Based on the systematic review (11) and the APSA guidelines for surveys (not exclusively regarding pediatric population), together with a summary of the characteristics of the surveys used in the healthcare field (20), a comprehensive guideline to be used for developing our ad-hoc survey was developed.

The first step in the survey design process is the clear definition of the research question and the selection of an appropriate study design, either descriptive or analytical. In the present study, a descriptive design was adopted, as the survey does not test a predefined hypothesis but aims to systematically collect and report data to identify general trends, as well as the incidence and prevalence of outcomes of interest.

The presented survey focuses on patients' needs, expectations, and experiences concerning both traditional rehabilitation practices and rehabilitation with advanced robotic and digital technologies. Basing on the target population which is represented by patients with a wide span of ages and different clinical pictures, existent surveys regarding patients' rehabilitative needs were analyzed: they are focused on specific diseases, in particular stroke (21–23), multiple sclerosis (24, 25) or other specific pathologies. Moreover, they involved a limited number of respondents and for the majority of questionnaires they referred to usability (26), and above all to technologies using virtual reality (25–27). The idea of this survey is instead to focus on the acceptance, barriers, satisfaction and willing to use different RADT for a wide variety of clinical conditions and by specifying the rehabilitative domain, as well as register the current traditional treatment carried out in alternative or in association to the use of RADT.

To ensure inclusivity and conceptual consistency, the International Classification of Functioning, Disability and Health (ICF) was adopted as the guiding framework. Developed by the World Health Organization (WHO), the ICF provides a standardized and internationally recognized classification system for health and disability. It supports a shift from a purely biomedical perspective to a biopsychosocial model, emphasizing the interaction between health conditions and contextual factors, including personal and environmental influences. By integrating biological, psychological, and social dimensions, the ICF enables a comprehensive understanding of functioning and disability (28, 29).

In line with the ICF framework, the survey presented in this paper has been designed to capture a multidimensional understanding of rehabilitative care and needs. The explored domains extend beyond physical and motor aspects to include social and environmental factors, reflecting the ICF's biopsychosocial approach. This allows for a more comprehensive evaluation of both traditional and technological interventions, enabling the study to provide insights into the broad spectrum of rehabilitative care. By using the ICF as a guiding framework, this survey aligns with the latest advancements in health research, offering a structured and universally applicable approach to understanding health, disability, and functioning and, in turn, patients' needs.

To make the ICF more practical for everyday use, the World Health Organization (WHO) and the ICF Research Branch (https://www.icf-research-branch.org/icf-core-sets/category/8-neurological-conditions) developed a method for creating the ICF Core Sets, based on a rigorous scientific process, including preparatory studies and input from a multidisciplinary team of experts. These Core Sets are designed to simplify the description of functioning in various contexts, such as clinical practice, by offering lists of key categories relevant to specific health conditions and care settings.

To ensure inclusivity and relevance in the developed survey, ICF Core Sets have been used, focusing on elements shared among various pathologies. The ICF Core Sets relevant to the clinical conditions encompassed by the Initiative were analyzed, drawing on resources such as those available at the ICF Research Branch (https://www.icf-research-branch.org/icf-core-sets/category/8-neurological-conditions). This involved a detailed analysis of all relevant Core Sets, including both disease-specific core sets and the generic rehabilitation ones.

In accordance with the COSMIN framework, the development of the survey was grounded in a clear definition of the construct of interest and its underlying domains. Using the ICF as the conceptual framework, all identified rehabilitation needs across different areas of functioning (such as motor, cognitive, etc) were systematically listed and assigned to overarching domains reflecting the primary construct they were intended to measure (e.g., “solving problems” within the cognitive domain).

To support content validity, a structured mapping process was conducted following the identification of individual rehabilitation needs. A dedicated table was developed to verify the relevance and representativeness of each ICF Core Set with respect to the specific clinical profiles addressed by the Fit4MedRob project: Cerebral Palsy, Multiple Sclerosis, Parkinson Disease, Post-Stroke, Acquired Brain Injury, Spinal Cord Injury, Peripheral neuromuscular diseases, Post-Oncological Surgery. When disease-specific ICF Core Sets were unavailable, the generic Core Set was applied. Throughout this process, iterative consultations with healthcare professionals and patient and parent associations were held, both in person and online, to ensure that all relevant aspects of functioning were adequately covered and that no important domains were omitted.

Informed by multidisciplinary stakeholder input, the most clinically relevant and prevalent functional domains were selected, ensuring alignment with both the target populations and the intended functional objectives of the technologies under investigation.

Subsequently, ICF categories were grouped into higher-level domains based on conceptual similarity, in line with COSMIN recommendations for domain structuring. For example, ICF categories related to walking and moving with or without assistive devices were grouped under the “mobility function” domain, capturing different modes of movement across environments. Similarly, categories describing changes in body position (e.g., sit-to-stand transitions) were included within the “postural function” domain.

Following this domain-structuring process, six main domains were identified:

mobility function;

At various stages of the process, patients' and caregivers' associations were asked to join the process of surveys' creation. Professionals within the Fit4MedRob Consortium and patients' and caregivers' associations met in person and during online meetings, to collaborate in the co-creation of the surveys.

All individuals involved in the process were asked to review the survey to ensure clarity of the content. This was achieved by sharing a working draft of the questionnaire in a collaborative document, allowing comments, suggested modifications, and insights to be discussed and incorporated into the final version, which was newly tested in the pre-testing phase by various Fit4MedRob personnel, both clinically and non-clinically oriented (details below).

In details, the role of the representative of the associations was to ensure that all the specific needs were represented, by eventually integrate the ICF Core sets, moreover they checked the language of the survey to be inclusive and respectful. When the survey was developed as a draft, they tested it with the aim of verified its feasibility in terms of time-consuming effort and this contributes to the ramification architecture of the survey (described below) for saving time and focus on the most important rehabilitation domains. Finally, this shared process helped in raising awareness of the importance of data collection for the further disseminative phase.

In summary, the development of the survey followed a structured workflow informed by the Donabedian criteria (Structure-Process-Outcome) and guided by COSMIN (Consensus-based Standards for the selection of health Measurement INstruments) criteria and in particular:

Structure: According to the Donabedian framework, the structural phase involved defining the resources, expertise and context required for survey development. An interdisciplinary team composed of clinicians, rehabilitation engineers, and researchers collaboratively identified the key domains of interest. Existing literature on technology adoption, rehabilitation robotics and user needs analysis was reviewed to ensure that the survey covered relevant constructs.

To allow respondents to freely express their opinions the survey was anonymous. On the other hand, demographic characteristics are pivotal in the data analysis to stratify respondents in homogeneous categories; for this reason, a first general section has been planned with some generic questions about individual characteristics, such as the age range and referral clinical center (data that cannot lead back to the patient's identity).

Then, following the ICF framework, specific needs associated with various functional domains were categorized, aligning each need with its overarching domain and grouping questions about the same topic together, as suggested by the literature (30):

Mobility. It aims to define different mobility situations related to various movement conditions, such as walking forward, backward, or laterally, moving on different surfaces (indoors and outdoors), navigating or overcoming obstacles, climbing and descending stairs, running, and jumping. It also includes movements using specific devices designed to assist with movement, such as manual or electronic wheelchairs, walkers, and canes.

Each of the six domains was investigated by a 5-point Likert scale analysis, to gather the following key parameters:

Independence: assessment of patient's independence and need for assistance or assistive devices, based on common classification systems like the Gross Motor Function Classification System (for movement abilities) and the Manual Ability Classification System (for hand use).

Then, in the final part of the questionnaire, at the end of all six domains and the economic section describe d below, subjects are asked to rank the domains in order of importance, i.e., to list them hierarchically from a subjective perspective.

Guided by the ICF framework, which emphasizes inclusivity across biological, social, and environmental domains, and considering that the target population comprises both children and adults with disabilities, different versions of the survey were developed. This allowed to assess the use of rehabilit ation technology across different age ranges and without excluding any group of participants with sensory, motor, and/or cognitive disability. More precisely, the following three surveys have been created:

Survey for collaborative adult patients, able to answer independently;

The adaptations have been made with the purpose of maximizing the survey's success and ensure respondents provide honest and meaningful answers. In particular, based on the principle of clarity and precision (33), vague terms like “usually” or “few” were avoided, opting instead for phrases that are temporally specific, such as “once a week” or “at least once a year”. Furthermore, visual aids were incorporated to clarify questions and enhance comprehension. In addition, an accessible language has been used: clear and widely understood terms were selected, considering their frequency of use in everyday language, to ensure accessibility and inclusivity. By designing the survey this way, the ICF's principles were followed, making sure it captured the views of the different characters involved in rehabilitation, without leaving anyone out.

Regarding the formulation of the questions, a 5-point Likert scale was selected to collect precise and detailed data (13).

Then, with the aim of avoiding excessive length of the survey, questions were structured with a ramification logic. Although all questions were mandatory, many were displayed only if specific conditions were met. Specifically, following the consent request, the first section collected general demographic information using closed-ended questions to ensure respondent anonymization while still capturing key characteristics such as age range. The collected information included age range, gender, clinical condition (selected from a predefined list with an “other” option for further specification), rehabilitation setting in which the child was receiving care, region of residence within Italy, and the child's reported familiarity with technology.

For each of the six domains, respondents were asked to rate the child's level of independence and the impact of limitations within that domain using 5-point scales. Subsequently, current rehabilitation treatments received during the previous year were investigated. First, traditional rehabilitation treatments (i.e., without the use of technological solutions) were assessed using five response options: (i) no, and this is acceptable; (ii) no, but the respondent would like to; (iii) partially (no more than once per week); (iv) regularly (twice per week or more); and (v) extensively (intensive treatment). For the latter three options, respondents were also asked to report their level of satisfaction and the treatment setting (inpatient, day hospital, or home-based). For each setting, response options included: (i) no, and this is acceptable; (ii) no, but the respondent would like to; (iii) yes, but the respondent is not satisfied; and (iv) yes, and the respondent is satisfied.

In cases where no traditional rehabilitation treatment was reported, the same set of questions was applied to technology-based rehabilitation. When a technological intervention was ongoing, respondents were additionally asked to specify the name(s) of the devices used. At the end of each domain section, and whenever a technological treatment was present, respondents were asked about their willingness to receive more information about rehabilitation technologies.

This conditional logic was applied consistently across all six domains, so that questions were displayed only when relevant (i.e., in the presence of ongoing treatment), thereby optimizing questionnaire completion time and respondent burden. This structure, developed with valuable input from parent associations, was designed to collect detailed, domain-specific information while minimizing questionnaire length and effort.

After completing the six domains, respondents were asked to report the economic impact of rehabilitation treatments over the previous year, using five cost ranges (0; <€500; €501–€2,000; €2,001–€5,000; >€5,000).

Finally, in the concluding section, respondents were asked to rank the six domains according to perceived priority, and an open-text field was provided to collect additional comments.

The survey is structured across multiple pages, with a progress bar displayed to inform respondents of their completion status.

The survey structure is summarized in Figure 1.

As suggested by existing guidelines, the surveys were pre-tested with various Fit4MedRob personnel, both clinically oriented (such as physicians or therapists) and non-clinically oriented (such as engineers). Specifically, the surveys were tested by 5 physicians, 8 physical therapists, and 4 biomedical engineers. Personnel was selected to represent a range of backgrounds and expertise. Additionally, similar feedback was sought from patients' and caregivers' associations regarding the readability, length, and potential misunderstandings of survey contents.

As a result, minor revisions to the wording were implemented, including correction of typographical errors and rephrasing to enhance clarity. In addition, the completion time was tested and estimated to range from approximately 20 to 35 min, depending on the branching logic activated by respondents' answers.

Among the potential approaches to administering a survey, such as face-to-face interviews, telephone interviews, self-completion questionnaires and computer-assisted methods (20, 34), there is no clear evidence that responses to sensitive questions differ between the various administration methods and no single method consistently outperforms all others. In the described Initiative, for a self-completion approach, utilizing an internet-based tool (Microsoft Forms) was chosen to ensure accessibility for all users via their personal smartphone, computer or tablet. The use of technologies which includes websites and smartphones has been suggested as increasing engagement of patients in goal-setting and definition of the rehabilitation content (35). This also facilitates the completion of a comprehensive survey avoiding taking a significant time and guarantees more freedom in disclosing personal feelings and impressions. An internet-based tool allows choosing a convenient time for starting and completing the survey and avoiding the challenge of obtaining responses on sensitive topics in presence of an interviewer. Thus, an online tool also facilitates the dissemination of the survey. In fact, distributing survey links via social media accounts of individual users and organized e-groups with an interest in specific health issues may increase the engagement and accuracy of responses (36). Institutional channels of all the clinical centers involved in the Fit4MedRob Consortium and of the several patients' and caregivers' associations were identified. Additionally, social media of all the clinical centers involved in the Fit4MedRob Consortium were used and in-person events (such as the Fit4MedRob conferences showing ad-hoc QR codes) were organized, to facilitate engagement.