According to Taiwan's People with Disabilities Rights Protection Act promulgated in 2007 (1), the local government in Taiwan should issue the disability identification and provide welfare and services based on the framework of the International Classification of Functioning, Disability, and Health (ICF) (2, 3). Taiwan is one of the pioneer countries to use the ICF Chapter code as a basis for the classification of disability and for the application of ICF in the disability eligibility determination system (DEDS) (4–7). In Article 5 of the People with Disabilities Rights Protection Act, people with disabilities refer to those who, with the following deviation or loss resulting from physical or mental impairments, are limited or restricted to be engaged in the ordinary living activities and participation (AP) in the society; and they, after processes of evaluation and assessment by the committee composed of professionals from medicine, social work, special education, and employment counseling and evaluation, can be regarded as suffering one of the following malfunction categories and issued a disability identification. The eight disability types are listed as follows: (1) Mental functions and structures of the nervous system; (2) Sensory functions and pain, the eye, the ear, and the related structures; (3) Functions and structures of/involved in voice and speech; (4) Functions and structures of/related to the cardiovascular, hematological, immunological, and respiratory systems; (5) Functions and structures of/related to the digestive, metabolic, and endocrine systems; (6) Functions and structures of/related to the genitourinary and reproductive systems; (7) Neuromusculoskeletal and movement related functions and structures; (8) Functions and related structures of the skin. The eight disability types or malfunctional categories corresponding to the eight chapters of body functions and structures of the ICF. The statement of the Article 1 of that act also emphasizes the focuses of societal participation in Taiwan government services. The societal participation therefore is one major outcome of welfare and services for people with disabilities. Besides, the ICF proposes that disability, or participation restriction, is the results of interaction among disease, body function and structures, environmental factors, and personal factors (2). In Taiwan's past disability evaluation system, the disability identification was issued after physician diagnosis and impairment examinations only. In the current DEDS system, the disability identification will be issued after a more comprehensive evaluation of body function, AP, and environmental/personal factors and needs assessment by a team (4–7). To assess the status of AP for adult applicants, the World Health Organization Disability Assessment Schedule 2.0 (WHODAS 2.0) (8) was then adapted by the ICF team.

After 5 years' preparation, the ICF-based DEDS have launched nationwide since July 2012, and the disability identification is issued based on the results of the ICF-based disability evaluation by a medical team from the authorized hospitals and on the results of the needs assessment from the local social welfare department (4, 5). The content of the disability evaluation includes tests related to body function and structure codes as well as AP components of the ICF. To assess the status of AP in the ICF-based DEDS, the ICF taskforce group have developed the Functioning Disability Evaluation Scale (FUNDES) since 2007 in Taiwan (5–7). The FUNDES has adult version (FUNDES-Adult) and child version (FUNDES-Child). The FUNDES-Adult has been developed with bilingual translation based on the 36-item interviewer-administered version of the WHODAS 2.0 (8), the mobility trial version for functional assessment (9), and environmental ICF codes (10) that designed and revised at the evaluation tool development phase for the DEDS (6). The FUNDES-Child has been developed based on the Child and Family Follow-up Survey (CFFS) (11).

To take into account, the localization and instrument validity of nationwide application, the changes and revisions of the FUNDES has been on-going based on the feedback of the field testers and experts or from the data analyses nearly every year. Several versions of the FUNDES have been developed (6, 12). Some psychometric properties of the previous FUNDES-Adult, version 5 (13, 14) and FUNDES-Child version (11, 15–17) have been examined and published. The previous construct validity of the FUNDES-Adult version only included the items of 36 items which are the same as WHODAS 2.0 and to follow the same 6-domain structures (13, 18). However, during the revision process, the measurement items have been added for collecting more AP information for people with disability and we did not re-check the psychometric properties of all items in the FUNDES-Adult yet. With the changes in the purpose and background of the assessment tools used, the factor structure of the inspection tools is very much necessary (19), which can assist us in the understanding and analyzing intervention impact of the disability.

This study aimed to examine the factor structures of AP part of the seventh version of the FUNDES-Adult (FUNDES-Adult, version 7) which was revised base on the Taiwan culture and for the purposes of the disability eligibility determination. The FUNDES-Adult, version 7 has applied to the DEDS for a long time, and it is also the basis and reference basis for all future versions and with large number of accumulated cases. Although we also published and used the eighth version later, this version differs only in typography. In the seventh version, we put all the information in one book, including ICF introduction, detailed description of the content and psychometric properties of the FUNDES-Adult and FUNDES-Child, test items, evaluation forms, and flash cards. In the eighth version, there are one FUNDES Manual and one FUNDES Item Booklet. The streamlined FUNDES Item Booklet was designed to assist the testers to evaluating the applicants in the field directly, which made it was not necessary to take the whole heavy FUNDES manual. All questions and measurement procedures of the FUNDES-Adult, version 7, were the same as that of the eighth version, so we have chosen the FUNDES-Adult, version 7, as the target of this study.

Each item of the Domain 1 to 6 of the FUNDES-Adult, version 7, has two dimensions—performance and capability—to measure AP in daily life over the previous 30 days (13, 14, 18). The performance refers to the extent of restriction on participation in daily life and the qualifier of performance is described as what an individual does in his or her current environment. Since the current environment always includes the overall societal context, performance can also be understood as “involvement in a life situation” or “the lived experience” of people in their actual context. As the assistive devices and other persons' assistance changed the performance difficulty level of persons with disabilities (20, 21), for the disability eligibility determination, the FUNDES team decided to add items with capability dimension in the FUNDES-Adult (5). The capability refers to the extent of restriction on participation in a real environment without assisting by any assistive device or persons. The capability dimension therefore captures the extent of difficulty in daily life without environmental supports (13). Therefore, the purpose of this study was to examine the factor structures of two dimensions of the FUNDES-Adult. We expected that the factor structures of performance and capability dimensions of the FUNDES-Adult, version 7, would be different.

This secondary data analysis study was part of a larger national survey conducted in Taiwan by the DEDS team (7). The present study was approved by the Research Ethics Committee of the Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation (IRB104-04-A;IRB107-46-B) and Joint Institutional Review Board, Taipei Medical University (TMU-JIRB), Taiwan. The de-identified data were retrieved from the database of the Taiwan Databank of Persons with Disability (TDPD) that included 144,850 adult-times who received the DEDS assessment in 201 authorized hospitals from November 2013 to January 2015 (6, 11, 15, 22, 23).

Certified testers associated with authorized hospitals in Taiwan administered the FUNDES-Adult, version 7, by interviewing the applicant's caregivers or him/herself (24). The certified testers were professionals licensed as physical therapists, occupational therapists, speech therapists, social workers, clinical psychologists, counseling psychologists, nurses, audiologists, special educators, and vocational evaluators. To ensure the number and quality of FUNDES testers, training programs were funded by the Taiwan central government with recruitment of all these licensed professionals, especially those in the DEDS hospitals. The training programs for certified testers covered the procedures of the DEDS and regulations (30 min), introduction to ICF and ICF-CY (30 min), introduction to assessment instruments [FUNDES-Adult (60 min), FUNDES-Child (40 min)], practice of assessment instruments (200 min), and the web-based platform for entry and storage of data (30 min). At the end of each training course, a paper-and-pencil test was administered to certify the attending professionals (24). By the end of 2014, there were about 7,700 certified testers in Taiwan (12). Names, identification (ID) number and other related information of all certified FUNDES testers are kept in the FUNDES tester personnel dataset for manpower quality control in the DEDS (12).

The population was 88,124 adults with disabilities. Their mean age was 53.02 ± 16.84 years old, ranged from 18 to 110 years. The EFA of psychometric properties of the FUNDES-Adult, version 7, was examined based on 8,730 adults with disabilities (aged 52.98 ± 16.81 years), and the other 500 people with disability (aged 54.30 ± 16.81 years) were used for CFA (Table 1). There were no significant differences in gender, age, disability type, and severity of disability between the EFA and CFA samples (p > 0.05) (Table 1). For disability type of our samples, most of them were with impairment of “mental functions and structures of the nervous system.” We compared the samples between those with and without that particular disability type, respectively, there was no significant differences in every type (p > 0.05).

Participation is one of the most significant outcomes of rehabilitation, social, and educational interventions (81). Using a large nationwide DEDS sample, the results of this study provided evidence of construct (structural) validity of the AP part of FUNDES-Adult by using both EFA and CFA for adults with disabilities and aged more than 18 years. The 5-factor structures of two dimensions of the FUNDES-Adult, version 7, are similar to our previous finding of FUNDES-Adult, version 5 (13). The items loading on each of the five factors of the performance dimension reflected the following domains: Performance of learning and interaction, Performance of mobility and self-care, Performance of housework, Impact of health on participation performance, and Performance in work/school; and that of capability dimensions are: Basic capability, Capability of housework, Capability of learning and interaction, Impact of health on participation capability, and Capability in work/school. This study also confirmed the second-order factor structure of the FUNDES-Adult. However, the items of each factor of two dimensions were different slightly and the names of the five factors were also different. The hypothesis of this study was partially supported. In Taiwan DEDS, the 5 domain scores and an overall disability score in two dimensions could be computed, each ranging from zero (indicating no perceived disability) to 100 (indicating maximum perceived disability). These disability scores are important for disability practices, research, and policies in the future in Taiwan.

Although Üstün et al. proposed the WHODAS 2.0 36-item version has a second-order factor structure in all cultures and populations tested (82) and it shows a mild correspondence with the theoretically six-domain structure. Such factor structure also found in the performance dimension of the FUNDES-Adult of persons with spinal cord injuries (18). However, the six-domain structure did not be confirmed by some previous studies (31, 58, 68, 70, 83). The possible reasons for somewhat different factor structures among the studies include differences in samples, influence of cultural and language differences, item numbers, and testing procedures between the initial WHODAS 2.0 and the Chinese FUNDES-Adult.

For example, the answers of “D3.4 Staying by himself or herself for a few days” in the WHODAS 2.0 were usually “not applicable” in Taiwan disability population. One of the reasons may be that most Chinese families are seldom core families and for those with disabilities are usually cared by other family members. They do not have the opportunities to be alone in our daily life in the past 30 days. These results are similar to the spinal cord injury studies which used the WHODAS 2.0 (65) or the FUNDES-Adult (18). Nielsen et al. (71) also found that aged persons discharged from emergency department in Denmark had more than 15% missing data in item D3.4. The item D4.5 “Sexual activities” was always missing due to the conservative culture and be refused to answer or the answer from the attitude of the respondent was not highly credible (71). The rating problem of the item D4.5 was also mentioned and suggested a cultural problem (68, 70). Therefore, we deleted this item in the FUNDES-Adult, version 10 (84).

For developing an ICF-based evaluation tool in Taiwan's DEDS, the FUNDES team has done some pilot studies (9, 85–87) and conducted a literature review for reliabilities and validities of the WHODAS 2.0 at the very beginning (87). The team then revised each item of the WHODAS 2.0 to be with performance dimension and capability dimension. Each item with performance dimension is almost just translated from the original English, with more detailed description of each item and adapting Chinese culture in the FUNDES Manual (27). For example, most male Taiwanese answered the item D5.1 (“Taking care of household responsibilities”) with “not applicable” (9), because male Taiwanese thought household activities were women's responsibilities. Therefore, the D5.1 question has been changed to” Taking care of household issues and family members.” A lot of examples are provided in the manual, including managing finances, car and house repairs, disciplining children, water flowers, etc. Some items that frequently confused testers were also revised based on the feedback of qualified testers. For example, item D1.4 (Learning a new task; for example, learning how to get to a new place?) was usually interpreted as testing the mobility problem due to the example sentence of “learning how to get to a new place”. We then revised the question of that item as “Learning a new task, for example, learning how to get to a new place and learning to use new daily living necessities and skills?”

For using the FUNDES-Adult as disability eligibility determination tools, the FUNDES team thought only the performance dimension is not enough. If performance dimension is the only criteria, one disabled person could possibly be deprived the disability qualification because of appropriate services and environmental support that leading to little restriction. We then add capability dimension to each item to measure the problem the applicants had without adaptive devices and others' assistance. Besides, to increase the better understanding of the capacity of the participants, the testers always started to test Do8 Motor action after demographic data collection. The process and results of the direct testing of the Do8 Motor action could help tester to deal with few applicants who responded most items with scores of 4 (profound) deliberately. The results of the Do8 could also to be used as later needs assessment related to mobility devices.

Using the EFA, Yen et al. found capability dimensions of the FUNDES-Adult, version 5, have five factors for 5,736 adults with disabilities (13); however, Chiu et al. found 6 factors in 521 adults with spinal cord injuries (18). As we mentioned before, differences of sample characteristics are possible reasons. Hence, we do not compare factor structures in different samples, disability group vs. spinal problems. We compared the factor structures between the FUNDES-Adult, version 5, and the FUNDES-Adult, version 7, further. It was found that the names of the five factors were the same, and only one item loaded in different factors. For FUNDES-Adult, version 5, the item D6.1 (How much of a problem did you have joining in community activities?) cross-loaded on three factors, Impact of health on participation capability factor (Factor loading = 0.577), Capability of learning and interaction factor (Factor loading = 0.359), and Capability of housework factor (Factor loading = 0.348) (13). The number of cross loading items might be due to the partial conceptual overlap between some aspects of the different domains of the WHODAS 2.0 (28). For FUNDES-Adult, version 7, the item D6.1 loaded only in the capability of housework factor (factor loading = 0.493). Therefore, the second-order 5-factor structures of two dimensions of the FUNDUS-Adult, version 7, was confirmed.

Besides the correlation matrix of the factor levels, we also calculated the omega reliability coefficients based on the formula proposed by McDonald (88, 89) and the estimated omega reliability coefficients were the average of the factor loadings in the Figures 1, 2 of this study (90). It showed that omega reliability coefficients were 0.78 and 0.80 for performance and capability dimension, respectively. Those are within the well-structured indicator of omega range of 0.75–0.83 (90).

Because most of the Do6 items of the FUNDESA-Adult, version 7, were loaded on a factor named as “Impact of health on participation performance. The modifications to the WHODAS 2.0 for the World Mental Health Surveys did not use original Do6 items (91). The FUNDES team then decided to add five new items and to keep the item D6.1 of the FUNDES, version 7, in the new Domain 6 (Societal participation) of the FUNDES-Adult, version 10. Then, we renamed the original Do6 as Do7 (Impact of health on participation) with seven items. The new five items of the Do6 (Societal participation) of the FUNDES-Adult 10th are derived from the Participation Measure−3 Domains, 4 Dimensions (PM-3D4D) (92, 93). The PM-3D4D is a 19-item measure that was developed on the basis of the conceptual model of participation (93). It is designed to evaluate participation in 3 domains—Productivity, Social, and Community—across 4 dimensions—Diversity, Frequency, Desire for change, and Difficulty. We selected and revised five items of Community domain of the PM-3D4D and Item D6.1 to form Do6 (Societal participation) of the FUNDES, version 10 (84).

Disability data is very important for the disability-inclusive development and for estimating disability prevalence (94). Guidebook published by the World Bank supports the implementation of the Washington Group Short Set (WG-SS) in multi-topic household surveys (95). The six questions of the WG-SS include seeing, hearing, walking, cognition, self-care, and communication. In the FUNDES-Adult, there were items related to walking, cognition, self-care, and communication functioning. Therefore, the FUNDES team have added seeing and hearing items in the FUNDES- Adult, version 10, and FUNDES-Child, version 10. The FUNDES, version 10, started to be used on 1 January 2022. Its psychometrics will be examined in the near future.

This study demonstrates the factor structure of the FUNDES-Adult, a modified assessment tool from the WHODAS 2.0, could be used as an ICF-based measure of AP successfully in the DEDS. The Taiwan's DEDS took the results of the FUNDES as references for the first stage of needs assessment to determine the supports related to parking space for persons with disabilities, transportation supports, and necessary accompany, and RehabBus services (a public transportation services to people with disabilities) (12, 85). The possibility of adjusting the disability grading by the FUNDES has been proposed. Right now, the disability grading or severity of the disability identification is decided mainly by the results of body structure and function assessment due to political and cultural reasons. The previous studies demonstrated there were significant correlations between body structure and function scores and FUNDES-Adult scores or WHODAS scores in each type of diagnosis or disability (30, 50, 51, 60). However, the FUNDES-Adult score of the extreme severe hearing impairment was lower than that of the mild stroke (22). That means that the disability grading based on body structure and function scores only is not fair. The impacts of combining body structure and function scores and FUNDES score on disability grading adjustment are under investigation. We expect the 5-factor structures of the FUNDES-Adult, version 7, could be used as references to adjust the disability grading when we collect enough data of the FUNDES, version 10, in the near future. The application of the FUNDES by clinicians or social welfare service providers to enhance the social participation and outcome evaluation tools for people with disabilities will be presented. However, the FUNDES-Adult requires additional testing and validation. For example, the WHODAS 2.0 shows as a valid instrument to assess functioning differences related to the clinical impact classification level in subjects with chronic obstructive pulmonary disease (36) and has the potential to become a patient-reported outcome measure (20). For persons with depression, most performance on the WHODAS 2.0 were improved after 2-year follow-up (40). We need to investigate the progress of other health conditions. Besides, the ICF-based collaborative problem-solving model could be used in the family-centered early intervention service process and the person-centered services (85, 96). We hope the main theme of ICF, enhancing the full participation of people with disabilities in society, could be reached through the application of the FUNDES.

The data analyzed in this study is subject to the following licenses/restrictions: The data set is an official data bank. Researchers can access the de-identification data through an application process. Requests to access these datasets should be directed to Head of the Department of Health and Welfare Services, Taiwan, through send the application letter to Taiwan Society of ICF, taiwansocietyicf@gmail.com and the website of the TSICF is http://www.icf.org.tw/.

The studies involving human participants were reviewed and approved by the Research Ethics Committee of the Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation (IRB104-04-A;IRB107-46-B) and Joint Institutional Review Board Taipei Medical University (TMU- Joint Institutional Review Board), Taiwan. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

C-FY and H-FL provided concept/idea/research design, data analysis, and writing. T-YC, W-CC, and T-HL provided data collection and data set maintenance as well as data analysis assistance. H-FL and T-HL provided project management. facilities/equipment, institutional liaisons, and administrative support. W-CC, B-SC, T-FW, and S-JL assisted testers training and data checking. All authors reviewed the manuscript before submission. All authors contributed to the article and approved the submitted version.

The study was funded by Ministry of Health and Welfare/Taipei, Taiwan (M05F4145, M06F5054, M07F5223, M08F3147, M09F4089, and M1006158).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.