The Impact of Adaptive Seating Devices on the Lives of Young Children With Cerebral Palsy and Their Families

Participants 

We invited parents and their young children who were clients of our facility and 2 other regional children's treatment centers. Children who participated had a primary diagnosis of CP with function categorized as GMFCS level III or IV14 and were between 2 years 6 months and 7 years 6 months of age. (Children with GMFCS level III or IV CP can sit upright independently or with support, but generally need pelvic or trunk support or an adaptive seating device to optimize hand function. Children with level III or IV function either need an assistive device to walk or have limited independent mobility in a manual wheeled mobility device.)

We conducted screening interviews to identify and recruit parents who were primary caregivers of the child with CP. We defined a primary caregiver as an adult who provided at least 10 hours of direct supervision per day as determined by self-report. Furthermore, we recruited only parents whose child did not use specialized postural control devices at home for floor sitting, chair sitting, and toileting activities. We mailed eligible and interested parents a letter to explain the project protocol and their roles as a research participant. Caregivers who agreed to participate in the 12-week long home trial provided signed consent.

We expected to have difficulty recruiting 30 families from our facility alone, so we involved 2 other children's centers to increase our participant pool. We adopted 2 different recruitment strategies: one for our facility and another for the other 2 centers. At our institution, we reviewed our electronic medical records and identified 155 children who met the initial inclusion criteria. We mailed a brief introductory letter about the study to the parents of these children and then telephoned families 1 to 2 weeks later. A randomized selection process provided each family an equal opportunity of being contacted. Of the 143 families we contacted, 85 families did not satisfy the screening requirements, 14 families declined because of scheduling conflicts, 9 families met the screening criteria but later decided not to participate after receiving the detailed study information letter, 5 children were deemed clinically inappropriate for the study devices as judged by the research occupational therapist during the first home visit, and 25 families participated in the full home trial.

To maintain the confidentiality of families at the other 2 centers, site clinicians reviewed their own medical records to identify children who met our age and diagnostic criteria and identified 46 potential families. Site administrative staff mailed these parents an introductory letter inviting them to contact our study coordinator if interested. Seventeen parents contacted our coordinator; 11 of these did not meet the initial screening criteria, 1 child was judged clinically inappropriate at the first home appointment, and 5 families participated in the trial.

Main Outcome Measures 

We previously developed the multidimensional, parent-report FIATS to detect the impact of assistive technology device use on the lives of children with physical disabilities and their families. The FIATS measures this impact by the contribution of 8 related constructs (subscales) that include child autonomy, caregiver relief, child contentment, doing activities, parent effort, family and social interaction, caregiver supervision, and safety. These constructs tap into aspects of child and family life that assistive technology devices may influence, such as the degree to which a child can perform activities independently (autonomy), interacts with others (family and social interaction), and requires attention from family members (supervision).

Parents use the FIATS to indicate the degree to which they agree or disagree with items on a 7-point Likert scale. The FIATS also contains items that contribute to a ninth independent subscale (technology acceptance) to measure parents' general receptiveness to assistive technology devices for their children. We modeled this subscale as a separate moderating construct that may temper the impact of technology on family life. Overall, the 9 subscales contribute 64 items to the FIATS.

Scoring on the FIATS is calculated by the sum of the means of the 8 related subscales. Because the range of each mean subscale scores is from 1 to 7, the overall range of FIATS scores is from 8 to 56. Lower FIATS scores are associated with lower child and family functioning on these dimensions. Because we designed the measure to detect changes in important aspects of family life that could be influenced by the introduction of seating devices, higher change scores suggest an overall positive impact on child and family life as defined by these constructs, whereas lower change scores suggest a negative effect on child and family life.

The FIATS has good content validity and face validity15 and acceptable internal consistency (α=.94) and test-retest reliability intraclass correlations (ICCs=.92; 95% confidence interval, .86–.95) for a 2- to 3-week retest period when used with families of young children with CP. Furthermore, the FIATS 9 subscales have Cronbach alphas between .64 and .92, and ICC point estimates of test-retest reliability between .77 and .92.16

Because the FIATS was an emerging measure of the impact of assistive technology devices on the lives of children and their families, we administered the standardized IFS to answer our research question and test the association between the 2 measures. The IFS measures parents' perceptions of the psychologic and social consequences of having a child with a chronic disability. It is a 15-item, single-factor measure with a 4-point response scale ranging from “strongly disagree” to “strongly agree.” The IFS had good internal reliability (α≥.88) and acceptable construct validity when completed by caregivers of young children with chronic physical conditions such as CP.17, 18 Scoring is calculated by the sum of the items' scores, with a total scoring range of 15 to 60. Lower IFS scores indicate lesser parent-perceived psychosocial consequences of having a child with a disability.

We selected the IFS to administer with the FIATS because it measured a related construct, had acceptable psychometric properties, and had been used in other research studies with families of children with CP. Parents completed the IFS at the end of each stage of the study. Because the original version of the IFS was developed in 1978,18 we made minor modifications to modernize its item wording. We changed “child illness” to “child condition” on 7 items that referred to the child with CP and changed an adjective on 1 item to ask the parent to compare their child to a “typical” child rather than a “normal” child. We retained the original wording on all other items. We confirmed that the internal consistency of the modified IFS for the first test administration (n=30) was acceptable (α=.94) in our study.

Secondary Outcome Measures 

At the first appointment and every other week during the trial, the research therapist either met with parents or contacted them by telephone to complete the home activity log, a semistructured interview guide developed for this study. The purpose of the log was to obtain parents' reports of the types and impact of activities in which their child engaged during the week. We developed the home activity log to capture child activity reports that were organized into occupational performance domains of self-care, productivity, and leisure. Parents reported how seated play, personal care, and recreational activities or behaviors changed over the week preceding the interview. During the intervention phase, the therapist asked the parent to tell her whether the study devices contributed to any changes in child and family activities. The therapist made field notes during the interview to record the perspectives of parents. We collected this additional information to help interpret the FIATS and IFS scores.

The therapist administered other standardized measurement scales during the home appointments. However, the primary purpose of these other scales was to explore the effect of the intervention on each child's functional performance, parent satisfaction, and caregiver assistance using individualized goals set by the parent. We focus on the main outcomes of the study here and discuss the outcomes of these individualized goals elsewhere.

Protocol 

One of 2 experienced occupational therapists working as research therapists for the study conducted interviews in the homes of participating families at 4 different times over 3 phases that lasted 12 weeks. The baseline phase was 3 weeks long, the intervention occurred over the following 6 weeks, and the final postintervention (return to baseline) phase lasted 3 weeks.

At the start of the baseline phase, the therapist gathered basic demographic information such as the number of family members, the school schedule of the child, and the number of hours the parent worked out of the home. The therapist used a demographic interview form developed for the study.

Parents completed the FIATS twice during the initial baseline phase: once at the first home visit and again 2 to 3 weeks later. Immediately after the second FIATS administration, we provided 2 postural control devices, the Flip2Sit activity seata for floor sitting and table level activities, and the Aquanaut toileting systema for toileting and grooming in the bathroom, for use by the child with CP. Clinicians and parents reported in earlier pilot studies that both devices provided functional support for a variety of home activities for young children with CP.19, 20 We also provided families a simple, self-standing breakfast trayb so children could play on an elevated surface while they sat in the activity seat at floor level.

At the start of the intervention stage, the research therapist observed the child and used her clinical judgement to evaluate the suitability and safety of the products, installed a universal toilet seat adapter on the family's toilet, and showed parents how to use the products with their child. The therapist provided the products' owner's manuals and asked parents to read them to be sure they understood how to use the devices.

The therapist administered the FIATS a third time 6 weeks later at the end of the intervention stage. Based on previous experience measuring the impact of specialized seating devices on child functioning,8, 9 we chose a 6-week period to allow adequate time for the family and child to adjust and develop regular routines using these devices. At the end of this stage, the therapist reclaimed the products and met with the parent 3 weeks later to administer the FIATS a fourth and final time. We offered both devices, free of charge, to the families after the measures were administered at the end of the study.

Statistical Analysis 

The first 2 administrations of the FIATS during the baseline phase were used in our larger study to estimate the test-retest reliability of the FIATS. To reduce the number of post hoc paired t tests, we estimated the effect of the study devices by comparing the mean scores for the FIATS and IFS for the second, third, and fourth test administrations by using a 1-way repeated measures ANOVA. We used a type I error probability of alpha equal to .05 (2 sided) for the repeated-measures ANOVA and a Bonferroni correction for multiple statistical testing during post hoc analyses.

The extent to which any measured effects are meaningful for families is arguable because no criterion standard methods or instruments exist for estimating a minimally important difference.21 As recommended for quality of life studies, we considered an important effect size to be a magnitude of at least 50% of the SD of the mean difference scores.22, 23

We lacked the statistical power to make inferences about the mean difference scores for each of the 8 subscales and the technology acceptance subscale of the FIATS. Hence, we conducted exploratory analyses of trends in the mean subscale scores over the 12-week study period.

Descriptive statistics from the demographic form provided a profile of participants, and a descriptive review of entries on the home activity log provided contextual information for our primary findings.

Demographics 

Twenty-nine mothers and 1 father participated in the study. Twenty-five parents had at least 2 children, 1 of whom was a child who met our inclusion criteria.

The mean age of the 30 children with CP was 4 years 6 months (range, 2y 6mo–6y 7mo). Twenty-four children attended school or daycare during the week. Of these, 11 attended school or daycare for 5 days per week, 11 attended for 3 or 4 days per week, and the remaining 2 children attended 2 or fewer days each week. Eleven children attended kindergarten, 8 attended nursery school, and 2 attended first grade. The remaining 9 children were cared for at home or in a daycare during the week and were too young to attend school.

Most children used some form of assistive technology device in the home. The most common device was a wheelchair or stroller with specialized seating. When not in the wheelchair, children were typically supported by an assortment of pillows, modified juvenile equipment (eg, a highchair or car seat), or family members. Homemade devices, such as modified potty seats and corner seats were less common, but were used occasionally to provide alternate positioning for children in the home.

Seventy-seven percent of the families had 5 or more people living in their homes. Although 15 parents who participated worked outside of their homes, only 4 had full-time jobs. All 12 families who had in-home support received 7 hours or less of caregiver assistance per week.

Effect of Intervention 

Because the ANOVA's assumption of sphericity was not met, we adjusted the degrees of freedom by applying a Greenhouse-Geisser correction and found significant mean differences among the overall FIATS scores (F1.4,40.6=19.25, P<.001). Post hoc testing confirmed significant mean differences in overall FIATS scores between baseline and intervention and intervention and postintervention phases. The magnitudes of our mean difference scores were 75% and 91% of the SD of mean change scores (table 1).

Table 1. Descriptive Statistics and 97.5% Confidence Intervals for Mean Score Differences for the FIATS at the Baseline, Intervention, and Postintervention Phases

	Phases	Mean ± SD	Absolute Difference of Means ± SD	Ratio of Difference to SD of Difference	97.5% CI of Difference
	Baseline	32.3±6.3
			4.0±5.3	.75	1.7–6.3
	Intervention	36.3±6.0
			5.2±5.7	.91	2.7–7.6
	Postintervention	31.2±6.3

Abbreviation: CI, confidence interval.

Table 2 provides the descriptive statistics for each test administration of the IFS. Tests of within-subject effects yielded no significant difference among the IFS mean difference scores (F2,58=1.63, P=.20), with an observed power of 33% and a corresponding type II error probability of beta equal to 67%. However, we found significant negative correlations (–.68≤r≤–.54) between the FIATS and the IFS total scores at the ends of week 3 (end of baseline stage), week 9 (intervention stage), and week 12 (end of postintervention stage) (table 3).

Table 2. Descriptive Results for IFS at the Baseline, Intervention, and Postintervention Phases

	Phases	Min	Max	Mean±SD
	Baseline	17.0	50.0	35.8±9.8
	Intervention	18.0	52.0	36.1±9.4
	Postintervention	22.0	54.0	37.5±8.8

Table 3. Correlations (Pearson's r) Between the FIATS and IFS Scores at the Baseline, Intervention, and Postintervention Phases

	Phases	Baseline IFS	Intervention IFS	Postintervention IFS
	Baseline FIATS	−.54†	−.55†	−.64‡
	Intervention FIATS	−.44⁎	−.54†	−.52†
	Postintervention FIATS	−.59†	−.57†	−.68‡

⁎ P<.05. † P<.01. ‡ P<.001 (2-tailed).

Figure 1 provides a plot of the mean scores for the FIATS subscales over the entire study period. Steeper positive slopes indicate a greater positive impact on child and family life between administration periods, whereas steeper negative slopes suggest a greater negative effect on child and family life between periods. Higher mean values indicate greater subscale contribution to the overall impact on families as measured by the FIATS.

View Large Image Download to PowerPoint

Fig 1. Mean scores at the baseline (weeks 1 and 3), intervention (week 9), and postintervention (week 12) phases for the technology acceptance and FIATS subscales.

Our descriptive analysis of the home activity log interviews suggested that 26 children used the activity seat and 24 children used the toileting system over the entire intervention period. Twenty-seven parents said that the use of 1 or both study devices increased the quantity and/or quality of their child's activities at home. For 8 families who did not use both study devices over the entire period, their principal reasons were that 1 or both devices did not provide sufficient postural support for their child and/or their child rejected the device.

Twenty-five parents reported that their children were happier, more autonomous, and/or able to interact more with siblings and other family members while using 1 or both of the study devices. Nearly three quarters of these parents noted that their children's activity levels reduced after the devices were returned. During both the baseline and postintervention stages, almost half of the parents said that their children lied prone or supine on the floor or sat with the support of other family members. Most of these parents suggested that playing with their children in these positions was more difficult and occurred less frequently.

Parents said that their children used the activity seat for productivity and leisure activities primarily, whereas the children mostly used the toileting system for self-care activities. Nineteen parents said that the duration, quality, and/or variety of play increased when their child used the activity seat. Seated activities reported only during the intervention period included coloring, playing with siblings, eating at the dinner table with family, watching television, and having a tea party with siblings.

Twelve parents reported that their children were unable, unwilling, or not ready to have a bowel movement while sitting on the toileting device during the intervention stage. These parents said that their children used the toileting product as a positioning device for playing or grooming activities in the bathroom. Because about one third of the children had not started toilet training when they were enrolled in the study, parents reported more supervision of their children in the bathroom during the intervention stage than during the other 2 stages. Three parents said that their children who began to use the toilet during the intervention phase continued to ask to use the toilet after we withdrew the study devices.

Twenty-five of the 30 participants accepted our offer of one or both devices at the end of the postintervention phase.

Study Limitations 

Although we showed important effects on family life because of the intervention, our study had limitations. Our study design was subject to more measurement error than more rigorous methodologies such as crossover designs or randomized controlled trials. Our sample size was sufficiently powered to detect an effect caused by the introduction of the study devices using the FIATS. However, it was neither large enough to make inferences about the effect of the technology as measured by the FIATS subscales nor sufficient to judge the parent-perceived effect of the devices on psychosocial outcomes as measured by the IFS.

Participants in this study were mainly mothers of young children with CP, so mothers of children with different types of disabilities and other caregivers may have different views of the impact of assistive technology devices on family life. Moreover, we recruited parents whose children did not use specialized seating devices for toileting and sitting on the floor or at a table, but most families we contacted had children who used a device for at least one of these activities. Consequently, our sample may not be representative of the general population of families who have young children with CP.

Parents may have responded in ways that reduced the measured effect within dimensions that reflected on their abilities as parents. This self-enhancing bias may explain why we found lesser contribution toward assistive technology impact from dimensions such as caregiver relief and parent supervision. Conversely, parents who participated may have responded in ways that made it seem that the assistive technology impact was greater than the actual effect on family life. However, it could be argued that we should have also found significant changes in the mean ratings of the IFS if such a systematic bias existed.