Validity, Reproducibility, and Minimal Detectable Difference of the Functional Upper Extremity Function Test - Simplified Version - for Adults With Moderate to Severe Asthma and COPD

Abstract

BACKGROUND:

Upper-limbs (ULs) functional tests which are valid and reliable for individuals with chronic respiratory disease (CRD) are scarce. The aim of this study was to investigate the intra-rater reproducibility, validity, minimal detectable difference (MDD), and learning effect of the Upper Extremity Function Test - simplified version (UEFT_S) functional test and to characterize its performance for adults with moderate-severe asthma and COPD.

METHODS:

The UEFT_S was performed twice, and the number of elbow flexions in 20 s was the outcome. In addition, spirometry, 6-min walk test (6MWT), handgrip dynamometry (HGD), and usual and maximum timed-up-and-go tests (TUG_usual and TUG_max) were also performed.

RESULTS:

Eighty-four individuals with moderate-severe CRD and 84 control individuals matched by anthropometric data were analyzed. Individuals with CRD presented better performance in the UEFT_S than controls (P = .023). UEFT_S correlated significantly with HGD, TUG_usual, TUG_max, and 6MWT (P < .047 for all). The test-retest intraclass correlation coefficient was 0.91 [0.86–0.94], and the MDD was 0.4%.

CONCLUSIONS:

The UEFT_S is a valid and reproducible tool to assess the functionality of the ULs in people with moderate-severe asthma and COPD. When applied in the modified form, the test can be considered simple, fast, and inexpensive, with an easy outcome to interpret.

Introduction

Common lung diseases such as asthma and COPD are often associated with a sensation of dyspnea^1,2 and muscle fatigue,^3-5 and these are the factors that commonly limit activities of daily living (ADLs).⁶ Considering that 80% of ADLs are performed with upper limbs (ULs), there is a gap in the literature regarding tools for its evaluation.⁷ When people with COPD and asthma performed activities with ULs without support, they lead to a reduction in inspiratory capacity and thoracoabdominal asynchrony;⁸ and their consequences are the increase in ventilatory demand, leading to dyspnea and physical inactivity.⁹ Consequently, this overload leads to lower mechanical performance of ULs and symptoms of exertion.^4-10 Therefore, assessment tools are needed to identify the physical-functional impairment of the ULs of patients with chronic respiratory diseases.⁷

Recently, clinical field tests have been studied to identify functional limitations of the ULs. Among these instruments, we highlight the 6-min pegboard and ring test (6PBRT)¹¹ and the unsupported UL exercise (UULEX),¹² which are specific for UL evaluation. In addition, the grocery shelving task (GST),¹³ Glittre ADLs,^14,15 and the Londrina ADLs Protocol^16,17 tests have a multitask characteristic and also reflect UL activities. Some of these tests require a long time for application, and others may require a large physical space for their performance, which may hinder their clinical applicability.

Another clinical field test known as Upper Extremity Function Test (UEFT)¹⁸ stands out for requiring only 20 s of movement by the subject. This is a UL assessment method consisting of kinematic and kinetic parameters of elbow flexion, a way to assess ULs at different environments.⁸ This test was originally developed to be performed with 2 accelerometers with triaxial gyroscope sensors coupled to the dominant UL (DUL). It is a valid test to assess functional capacity in COPD;¹⁸ it also identifies frailty in older adults¹⁹ and is used in screening for cognitive impairment such as Alzheimer’s.²⁰ However, the accelerometer with gyroscope may not be accessible to all professionals and make it difficult to use this test in clinical practice. Therefore, we hypothesized that the UEFT, in a simplified version without the accelerometer in DULs, could also be valid for individuals with asthma and COPD.

Therefore, the purpose of this study was to simplify the UEFT, which had as main outcomes the objective kinetic and kinematic measures such as slowness (degree[s]), range of motion (degrees), ascent time (s/100), velocity variability (%), and elbow flexion (number of repetitions). With the UEFT - simplified (UEFT_S), the main outcome was the number of elbow flexions in 20 s, and the objective was to investigate whether this test predicts functionality of individuals diagnosed with moderate-severe asthma and COPD (convergent and discriminative validation). Additional objectives were to investigate the intra-rater reproducibility, the minimal detectable difference (MDD), and the learning effect of this functional test in this population.

QUICK LOOK.

Current knowledge

Changes such as bronchial remodeling, lifestyle changes and loss of functionality can lead to loss of autonomy and reduced quality of life in patients with chronic respiratory disease. Therefore, it is important to assess activities of daily living using a protocol comprised of simple upper limb (UL) activities that are frequently used in these subjects’ daily lives.

What this paper contributes to our knowledge

The Upper Extremity Function Test - simplified version is valid to assess UL functionality in adults with moderate-severe asthma and COPD. The study analyzed psychometric properties of the test, which demonstrated excellent intra-examiner reliability and a learning effect of 0.4% in asthma and COPD.

Methods

Study Sample and Design

A cross-sectional study was conducted with adults and elderly people treated at the Department of Pulmonology of the University Hospital of Londrina diagnosed with asthma or COPD. In addition, snowball sampling was used accompanied by the distribution of informative flyers in basic health units in the city (Londrina, Brazil) and dissemination in social media inviting individuals to participate in the study. People diagnosed with asthma according to the Global Initiative for Asthma (GINA)² and COPD according to the Global Initiative for Chronic Obstructive Lung Disease (GOLD)¹ were included. It is noteworthy that the individuals with COPD included in this study came from another research project that offered physical training for 3 months, but data from baseline assessment were collected prospectively for the present study.

Inclusion criteria were > 18 y, under medical treatment for at least 6 months, clinical stability (ie, no flare-up symptoms or medication increments for at least 30 d), absence of any regular physical training in the last year, and absence of limiting cardiovascular and/or musculoskeletal diseases. Exclusion criteria were mild asthma (steps 1 and 2) due to less functional impairment, inability to perform the proposed tests, exacerbation or change in medication in the evaluation period, diagnosis of other lung disease(s), and refusal to continue evaluations for any reason. Data collection took place between March 2019–December 2020. The study was approved by the ethics committee of universities Unopar (2.09.139) and UEL (1.730.247); all participants signed an informed consent.

To verify the discriminative validity of the UEFT_S, retrospective data from individuals who participated in another study, carried out between July 2018–September 2019, with the objective of creating reference values for functional tests in adults, were also analyzed (Study approved by the CEP of the University Pitágoras Unopar, Londrina, Brazil; number 2.702.745). These volunteers had normal lung function (FEV₁/FVC) ≥ 70% and were healthy based on individual clinical report. The control sample was matched for sex, age, weight, height, and body mass index, all of which performed the tests in the same protocols.

Assessments

All participants in the present study were initially evaluated with anthropometric data measurements to characterize the sample. The assessment of pulmonary function was performed according to the international standards proposed by Miler et al²¹ and reference values for the Brazilian population.²² Subjective assessment for symptoms of anxiety and depression was performed using the Hospital Anxiety and Depression Scale;²³ level of sensation of dyspnea in daily life with the Modified Medical Research Council²⁴ scale; in addition to specific questionnaires for assessment of asthma control by the Asthma Control Questionnaire,²⁵ clinical impact of COPD symptoms assessed by the COPD Assessment Test,²⁶ and London Chest ADL (LCADL).²⁷ Finally, peripheral muscle strength was assessed with a hydraulic dynamometer (Jamar, Patterson Medical, Warrenville, Illinois). The participant remained in a sitting position, was instructed to keep shoulders in a neutral position and elbow at 90 degrees, and the grip strength in the dominant hand was considered. This test was performed 3 times, and the best result was considered for the analysis.²⁸

Functional Tests

To assess the functional exercise capacity, the individuals performed two 6-min walk test (6MWT) in a 30-m corridor following the recommendations of the American Thoracic Society and European Respiratory Society,²⁹ and the predicted value of walking distance was calculated according to the reference values proposed by Britto et al³⁰ for the Brazilian population. Subjects performed 2 tests, and the best performance was considered.

Individuals also performed the timed-up-and-go (TUG)³¹ test, and the test execution time was measured. Participants were instructed to get up from a chair, walk a distance of 3 m, turn to return, walk to the chair again, and sit down firmly. The stopwatch started counting time from the command to go and was stopped when the participant’s back was fully supported by the dorsal column on the back of the chair. Two protocols of this test were performed with the individuals walking at their usual pace (TUG_usual) and also at the fastest pace possible (TUG_max).³² Subjects performed 2 tests, and the best performance was considered.

Finally, a new proposal for the functional test of ULs (UEFT_S) was performed. In this test, the individual was instructed to perform elbow flexion and extension movements on the DUL for 20 s, in full range of motion, repeatedly as fast as possible in a sitting position and keep the trunk leaning against the chair during the execution. The stopwatch started counting the time from the command to go and was stopped after 20 s. The test started with the individual seated with the elbow semi-extended, receiving the following instructions: “When I say go, bend the elbow completely and then stretch it completely repeatedly as fast as possible for 20 s. I will count the number of times your elbow will bend.” Prior to the official test, participants performed a practice trial on their non-dominant side to familiarize themselves with the protocol, and they were not encouraged during the task. The evaluator counted the number of times the individual performed the flexion movement completely. The test was performed twice by the same evaluator, immediately before the performance of each 6MWT, with a time interval of 30 min between them; the recovery of symptoms was assessed by evaluation of the dyspnea symptoms scale (Borg D) and limb fatigue (Borg F), S_pO2, and heart rate before starting the second test.³³

Data Analysis

Data were expressed as absolute and relative frequency and median (interquartile range). Data distribution was verified by the Shapiro-Wilk test. Spearman correlation coefficient and the Mann-Whitney U test were used to investigate the convergent and discriminative validity, respectively. Our hypothesis was that the performance of UEFT_S would moderately correlate (r ≥ 0.40) with the 6MWT, TUG_usual, TUG_max, and peripheral muscle strength (convergent validity) and to differentiate from a group of individuals without pulmonary impairment (discriminative validity). The intra-rater reproducibility was investigated by means of reliability analysis with the intraclass correlation coefficient (ICC) with a 95% CI describing upper and lower limits. ICC values were interpreted according to the following cutoff points: < 0.50 = poor reliability, 0.50–0.75 = moderate reliability, 0.75–0.90 = good reliability, and > 0.90 = excellent reliability.³⁴ Visual analysis of agreement was also presented in the Bland-Altman³⁵ graph. The Wilcoxon test was used to compare performance between the 2 test runs.

In addition, the measure of variability was determined by calculating the absolute and relative standard error of the mean (SEM and SEM%, respectively) using the formulas (SEM = SD [second − first test]/√[1 − ICC]; SEM% = [SD first + second test/average of the first + second test] x 100) and the MDD absolute and relative values (MDD%) with the formula (MDD = 1.96 * SEM * √2; MDD% = [MDD/average of the first + second test] *100); the learning effect of UEFT_S was also calculated (learning effect = [first − second test]/second test; learning t percentage = [first − second test]/first test] * 100).^36,37 The statistical significance considered was P < .05.

Results

From the 236 individuals diagnosed with asthma who were eligible for the study, 79 were assessed, but only 50 were analyzed. Whereas from the 161 individuals with COPD eligible for the study, 96 were assessed for evaluation, and only 34 were analyzed. The reasons for the exclusions are shown in Figure 1. The sample consisted of mostly middle-aged, overweight women with pulmonary function with air flow limitation. All individuals with asthma between step 3–5 with treatment levels were classified as moderate to severe according to GINA² and those diagnosed with COPD classified according to GOLD II–IV, with 55% in moderate severity.¹ Other characteristics of the sample are described in Table 1. Discriminative validity of the UEFT_S showed a lower performance in the general sample when compared with the control group. Both groups also performed worse on the UEFT_S than individuals with normal lung function when compared separately (asthma P = .02; COPD P < .001). Results with separate samples are available in the online supplement (See related supplementary materials at http://www.rcjournal.com).

Fig. 1.

Flow chart. UEFT_S = Upper Extremity Function Test - simplified; GOLD = Global Initiative for Chronic Obstructive Lung Disease.

Table 1.

Characteristics of Subjects With Asthma and COPD Compared With the Control Group

The performance of the first UEFT_S (number of elbow flexions) correlated (convergent validation) with the 6MWT, both TUG test protocols, and handgrip strength, is shown in Figure 2. Furthermore, the UEFT_S correlated with the LCADL domains (personal care: r = −0.45, P < .001; leisure: r = −0.38, P < .001; domains of domestic and physical activities: r = −0.28, P < .001; and total score: r = −0.22, P = .040). The correlations remained similar when we analyzed the data using the best performance of 2 tests in UEFT_S.

Fig. 2.

Correlation of Upper Extremity Function Test - simplified with functional tests and peripheral muscle strength. 6MWT = 6-min walk test; UEFT_S = Upper Extremity Function Test - simplified; TUG = timed-up-and-go test.

The UEFT-S proved to be a reproducible test, and in Table 2, the values of ICC, SEM, MDD, and learning effect are described. The test-retest can be seen in the visual analysis by Bland-Altman in Figure 3. Despite a mean value close to zero, we observed a large variability in the 95% CI. Interestingly, we observed that 36 (42%) participants presented a difference > 2 repetitions (more or less) between the first and the second tests. The comparison between the first and second test showed that there was no significant difference in the number of elbow flexions in 20 s (learning effect) for the group of participants, although 36% of participants had better performance in the second test (Fig. 4).

Table 2.

Intra-Rater Reliability, Standard Error of the Mean, Minimal Detectable Difference, and Learning Effect

Fig. 3.

Bland-Altman intra-rater visual analysis of the Upper Extremity Function Test - simplified. UEFT_S = Upper Extremity Function Test - simplified.

Fig. 4.

Comparison of functional performance between the first and second Upper Extremity Function - simplified. UEFT_S = Upper Extremity Function Test - simplified.

Discussion

The present study showed that the UEFT_S is valid for assessing functional capacity in adults with moderate-severe asthma and COPD. It is also able to discriminate these populations. We also suggest familiarization on the non-dominant side before performing the protocol. Two tests were performed with a minimum interval for symptom recovery, and the best test was considered for individual interpretation.

It is known that the ULs are intimately involved in functional activities especially to perform ADL tasks.^38,39 Considering this information, the use of adequate instruments for evaluation of ULs allows the rater to elaborate adequate physical-functional diagnoses and distinguish limitations of each participant.⁴⁰ Therefore, investigating reproducibility and validity are traditional domains for utilizing a new assessment proposal.^35,41

Reliability, which is also known as reproducibility , is the ability to obtain similar results with repeated assessments.⁴¹ This study demonstrated excellent intra-rater reliability in the UEFT_S (ICC = 0.91).⁴² Previous studies assessing the UULEX reproducibility also demonstrated excellent results, with ICC = 0.88 in young people and 0.82 in older adults from Portugal.⁴³ Very similar results were found in Brazil with healthy adults aged 31–80 y (ICC = 0.85).⁴⁴ The same happened for people with rheumatoid arthritis (ICC = 0.95).⁴⁵ The UULEX test assesses maximum exercise capacity of ULs without supporting the upper extremity, and all these results corroborate with our findings. The 6PBRT, which assesses the functionality of the ULs in COPD, presented an ICC = 0.91,¹¹ similar to our results.

A better performance in the second test was observed in 36% of the participants with chronic lung disease evaluated in this study. However, there were no significant differences between the first and second UEFT_S (Fig. 4) when considering the group of participants. One way to assess absolute reproducibility can be performed by the limits of agreement derived from the Bland-Altman³⁵ chart. It characterizes the 95% CI around the mean of the test-retest difference. Observing Figure 2, it seems reasonable to indicate that in individual assessments the test can show an important variation in results, which could mask some underestimations, especially in longitudinal studies. It is, therefore, suggested to familiarize the non-dominant side and perform 2 tests (on the dominant side) to ensure better performance of the UEFT_S.

In this study, the MDD represents the minimum amount of change in the performance of the test, which will not be due to the chance of measurement variation. The less the MDD the more sensitive the instrument is.³⁶ Findings from this analysis indicate low variability and learning effect in the general sample when compared to the MDD% of 14 in the TUG test at usual speed of people with COPD.³² For the UULEX test, a low value of SEM% = 0.29 and an intra-rater MDD = 0.80 s were found in participants diagnosed with rheumatoid arthritis; moreover, in individuals with asthma, Oliveira et al⁴⁶ calculated MDD varied between 15–31% and SEM% from 11–30% in functional tests of lower limbs. Thus, those authors suggest the application of 2 tests in this population. These results are similar to the present study, which indicated high variability when the UEFT_S was performed by individuals with asthma and COPD.

This study proposed a simple, low-cost, and quick functional test to assess the UL functional capacity of adult individuals with moderate to severe asthma and COPD. The UEFT_S showed a difference in performance between individuals without pulmonary impairment and those with airway obstruction (P = .01), in agreement with the validation study by Thoosizadeh et al,¹⁸ of the original UEFT in individuals with COPD, which found a correlation of r = 0.50 (P = .040) between the 6MWT and the number of elbow flexions in the UEF test. Other objective parameters expressed through the accelerometer in the UEFT protocol indicated significant correlations with flexibility, power, and speed, ranging between 0.44–0.62 (P = .054) for people with COPD.¹⁸

Podsiadlo and Richardson³¹also found moderate correlation (r = −0.55 and −0.51) in the TUG validation study in older adults. The same test in 2 protocols was recently validated for individuals with asthma (TUG_usual r = −0.62 and TUG_max r = −0.61) and correlated with the 6MWT.⁴⁶ For people with COPD, the UULEX test correlated with physiological variables such as heart rate, oxygen consumption, carbon dioxide production, minute volume, dyspnea, and fatigue (r > 0.70, P = .050).¹² Furthermore, performance time was correlated between the UULEX test and the supported UL exercise in its validation (r = 0.56).⁴⁷ Another UL test, the GST, that resembles UL movements performed without support, as in the UEFT_S, also stands out for being valid, strongly correlating with the UULEX and cardiorespiratory physiological responses of the test (0.69 > r < 0.85).¹³

Our hypothesis to explain weak correlation of the UEFT_S with handgrip strength was expected, as it was also described by Puhan et al⁴⁸ (r = 0.45) between the sit-to-stand test and handgrip strength. Thoosizadeh et al¹⁸ found correlation between 0.30–0.52 for the kinetic and kinematic outcomes of the UEFT system and respiratory and peripheral muscle strength in people with COPD. The same is observed in patients with asthma. Low values of correlation have also been reported between lower-limb functional tests and handgrip (for ULs), with values ranging between 0.34–0.44.⁴⁶ These values agree with the findings of the present study. The UEFT_S does not seem to reflect UL muscle strength but better reflects the functional mobility of these individuals, as assessed with the TUG functional test.

Limitations

The authors consider that the absence of other UL tests, such as UULEX or 6PBRT, is an important limitation in the present study protocol, and it is still unknown whether the UEFT_S applied inter-rater would have the same effect. Unfortunately, for logistical reasons, the included sample did not perform these assessments and did not have a second evaluator. However, the literature describes tests that involve walking in patients with chronic lung diseases to validate new functional assessment instruments for ULs.^16-18 We also highlight the need for a study to analyze the test responsiveness. Finally, we recognize the limited number of subjects with COPD and how this affects the generalization of the results by not allowing proper subgroup analysis (eg, according to disease severity). As strengths, we propose the use of a simplified version of a functional test to assess the ULs of patients with moderate to severe asthma and COPD. UEFT_S has the potential for wide applicability, considering it is a low-cost, fast, simple, and easy-to-understand functional test.

Conclusions

The UEFT_S is valid for evaluating the functional capacity of adult ULs with moderate-severe asthma and COPD. This test presents correlations with the 6MWT, TUG_usual, TUG_max, handgrip strength, and domains of the LCADL scale. Furthermore, we suggest a familiarization in the non-dominant side and that 2 UEFT_S with a minimum interval between them should be preferred. As outcome for clinical practice, the greater number of elbow flexions of the DUL should be considered. This UL test can be an extremely fast, practical, simple, and low-cost clinical field test alternative, which enables applicability in a wide evaluation scenario to identify physical-functional limitation of ULs.