Historically, disability evaluation, whether for benefit determination, programmatic evaluation, or clinical purposes, has focused on performance-based testing due to the general underlying belief that it offers greater value than other forms of measurement such as patient-reported outcome measures (PROMs) (1, 2). However, for research purposes, there has been an overwhelming uptake of PROs with considerable investment in their development, validation, and interpretation in the last decade (3, 4). A substantial body of research has demonstrated the correspondence between performance tests and self-report as well as the conditions under which this correspondence breaks down (5–8). In addition, the advent of symptom-validity tests has helped to address concerns about the intentional misreporting of function for personal gain by offering a way to flag cases of concern (9). This raises questions about why there has not been a better uptake of PROMs for the purpose of disability evaluation. Thus, the goal of this perspective is to elucidate the issues relevant to the use of PROMs in disability evaluation.

The evaluation of disability varies widely based on the purpose of the assessment (10, 11). The selection of the best measurement tool to address a specific scientific question starts with clearly identifying the purpose of measurement. Yet, both clinicians and researchers often gravitate toward discussion of tool availability and features (including psychometric properties) without adequate consideration of what they are trying to accomplish. For instance, if you want to know the number of people who could potentially benefit from a new walker, you need to measure aspects of function among adults who are potential users of the device. Not only would you need to know the number of people with walking difficulties, but if arm use was required for this device, then you would also need to know about arm function and potentially the ability to grip the device. You may need to know about balance and the cognitive ability to operate features of the device. Thus, you need very specific information about a broad range of functioning in the population of interest. In contrast, if you wish to know the population of people who could potentially benefit from civil rights legislation to reduce barriers for people with disabilities, such as the Americans with Disabilities Act, you would want to capture the broadest group of people possible who could be affected by such legislation, including those with difficulties hearing, seeing, thinking, moving, doing everyday activities, and participating in everyday tasks and social activities.

As part of a collaboration with the Social Security Administration, we were tasked with identifying efficient ways to assess work disability. Since contemporary concepts of disability characterize it as the outcome of the interaction between the capabilities of individuals and the environment in which they are operating, the evaluation of work disability requires the examination of all aspects of functioning of the person (at the person level as opposed to the cellular, organ, or body system level) in relationship to the demands of various jobs (12). Thus, by clearly identifying the specific purpose for disability evaluation, the choice of measure, be it self-reported or performance-based, becomes clearer, which simplifies the task of measure selection.

Once measurement options have been narrowed by clarifying its purpose, then other aspects of measurement selection come into play such as content coverage, administration mode (paper/pencil, electronic, interview, telephone), other aspects of administration (observed performance vs. self-report, use of proxy respondents), administration time, cost of administration, the need for trained administrators, and psychometric properties of the instruments. Depending on the situational constraints, more than one type of measure may be selected to best address measurement needs.

One crucial aspect of measure selection that is often overlooked is whether the measure has been tested in real-world situations and if score interpretation has been developed and supported. What good is it to measure the work-related functional capabilities of individuals if the level of ability needed for each functional domain relative to a given job is not known? Real-world studies that provide a foundation for clinical and population-based score interpretation are crucial to selection and use of measures for disability evaluation. Examination of whether normative values have been established, the method for establishing the normative values, the basis for establishing cut-points or thresholds, and comparability of scores over time and across population, are essential considerations of score interpretability.

In the context of measuring functional abilities for the purposes of disability evaluation, assessment, and treatment planning, comprehensive measurement relies on obtaining both the subjective experience of individuals, typically collected via self-report measures, and objective indicators of functioning, which usually rely on performance-based assessments (12). Each measurement approach offers distinct advantages and limitations, which must be considered when interpreting outcomes in the context of disability assessment (13, 14). Performance based tests are especially valuable in clinical settings where decisions about rehabilitation or functional limitations must be grounded in demonstrable performance, however the degree to which that translates into real world functioning in daily activities such as work is limited based on previous research findings (15). Furthermore, when measuring the same underlying construct, self-reported measures and performance-based assessments scores have been found to be often moderate to highly correlated (16). A primary advantage of self-reported measures is their capacity to capture the subjective experience of disability. These self-reported instruments provide insights into how individuals perceive their limitations and how these limitations affect their quality of life and daily activities (17).

To illustrate these advantages and opportunities for improvement in the context of disability assessment, we will draw upon previous research examining the reliability and validity of the Work Disability Functional Assessment Battery (WD-FAB) (18–22). This body of work provides illustrative examples of the advantages and disadvantages of using a robust self-report-based tool to characterize multiple dimensions of physical and mental health function relevant to disability vs. performance-based assessments of disability among populations with underlying physical and/or mental health conditions. The WD-FAB is a standardized self-report measurement tool using item response theory and computer adaptive testing to characterizes physical and mental functional relevant to work disability along eight key dimensions: Basic Mobility; Upper Body Function; Fine Motor Function; Community Mobility; Mood & Emotions; Resilience & Sociability; Self-Regulation; and Communication & Cognition. The WD-FAB also includes an opt-in scale for individuals who use wheelchairs to capture mobility-related tasks. The WD-FAB draws upon a pool of over 300 items to generate scores for all 8 scales in 15–25 min. WD-FAB scores allow direct comparisons between individuals with self-reported activity limitations with the general population of working age adults and are sensitive to detecting change in function over time. See Table 1 for a description of the WD-FAB measurement domains and descriptions of each scale.

Recent evidence has emerged to support the collection of both performance-based and patient-reported outcomes measures (PROMs) to evaluate disability across domains at the clinical level. It has been shown that complementary rather than redundant information about a person's functional abilities is captured by implementing both performance-based and PROMs (30, 31). For example many PROMs have been validated with minimal clinically detectable change scores affording standardized and objective longitudinal tracking in the clinical setting (32, 33). The opportunity for improved clinical outcomes by employing PROMs arises from shared decision-making, furthering specificity of care needs and thereby patient engagement in the current era of value-based care. Incorporating PROMs in parallel with performance measures can serve the current triple aim of healthcare if carefully and conscientiously implemented to assess both the individual and population levels. PROMs have been able to capture information affecting overall function previously undisclosed by examining performance-based measures alone (34, 35). Patient perception of performance includes longitudinal retrospection of activity whereas clinical performance measured in a structured, point-in-time environment is limited in its wider applicability to daily function (36). Integrating the two measurement approaches potentially provides insight into whole person function and which domains of function require treatment for return to activity that is meaningful to the patient and measurable by the clinician.

Data from PROMs and evidence-based performance measures is ideally captured at the outset of care to achieve mutually defined goals and expectations for care. PROMs integrated for clinical care have also demonstrated responsiveness over time to further provide guidance for ongoing treatment management (37, 38). Furthermore, PROMS are generally easier and cheaper to obtain, thus more likely to support longitudinal measurement. Given the current emphasis on patient centered care, collecting both relevant patient centered PROMs and performance-based measures in an efficient and timely manner presents challenges in the clinical setting. PROMs can be captured prior to the clinical visit in the EMR portal, via telephone interview, or at the point of care. Considerable resources, including previously implemented software, electronic literacy, health literacy, labor, and proxy availability from the health care setting, provider, and patient are required for PROM integration. Capturing domain information in computer adaptive testing (CAT) tools such as the WD-FAB can ultimately reduce provider and patient burden while improving more specific and individualized data for patient education, shared decision making, goal setting and treatment planning (38). Analysis and interpretation of both PROMs and performance measures at the point of care requires significant clinician focus in busy clinical environments (39). Limited provider time for interpretation and integration of findings from both the PROMs and performance measures is common but necessary to facilitate patient engagement and potentially facilitates self-management of function. Furthermore, when findings from PROMs and performance-based measures are discordant, further analysis may be necessary to specify an efficient and effective treatment plan (40).

Overall, self-report tools such as the WD-FAB provide a way to efficiently collect systematic and comprehensive information about individuals' experiences of their functioning. Self-report measures such as the WD-FAB are not designed to replace medical evidence but are intended to be used to complement and strengthen clinical performance-based measures of health and disability by providing valuable insights in better understanding disability not only as a biomedical issue but as a lived experience. Furthermore, discrepancies between the two types of assessments could prove to be diagnostically meaningful and indicative that further psychological evaluation for complex co-occurring physical and mental health conditions is required. Ultimately selecting between performance-based measures and self-reported measures of health, function and disability requires careful attention to the patient's individual needs, the clinical or research setting, and the practical resources available. Table 2 summarizes the overall strengths and limitations of each measure to consider when thinking about which to use in practice.

Different measures offer different perspectives on health, function, and disability, each with their own benefits and tradeoffs. Performance-based testing supposedly provides a more objective perspective, but it is limited by the environment in which the testing is conducted, which does not often reflect the complexity or variability of real-world situations. Patient-reported outcome measures (PROMs) can be more comprehensive and more closely reflect respondents lived experiences, although there are sometimes concerns around reporting bias, which can be at least partially addressed through measure validation. In the context of disability evaluation, one key consideration is patient burden. PROMs can take less time to complete and can be completed in patients' own environments. This reduces the time, effort, and resources a patient would be required to put forth to make it to a testing site. Ultimately, it is important to use an approach that provides the most relevant measure to the disability evaluation purpose.