Psychometric properties of the PEDI-CAT for children and youth with spinal muscular atrophy

1 Introduction

Spinal Muscular Atrophy (SMA) is a neuromuscular disorder and is most often a result of a homozygous deletion in the SMN1 gene [1]. The SMN1 gene codes for the production of the SMN protein which maintains normal function of the motor neurons. Insufficient levels of the SMN protein leads to loss of motor neurons in the spinal cord and brainstem and produces muscle weakness and compromised respiration. The three most common types of SMA vary in presentation and severity. SMA type I is the most severe with onset of symptoms during the first six months of life [2, 3]. Typically, children with SMA type I have severe muscle weakness, hypotonia, difficulty with breathing and feeding, and do not develop motor skills such as sitting or crawling [4–6]. Children with SMA type II are usually diagnosed by age 2 years and initially can maintain sitting independently and often use a wheelchair for functional mobility [3, 6]. SMA type III is a milder phenotype and is usually diagnosed by age 3 years. However, in some cases it doesn’t present until later in childhood. Children with SMA type III can walk independently although some may lose the ability to walk as they get older [3, 6]. Ultimately, all individuals with SMA will experience some level of difficulty with mobility and activities of daily living and require the support of a rehabilitation team [5–7].

The natural history of disease progression in SMA has been documented at the capacity level [6–12]. Capacity tests are commonly used to record what a child can do in a clinic or test setting [13]. Several capacity-based motor tests have been validated for individuals with SMA including the Children’s Hospital of Philadelphia Infant Test of Neuromuscular Disorders [14, 15], Test of Infant Motor Performance Screening Items [16], Hammersmith Functional Motor Scale-Expanded [17, 18], Revised Hammersmith Scale [19], Motor Function Measure [20], Revised Upper Limb Module [21], and 6-Minute Walk Test [22, 23]. These capacity-based measures are critical when evaluating natural history [13] but they do not describe typical motor function in natural settings such as in the home, school, and community.

Performance-based tests provide information about what an individual does in their natural settings [13]. Differences in motor capacity and motor performance have been documented for children with cerebral palsy. This information is useful when designing interventions [24]. Very little information is available on typical functional performance of individuals with SMA in their natural settings [25]. The Pediatric Evaluation of Disability Inventory-Computer Adaptive Test (PEDI-CAT) is a performance-based outcome measure completed by caregiver report. It is designed to identify functional delay or limitations and to measure change in function in children and youth with disabilities up to 21 years of age. The PEDI-CAT has been validated for children with developmental disabilities [26–29]. The full item bank properties of the Mobility and Daily Activities domains of the PEDI-CAT were evaluated specifically for children and youth with SMA [30]. There appear to be no other studies that have been published on the psychometric properties or feasibility of using the PEDI-CAT in children and youth with SMA.

In December 2016, nusinersen (Spinraza®) was approved by the Food and Drug Administration (FDA) for the treatment of infantile-onset and childhood SMA [31]. In May 2019, onasemnogene abeparvovec-xioi (Zolgensma®) was approved to treat children less than two years of age with SMA [32]. In August 2020, the orally administered small molecule risdiplam (Evrysdi^TM) was approved by FDA for all patients with SMA older than 2 months of age [33]. Improvements in motor capacity and survival have been reported following these therapeutic interventions [33–41]. It is the experience of these authors that parents have reported improvements in their child’s typical function following use of nusinersen, risdiplam, or onasemnogene abeparvovec-xioi; yet, some of these improvements are not captured by capacity measures.

With new pharmacological therapies and rehabilitation interventions in development, valid and clinically feasible performance-based measures are needed to document typical function in home, school, and other community settings in individuals with SMA. These measures can serve as clinical trial endpoints and provide justification to consumers and insurance companies for these potentially costly interventions. The purpose of this study is to examine the psychometric properties of the PEDI-CAT Daily Activities and Mobility domains in a sample of children and youth with SMA. Specifically, this study evaluated the: 1) convergent validity (ability of the two domains to measure the construct of motor severity), 2) discriminative properties (ability to detect limitations in motor skills compared to typically developing children and youth), 3) floor and ceiling effects, and 4) clinical utility of the PEDI-CAT.

2 Methods

3 Results

Demographic information for the sample is provided in Table 2. This data set included 96 participants for the Daily Activities domain and 95 for the Mobility domain. The PEDI-CAT was completed by parents (n = 85), other caregivers (n = 2), or by clinicians in conjunction with the child’s parents (n = 9). Most tests (96.9%) were administered in English with the rest administered in Spanish. Descriptive statistics for PEDI-CAT scores for the Daily Activities and Mobility domains and the Wheelchair subdomain are provided in Table 3 by SMA type and motor classification. Significant differences in the ages of participants were found by SMA type, with SMA type III participants being significantly older than SMA type I or SMA type II participants (p = 0.001). No statistically significant age differences were found among the three motor classification groups.

Rehabilitation professionals need valid and clinically feasible outcome measures that can be used to identify delays or limitations in functional performance and to document the function of children and youth with SMA of varying ages and abilities. This study demonstrates that the PEDI-CAT has good psychometric properties to support its application as a performance-based caregiver-reported outcome measure in children and youth with SMA [37].

The PEDI-CAT Mobility and Daily Activities domains were able to identify differences in functional skills for participants with different SMA types and motor classifications, demonstrating convergent validity. This is clinically important as children and youth with SMA differ in functional abilities across types and motor classifications. In addition to SMA disease factors, age can impact scaled scores since older children without disabilities have higher levels of function and therefore have higher scaled scores than younger children. Participants with SMA type III were significantly older than participants with SMA type I or II. However, no statistically significant age differences were found among the three motor classification groups which supports the ability of the PEDI-CAT to measure differences in Daily Activities and Mobility function according to SMA motor severity.

The PEDI-CAT was able to detect limitations in mobility function (more than 2 standard deviations below the mean) in 89.5% of the sample and mild limitations (1-2 standard deviations below the mean) in 8.4% of the sample. Therefore, 97.9% of the sample of children and youth with SMA had some limitations in mobility function. This result is consistent with the natural history of motor limitations in SMA [7–11], demonstrating good discriminative properties of the Mobility domain of the PEDI-CAT. Limitations in function were also identified in the Daily Activities domain. Just over half of the sample (53.1%) had a limitation in daily activities function and 20.8% had mild limitations. This is consistent with the typical pattern of weakness found in SMA with more involvement in the lower extremities than in the upper extremities [12]. Preservation of wrist and finger flexors and extensors and biceps muscle function [12] may allow children and youth to have the ability to perform daily activities skills such as brushing teeth, cutting fingernails, and using a computer keyboard; but these children or youth may still experience limitations in mobility activities of walking up stairs or running.

Floor effects were present for a small group of participants (≤10.4%) in this sample. A few participants with SMA type I or Non-sitters received the lowest possible score on the Daily Activities and Mobility domains. This was due to the content and level of difficulty of the items. The easiest items in the Daily Activities domain are related to swallowing liquids and chewing food [42]; however, these items typically are difficult or impossible for children with SMA type I and Non-sitters [6] The Mobility domain addresses basic movement skills such as turning head in supine, lifting head in prone, and rolling. These activities are difficult for children with SMA type I or Non-sitters [4]. These findings are similar to those of a study that used Rasch methods to calibrate the PEDI-CAT Daily Activity and Mobility full item banks [30]. A lack of easier items for individuals with minimal motor function could potentially limit test sensitivity to small yet important changes after interventions. To improve responsiveness of the PEDI-CAT, items which are meaningful to children and caregivers could be added.

Some of the parent-suggested items address functional capacity but could be represented at a performance level of measurement. Examples of performance level items include maintaining their head in the upright position in supported sitting in the classroom while listening to the teacher, operating a computer using eye gaze or touch screen to complete homework tasks, or moving around in the community with friends using a powered wheelchair.

A ceiling level score on the Daily Activities domain was only observed in one 21-year-old participant with SMA non-ambulatory type III (classified as a Sitter) who was independent with activities of daily living including dressing, eating, and bathing. He presented with proximal upper extremity weakness but had better distal strength. He compensated for his weakness by using adaptive equipment or modifying tasks. Also, he demonstrated good functional performance in his home environment. Even though there was only one participant who had a ceiling level score, it suggests that the Daily Activities domain may not capture disease burden for older individuals with SMA Type III. Therefore, harder items could be added to improve responsiveness.

Thirty-nine children in this sample were power wheelchair users and 31 were manual wheelchair users. However, the wheelchair items were only administered for children who had some ability to self-propel (n = 15). The Wheelchair subdomain only contains 12 items and excludes power wheelchair tasks, limiting its ability to record a wide range of function and potentially limiting its ability to measure changes in wheelchair function. Physical therapists frequently evaluate equipment needs, consult with families about equipment use, and teach children how to use power wheelchairs. Therefore, items on positioning and sitting tolerance as well as independence with power mobility are missing content that could be added to the wheelchair subdomain to improve its ability to document function related to wheelchair use.

Overall, the clinical utility of the PEDI-CAT was quite good as evidenced by the short administration times. The PEDI-CAT was completed by caregivers during clinic visits while children and youth were participating in capacity measures or being seen by specialists. Clinic physical therapists reported that in some cases, caregivers stopped answering questions on the PEDI-CAT to observe their child or to discuss rehabilitation needs with the team. These times might be higher than if the caregiver completed the test without any distractions.

Caregivers mostly provided positive feedback as to the meaningfulness of the items on the PEDI-CAT and their willingness to complete additional tests. Switching from the Content-Balanced to the Speedy administration option would improve administration time without sacrificing score precision. Also, it might improve parent feedback on item relevance since a different algorithm is used to efficiently select 15 or fewer items. Anecdotally, therapists reported that they selected the Content-Balanced option because they wanted to refer to the item map. Toward the end of the study, therapists determined that the Content-Balanced option was not helpful for children and youth who were non-ambulatory or diagnosed with SMA I or II since the content areas are hierarchical and many of the items were too hard for participants and did not need to be administered. Use of the Speedy option prevents or decreases the number of irrelevant items especially for children and youth with lower function since item selection is not made according to content areas and item level of difficulty, but strictly by item level of difficulty. The newer versions (version 1.3.9 and later) of the PEDI-CAT software include an item map for both the Speedy and Content-Balanced options [42].

In summary, the PEDI-CAT is a valid, discriminative, and feasible performance-based caregiver-reported outcome measure for children and youth with SMA. This study supports the clinical application and use of this outcome measure for research purposes. Future work should include prospective collection of longitudinal data on the PEDI-CAT in all SMA types to study its sensitivity to change and potential responsiveness to effective interventions. Adding easier but functionally relevant items to the PEDI-CAT would reduce the floor effect for children and youth with SMA type I. Also, it would be helpful to explore the addition of power wheelchair and positioning items to the wheelchair subdomain to improve its applicability to a wider range of wheelchair users.