another variable critical of design and methodology involves operationalizing all variables associated with the treatment. Therefore providing a complete description of program operation to determine content, strategies and circumstances helps to contribute success to the intervention itself. This involves detailing program philosophy, objectives, procedures, curriculum, functioning and strategies (Bricker & Litman, 1982). It also involves defining criteria for inclusion by describing research samples more systematically and carefully. Specifying and documenting intervention parameters that include details of intervention duration, intensity, subject characteristics (age, gender, race, SES, diagnostic characteristics/ co morbidity), location of treatment, the setting and family dynamics and family role(s) is also important (Kendall et al., 1999).
When describing the intervention, it is also important to describe the type of research design, outcome measures and procedures. It is also essential to describe criteria for therapist credentials and their involvement in the therapy including training procedures (Guaralnick & Bennett, 1987; Green, 1996; Dunst & Snyder, 1986; Adelman, 1986; Bricker, 1981; Howlin, 1998).
( c ) Evaluative Methods
The third component essential to the success of an intervention program involves its evaluative process. A thorough program evaluation can identify effectiveness of service delivery and determine whether objectives were met and the extent of positive effects from program delivery. It is essential to measure change in areas of cognitive functioning, but also in areas that include pro-social competence, social functioning and parental and familial functioning (Kendall et al., 1999; Ozonoff & Cathcart, 1998). Multiple tools used to measure change are required in order to objectively assess gains. Measures may include observation (in various settings/ direct with the child or indirect via video), anecdotal, standardized performance tests, ratings (parental and educators), and questionnaires. Kendall et al., (1981), suggests a multi-method assessment to measure change on two levels (a) the specifying level: that refers to specific learned skills demonstrated during formal assessments. (b) Impact Level: that refers to the global effects of treatment gains on generalized functioning.
Outcome evaluation and follow up studies support the efficacy of effectiveness interventions especially when clinical significance is obtained with long-term follow up. Therefore time of outcome and follow up evaluations affect statistical significant results. Outcome results assess developmental gains but are also important to assess reductions in autistic symptomology and type of school placement. Outcome valuation is dependent on design methods and design methods are dependent on elements that direct intervention. These three variables are critical in designing an intervention program and assessing effectiveness of therapy on children with autistic spectrum disorder.
