To test the Norwegian version of the Canadian Occupational Performance Measure (COPM) for validity, responsiveness, and feasibility in patients with hand osteoarthritis.
To test the Norwegian version of the Canadian Occupational Performance Measure (COPM) for validity, responsiveness, and feasibility in patients with hand osteoarthritis.
Seventy-nine patients completed a COPM interview and several self-reported health status questionnaires, including Arthritis Impact Measurement Scales 2, modified Health Assessment Questionnaire; Western Ontario and McMaster Universities Osteoarthritis Index, and the Australian/Canadian Osteoarthritis Hand Index. Rescoring of the same instruments was performed 4 months later, after an intervention.
The COPM detected a great variability of occupational performance problems. The hypotheses for testing validity were confirmed. Mean COPM change was 1.51 (P < 0.001) in performance score, and 2.22 (P < 0.001) in satisfaction score. The median time spent on the COPM interview was 30 minutes (range 10–70 minutes). The patients found the questions easy to understand, but 37% reported problems performing the scoring procedure.
The Norwegian version of the COPM is a valid and responsive instrument for use in clinical practice in osteoarthritis patients. It may serve as an instrument to promote a client-centered approach and as a supplement to other health measures in the planning and evaluating of interventions. Feasibility regarding scoring needs to be improved.
In the last decades there has been a change in the health care system, with an increased emphasis on client participation in the treatment or rehabilitation process. Improved occupational performance and community participation have also, to a larger extent, become important goals and outcomes of rehabilitation (1, 2). Patients with rheumatic diseases usually experience limitations in performing activities of daily living. The kind of occupational performance problems the individual will experience is to a large degree influenced by the patient's disease, personality, age, social situation, and the physical and cultural context (3). Intervention should be directed toward the problems that the patients regard as most important. Thus, there is a need to develop tools that describe the patients' problems within a wide range of occupations, to know which of these problems cause greatest concerns to patients, and to measure how they experience and rate their performance of these activities.
In the field of rheumatology, the issue of including the patient perspective in clinical practice and research has been discussed for some years, and was one of the main topics of the Outcome Measures in Rheumatology Clinical Trials (OMERACT) conference in 2000 (4). To date, rheumatic patients' preferences in activities of daily living have been addressed in a few outcome measures; e.g., the McMaster Toronto Patients Preference Disability Questionnaire (5), and the Problem Elicitation Technique (6).
The Canadian Occupational Performance Measure (COPM) (7) is a generic and individualized measure designed to detect changes in patients' self perception of occupational performance and satisfaction with performance over time. The assessment is theoretically based on the Canadian Model of Occupational Performance, in which occupational performance is defined as, “consisting of self-care, productivity, and leisure; being influenced by the environment, one's social roles and one's developmental level; being client-defined; and consisting of both a performance (objective) dimension and a satisfaction (subjective) dimension” (8).
COPM is currently used in >25 countries in the world and has been translated to many languages, including Icelandic, Danish, and Swedish. The psychometric properties of the instrument have been established in some groups of patients, including patients with chronic pain (9) and disabled individuals living in the community (8). In a recent study, construct validity was tested by comparing COPM performance scores with Health Assessment Questionnaire (HAQ) scores in a group of patients with rheumatoid arthritis (10). The results showed a significant correlation between the COPM scores and the HAQ scores for similar activities.
The COPM manual reports that the intraclass correlation coefficients for test-retest reliability of performance and satisfaction scales range from 0.75 to 0.89 in a mixed group of patients, and concludes that COPM is a reliable measure (7). COPM has also revealed high responsiveness for measuring change in occupational performance (9, 11, 12).
Several studies of feasibility report that patients and therapists in general find the administration of COPM easy, and that the process seems to identify a wide range of occupational performance problems. It may serve as a basis for the process of establishing targeted outcomes, planning further intervention, and evaluating the effect of treatment or rehabilitation (8, 9, 12–14).
The COPM was recently translated to Norwegian (15). The OMERACT filter for outcome measures in rheumatology recommends that all outcome measures should be evaluated according to the criteria of truth, discrimination, and feasibility (16). In this study, the psychometric properties of the Norwegian version of the COPM were tested for validity, responsiveness, and feasibility in a group of patients with hand osteoarthritis.
This study is part of a larger study aimed at describing clinical and functional consequences of hand osteoarthritis, and has been approved by the regional Ethical Committee. Participants were recruited among patients previously referred to a rheumatology outpatient department. All patients diagnosed with hand osteoarthritis within the 2 last years were invited to participate in the study. Of those willing, the first 100 were consecutively included in the COPM study. Inclusion criteria were age between 50 and 70 years, ability to communicate well in Norwegian, and osteoarthritis of the hand according to the classification criteria of the American College of Rheumatology (17). Patients with no described or prioritized occupational performance problems in COPM, with functional problems mainly related to diseases other than osteoarthritis, or with cognitive deficits affecting the interview or scoring process were excluded.
The study was performed with a longitudinal design. The baseline data collection comprised a comprehensive clinical examination by a rheumatologist, including completion of several self-reported health status questionnaires. Within 3 weeks after this examination, the patients completed the COPM interview and scoring with an occupational therapist. The time spent on interview and scoring was recorded in minutes. After the interview, the patients were asked about their impression of the interview, if they found the questions easy or difficult to understand, and how they experienced the scoring procedure.
The data collection was followed by occupational therapy intervention. Depending on the problems described and prioritized by the patient, the intervention could be information about osteoarthritis, teaching of energy conservation and alternative working methods, providing appropriate technical aids and orthotics, or discussing alternative activities that could replace former valued occupations when these had become too difficult for the patient to perform. The therapist also noted whether the information revealed during the COPM interview was helpful in goal formulation, planning, and monitoring of the treatment process.
The second data collection was performed 4 months after baseline and comprised a rescoring of the COPM and completion of the self-reported health status measures. Neither the patient nor the observer had access to the baseline scores. All the interviews and rescorings were carried out by the same therapist.
In COPM, the patients define their occupational performance problems within 3 areas of self care, productivity, and leisure during a semistructured interview (7). The patients are asked how important it is to be able to do each of these activities, and each performance problem is rated for importance on a scale of 1–10, (1 = not important at all; 10 = extremely important). Finally, the 5 most important activities are rated for performance (1–10 where 1 = not able to do it) and satisfaction with performance (1–10 where 1 = not satisfied at all) by asking the patient to rate the way he or she does this activity now, and how satisfied he or she is with this performance. Total performance and satisfaction scores are calculated by dividing the sum of the scores by the number of problems. Change in performance and satisfaction may then be measured by rescoring the prioritized problems after an agreed period of time. A change of 2 or more is regarded as a clinically important difference (7).
The modified HAQ (MHAQ) (18) is a modified version of the Standford Health Assessment Questionaire and was developed to measure disability in patients with arthritis. In MHAQ, patients rate their difficulties in performing 8 activities of daily living on a scale ranging from 1 (without difficulty) to 4 (unable to do).
Arthritis Impact Measurement Scales 2 (AIMS2) (19) is a revised and expanded version of AIMS, developed as a multidimensional instrument to measure health-related quality of life in arthritis patients. AIMS2 consist of 12 scales: mobility, walking and bending, hand and finger function, arm function, self care, household activities, social activities, support from family and friends, arthritis pain, work, level of tension, and mood. In question 60, the patients are asked to check 3 of these 12 areas in which they would most like to experience improvement in health (20).
The Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (21) is a disease-specific measure of pain, stiffness, and physical disability in patients with osteoarthritis in the hips or knees. The Australian/Canadian Ostearthritis Hand Index (AUSCAN) (22) measures pain, stiffness, and physical disability in patients with osteoarthritis in the hands. The patients rate their symptoms and physical disability on a scale of 0–4, with lower scores indicating lower levels of symptoms and disability.
The patients' global assessment of disease activity was recorded on a 100-mm visual analog scale, ranging from 0 (indicating low activity) to 100 (indicating high activity).
Content validity is concerned with the sampling adequacy of items for the construct that is being measured, and is always based, to a large degree, on clinical judgment (23, 24). Regarding COPM, the following questions were asked: Are the items or questions in COPM appropriate to identify occupational performance problems? and Do they adequately measure all dimensions of occupational performance? The analysis was based on what kind of occupational performance problems the patients listed during the interview, and the distribution of problems in the 9 categories in the COPM.
Criterion validity may be tested by looking at how a measure is systematically related to other tests that measure the same construct (25). Criterion validity in COPM was assessed by evaluating correspondence between patient's priorities for improvement in health (as rated in AIMS2) and the problems prioritized in COPM, and by looking at the correlation between COPM scores and the scores from MHAQ, AUSCAN, and WOMAC. A moderate (30–70%) to high (>70%) correspondence was expected for AIMS2 and COPM, whereas a low (0–0.3) to moderate (0.3–0.7) correlation was expected regarding COPM and the self-reported health status measures. The rationale for the last hypothesis was that although COPM is constructed as an individual measure, rating activities that are both important and difficult for the patient to do, the other health status measures primarily are constructed to measure function at a group level, thereby often containing activities that are not necessarily important or difficult for the patient to perform.
When evaluating correspondence between AIMS2 and COPM, the following procedure was conducted: The prioritized COPM problems were categorized into one of the categories in AIMS2. Correspondence between AIMS2 and COPM was calculated as percentage of the AIMS2 priorities overlapping with COPM problems. However, the areas support from family and friends, arthritis pain, level of tension, and mood had to be excluded because there are no activities that correspond directly with these categories.
Construct validity describes to what degree the results of an instrument agree with hypotheses previously described regarding the construct being measured (26). The hypothesis in this study was that disease activity is one factor determining occupational performance and satisfaction with performance, and was tested by using Pearson's correlation coefficients to calculate correspondence between patients' global assessment of disease activity and COPM scores for performance and satisfaction at baseline.
Responsiveness describes how well an instrument is able to measure changes within an individual over time and should be determined for all instruments used in an evaluative manner (27). A paired sample t-test was applied for testing changes in occupational performance and function between initial scoring and reassessment. Standardized response mean (SRM) was calculated for each instrument by dividing the mean change score by the standard deviation of the mean change score. SRM and changes in COPM scores were compared with SRM and changes in functional scores in MHAQ, WOMAC, and AUSCAN.
Feasibility was considered by looking at the time spent on the COPM, the patient's comments regarding the interview and scoring process, and whether or not the results from the assessment could serve as a basis for planning and evaluating the intervention (27). The comments from patients regarding the interview and scoring process were written down and later grouped and categorized.
For the statistical analysis, SPSS for Windows (version 10.1; Chicago, IL) was used. Although MHAQ, AUSCAN, WOMAC, and COPM are categorical measures, parametric tests were used because the results met the necessary assumptions of normal distribution, and the possibility of calculating means (24). The results from patients who had changed medication or undergone surgical treatment during the intervention period were included in the testing of validity and feasibility but excluded from the analysis of responsiveness. All P values <0.05 were considered significant.
Of 100 patients, 79 completed the initial assessments. Six patients did not show up for the COPM interview and 15 were excluded due to the following reasons: no described or prioritized occupational performance problems (n = 13), functional problems mainly related to cerebral palsy (n = 1), and cognitive deficit (n = 1). The demographics of the patients are presented in Table 1. In general, the excluded patients had a shorter disease duration and better function than the patients in the test sample.
|Test group n = 79||Excluded n = 15||P|
|Age, mean ± SD years||63.2 ± 5.4||61.5 ± 5.2||0.32|
|Disease duration, mean ± SD years||14.3 ± 6.9||12.0 ± 8.2||0.39|
|Comorbidity, no. (%)||40 (50)||3 (20)||0.09|
|Living alone, no. (%)||28 (35)||4 (27)||0.74|
|Still working, no. (%)||23 (29)||6 (40)||0.50|
|MHAQ, 1–4, mean ± SD||1.6 ± 0.4||1.1 ± 0.2||<0.001|
|WOMAC, 0–4, mean ± SD||1.5 ± 0.9||0.5 ± 0.5||<0.001|
|AUSCAN, 0–4, mean ± SD||2.0 ± 0.8||0.7 ± 0.3||<0.001|
|Disease activity, 0–100, mean ± SD||47.8 ± 22.0||20.6 ± 15.1||<0.001|
Of the total test group, 65 patients completed the reassessment of COPM. Reasons for not performing the reassessment were surgical treatment (n = 6), changed medications (n = 3), unable to meet due to other disease (n = 2), and dropout with no reason given (n = 3).
The patients listed a total of 874 occupational performance problems, and prioritized 378 of these (Figure 1). The occupational performance problems represented 161 different activities, of which opening bottles and jars was most frequently reported (n = 50), followed by wringing out wash cloths (n = 49), rising from chairs (n = 30), walking stairs (n = 30), and buttoning (n = 28). All problems were easily categorized in to 1 of the 9 subcategories in the COPM. The baseline COPM performance score was 4.34 (SD 1.59) and the mean satisfaction score was 3.86 (SD 1.74).
The patients listed a total of 127 priorities for improvement in health in AIMS2, with the categories hand and finger function (n = 37) and household activities (n = 30) as the most frequently prioritized, and work (n = 6) and social activities (n = 5) as the least frequently listed categories. The prioritized problems in COPM corresponded with 91 of these, presenting an overlap of 72%. The correlations between COPM scores and MHAQ, WOMAC, and AUSCAN in general were low (Table 2).
|COPM performance||COPM satisfaction|
In the analysis of criterion validity, the correlation coefficients (Table 2) indicate a small, but significant, relation between disease activity and COPM performance and satisfaction.
As shown in Table 3, there was a slight, but statistically significant, deterioration in the MHAQ score, with corresponding, but not significant, changes in the WOMAC and AUSCAN scores. There were positive changes in the total COPM scores for both performance and satisfaction with performance.
|Mean change||95% CI||P|
|WOMAC function||−0.09||−0.24, 0.04||0.17|
|AUSCAN function||−0.16||−0.33, 0.01||0.06|
|COPM performance||−1.51||−2.04, −0.98||<0.001|
|COPM satisfaction||−2.22||−2.80, −1.63||<0.001|
Of the 65 patients in the responsiveness analysis, 24 had a positive change of ≥2 points in both total COPM performance and satisfaction scores, and none had a negative change of ≥2 points in both these scores. Of the 299 rescored activities, there was a positive change of ≥2 points of both performance and satisfaction in 113 activities, and negative change of ≥2 points in 14 activities. Five of the 24 patients with positive changes of ≥2 points in total scores had a positive change in all their prioritized activities.
The SMR (Figure 2) was large (>0.7) for the COPM scores, moderate for MHAQ (0.3–0.7), and small (<0.3) for WOMAC and AUSCAN. Of the 4 instruments, COPM had numerically the greatest ability to detect change in function, with the satisfaction score being most sensitive.
The average time spent on the interview and scoring was 30 minutes (range 10–70 minutes). Both patients and therapist found the amount of applied time reasonable. The patients all found the questions easy to understand, but 29 patients (37%) reported problems performing the scoring procedure. The most frequent reason given for this was that they were not used to quantifying their situation in numbers. Some patients thought the performance scoring was difficult due to their fluctuating condition, and the scores could therefore change from one day to another. Others were afraid that their low ratings would be interpreted as complaining. Some patients even inverted the scale; as they thought a score of 10 on performance meant not able to do it, when it meant able to do it extremely well.
A frequent comment was that the interview and scoring helped to concentrate and focus on what was important. Some of the patients also stated that experience with similar rating procedures made the COPM scoring easier to complete.
This study highlights that COPM is a highly responsive instrument for measuring changes in function. The individual design of the instrument is probably the main reason for this, giving significant information about which occupational performance problems the patient experiences as most important to address in intervention, thereby assuring that a change is likely to be noticed. In this way, the COPM also provides information that is complementary to other health measures frequently used in clinical research.
In general, the results confirm the validity of the COPM in patients with hand osteoarthritis. The distribution of problems in the 9 categories in COPM corresponds with results from a study by Yelin et al (28), who found that patients with osteoarthritis experienced most losses in performance of household chores, shopping and errands, and leisure, and that they also spent more time than healthy persons in personal care and hygiene. The great variability in the activities described indicates that COPM identifies a wide range of occupational performance problems within different areas. The low number of described problems in the categories work and play/school is probably explained by age, and the fact that only 29% of the patients were still working.
Criterion validity was supported by a high correspondence between patient's preferences for improvement in health in AIMS2 and the prioritized problems in COPM, and also by the parallel in the areas not prioritized, namely work and social activities or socialization. As expected, there were low correlations between the scores from MHAQ, AUSCAN, and WOMAC and the COPM scores, indicating that COPM measures function from a different aspect.
The testing of construct validity supports the hypothesis that disease activity is one important factor determining occupational performance and satisfaction with performance. However, the relatively low correlation also point to the fact that occupational performance is a complex phenomenon, with many factors contributing to the process and experience. Some of these may be the varying complexity of different activities, the working technique used by the individual, and the context in which the activity takes place. Future studies should investigate these associations to increase our knowledge of which factors to address in treatment and rehabilitation.
In the present study, there were significant positive changes in both the performance and the satisfaction scores in COPM. The greater change in satisfaction score than in performance score is similar to findings from other COPM studies (29) and suggests that, for example, changed attitudes and adjusted expectations to performance of various tasks may have increased the patients' satisfaction, even if the performance remained unchanged. A limitation to the study is that the COPM interview and the following intervention were carried out by the same therapist. Some of the improvement demonstrated may therefore be explained as an eager to please effect. However, the variations in scores between the patients and within the activities prioritized by each patient indicate that this is not the only reason, because an eager to please effect would be expected to be distributed more evenly among the patients and activities.
Beaton et al (30) describe how an experience of being better may not only reflect a change in the state of the patient's disorder, but also an adjustment in life, an adaptation to living with the disorder, or a combination of any of these. They suggest that outcome measures should allow patients to generate their own item content to express how the disorder has affected them. The individual design of the COPM ensures these recommendations, thereby promoting responsiveness. When comparing SRM across different instruments, COPM satisfaction is clearly the most sensitive score, followed by COPM performance and MHAQ.
Studies of persons with rheumatoid arthritis have shown that declines in the ability to engage in valued activities significantly increase the risk of developing depression (31) and that satisfaction with abilities seems to mediate the relationship between occupational performance and an increase in depressive symptoms (32). To be able to address this in care, it is important to assess not only functional decline, but also which activities are most valued by the individual and how the limitations are experienced. Interestingly, both an early version of MHAQ (18) and the revised AIMS2 (19) include satisfaction with function as a dimension of health, but the use of these satisfaction measures has been limited (4). However, the results from the present study imply that measuring satisfaction provides sensitive, valid, and clinically important information.
Regarding feasibility, the patients' responses to the COPM interview in general were positive, indicating that the process of describing and prioritizing activity problems provides a good basis for further planning and intervention. However, 37% of the patients experienced the scoring procedure as difficult, a problem also recognized in other studies (8, 29, 33). The fluctuating state of osteoarthritis was one reason given for this, a problem that probably is common for most patients with rheumatic diseases regarding numerical scoring of health status and functional problems. The 10-point COPM rating scale may have increased these problems because it presents a relatively high number of scoring options. A reduction to, e.g., a 5-point scale could be an alternative that merits testing and validation. The age of the patients may have contributed to the scoring problems because difficulties with numeric scoring procedures are known to increase with age (34, 35). Elderly people with long experience from a health care system based on a traditional medical model may also expect the health professional as the expert to be the one who should decide which problems require intervention. The patients' concerns about being perceived as complainers and some of their difficulties in the numeric scoring could be understood as anxiousness of giving wrong answers and may reflect challenges connected to the role as an active participant in the treatment process. Further testing of COPM in younger patients and in patients with other rheumatic diseases may help clarify these aspects.
In summary, the results indicate that COPM is a valid and responsive instrument for use in clinical practice with patients with hand osteoarthritis. Although the scoring process was somewhat difficult for some of the patients, the study also confirms the clinical feasibility of the COPM in the process of goal setting, planning, and monitoring intervention. The COPM may therefore serve both as an outcome measurement and as an instrument to promote a client-centerd approach in various clinical settings.