Abstract
Background Physical functioning of patients with hip or knee osteoarthritis is measured by self-report questionnaires and performance-based tests. However, performance-based tests often are not feasible.
Objective The aim of this study was to develop a computer-administered questionnaire (ie, the Animated Activity Questionnaire [AAQ]) to measure physical functioning in patients with hip or knee OA. By showing animations of activities, the influence of the patient's own reference frame is minimized. The AAQ measures the same aspects of physical functioning as performance-based tests do.
Design This was a development and preliminary validation (cross-sectional) study.
Methods A pilot version of the AAQ was developed using motion capture to analyze the movement of a person performing 7 daily activities. Different animations of the same activity were made with 2 to 5 levels of difficulty. For each activity, participants were asked to choose one animation that best corresponds to their own way of performing the activity. A preliminary validation study was performed to compare the AAQ with validated self-report questionnaires (Knee Injury and Osteoarthritis Outcome Score, Hip Disability and Osteoarthritis Outcome Score, and questionnaires on walking, stair climbing, and rising and sitting down) and performance-based tests (walking, Timed “Up & Go” Test, Timed Stair Test) in 33 patients with hip or knee osteoarthritis.
Results As expected, the AAQ showed a correlation above .70 (.79, 95% confidence interval=.61–.89) with the total score of the performance-based tests. On the subscore level, the results were partly as expected. Fifty-eight percent of the participants preferred the AAQ over self-report questionnaires and performance-based tests.
Limitations The findings need to be replicated in larger samples of patients because the sample size of the study was rather small.
Conclusion The AAQ might be a good alternative for measuring physical functioning of patients with hip or knee osteoarthritis. The AAQ can easily be adapted for use in other patient populations. However, further development and validation are needed.
The validity of self-report questionnaires versus performance-based tests to assess physical functioning in patients with hip or knee osteoarthritis (OA) and other patient groups has been under debate for a long time.1–16
A large number of studies showed low to moderate correlations between self-report questionnaires and performance-based tests.4,8,11,14–22 There seems to be consensus that physical functioning can best be measured by a combination of self-report questionnaires and performance-based tests because they measure different aspects of physical functioning.23
Self-report questionnaires measure the degree of difficulty patients perceive in performing daily activities. The perceptions of patients, however, are determined not only by the severity of their disease but also by a number of personal and environmental factors. It has been shown that patients use different reference frames when they respond to questions such as “What degree of difficulty do you have ascending stairs?” (a question from the Knee Injury and Osteoarthritis Outcome Score [KOOS]24). Their responses will depend on the comparison group they have in mind. Some patients compare their current performance with how they performed before the onset of their disease or how they performed some time ago, whereas others compare themselves with peers who are healthy.25 Their responses also will depend on what kind of stairs they have in mind26 as well as the importance of stair climbing in their daily life. Patients' responses are influenced by the confidence they have in their own ability to climb stairs.13 Cognitive impairments,3 cultural factors, comorbidity, age, and education level also play a role in answering these and other kinds of self-report questions.1,6,9 Finally, responses to self-report questions are determined by the patients' interpretation of the word “difficulty.” Patients may consider limitations in performing daily activities, but they also may consider other factors such as pain or fatigue. It has been shown that self-report questionnaires are more influenced by pain and fatigue compared with performance-based tests.12,19,21,23,27,28
Because of the influence of these personal and environmental factors on self-reported physical functioning, it is recommended to use performance-based tests in addition to self-report questionnaires. These tests often are taken as a starting point for treatment, such as exercise therapy.
Performance-based tests, however, have a number of disadvantages. They measure physical functioning in an artificial situation, which may provide little information about how a patient performs in his or her own environment. Furthermore, they capture only a snapshot of what a patient can functionally do. Performance-based tests can be influenced by short-term impairments, observer bias, stimulation by the tester, and the patient's motivation to perform the test.7,9 The greatest problem with performance-based tests, however, is that they require personnel, facilities, and the physical presence of patients, which means they are time-consuming, expensive, and a burden for patients. Performance-based tests, therefore, are often considered unsuitable for use in large-scale studies.
A solution to these requirements of performance-based tests might be a computer-administered animation questionnaire.26,29 On a computer, patients watch an animated video in which a dummy (an animation of a person) performs an activity such as stair climbing. Patients are asked whether they can perform the activity in a similar way. By showing animations, the influence of the patient's own reference frame is minimized.29 It is expected, therefore, that an animated questionnaire measures the same aspects of physical functioning as measured by performance-based tests. Thus, a computer-administered questionnaire can potentially be a suitable alternative for performance-based tests for large-scale studies.
The aim of this study was to develop and validate a pilot version of a computer-administered questionnaire (ie, the Animated Activity Questionnaire [AAQ]) for measuring physical functioning of patients with hip or knee OA. Because we examined a newly proposed set of measurements, our validation study was considered preliminary. We did not conduct formal hypothesis testing but rather were interested in general associations among measures that would demonstrate consistent patterns of associations. These data would allow for better estimation of sample size for a more definitive study. We expected that scores on the AAQ would correlate highly with the results of performance-based tests and moderately with scores on self-report questionnaires. We also expected that the correlations between the AAQ and the performance-based tests would be greater than the correlations between the self-report questionnaires and the performance-based tests. Finally, we expected that the AAQ would be less influenced by pain than the physical functioning subscales of the self-report questionnaires.
Method
Development of the AAQ
Our aim was to develop a pilot version of a computer-administered questionnaire to measure physical functioning in patients with hip or knee OA. Physical functioning was defined as an individual's ability to perform activities, consistent with the definition of “activity limitations” in the International Classification of Functioning, Disability and Health (ICF).30 We intended to develop a disease-specific questionnaire, which means that the activities to be included should be activities that patients with hip or knee OA may find difficult to perform because of their hip or knee OA.
For the pilot version of the AAQ, we selected 7 activities that patients with hip of knee OA often consider difficult to perform and that are included in most self-report questionnaires on physical functioning: ascending and descending stairs, sitting down and standing up from a low couch, sitting down and standing up from a toilet, and walking. For each activity, 2 to 5 levels of difficulty were defined, for which an animated video was made (Tab. 1). For stair climbing, for example, difficulty levels were described as “easily” (without difficulty), with a little difficulty (eg, climb stairs using a handrail), or with very much difficulty (eg, climb stairs by pulling up the other leg before taking a new step). In total, 23 animated videos were made.
Activities With Levels of Difficulty (Response Options) of the Items in the Animated Activity Questionnaire (AAQ)
The animations were made by a specialized company (Re-lion BV, Enschede, the Netherlands) involved in the development of computer gaming. Motion capture was used as a technique to transform movement of a person into animations. A physical therapist with extensive experience in the treatment of patients with hip and knee OA (W.F.P.) acted as a patient. The actor wore a special suit with motion sensors while performing the activities. Based on the analysis of the data from the motion sensors, an animation was made for each activity. We decided to show a neutral dummy in the animations instead of a real person to minimize the influence of the patient's reference frame (eg, with regard to age). Also, the materials (eg, stairs, toilet) were conceptually visualized to minimize any context effects and standardize item interpretation. The animations last between 3 and 22 seconds. An example of an animated video is available on our website31 and online at ptjournal.apta.org. A screenshot is presented in Figure 1.
An example of an animated video for the Animated Activity Questionnaire (AAQ).
An example of a screenshot and a question from the Animated Activity Questionnaire (AAQ).
The AAQ was constructed in Word, including questions with links to the videos. For each of the 7 activities, participants were asked, “How do you perform this activity? Choose the animation that best represents your way of performing the activity.” The AAQ did not have a specific reference frame (participants may have answered how they usually perform the activity or how they would do it at that moment). Participants could click on the links to see an animation. Each animation could be watched as often as the participants wanted. An example of a question is given in Figure 1. A researcher (C.C.) was present to start up the computer, but the participants completed the AAQ by themselves, without help or interference of the researcher.
Each item of the AAQ was scored from 1 to 2, 1 to 3, or 1 to 5, depending on the number of videos per item. The item scores were subsequently transformed into scores from 0 to 4, and a total score was calculated by summarizing the transformed scores of the 7 questions. In order to compare with the performance-based tests (described below), 3 subscores also were calculated: a subscore for stair climbing (2 items on ascending and descending stairs), a subscore for rising and sitting down (4 items on sitting down and standing up from a low couch and sitting down and standing up from a toilet), and a subscore for walking (1 item on walking). For comparison with scores from self-report questionnaires (described below), the total score and the subscale scores were finally transformed into scores from 0 to 100 (higher scores indicating worse physical functioning).
Participants
For the preliminary validation study, patients with hip or knee OA were recruited from 2 centers: a rehabilitation center (Reade, Centre for Rehabilitation and Rheumatology) and an orthopedic surgery department (Onze Lieve Vrouwe Gasthuis [OLVG]) in Amsterdam, the Netherlands. Patients were eligible for the study if they were diagnosed with hip or knee OA (according to the American College of Rheumatology criteria).32,33 Patients were included who were not yet in need of surgery, scheduled for surgery, or had undergone surgery, resulting in a wide range of ages and disease severity. At Reade, a convenience sample of patients was selected who were currently under treatment. At OLVG, patients were selected who were scheduled for a visit at the outpatient clinical in the next 1 to 2 weeks. Thirty-three patients were included in the study. Characteristics of the participants are presented in Table 2.
Participant Characteristics (N=33)a
Procedure
Included participants were asked to complete a self-report questionnaire at home during the week before a scheduled visit to their physical therapist or orthopedic surgeon. This is common practice in the participating centers. At their visit, participants were asked to complete the AAQ. After completing the AAQ, they were asked to perform 4 standardized performance-based tests. Next, participants were interviewed about the comprehensiveness and user-friendliness of the AAQ, whether they recognized themselves in the animations, and whether they missed any levels of difficulty in the animations (eAppendix). Finally, they were asked for their preference regarding the self-report questionnaire, the AAQ, or the performance-based tests and which of the 3 test forms took the most time to complete.
Questionnaires
Participants completed 4 questionnaires. First, the Knee Injury and Osteoarthritis Outcome Score (KOOS) or the Hip Disability and Osteoarthritis Outcome Score (HOOS) questionnaire was completed.34,35 The KOOS and HOOS consist of 5 subscales measuring pain (KOOS, 9 items; HOOS, 10 items), symptoms (KOOS, 7 items; HOOS, 5 items), activities of daily living (ADL) (KOOS and HOOS, 17 items), sports and function (KOOS, 5 items; HOOS, 4 items), and knee-related or hip-related quality of life (KOOS and HOOS, 4 items). The ADL subscale is the same as the physical functioning subscale of the Western Ontario and McMaster University Osteoarthritis Index (WOMAC) questionnaire,36,37 on which the KOOS and HOOS are based. Each question is rated on a scale from 0 to 4. A normalized score from 0 to 100 (higher scores indicating more symptoms or worse functioning) was calculated for each subscale. Missing values were treated according to the KOOS or HOOS manual: 1 or 2 missing values are substituted with the average value for that subscale; if more than 2 items are omitted, the response is considered invalid, and no subscale score is calculated.
Second, participants completed 3 questionnaires measuring limitations in stair climbing (Climbing Stairs Questionnaire, 15 items),38 limitations in rising and sitting down (Questionnaire Rising and Sitting Down, 39 items),39,40 and limitations in walking (Walking Questionnaire, 35 items).41 These questionnaires were chosen in addition to the HOOS and KOOS because they provide a more detailed and more precise assessment of these specific activities. Each question is rated on a scale from 1 to 7. Missing items were imputed by the mean item score of the patients who completed that item. A normalized total score (0 indicating no limitations and 100 indicating most limitations) was calculated for each questionnaire. To compare with the AAQ, an overall physical functioning score also was calculated by taking the mean of the total scores of the 3 questionnaires. This will be referred to as the “total score specific limitations.”
Performance-Based Tests
Participants completed 7 performance-based tests (which were as similar as possible to the tasks included in the AAQ). The Timed Stair Test (TST) consists of 4 subtasks: (1) standing up and walking, (2) ascending stairs, (3) turning and descending stairs, and (4) walking back, turning, and sitting down.42,43 We measured the time of the second and third subtasks (ascending and descending stairs). With the Timed “Up & Go” Test (TUG), we measured the time a person takes to get up from a chair, walk 3 m, turn round, walk back, and sit down on the chair again.44,45 With the Timed Chair Stand Test (TCT), we measured the time (in seconds) a person takes to get up from a chair and sit down again 5 times as fast as possible.46 In addition, we measured time getting up from a low couch, time sitting down on a low couch, and time walking 20 m. The performance-based tests were performed as described in the literature. Participants were not allowed to use their hands for the TCT; they were allowed to use the handrail for the TST. The participants were not given a practice trial because it was considered too burdensome for the patients. The 7 performance-based tests were measured in seconds (starting on the investigator's command) and simply summarized into 3 subscores: a subscore for stair climbing (ascending and descending stairs) and subscores for rising and sitting down (TUG, TCT, and sitting down and standing up from a low couch). The 3 subscores were summarized into 1 total performance score. Higher scores indicate worse functioning.
Data Analysis
Scores of the HOOS and KOOS were combined into HOOS/KOOS subscores. Mean (SD) and median (minimum-maximum) scores were calculated for all scales. The total score of the AAQ was correlated with the HOOS/KOOS ADL score, the HOOS/KOOS sport score, the HOOS/KOOS pain score, the total score specific limitations, and the total performance score. The task-specific subscores for rising and sitting down, stair climbing, and walking of the AAQ were correlated with the scores of the Questionnaire Rising and Sitting Down, the Climbing Stairs Questionnaire, the Walking Questionnaire and with the subscores for rising and sitting, stair climbing, and the walking item of the performance-based tests. We expected that scores of instruments measuring the same concept should correlate highly, whereas scores of instruments measuring (slightly) different concepts should correlate moderately to low. We also expected that the AAQ would be less influenced by pain than the physical functioning subscales of self-report questionnaires. The relationships among the concepts measured by the different instruments are shown in the eTable. The following pattern of associations was expected:
The total score of the AAQ correlates highly (at least .70) with the total score of the performance-based tests. This association would support the idea that the AAQ can potentially be a suitable alternative for performance-based tests for large-scale studies.
The total score of the AAQ correlates moderately (.30–.70) with the HOOS/KOOS ADL score, the HOOS/KOOS sport score, and the total score specific limitations.
The subscores of the AAQ correlate highly (at least .70) with the corresponding subscores of the performance-based tests.
The subscores of the AAQ correlates moderately (.30–.70) with the scores of the Questionnaire Rising and Sitting Down, the Climbing Stairs Questionnaire, and the Walking Questionnaire.
The correlation between the AAQ and the performance-based tests would be greater (at least .10) than the correlations between self-report questionnaires and performance-based tests.
The correlation between the HOOS/KOOS pain subscale and the AAQ would be lower (at least .10) than the correlation between the HOOS/KOOS pain subscale and the self-report physical functioning scales (HOOS/KOOS ADL score, the HOOS/KOOS sport score, and the total score specific limitations).
Because of the preliminary nature of our analyses and small sample sizes, no tests for statistical significance were performed. All correlations were calculated as Spearman correlations because some variables (eg, the performance-based score) were not normally distributed. Two-sided 95% confidence intervals (95% CI) were calculated.
Role of the Funding Source
This study was financially supported by a grant from the Anna Foundation, Leiden, the Netherlands. The sponsor had no role in the study design; collection, analysis, or interpretation of data; writing of the manuscript; or the decision to submit the manuscript for publication.
Results
Table 3 presents the correlations among the total scores of the AAQ, the self-report questionnaires, and the performance-based tests. The correlation of the total score of the AAQ with the total score of the performance-based tests was above .70 (.79, 95% CI=.61–.89) as expected. The total score of the AAQ correlated moderately, as expected, with the HOOS/KOOS ADL score (.60, 95% CI=.32–.78) and the HOOS/KOOS sport score (.68, 95% CI=.44–.83) but higher than expected with the total score specific limitations (.90, 95% CI=.81–.95). The correlations among the task-specific subscores (stair climbing, rising and sitting, walking) are presented in Table 4.
Spearman Correlations (95% Confidence Interval) Among the Total Scores of the Animated Activity Questionnaire (AAQ), the Self-Report Questionnaires and the Performance-Based Testsa
Spearman Correlations (95% Confidence Interval) Among the Subscores of the Animated Activity Questionnaire (AAQ), the Subscores of the Self-Report Questionnaires, and the Subscores of the Performance-Based Tests
The correlations of the subscores of the AAQ with the subscores of the performance-based tests were at least .70, as expected, for stair climbing (.79, 95% CI=.61–.89) and rising and sitting down (.70, 95% CI=.47–.84) but lower for walking (.60, 95% CI=.32–.78).
The correlations of the subscores of the AAQ with the scores of the Questionnaire Rising and Sitting Down, the Climbing Stairs Questionnaire, and the Walking Questionnaire were moderate, as expected, for stair climbing (.66, 95% CI=.41–.82) and walking (.63, 95% CI=.37–.80) but higher than expected for rising and sitting down (.86, 95% CI=.73–.93).
The correlations between the AAQ and the performance-based tests (.79, 95% CI=.61–.89) were greater (at least .10) than the correlations between the HOOS/KOOS ADL and sport subscales (.56, 95% CI=.27–.76 (Fig. 2), and .66, 95% CI=.41–.82, respectively), as expected. However, the correlations between the total score specific limitations and the performance-based tests was higher than expected (.84, 95% CI=.70–.92).
Correlations between (A) the Animated Activity Questionnaire (AAQ) and performance-based tests and (B) the Hip Disability and Osteoarthritis Outcome Score/Knee Injury and Osteoarthritis Outcome Score (HOOS/KOOS) activities of daily living (ADL) subscale and performance-based tests.
The correlations between the HOOS/KOOS pain subscale and the AAQ (.49, 95% CI=.18–.71) were lower than the correlations between the HOOS/KOOS pain subscale and the HOOS/KOOS ADL score (.90, 95% CI=.81–.95), the HOOS/KOOS sport score (.62, 95% CI=.35–.79), and the total score specific limitations (.62, 95% CI=.35–.79), as expected.
In the interviews, 30 of the 33 participants (91%) indicated that the AAQ was completely comprehensible. Twenty-five participants (76%) considered the animations completely clear, and 8 (24%) considered the animations mostly clear. In 87% of the cases, the participants recognized themselves completely or mostly in one of the animations of the 7 activities. Eleven participants stated that they would like to distinguish good days from bad days. Two participants missed a level of difficulty (using support such as from a grip, wall, or sink) for rising from and sitting on the toilet. Nineteen participants (58%) preferred the AAQ over the other methods (13 younger individuals [≤60 years of age] and 6 older individuals [>60 years of age]). Eleven participants (33%) preferred the performance-based tests (2 younger individuals [≤60 years of age] and 9 older individuals [>60 years of age]) because in those cases a health care worker was present to observe how they performed. Some participants thought that this approach would give a better picture than completing a questionnaire. None of the participants preferred the self-report questionnaires.
Discussion
We developed the AAQ for patients with hip or knee OA to measure the same aspects of physical functioning as performance-based tests, but using a computer-administered animation questionnaire instead of time-consuming tests. A preliminary validation study also was performed. We expected that scores on the AAQ would correlate highly with the results of performance-based tests and moderately with scores on self-report questionnaires. The results were mostly (75%) consistent, with the general pattern of associations that we expected. Our data also showed that the AAQ was less influenced by pain than the physical functioning subscales of the self-report questionnaires.
Recently, Marsh et al26 and Rejeski et al29 developed a video-animated tool for assessing mobility. They found that the video-animated tool correlated equally high (about .60) with self-reported mobility as with performance-based tests. There are some differences between the method used by Marsh et al and Rejeski et al and the method used in our study. We used different animations per activity for different levels of difficulty, whereas Marsh et al and Rejeski et al used one animation per activity with multiple response options. Marsh et al and Rejeski et al only minimized the reference frame of the question, whereas we also minimized the reference frame of the response options. It might be useful to compare different methods and different kinds of video animations in future research to develop an optimal animation questionnaire.
Some of the correlations of the AAQ with the Climbing Stairs Questionnaire,38 the Questionnaire Rising and Sitting Down,39,40 the Walking Questionnaire,41 and the total score specific limitations were striking. The correlation between the AAQ total score and the total score specific limitations was higher than expected (.90, 95% CI=.81–.95). A similar result was found for the task of rising and sitting down (.86, 95% CI=.73–.93). The specific limitations questionnaires correlated highly with the performance-based tests (Tab. 3). An explanation for these findings might be that the specific limitations questionnaires include statements about what patients actually do (eg, “I climb stairs, but it takes me longer”), instead of asking whether patients have difficulty performing activities. Therefore, the specific limitations questionnaires might be less influenced by the patient's reference frame. Thus, in contrast to other self-report questionnaires, such as HOOS/KOOS, the specific limitations questionnaires appear to measure highly similar aspects of physical functioning as performance-based tests. It could be argued that this similarity means that there is no need for the AAQ because the specific limitations questionnaires seem to measure the same construct. However, the specific limitations questionnaires consist of 89 items, whereas the AAQ consists of only 7 items. No participants preferred the self-report questionnaires. However, this result may be biased by the large number of questionnaires that we used in the context of this study. Nevertheless, 58% of the participants had a clear preference for the AAQ, mostly younger individuals (≤60 years of age). Older participants more often preferred the performance-based tests. No patients preferred the self-report questionnaires. We conclude, therefore, that the AAQ has great potential as a new measure for assessing physical functioning in patients with hip and knee OA.
The AAQ is a relatively inexpensive and easy method for measuring similar aspects of physical functioning as performance-based tests, which makes this method suitable for large-scale studies. It took about 15 minutes to complete the AAQ, about the same amount of time as needed for the performance-based tests. This finding could potentially reduce the need for on-site testing (which reduces costs) and may reduce dropout rates in longitudinal studies. More than half of the participants preferred the AAQ over the other methods. However, it cannot be ruled out that patients had a tendency to please the researcher. This tendency plays a role in both the AAQ and the performance-based tests and is difficult to avoid in this kind of study.
The AAQ is a self-report instrument that can be administered on a computer, on a tablet, or via the Internet. It also may be possible in the future to develop a computer-adaptive test (CAT). A CAT is a computer-administered questionnaire in which questions are selected from a large pool of questions (called an “item bank”) and administered in a sequence and length determined by a computer, based on answers to previous questions. There is increasing attention in the field of measurement for CATs because they offer precise measurement with a limited number of items.47 A first prototype CAT for measuring physical functioning in the field of lower extremity OA has recently been developed.48 The AAQ also could be used for a CAT in the future, using videos as an item bank.
A computer animation questionnaire can easily be adapted for use in different countries because the translation of the questions is very straightforward and the visualization itself does not need translation. A computer animation questionnaire also can easily be adapted for use in other patient groups (eg, patients with low back pain) by adding or removing animations.
However, before the AAQ can be used in patients with hip or knee OA in research or clinical practice, more research is needed for several reasons. First, the content validity of the AAQ needs to be improved by adding more levels of difficulty to some items and adding more activities that are relevant for patients with OA. In this pilot version, only 7 activities were included. More animations of important activities such as getting in and out of a car, dressing, and performing household activities should be developed. The selection of activities should be developed using a combination of patient input (focus groups), the International Classification of Functioning, Disability and Health (ICF) Core Set for OA,49 other instruments, and health professional experts. No indications were found to change the wording and response options of the questionnaire. Second, participants indicated that for some activities (eg, rising and sitting down on the toilet), additional levels of difficulty (eg, a level including support from a grip, wall, or sink) are required, which also can be obtained from focus groups with patients. Third, our findings need to be replicated in larger samples of patients because the sample size of our study was rather small (N=33) and new animations are being developed.
The differences between the correlations were rather small (which was expected because the constructs being measured are quite similar), which means that small changes may lead to other conclusions. Based on the observed correlation of .79 (95% CI=.61–.89) between the total score of the AAQ with the total score of the performance-based tests, we calculated that a sample size of 105 patients is needed for the lower band of the confidence interval to become >.70. In future studies, the AAQ and other measures should preferably be completed on the same day and refer to the same time period because patients indicated they have good days and bad days. Furthermore, it may be interesting to examine whether the correlations of the AAQ with other measures are different in patients with more or less pain. In a previous study, we showed that correlations between performance-based and self-reported physical functioning were higher in patients with less pain compared with patients with more pain.12 Finally, a more detailed assessment of validity is desired (eg, by comparing scores on the AAQ with self-reported or observed limp and reciprocal stair climbing). Also, other measurement properties of the AAQ, such as test-retest reliability and responsiveness, should be assessed. A new study (funded by the European League Against Rheumatism [EULAR]) is currently being performed in which additional animations are being developed. The measurement properties of the new AAQ will be assessed in patients from 6 European countries.
One should keep in mind that the AAQ may solve some of the problems of self-report questionnaires and performance-based tests, but it cannot solve all limitations. For example, like self-report questionnaires,3 the AAQ may not give accurate results in people with impaired cognition, depression, poor memory, or a poor self-image.
In conclusion, the results of this study support the idea that the AAQ measures the same aspects of physical functioning as performance-based tests and that it may be a suitable alternative for use in large-scale studies. Moreover, the AAQ is very welcomed by patients. These results encourage further development and validation of the AAQ.
Footnotes
Each author participated sufficiently in the work to take public responsibility for appropriate parts of the content of the manuscript. Dr Terwee, Dr Roorda, Dr Harlaar, and Professor de Vet contributed to the conception and design of the study. Ms Coopmans, Mr Peter, Dr Poolman, and Dr Scholtes contributed to the acquisition of data. Dr Terwee, Ms Coopmans, Dr Roorda, and Professor de Vet contributed to the analysis and interpretation of the data. Dr Terwee and Ms Coopmans drafted and revised the manuscript, and all authors critically commented on the drafts and provided important intellectual content. All authors approved the final version. Dr Terwee takes responsibility for the integrity of the work as a whole, from inception to finished article.
This study was approved by the Medical Ethics Committee of VU University Medical Center and was conducted in accordance with the Good Clinical Practices protocol and Declaration of Helsinki principles.
The study was financially supported by a grant from the Anna Foundation, Leiden, the Netherlands.
- Received December 3, 2012.
- Accepted September 4, 2013.
- © 2014 American Physical Therapy Association
References
Issue highlights
