Abstract
Using the Upper Extremity Fugl-Meyer Assessment (UE–FMA) as a comparison, the Functional Upper Extremity Levels (FUEL) is a valid tool to assess the upper extremity in an acute neurological population and supports the value of the FUEL’s application in clinical poststroke care.
Stroke is a chronic health condition that leaves many individuals with residual disability and participation limitations. Upper extremity (UE) hemiplegia, a motor deficit frequently present poststroke, severely limits function (Lang et al., 2013; Woytowicz et al., 2017) and often compromises engagement in meaningful occupations (Nilsen et al., 2015; Sethy et al., 2018). Assessment of UE function and progress is, therefore, a central aspect of poststroke care.
Occupational therapists are trained to analyze UE neurological dysfunction and its impact on participation in activities of daily living (ADLs; American Occupational Therapy Association [AOTA], 2020). Evidence-based practice requires that clinicians have a clear rationale for selecting valid, reliable, and accurate assessment tools that substantiate the effectiveness of interventions and optimize functional outcomes (Rao, 2012). Furthermore, third-party payers require clinicians to use evidence-based measures with sound psychometric properties to justify reimbursement.
Many neurological UE assessments are used in research and the clinic. Finding the best evidence-based assessment tool for a client, however, can be a daunting task. Existing neurological UE scales have shortcomings that make it difficult for clinicians to document the client’s functional deficits in the performance of self-care tasks (Houwink et al., 2011). Several neurological UE assessments primarily assess impairment, such as the Upper Extremity Fugl-Meyer Assessment (UE–FMA; Fugl-Meyer et al., 1975) and Box and Block Test (Mathiowetz et al., 1985), but do not address function. Assessment tools that focus on function, such as the Action Research Arm Test (Yozbatiran et al., 2008) and Wolf Function Motor Test (Wolf et al., 2001), can be costly and time consuming and require extensive training. Although these assessments are important for research purposes, they are often not used in the clinic (Connell & Tyson, 2012; Lang et al., 2013).
Classification systems are useful to clinicians because they allow for rapid evaluation and meaningful communication of the effects of stroke (Woodbury et al., 2013). Unfortunately, not many UE recovery assessments meet these requirements. The Chedoke–McMaster Stroke Assessment, for example, classifies the effects of stroke at the impairment and activity level but requires 45 to 60 min to administer (Gowland et al., 1993). The UE–FMA takes about 20 to 30 min to administer, but it classifies UE recovery into only mild, moderate, and severe impairment categories. The Australian Therapy Outcome Measures for Occupational Therapy is a comprehensive classification system that reflects the client’s impairment and functional status, using numeric scoring, but requires extensive training to perform (Unsworth & Duncombe, 2004). Therapists are in need of an evaluation tool that is quick to administer, reliable, valid, functional, and low cost. This tool would allow clinicians to simply and effectively convey the status of the complex neurological UE to the rehabilitation team and further delineate the value of occupational therapy to third-party payers (Ali et al., 2013). The objective of our research study is to establish convergent validity for the Functional Upper Extremity Levels (FUEL), a function-based upper limb measure, with the UE–FMA, the gold standard for upper limb motor recovery poststroke.
Upper Extremity Fugl-Meyer Assessment
The UE–FMA is one of the most commonly used impairment-based motor recovery assessments (Baker et al., 2011; Fugl-Meyer et al., 1975). It is divided into four sections: UE, wrist, hand, and coordination. During administration, the therapist asks the client to perform specific movements (e.g., shoulder flexion and abduction) or motor skills (e.g., pincer grip on a pencil). Performance on the 33 items of the UE–FMA is scored on a 3-point Likert scale (0 = no movement, 1 = partial movement, 2 = full movement), with a maximum score of 66. Higher scores indicate greater levels of motor recovery. Although the UE–FMA is useful in its ability to measure specific movements and categorize them (Bushnell et al., 2015), it does not assess the ability to use the affected arm functionally during ADLs (Sabari et al., 2014).
Functional Upper Extremity Levels
The FUEL is a classification system that illustrates and simplifies UE motor performance of the hemiplegic arm during function (Gillen & Nilsen, 2020; Napoleone et al., 2019; Sabari et al., 2014; Van Lew et al., 2015). The operational definitions for each FUEL level were developed by occupational therapists with expertise in the field of neurological rehabilitation and based on anecdotal clinical descriptions of the neurological UE used in practice. The FUEL objectifies and organizes the factors that influence UE function to create a comprehensive ordinal classification system (Van Lew et al., 2015; Wilson, 1980).
The FUEL classifies the UE on the basis of motor ability and how the UE is incorporated into an activity using the following levels: ▪ Nonfunctional: The involved arm or hand is not incorporated. ▪ Dependent stabilizer: The involved arm or hand is placed into a position for function (e.g., to stabilize an object) by the less involved arm. ▪ Independent stabilizer: There is adequate movement to place the involved arm or hand into a position for function (e.g., to stabilize an object). ▪ Gross assist: The involved arm or hand is actively integrated into simple functional activities, including gross grasp. ▪ Semifunctional assist: The involved arm or hand is used actively to assist in accomplishing functional tasks, including isolated finger movement. ▪ Proximal dependent semifunctional assist: The involved arm or hand is a semifunctional assist level, but there is insufficient proximal movement (shoulder or elbow) for function. ▪ Functional assist: The involved arm or hand is actively used during all tasks but may require increased effort, time, or both. Fully functional: The involved arm or hand has returned to complete function, with no detectable difference in ability noted between both arms or hands (Figure 1).
Functional Upper Extremity Levels (FUEL) assessment.
To further classify the UE, the FUEL uses qualifiers similar to the “Qualifier Scale” in the World Health Organization’s (2001)
International Classification of Functioning, Disability and Health. The FUEL uses three of these qualifiers to provide more information to identify how much cueing is needed to optimize motor ability. The qualifiers are as follows: ▪ amount of cuing (i.e., no cuing = 0% of the time, minimal cueing = 25% of the time, moderate cueing = 50% of the time, maximal cueing = 75% of the time) ▪ type of cueing (i.e., verbal, tactile, demonstrative) ▪ clinical reason for cueing (i.e., impairments in sensation, perception, vision, cognition, motivation/behavior).
These qualifiers do not change the FUEL level. They differentiate the amount or type of cues required to maximize UE function that may otherwise be limited by nonmotor impairments. A therapist would document “the client is able to incorporate her affected UE as a gross assist with minimal verbal cues, secondary to left body neglect.” This example includes the FUEL level, the amount and type of cueing, and the clinical reason for the cues. When the qualifiers are combined with the client’s FUEL level, they more precisely and comprehensively describe UE function.
The FUEL has many advantages that make it a quick, practical, and easy-to-use classification system. The FUEL requires minimal training (Van Lew et al., 2015) and can be simultaneously administered while the client is performing routine ADLs (e.g., grooming, meal preparation, laundry). No additional time is required other than the time it takes for clients to perform their ADLs. This can be completed in as little as 5 min. The FUEL is cost-effective, requiring only everyday objects typically used during daily therapy sessions. As an occupation-based measure, the FUEL allows for clients to use their personal belongings, which are nonstandardized self-care items. This process is similar to the Self-Care Section GG of the Inpatient Rehabilitation Facility Patient Assessment Instrument (IRF–PAI; Centers for Medicare & Medicaid Services, 2021), which replaced the FIM®), in which clients also use personal belongings or ADL items provided by the facility (e.g., toiletries, clothing, utensils).
After observing the client perform ADLs, a FUEL level is assigned that reflects the client’s usual abilities for each task. For both the IRF–PAI and the FUEL, these natural variations in task requirements in a real-life setting are what help show the amount of assistance required and UE function of the client. This is in contrast to the use of standardized items, which would not necessarily reflect a client’s real-world performance with items of variable sizes and shapes.
Method
Research Design
The objective of the study was to establish a level of convergent validity for the FUEL using correlation statistics comparing the FUEL with the UE–FMA. Retrospective chart reviews of all clients admitted to a stroke unit at Burke Rehabilitation Hospital were performed, whereby initial and discharge evaluation scores for the UE–FMA and FUEL were collected. As part of the usual standard of care on the unit, occupational therapists administered the FIM, the FUEL, and the UE–FMA on both admission (Day 2) and discharge (within 3 days of discharge date). While the FIM was administered, the FUEL level was determined simultaneously, through the observation of container management or tool use during eating (e.g., utensils) and grooming (e.g., toothbrush) activities. The UE–FMA was performed according to standard methods.
The data included in the study consisted of the initial FUEL and UE–FMA scores and discharge FUEL and UE–FMA scores for all participants. The researcher coded FUEL scores as follows: nonfunctional = 0, dependent stabilizer = 1, independent stabilizer = 2, gross assist = 3, proximal dependent semifunctional assist = 4, semifunctional assist = 5, functional assist = 6, and fully functional = 7. UE–FMA scores were scored using the 0–6 range.
Ethical clearance was obtained from the Albert Einstein Medical College and University of St. Augustine for Health Sciences Institutional Review Boards. Informed consent was not necessary because both tools are used as the standard of care at the hospital.
Participants
Participants were obtained from a convenience sample that included clients with a stroke diagnosis admitted to the stroke inpatient rehabilitation unit at Burke Rehabilitation Hospital between January 2017 and June 2019. Participants were excluded if the FUEL or UE–FMA data were missing from their electronic medical record at initial or discharge evaluations.
Outcome Measures
The current standard of care for this hospital unit includes the administration of both the UE–FMA and FUEL assessments for clients with stroke. Both assessments are administered on initial and discharge evaluations.
Functional Upper Extremity Levels
Content validity for the FUEL was established through expert evaluation of the levels, definitions, and qualifiers (Van Lew et al., 2015). Five external experts assessed the tool on the following constructs: practicality, clarity, and simplicity. A Fleiss’s κ statistic for multiple raters using ordinal data indicated an interrater reliability of .754, showing substantial agreement among raters (Van Lew et al., 2015).
Upper Extremity Fugl-Meyer Assessment
The UE–FMA is often identified in stroke recovery research as the gold standard (Baker et al., 2011; Page et al., 2015; See et al., 2013) because of its strong psychometric properties (Page et al., 2015). The UE–FMA has excellent inter- and intrarater reliability (intraclass correlation coefficient [ICC] = .98–.99) and validity (rs = .74–.93, p < .0001), with a minimal detectable change of 3.2 points (See et al., 2013).
Analysis
Data were analyzed with IBM SPSS Statistics (Version 25). A Pearson correlation coefficient was performed to identify the relationship between the FUEL and UE–FMA for both initial and discharge evaluation scores to support convergent validity for the FUEL. The Shapiro–Wilk test was used to confirm normality of the four data sets.
Results
Participant Demographics
Records for 504 participants were reviewed, 212 of which were excluded because of a nonstroke diagnosis or missing data in the electronic medical record. The sample thus included 292 clients with stroke (48% female), with a mean age of 71.3 (SD = 12.9). Types of stroke were ischemic (n = 254), hemorrhagic (n = 28), or unknown (n = 10). The participants with an ischemic stroke were further categorized into infarction (n = 28), embolism (n = 66), occlusion–stenosis (n = 70), and thrombosis (n = 90). Forty-four percent had right-sided paresis; 41%, left-sided paresis; 3%, bilateral paresis; and 12% other. The mean initial UE–FMA and FUEL scores averages were 40.3 (SD = 23.3) and 4.14 (2.63), respectively. The mean discharge UE–FMA and FUEL score averages were 45.83 (SD = 22.22) and 4.81 (SD = 2.32), respectively.
Psychometric Properties
An analysis using Pearson correlation coefficient indicated a significant linear relationship between the UE–FMA and the FUEL scores for both admission (r = .929, n = 292, p < .001) and discharge (r = .943, n = 292, p < .001).
Discussion
The results showed that incremental improvements in UE function, as measured by the FUEL, were consistently correlated with greater UE–FMA scores. Both admission and discharge comparisons yielded a strong positive correlation between the two assessments (rs = .929 and .943, respectively), supporting strong convergent validity between the FUEL and the UE–FMA.
A previous study by Knutson et al. (2019) used the UE–FMA to examine convergent validity of the Stroke Upper Limb Capacity Scale (SULCS). The study showed a strong correlation between the UE–FMA and the SULCS and concluded that the SULCS is a good measure of upper limb capacity in chronic stroke patients. This research supports the use of the UE–FMA as a good comparison measure to establish convergent validity for the FUEL.
With their unique understanding of function, occupational therapists have the potential to make significant contributions to stroke rehabilitation research by creating function-based assessments (Ranford et al., 2019). Occupational therapists should also ensure these function-based assessments are valid, efficient, and reliable (Doucet, 2012; Rowe, 2013). The findings of the current research provide preliminary evidence that the FUEL is a valid function-based occupational therapy poststroke assessment. The FUEL’s ability to be simultaneously administered during a routine ADL, which does not require additional time or materials, makes it more efficient than other assessments. This is important, given the current health care environment, in which third-party payers are covering less time for inpatient and outpatient services. In addition, clients, policymakers, and regulatory organizations are increasingly seeking function-based results that facilitate occupation participation (Fleming-Castaldy & Gillen, 2013). The combination of this research and the prior research on the FUEL (Van Lew et al., 2015) supports using the FUEL in practice to provide a more functional and meaningful illustration of UE recovery poststroke.
Limitations
Varied training and clinical experience may have affected the results of this study. Staff turnover occurred during the course of the study’s time period, resulting in multiple trainers and trainees for the FUEL and UE–FMA.
Future Research
Ongoing research is needed to further analyze the psychometric properties of the FUEL. In addition, research is necessary to ascertain the FUEL’s utility in settings other than inpatient rehabilitation (e.g., subacute, outpatient, home care). Finally, although this study showed a positive correlation between the FUEL and the UE–FMA, more research would help to determine the FUEL’s utility as a potential stand-alone UE assessment poststroke. This could be accomplished through studies focusing on comparisons of the FUEL with established UE classification systems, which would potentially help to substantiate the levels of criterion and construct validity.
Implications for Occupational Therapy Practice
Results of this study have the following implications for occupational therapy practice: ▪ The FUEL exhibits high convergent validity in measuring upper limb recovery in the acute poststroke population. ▪ The FUEL is a function-based classification measure that can easily be used in the clinical setting.
Conclusion
Research continues to support the need for performance-based measures that are time efficient, reliable, and valid (Rowe, 2013). The addition of a function-based assessment helps meet therapists’ need to determine a client’s level of meaningful UE engagement in occupation (Rao, 2012). This research demonstrates the convergent validity of the FUEL, a function-based measure, compared with the UE–FMA, the gold standard of neurological UE assessment, which supports the use of the FUEL to capture UE functional recovery poststroke.