The integrated model of therapy service delivery is currently the preferred model for providing physical and occupational therapies in the school setting. Giangreco 1 defined integrated therapy as “the incorporation of educational and therapeutic techniques employed cooperatively to assess, plan, implement, evaluate, and report progress on common goals and needs.” These integrated therapeutic interventions occur during everyday routines in a child’s natural environment (ie, home, school, daycare center). 2 In 1977, Sternat et al 3 provided a description of two different models of school service delivery. They defined isolated therapy as therapy in which the child is removed from the classroom and treated in a separate therapy room. Additionally, they defined integrated therapy as therapy in which the child receives assessment and treatment in the classroom, with input from other school staff and the child’s parents. The child’s family participates, as actively as desired, in developing and implementing the child’s goals and objectives. 4
Since that time, integrated service delivery has been further characterized as more than treatment that takes place in the classroom since team collaboration is a necessary component of transdisciplinary planning and implementation. Rainforth and York-Barr 5 described this cooperative team approach to promote student achievement of functional skills. The Iowa State Department of Education also emphasized teamwork in their definition of the integrated model as one in which the therapist’s contact is with the child as well as the teacher, classroom or direct aides, and family. 6
Integrated therapy has been further conceptualized to describe various implementation strategies. McWilliam 7 described a continuum of consultative therapy services that ranges from totally segregated to fully integrated. This work articulated six dimensions and variations of service delivery models. Location of therapy service provision is the most visible dimension, as stated by McWilliam (ie, in or out of classroom), but other methods of integration can also be employed. For example, it is possible to remove a child from the classroom and still collaborate (integrate) with the classroom teacher on goals. Integration may also be attained when other children from the classroom are involved with the child in the therapeutic activity. This activity may occur in or out of the classroom. Dunn 8 further emphasized the significance of peer modeling and opportunities for children to practice in functional contexts using an integrated model of service delivery. When using an integrated approach, therapists observe classroom routines, identify a child’s motor needs for performing such activities, determine intervention strategies (which are shared with and taught to the classroom staff including teachers, aides, and other classroom-based providers), and write measurable goals to evaluate the child’s progress. The classroom staff collaborates with the therapist about instructional programming and developing discipline-free goals and objectives. These cooperative activities can facilitate improved motor outcomes; team members are able and encouraged to provide needed support or intervention (such as neurodevelopmental treatment (NDT) facilitation or physical cues) in the therapist’s absence. 2,9 Integrated approaches are used to encourage service provision for children in a functional context, in an environment where team members work together across traditional boundaries to offer a unified educational program. 10,11
Supporters of the integrated model of therapy service delivery have identified potential advantages of the model. York et al 2 contend that integrated therapy corresponds closely with PL 94-142, the original Education for All Handicapped Children Act, which mandates education for all children in the “least restrictive environment.” For a school therapist to help students achieve their maximum educational performance, he or she must observe how students function in their natural settings. These same authors proposed that the “least restrictive environment” should also apply to the location of related services provided in schools, not merely the school and/or classroom attended. 12 The integrated model is also consistent with provisions of the Individuals with Disabilities Education Act (IDEA), reauthorized in 1997.
In addition to citing the legal reasons that support integrated therapy, McEwen and Sheldon 13 advocated the model’s use for its support of motor learning principles, such as transfer of behavior, practice frequency, and part-whole practice. Giangreco et al 12 proposed that integrated therapy benefits the staff since, with a shared set of goals, all staff members become invested in problems, assisting each other when all members cannot be present. This may lead to more creative problem solving in addition to improved collaboration, knowledge, and skills of all the professionals working with the child.
Despite the rationale for the integrated approach to therapy, some challenges exist. Previously, struggles occurred related to the different backgrounds of the professionals involved (ie, educational or medical). 12–14 These differences in perspectives have diminished due to the current focus on functional outcomes for children, which are not discipline specific. The team, including the family, writes child goals and objectives, not therapy-specific objectives. 15 Furthermore, contemporary pediatric practice has incorporated the Disablement Model as a foundation for program evaluation and planning. This framework transcends practice settings, whether educational or medical. Methods to further enhance collaboration among professionals from different backgrounds and to prevent conflict that have been suggested are promoting open communication and providing time for team building. 5
Scheduling presents another challenge; a consistent scheduled therapy session (ie, day and time) does not allow the child and therapist to work on educational objectives under different conditions. The use of block scheduling has been suggested to provide flexibility, enabling therapists to assist students in various environments at various times of the day. 16
Although several authors have discussed the advantages and rationale of the integrated model of service delivery, 2,12,13,16 few empirical studies have been performed to date that have compared specific functional outcomes in preschool and school-age children exposed to integrated or isolated service provision. 17–19 McWilliam and Sekerak 20 surveyed physical therapists (PTs) practicing in early intervention settings to examine attitudes regarding in-class and out-of-class service delivery. The results indicated that therapists would use either model (or a combined model) depending on factors such as child characteristics, family preferences, individual treatment approaches (eg, NDT, motor learning), the specific service being provided, and the classroom placement. The integrated model was chosen most often.
Two similar studies showed no difference in outcomes achieved with the two different models, but classroom staff expressed a preference for the integrated model. 17,18 Cole et al 17 conducted a study to evaluate the effectiveness of in-class versus out-of-class physical and occupational therapy on motor skills of preschoolers with special needs. After a school year of receiving services, no significant differences were found between the two groups on the children’s standardized motor test scores, including the McCarthy Motor Scale and Battelle Gross and Fine Motor Scales. 17 A study by Dunn 18 examined differences between achievement of Individualized Educational Program (IEP) goals of 14 young children (preschool and kindergarten) receiving occupational therapy in a direct service or collaborative consultation model. Direct service in this study was provided in an isolated model for 60 minutes weekly. Therapists in the study’s collaborative consultation group did not give direct service to children but could provide a demonstration of techniques to the teacher, using predominantly verbal and written instructions. This collaborative consultation model is interpreted as an implementation of integrated service. Teachers and therapists in this group were required to spend 60 minutes weekly planning, observing each child’s behaviors, and discussing each child’s needs. No significant difference was found between the achievement of IEP goals in the two groups, but the teachers in the collaborative model attributed success of IEP objectives to the therapists in 60% of cases compared with 36% of cases for the direct-service therapists. Despite findings in these studies of no difference between groups, the preference for the integrated model by classroom staff is an important result. PTs and occupational therapists (OTs) are often members of an integrated IEP team, and the reported preference for the integrated model suggests benefits to the child that were not measured by IEP objectives or standardized test scores.
In another integrated model outcomes study, Karnish et al 19 studied improvements in the gait of three subjects with spastic quadriplegia (aged four, six, and 14 years), using the variables of upright posture, stride length, foot clearance, and speed. The authors of the study concluded that physical therapy provided in the natural setting is more effective than in the isolated setting for improving gait speed and quality. The authors also noted that routine obstacles were encountered in the natural settings that children had to negotiate and that distractions in the natural setting did not have an adverse effect on the child.
Limited evidence can be found in the literature related to motor and functional outcomes in preschool and school-age children who have received therapy services using integrated models. Conditions and variables that influence therapists’ selection of integrated or isolated models for school-aged children have not been explored beyond the work of McWilliam and Sekerak, 20 which focused on infants. The effectiveness of the integrated model for children with varying characteristics also has not been studied. Evidence that informs pediatric PTs and OTs about the benefits and limitations of integrated models is essential to validate current pediatric therapy practice, given the emphasis on inclusion and natural environments.
The purpose of this study was to gain information on the use of integrated models by pediatric PTs and OTs that provide therapy services to children in the greater Rochester, NY, community. The study also compared outcomes, specifically percentage of IEP motor and functional objectives achieved by subjects in the study.
All agencies providing therapeutic services to children in the region were identified. A nonrandom sample of 220 pediatric physical and occupational therapists working in early intervention, preschool, and school system settings was obtained through recruitment of participation of their employing agency. This participation was negotiated through a signed Letter of Cooperative Approval for those facilities without an Institutional Review Board. The therapists represented 11 facilities within a 30-mile radius of Rochester, NY, and included those providing service to children in urban, suburban, and rural settings. The Research Subjects Review Board of the University of Rochester Medical Center and the Human Subjects Review Board of Ithaca College approved the project.
The authors developed a questionnaire to gather information regarding opinions and the use of the integrated model by pediatric physical and occupational therapists in the greater Rochester, NY, region. Therapists also identified child characteristics (eg, age or diagnosis) and environmental variables (eg, teacher support) that influenced their selection and use of integrated models. They also reported their observations during their experience as pediatric therapists regarding the effects of environmental conditions (eg, space, appropriateness of the therapy activity in the context of the classroom’s naturally occurring activity), and child variables (eg, distractibility) on the effectiveness of the model. The questionnaire was distributed as a pilot study to a sample of medical center faculty and community-based supervising PTs and OTs. Some of the initial respondents were included in the final sample. The survey was revised and reviewed again.
Practice Setting: Caseload Characteristics and Amount of Integrated Service
The questionnaire sampled basic caseload information of pediatric physical and occupational therapists in the region, with regard to diagnoses and number of children served, age ranges, predominant service delivery model used, and opinion and preference of the type of service delivery model used. We requested responses based on the respondent’s “predominant” service model selection due to our belief that many therapists are “blending” integrated and isolated approaches, depending on specific child or environmental conditions or variables. Respondents were also asked to identify the philosophical framework of the agency or practice setting. For example, does the agency set policy regarding treatment in classroom settings? Do child needs or environmental variables (such as limited space in the classroom) determine whether an integrated model is best for each child? Therapists were also asked about their opinions regarding child characteristics that encouraged a better response to integrated therapy.
Review of IEP Outcomes: Integrated or Isolated Therapy Service
The second part of the questionnaire requested the participation of the therapist to recruit a subject for the outcomes part of the study. This outcomes portion involved retrospective review of IEPs on children for whom the treating therapists had used either an integrated or isolated service delivery model. Each therapist who participated in the questionnaire was invited to participate in the outcomes portion of the study. Of the 109 therapists who completed the survey (from the 11 participating agencies), 47 identified and contacted parents, and successfully obtained parent signatures for informed consent. This allowed retrospective review of each child’s therapy records. The children whose records were reviewed ranged in age from three years to 13 years 10 months. Of the charts of the 47 children identified for inclusion in the study, 41 charts were actually reviewed and included in the data analysis. The data for the remaining six subjects were not included for analysis for one of the following reasons: 1) incomplete data were provided or 2) described treatment model did not meet criteria for inclusion in integrated or isolated data groups.
Specifically, the authors gathered outcomes data on successful completion (ie, fully met the mastery criteria) of specific IEP objectives. The treating therapist developed the IEP objectives in collaboration with the educational team and family, which were agreed on at the time of the IEP meeting. They also made the selection of service model at the outset of the school year’s plan of care. The objectives were characterized as either gross motor or fine motor to allow a categorical comparison.
The criterion for placement of subjects into the integrated group was a majority of treatment (>50%) provided in natural settings. The criterion for placement into the isolated group was a majority of treatment (>50%) provided in settings other than the subject’s typical educational settings(s). The word majority was included in the definition for placement of subjects in groups because many therapists are blending their approach to service delivery. The percentage of IEP objectives achieved was the variable measured and described.
Comparisons were made between year-end outcome measures from 1996 to 1997 and 1997 to 1998 based on data collected through retrospective chart review from the records of children receiving either integrated or isolated therapy services in center-based settings. Comparisons were made between groups (integrated vs isolated) on successful achievement of IEP objectives.
Simple summary statistics (means and percentages) were calculated. Comparisons were made between PT and OT respondents. To test for statistical significance of any differences observed in the data between groups, the Mann-Whitney U test or Fisher exact test was used where appropriate. All tests were two sided and evaluated at an α level of 0.05 throughout.
Of the 220 questionnaires sent out, 107 (49%) questionnaires were returned, representing a total of 2,400 children who receive intervention from the responding therapists. Respondents included 45 PTs and 60 OTs. Two respondents did not identify their discipline, and the data from these two individuals were not included in any discipline comparison statistics (PT or OT). The settings in which services were provided included homes, center-based early intervention classrooms, preschools, day-care centers, and schools. The respondents worked in suburban, urban, and rural settings. Respondents were employed by two school districts, five regional agencies such as the Board of Cooperative Educational Services (BOCES), and the Association for Retarded Citizens, three voluntary agencies (eg, United Cerebral Palsy and an independent children’s center), and one private practice. The participating agencies in this study were within a 30-mile radius of the city of Rochester, NY.
Description of Regional Practice
For those participants whose caseloads were fully specified in the responses of the surveys, the percentages of current caseloads were calculated as a weighted average of the figures that each respondent reported; results are summarized in Table 1. Overall respondents reported that 24.7% of children on their caseloads were treated predominantly in their classroom or natural setting (ie, integrated setting). The percentage of children in a predominantly integrated setting reported by OTs was higher than that reported by PTs (27.8% for OTs, 20% for PTs). The therapists reported that 55.3% of the children in their caseloads were treated predominantly in isolated settings (eg, therapy rooms, free hallways). The percentage of children being treated in predominantly isolated settings by PTs (61.6%) was higher than that reported by OTs (50.1%). The remaining children (20%) were treated with a blended approach, not characterized as predominantly integrated or isolated. Differences in the caseload percentages between OTs and PTs were found to be statistically significant (p = 0.0001). It was possible that individual children for whom reports were obtained in this study received treatment from both a physical and occupational therapist. Since the survey responses were anonymous and the children categorically referenced in part A of the survey were not coded or identified, we did not know the number of children receiving both services.
Respondents were asked how many total hours per month were spent meeting with teachers for planning; these hours are displayed in Table 2. Forty-seven percent of therapists reported that less than one hour per month was spent for this purpose. Differences were noted in monthly time spent for planning by therapists using predominantly integrated models (mean, 4.1 hours) or isolated models (mean, 2.4 hours). These differences were found to be statistically significant (p = 0.0173). Several therapists remarked that there was not enough time in their schedule for this important activity (see Table 2).
Therapists’ attitudes were evaluated regarding integrated therapy using a five-point Likert scale (Table 3). More than 65% of respondents strongly agreed with the statement “Child needs should determine where service is provided.” Approximately 49% of respondents agreed and 37% strongly agreed with the statement “Child needs and other variables (space, equipment, schedule) should determine where service is provided.” Forty-three percent of respondents disagreed and 28% strongly disagreed with the statement “The facility and/or agency should set policy regarding where service is provided.” Approximately 38% of respondents disagreed with the statement “Classroom teacher preference should determine where service is provided,” and 26% of respondents were neutral on this statement. Comparison of OTs and PTs showed no significant differences in the way that these questions were answered, although we noted that differences in two of the four questions approached statistical significance (p < 0.1).
When asked to select a statement that best reflects their agency’s philosophical framework (Table 3), 66% of respondents chose the statement “Child needs and other variables (space, equipment, schedule) should determine where service is provided.” More than 27% of respondents chose the statement “Child needs should determine where service is provided.” The differences between OTs and PTs were not significant.
The therapists were next asked a question that assessed the respondents’ belief about the meaning of an integrated model approach to a plan of care for a child. They were asked to choose which of three possible definitions of integrated therapy most closely resembled their perception of integrated therapy. Eighty-nine percent of respondents chose the definition “Therapy provided during regular ‘program’ activities/schedule, with inter- or transdisciplinary interaction between staff.” We developed this definition with reference to the characteristics of the model as described by Giangreco et al. 12 The survey authors developed alternative definitions to characterize situations that might resemble an integrated model but would still not “integrate” with the child’s natural activities. These alternatives included “on-site therapy where child attends program” and “multidisciplinary treatment session.” One hundred percent of respondents stated that they would “use integrated models with those children for whom it is most beneficial, considering child need and environmental variables,” ie, using integrated approaches when optimally suited for the child or classroom characteristics. This response was chosen over alternatives that advocated using an integrated model with all children or using an isolated model with all children.
Caseload Representation: Diagnosis and Age
The total number of respondents to questions specific to interventions provided in center-based settings was 90. The children who were most frequently served with an integrated approach were children diagnosed with developmental delays/disabilities who represented 84% of the 90 responses (see Table 4). The largest proportion of children served with an integrated approach were between the ages of five and 10 years (62%), followed by children aged 10 to 21 years (46%) and children aged three to five years (43%).
Perception of Child Improvement
The respondents who indicated that they use an integrated approach for any children in their caseload were asked what percentage of children that they believe improve with an integrated approach. The category 70% to 75% was most frequently reported (25%) followed by 90% to 99% (19%). Eight therapists (9%) reported that they believed that 100% of children improve their function with an integrated approach.
Variables Influence Effectiveness of Approach
Some therapists who participated in the survey responded that they did not believe that diagnosis or age strongly influenced the effectiveness of integrated therapy. Twenty-six percent expressed the belief that cognitive, psychosocial, and environmental variables were more influential than age or diagnosis on the integrated model’s effectiveness. The psychosocial variables included such things as child behavior, cooperation, motivation, and attention span. Also, the child’s ability to accept treatment in an integrated setting was noted to have an influence on his or her participation. Some therapists noted that initial individualized (isolated) sessions allow the therapist and child to establish a rapport and the therapist to gain a “true sense of the child’s capabilities.” They stated that initial individual sessions supported eventual success with an integrated approach. The environmental factors included such things as classroom structure and activities that promote inclusion of therapy activities. Some specific advanced gross motor skills (eg, stair negotiation, use of playground/gym equipment) were noted to be difficult to integrate into a child’s regular activities, if the planned time and therapy activity were not appropriate to the planned classroom activity (eg, math, reading). Additionally, some respondents believed that the cooperation of teachers and aides was the key to successful outcomes.
Outcomes: Chart Review
Nearly half of the subjects (19 of 41, 46%) received both services (physical and occupational therapy), and the models used were different. For example, a child may have received physical therapy using an isolated model and occupational therapy using an integrated model. If data were complete in both areas, subjects were included in both data sets. Therefore, a total of 60 data sets were gathered and analyzed and included 25 instances of the integrated model and 35 instances of the isolated service delivery model. Objectives were categorized as either gross motor objectives or fine motor objectives. Examples of gross motor objectives include measures of balance, motor planning, stability/strength, gross motor developmental milestones, and range of motion. Examples of fine motor objectives include measures of sensorimotor processing, self-help, and bilateral hand skills. Accomplished objectives were calculated as a percentage of the total objectives for typical gross motor or fine motor area (including example objectives listed above). This percentage of objectives met is presented in Table 5. Differences between the groups defined by integrated and isolated were not found to be significant for either gross motor (p = 0.655) or fine motor (p = 0.567) objectives.
The responses of participating PTs and OTs shared certain themes and observations regarding integrated service delivery models. A majority of therapists characterized their primary treatment approach as isolated intervention. However, nearly 45% of therapists used a predominantly integrated or blended approach (ie, incorporating integrated and isolated sessions) to service delivery. Many described the use of both integrated and isolated approaches to therapeutic interventions, varying the model within the week or month. We believe that therapists incorporate integrated sessions whenever feasible, as part of “best practice” in compliance with the intent of IDEA. Related services that are “designed to meet their unique needs” in educational settings should not be prescriptive in the model or approach. Physical and occupational therapy interventions lose the potential for child-specific individualization if all therapy is provided in the same way. Therapists are presumably observing and acting on those conditions and variables, which, at times, may make an isolated approach more effective or more feasible. Furthermore, with McWilliam’s 7 identification of a continuum of therapy services, we can understand that it is indeed possible to “pull a child out” (of the classroom) and still provide integrated therapy service through a collaboration with the classroom teacher’s goals. Location and model do not necessarily describe the same thing.
The attitudes of therapists and agencies (as perceived by therapists) revealed discrepancies regarding the philosophical orientation toward the service delivery model. A majority of therapists reported that child needs should determine choice of approach or model. However, the therapists also reported that the agency’s philosophy was that the child’s needs and other environmental and mechanical variables (eg, space, schedule) should determine choice of model or approach. Potential scheduling and professional conflicts may arise during situations in which a discrepancy exists between team members on the choice of service delivery model. All therapists reported the use of an integrated approach when it was beneficial for the child. Even therapists who reported more frequent use of a traditional, isolated model indicated (through written comments) their incorporation of integrated components in their treatment plans, in an individualized manner, for individual children.
Respondents to this survey indicated that classroom staff is willing to work with therapists on classroom integration of a program of therapeutic interventions. However, reports of mutual team planning time suggest that nearly half of the respondents had less than one hour per month for this essential component. The importance of adequate time for team planning must be emphasized. Integrated programs do not just happen. They require a strong team approach, with frequent updates and interdisciplinary exchange of information to maximize program effectiveness. 21 The difference in reported planning time between groups (predominantly isolated or integrated) is significant. Severely limited planning time would be a constraint for those therapists and classroom teachers who want to develop schedules and strategies that promote integrated therapy throughout the child’s typical daily activities. Programs should include planning time needed for therapists and teachers to effectively implement integrated therapy approaches.
This study provides information on current therapists’ perceptions and attitudes toward the integrated model of service delivery and variables that influence the use and effectiveness of the model. Comparison of outcomes (successful attainment of gross and fine motor IEP objectives) revealed no significant differences between groups treated using different models. We noted that the types of objectives reported included a mixture of impairment- and function-related outcomes. Ketelaar et al 22 reported the benefits of a functional therapy program in which emphasis is placed on practicing functional activities versus emphasis on enhancing quality of movement. The mixed focus (ie, impairment and function), particularly for gross motor objectives, may have influenced our findings. Although improvement in impairment level goals might lead to functional improvement, this can only be known with certainty if functional measures are also used. Practice opportunities for motor learning are most available when functional objectives are targeted using integrated approaches; this presumably supports optimal outcomes for children on functional objectives. Several limitations of the study must be considered. We recognize that the treating therapists did not assign the type of service delivery randomly. Therapists also recruited children for the retrospective chart review. It is possible that bias occurred because therapists may have a preference for one model or may have recruited a child who had demonstrated particularly good or particularly poor progress on the measures used.
Amendments to IDEA in 1997 (PL 105-17) reiterated the expectation that services be provided in natural environments for children of ages birth through two years (part C of IDEA) and in the least restrictive environment for children three years through school ages (part B of IDEA). 23 Integrated and inclusive programs offer service delivery models that provide a service system response to that challenge. Cognitive, psychosocial, and environmental variables were reported to be more influential than age or diagnosis on the integrated model’s effectiveness. The psychosocial variables included such items as child behavior, cooperation, motivation, and attention span. Cognitive status and relative distractibility of the child were reported as factors to consider when choosing to intervene in the classroom. 23 The relevance of the child’s affective state to the effectiveness of the therapy session should serve as a valid defense for “individualizing” the care provided. For some children, initial individualized sessions may help to establish a rapport and determine a clearer sense of the child’s capabilities. Careful attention to both of these details should support eventual success with an integrated approach, given the assumption that the child’s needs are appropriately addressed using an integrated model. Some specific advanced gross motor skills including balance and locomotor tasks were noted to be difficult to integrate into a child’s regular activities if the planned time and activity for therapy were not appropriate for the planned classroom activity (ie, math or reading). Classroom schedules, routines, and activities have been reported to be a key to successful integrated intervention. 23 Additionally, it was believed by both OT and PT respondents that the cooperation of teachers and aides was essential for successful outcomes. Further research is strongly encouraged to determine whether particular variables may help select the most appropriate service delivery model for children. Differences between physical and occupational therapy services should be studied in terms of perceived benefits and actual child outcomes to determine whether integrated services are more effective in one discipline than in another. A cost/ benefit analysis of the variations along the continuum of service delivery models would also be useful. In the interest of “best practice” and the provision of exemplary physical and occupational therapy care to children, providers need to know the conditions under which integrated services across the continuum of implementation possibilities are justified and most appropriate.
The authors gratefully acknowledge the participating agencies and physical and occupational therapists, as well as participating parents, without whom the project could not have occurred. The agencies include Monroe BOCES #1, Monroe-Orleans BOCES II, Kids 1st, Rochester City School District, Mary Cariola Children’s Center, Greece Central School District, Finger Lakes United Cerebral Palsy, Pediatric Therapy Services/United Cerebral Palsy of Rochester, Livingston County ARC Children’s Services, Genesee Valley BOCES, and Wayne-Finger Lakes BOCES. The authors also appreciate assistance with preparation of the manuscript provided by Gregory Liptak, MD, MPH, and Deborah Nawoczenski, PhD, PT, in addition to technical assistance provided by Jeff Houck, PhD, PT.