Measuring functioning in patients with hand osteoarthritis—content comparison of questionnaires based on the International Classification of Functioning, Disability and Health (ICF)

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

1 Department of Internal Medicine III, Division of Rheumatology, Vienna Medical University, Vienna, Austria, 2 ICF Research Branch of the WHO Collaborating Center for the Family of International Classifications at the German Institute of Medical Documentation and Information (DIMDI) and 3 /label>Department of Physical Medicine and Rehabilitation, Ludwig-Maximilians University, Munich, Germany and 4 Department of Rheumatology, Leiden University Medical Centre, Leiden, The Netherlands

Rheumatology, Volume 45, Issue 12, December 2006, Pages 1534–1541, https://doi.org/10.1093/rheumatology/kel133

Cite

Tanja Stamm, Szilvia Geyh, Alarcos Cieza, Klaus Machold, Barbara Kollerits, Margreet Kloppenburg, Josef Smolen, Gerold Stucki, Measuring functioning in patients with hand osteoarthritis—content comparison of questionnaires based on the International Classification of Functioning, Disability and Health (ICF), Rheumatology, Volume 45, Issue 12, December 2006, Pages 1534–1541, https://doi.org/10.1093/rheumatology/kel133

Abstract

Objective . When selecting a questionnaire, researchers and clinicians need to know whether or not a questionnaire covers the relevant outcomes. The aim of this study was to analyse and compare the content of questionnaires that have been used to assess functioning in patients with hand osteoarthritis (OA) based on the International Classification of Functioning, Disability and Health (ICF).

Method . Questionnaires were identified in a structured literature search. All concepts included in the items of the questionnaires were linked to the ICF categories according to the 10 established linking rules by two health professionals. The degree of agreement between the two health professionals was determined by means of kappa statistic. On the basis of the linking, the content of the instruments was compared. For each concept, it was examined whether functioning is measured on the level of activity or participation or both activity and participation. Indicators for content density, content diversity and the percentage of linked ICF categories addressing participation were calculated.

Results . Health Assessment Questionnaire, AUSCAN, Cochin scale, FIHOA, SACRAH and AIMS2-SF were analysed. The result of the kappa statistic for agreement between the two investigators was 0.74. 163 concepts were identified in the 113 items of all instruments, which were then linked to seven ICF categories of the component body functions , 45 categories of the component activities and participation and six categories of the component environmental factors . AUSCAN and SACRAH had the lowest and AIMS2-SF showed the highest diversity ratio and the highest percentage of linked ICF categories that addressed participation.

Conclusion . When selecting instruments for comprehensive measurements of functioning in hand OA, researchers and clinicians are advised to include both one instrument with a low diversity ratio (for disease-specific aspects) and another instrument with a high diversity ratio (for broader aspects of functioning including some aspects of participation).

Introduction

Osteoarthritis (OA) is the most common joint disease, most frequently involving the joints of the hands [ 1 ] and is characterized by a progressive loss of articular cartilage, deformity, swelling, stiffness, deceased range of joint motion and periods of pain in and around the affected joints [ 2 , 3 ]. Consequently, hand OA leads to a reduction of grip strength [ 4 ], increasing difficulties when performing tasks of everyday life [ 5 ], loss of productive work time [ 6 ] and a decreased ability to perform manual activities [ 7 ]. Hand OA has an enormous socio-economic impact because it affects 60–70% of the population above the age of 65 [ 8 ] and in particular, women already above the age of 47 [ 9 ]. Since, almost 80% of the population can expect to live through most of their seventh decade of life, the socio-economic impact of OA is likely to increase even further in future [ 8 ].

Pain and loss of hand movement-related functions decrease the patients’ ability to perform manual tasks and diminish their quality of life. No causative treatment has been proven to influence progressive cartilage damage [ 10 ]. Thus, persons with hand OA are commonly referred to occupational and physical therapists to improve their performance in daily activities [ 11 ].

In order to measure the clinical effects of treatment, measurement of functioning is recommended in the literature by expert consensus [Outcome Measures in Rheumatology (OMERACT)–Osteoarthritis Reasearch Society International (OARSI)] for hip, knee and hand OA [ 12 ], as well as specifically for hand OA [ 3 ], although there is no consensus on how functioning should be measured. Various self-report questionnaires have been developed [ 13 , 14 ]. In contrast to the so-called hand function tests, which require trained observers and a specific setting in time and place, questionnaires may be considered more feasible in busy clinical settings because they do not need the presence of professional staff when administered. However, the questionnaires recommended in the literature [ 13 , 14 ] have considerable differences in length and item content.

In addition, it is not clear whether these questionnaires measure functioning on the level of activity or include both activity and participation. The International Classification of Functioning, Disability and Health (ICF) offers a comprehensive understanding of functioning. In the ICF, functioning is described as the complex interplay of the health components body functions, body structures, activities and participation and contextual factors , such as environmental and personal factors ( Fig. 1 ). Activity is defined in the ICF as the execution of a task or action by an individual, whereas participation is the person's involvement in a daily life situation [ 15 ]. Participation may thus be considered important from the perspective of the patients, because it refers to whether restrictions are experienced in daily life situations. It has not been analysed to date to what extent the recommended instruments measure functioning on the level of activity or participation, or both activity and participation.

The model of the ICF. The model shows the relationship between the ICF components body functions and body structures, activities and participation and the contextual factors: environmental factors and personal factors . Each component consists of chapters which then consist of categories. Within the component activities and participation, Chapter 4 Mobility includes the categories d440 Fine hand use (second-level) and d4100 Picking up (third-level) among many others.

Content validity is a most important issue that should be considered when a questionnaire is selected as an outcome measure. While Rasch analyses and other modelling approaches focus on the relative difficulty of items related to the scoring ability of the investigator and the discriminative capacity of the items, content validity is essential in order to ensure that an instrument measures all the relevant aspects of an outcome [ 16 ]. Researchers and clinicians then need to know whether or not a questionnaire covers the relevant study outcomes or endpoints. The ICF provides a comprehensive frame of reference, which allows comparing the content of health status instruments. On the basis of this comparison, researchers and clinicians will be able to select an appropriate instrument that covers the relevant outcomes.

ICF-based tools, the so-called ICF Core Sets, have been developed to map functioning specific to patients with a certain health condition from the comprehensive perspective of the ICF. The ICF Core Set for OA is a list of the ICF categories, which represent the typical problems of patients with hip, knee and hand OA. These categories were selected by expert consensus based on results of preliminary research, including a systematic literature review, a Delphi exercise and a cross-sectional study with patients [ 17 ]. The ICF Core Set should serve as a standard to map functioning in patients with OA. It has not been analysed to date which aspects of the ICF Core Set for generalized OA relate to hand OA specifically and are covered by the hand functioning questionnaires.

The aim of this study was to analyse and compare the content of questionnaires that have been used to assess functioning in patients with hand OA. The specific aims were (i) to compare the content of the instruments based on the linking with the ICF, (ii) to determine whether each concept in the area of activities and participation addresses either activity or participation, or both activity and participation and (iii) to determine which aspects of the ICF Core Set for OA are covered by the questionnaires.

Methods

Search strategy

A structured literature search was undertaken in summer and fall 2005. The following databases were searched using the keywords ‘instrument’, ‘measure’, ‘assessment’, ‘questionnaire’, ‘functional assessment’, ‘function’ and ‘hand OA’: CINAHL R (database) (1982–1998 and 1999–2005), PsychINFO (1988–2005), Embase (1988–2005) and Medline (1967–2005). Eligibility of the questionnaires was checked in three steps. In the first step, descriptive, evaluative and psychometric studies were selected. Case reports, economic evaluations, primary prevention studies and reviews were excluded. In the second step, studies and articles were selected that report the use of functioning questionnaires in hand OA. All questionnaires used were documented. In the third step, the following criteria for inclusion of the instruments were applied: questionnaires that (i) assess hand functioning, (ii) were specifically developed for OA of the hand, were condition-specific for hand OA or developed for arthritis in general, which have then been applied to patients with hand OA, (iii) were published in a peer-reviewed journal, (iv) exist in English version.

Generic questionnaires for upper-limb function were excluded from this analysis, because they are not commonly used in patients with hand OA and are thus not recommended in the literature for measurement of functioning in patients with hand OA [ 13 , 14 ]. Hand function tests were also excluded from the analysis, because they require trained observers and are often not feasible in busy clinical setting for screening or routine care. The present analysis thus specifically focused on questionnaires specific for patients with hand OA, which are feasible in a clinical setting for routine assessment.

Generic questionnaires that assess health status and health-related quality of life such as the SF-36, WHODAS or the EuroQoL were excluded, because this study focused on questionnaires that were designed to specifically assess functioning in hand OA or that were developed in other arthritides and adapted to patients with hand OA. The content of generic questionnaires that assess health status or health-related quality was explored by Cieza and Stucki [ 18 ] who linked these questionnaires to the ICF.

Linking to the ICF

In order to compare the content of the identified instruments based on the ICF and to examine the differences, the concepts contained in each item of the questionnaires were linked to the appropriate ICF category.

In the ICF classification, the letters b, s, d and e [standing for body functions (b), body structures (s), activities and participation (d) and environmental factors (e)], which refer to the components of the ICF, are followed by a numeric code starting with the chapter number (one digit), followed by the second level (two digits) and the third and fourth levels (one digit each). The component letter with the suffix of one, three, four or five digits represent the code of the so-called categories. Categories are the units of the ICF classification. Within each chapter, there are individual two-, three- or four-level categories. An example selected from the component activities and participation (d) would result in the following code: ‘ b2 Sensory functions and pain’ is the first level, ‘ b280 Sensation of pain’ represents the second level, ‘ b2801 Pain in body part’ corresponds to the third level and ‘ b28016 Pain in joints’ to the fourth level. At the end of each chapter, there are ‘other specified’ categories (uniquely identified by the final code 8) and ‘unspecified’ categories (uniquely identified by the final code 9).

Linking rules have been developed to link health-status measures to the ICF in a specific and precise manner [ 19 , 20 ]. On the basis of these linking rules, each item of an instrument should be linked to the ICF category that most precisely represents the item's contents. An item of a questionnaire can include more than one concept, thus the first step of the analysis was to identify the concepts in each item. A concept was defined as one separate meaningful entity, such as a body structure, a body function, an activity or a contextual factor. An example is the following item of the SACRAH questionnaire: ‘Please, assess the pain you had during the last 48 hours caused by your finger joint conditions. How severe was your pain during regular daily work?’. In this item, the following two concepts were identified: ‘pain in finger joints’ and ‘regular daily work’. The two concepts were linked to the following two ICF categories: b28016 Pain in joints and d850 Work & employment .

If the content of a concept was not explicitly named in the corresponding ICF category, the ‘other specified’ option at the third and fourth coding level of the ICF classification was linked and the additional information was documented. An example is the concept ‘turning the pages of the newspaper’, which was linked to the ICF category d4458 Hand and arm use, other specified because ‘turning pages’ is not explicitly named in the ICF. If the content of a concept was more general than the corresponding ICF category, the code of the higher level was linked. An example is the concept ‘needing help’ to get out of bed in an item of the AIMS2-SF, which was linked to Chapter 3 Support & relationships because ‘needing help’ is more general than the available second level categories in this chapter. According to another rule, if a concept was not contained in the ICF classification, this concept was assigned ‘nc’ (not covered) [ 19 ]. An example is the concept ‘inactivity’ (item: ‘pain in times of inactivity’) in the AUSCAN and the SACRAH, which was found to be not covered by the ICF and was therefore linked to ‘nc’.

In order to differentiate between activity and participation, it was examined for each concept linked to componen ( d ) (activity and participation component) whether the linked ICF category refers to a task (capacity–activity) or to life-involvement (performance–participation) according to the ICF model. Capacity refers to an individual's ability to execute a task or an action in a ‘standardized environment’ and performance describes what an individual does in his or her current environment [ 15 ] (p. 15). The judgement was made by two health professionals according to the description or definition of the item of the instrument in the literature.

The number of the concepts identified in each questionnaire and the ICF categories linked were reported both in total and separated by component. The ICF categories were then compared on the second level with the categories in the ICF Core Set for OA. In addition, content density and content diversity were calculated. Content density means the average number of concepts per item. A value of one indicates that one concept was contained in each item. A value exceeding one indicates that more than one concept is identified in some items. Content diversity refers to the number of ICF categories per concept. A value of one indicates that each concept was linked to a different ICF category. A value below one indicates that several concepts were linked to one and the same ICF category. In addition, the number of linked ICF categories that were attributed to participation was reported. For each linked category, it was examined whether the category was covered in the ICF Core Set for OA.

Accuracy and rigour of the analysis

Consensus between two health professionals (B.K., T.S.) was used to decide which concepts were identified in all items of the questionnaires and which ICF category should be linked to each concept. In the case of disagreement between the two health professionals, a third person trained in the linking rules was consulted. In a discussion led by the third person (A.C.), the two health professionals that linked the concept stated their pros and cons for the identification of a concept and for linking this concept to a specific ICF category. Based on these statements, the third person made an informed decision.

The degree of agreement between the two health professionals regarding the identified and linked concepts was calculated by means of the kappa statistic [ 21 ]. Values of kappa generally range from 0 to 1, whereas 1 indicates perfect agreement and 0 indicates no additional agreement beyond what is expected by chance alone. Kappa coefficients above 0.61 are regarded as good [ 22 ]. The data analysis was performed with SPSS 12.0.1 on a personal computer.

Results

Literature search

The literature search produced 37 hits in CINAHL, 115 in PsychINFO and Embase and 974 in Medline in the first step. In the second step, the selection criteria for the articles were applied and 27 articles and four abstracts were reviewed in detail. In the third step, six questionnaires were identified according to the selection criteria and included in the analysis.

Instruments

The following questionnaires were included in this analysis: the Health Assessment Questionnaire (HAQ) [ 23 , 24 ], the AUSCAN [ 25 , 26 ], the Cochin scale [ 27 , 28 ], the Functional Index of Hand OA (FIHOA) [ 29 , 30 ], the Score for Assessment and Qualification of Chronic Rheumatoid Affections of the Hands (SACRAH) questionnaire [ 31 , 32 ] and the Arthritis Impact Measurement 2 Short Form (AIMS2-SF) questionnaire [ 33 , 34 ]. The AIMS2-SF questionnaire includes 26 items and is the short form of the AIMS2 questionnaire, which has 78 items. The short form was selected because of the afore described feasibility focus of this analysis. The number of items of the questionnaires and the number of sub-scales or domains are presented in Table 1 .

Links between the ICF categories and the questionnaires

The table shows the categories of the ICF and the linking to the concepts included in the items of the instruments. To all concepts marked with (*) a performance perspective was attributed. To all other concepts we attributed an capacity perspective, except the concepts marked with (•), which were coded with both a performance and a capacity perspective. All marked categories with (▪) in the far right column are also included in the ICF Core Set for OA on the second level of the ICF classification. The ICF Core Set for OA is a short list of the most important ICF categories representing the typical spectrum of functioning problems in patients with hip, knee and hand OA [ 17 ].

Links between the ICF categories and the questionnaires

The table shows the categories of the ICF and the linking to the concepts included in the items of the instruments. To all concepts marked with (*) a performance perspective was attributed. To all other concepts we attributed an capacity perspective, except the concepts marked with (•), which were coded with both a performance and a capacity perspective. All marked categories with (▪) in the far right column are also included in the ICF Core Set for OA on the second level of the ICF classification. The ICF Core Set for OA is a short list of the most important ICF categories representing the typical spectrum of functioning problems in patients with hip, knee and hand OA [ 17 ].

As the AUSCAN questionnaire is not a public domain and was thus not available, the linking exercise was performed using the candidate items from the publication [ 26 ] instead of linking the items from the questionnaire.

Linking to the ICF

The result of the kappa statistic for agreement between the two investigators was 0.74 ( P < 0.0001). Thus, the kappa coefficient exceeded the value 0.61, which has been regarded as good. One hundred and sixty-three concepts were identified in the 113 items of the questionnaires, which were then linked to seven ICF categories of the component body functions , 45 categories of the component activities and participation and six categories of the component environmental factors . No concepts were linked to categories of the components body structures and personal factors . The following four concepts were not found to be covered by the ICF: the concept of ‘inactivity’ in relation to the assessment of pain in the finger joints in both the AUSCAN and the SACRAH, the concept of pain ‘during the night’ in the SACRAH and ‘accepting a handshake without reluctance’ in the FIHOA.

Table 1 shows the comparison of the concepts contained in the items of the instruments using the ICF categories as a reference and ordered by the component. The numbers in the table represent the frequencies with which the ICF categories were found to be addressed in the different instruments. A higher number indicates that several concepts from a specific instrument were linked to the same ICF category. For example, the category d4402 Manipulating was linked to the following five items in the SACRAH: ‘to fasten your bra/tie your tie’, ‘to button up and unbutton your shirt/blouse’, ‘to do up or undo a zip’, ‘to strike a match’ and ‘to handle paper-money’. Table 2 shows the items and the concepts of the SACRAH with corresponding ICF categories as an example for the linking exercise.

Linking of the items of the SACRAH questionnaire to the ICF

The table shows the linking of the items of the SACRAH to the categories of the ICF as an example. Every item of the instrument was linked to the appropriate ICF category. In the far right column, either a capacity (C) and/or performance (P) perspective was attributed to each ICF category that was linked.

Linking of the items of the SACRAH questionnaire to the ICF

The table shows the linking of the items of the SACRAH to the categories of the ICF as an example. Every item of the instrument was linked to the appropriate ICF category. In the far right column, either a capacity (C) and/or performance (P) perspective was attributed to each ICF category that was linked.

The total number of the ICF categories linked to the concepts of the questionnaires show considerable differences. The concepts of the AIMS2-SF and the HAQ were linked to 32 and 31 ICF categories, respectively, while the FIHOA and the AUSCAN were linked to seven and 10 ICF categories, respectively. Some categories were linked up to five times to the same ICF categories, for example the ICF category b28016 Pain in joints ( Table 1 ), indicating the existence of five items with similar content in one and the same questionnaire. Table 3 shows the total number of the items, concepts and linked categories as well as the ratios of content density and diversity. The HAQ shows the highest density ratio which means that in most items more than one concept was identified and linked. The AIMS2-SF has the highest diversity ratio indicating that most concepts were linked to different ICF categories ( Table 3 ).

Frequencies of items, concepts and ICF categories in relation to each other

The frequencies of the items and concepts from the questionnaires and ICF categories are reported in total numbers (percentages in brackets). The far right column indicates the total number of categories in the ICF as well as the number of ICF categories per component. The total number of 1454 ICF categories comprises 310 categories in the component body structures , 493 categories in the component body functions , 393 categories in the component activities and participation and 258 categories in the component environmental factors .

Frequencies of items, concepts and ICF categories in relation to each other

The frequencies of the items and concepts from the questionnaires and ICF categories are reported in total numbers (percentages in brackets). The far right column indicates the total number of categories in the ICF as well as the number of ICF categories per component. The total number of 1454 ICF categories comprises 310 categories in the component body structures , 493 categories in the component body functions , 393 categories in the component activities and participation and 258 categories in the component environmental factors .

‘Performance’ was attributed to six concepts of the AIMS2-SF which were linked to four ICF categories in the component activities and participation. Both ‘capacity’ and ‘performance’ were found to be represented in the AIMS2-SF, the HAQ and SACRAH ( Table 3 ). All other concepts were attributed a ‘capacity’ perspective only.

Eighty-six point two per cent of the linked ICF categories from the questionnaires are also covered in the ICF Core Set for OA ( Table 1 ). However, if all questionnaires are taken together, the ICF Core Set for OA includes nine additional categories in the component body functions , four in the component activities and participation and 13 additional environmental factors , which are not addressed in any of the questionnaires (data not shown).

Discussion

The first aim of the present study was to compare the content of the six different questionnaires in relation to the ICF. The areas covered by the questionnaires differ importantly. Both the AIMS2-SF and the HAQ were developed and used in different forms of arthritis, while the AUSCAN, the Cochin, the FIHOA and the SACRAH were developed specifically for patients with hand OA [ 23 , 26 , 27 , 30 , 31 , 33 ]. However, all six questionnaires were used in hand OA clinical trials and are recommended to measure functioning in hand OA [ 12 ]. As there is an important variance in the content of the questionnaires, it is important for clinicians and researchers who wish to select an instrument for measuring a specific endpoint to know which areas are covered by which instrument and which areas are not covered at all.

The AIMS2-SF has the highest diversity ratio (linked ICF categories per concepts) indicating that almost each concept was linked to a different ICF category. Both the SACRAH and the AUSCAN have considerably lower content diversity ratios. Thus, more concepts were linked to one and the same ICF category. An example is the ICF category b28016 Pain in joints to which five items of the AUSCAN and four items of the SACRAH were linked. Another example is the ICF category d4402 Manipulating to which three items of the AUSCAN, two items of the Cochin, four items of the FIHOA and four items of the SACRAH were linked. It could be argued that the content of these multiply linked items indicates that these measures cover problems specific to people in hand OA in much more detail, also accounting for the variability of problems across different situations than the questionnaires with a higher diversity ratio. However, some aspects of participation such as leisure activities which may be specifically important to patients with hand OA are covered only in the AIMS2-SF. When selecting an instrument for comprehensive measurement of functioning, researchers and clinicians may wish to include both one instrument with a low diversity ratio (measuring disease-specific aspects) and another instrument with a high diversity ratio (measuring broader aspects of functioning including some aspects of participation).

The HAQ has the highest content density ratio (concepts per item), which indicates that on an average more than two concepts are included in each item. This might have the implication that the HAQ is difficult to score for the patients because more than one concept has to be scored in one item. An example is the item ‘dress yourself including tying shoelaces and doing buttons’, which includes the following three concepts: ‘dressing’, ‘tying shoelaces’ and ‘doing buttons’. The patient's abilities to dress, to tie shoelaces and to do buttons may be different from each other, although the patient has to give one answer to this item.

Only two instruments, the HAQ and the AIMS2-SF cover environmental factors. The HAQ includes only ‘assistive products and devices’, while the AIMS2-SF also covers the social environment, namely attitudes of friends and immediate family members and social support and relationships.

The second aim was to determine whether each concept in the area of activities and participation addresses either activity or participation or both activity and participation. The ICF offers some possibilities to differentiate between activity and participation. One of these possibilities is to attribute a capacity or performance qualifier to each ICF category [ 15 ]. In the present study, the perspective of capacity or performance could be attributed to every item from the category d ( activity and participation ) of the ICF. Only the AIMS2-SF includes items that measure solely participation. An example is the item ‘How often were you on the telephone with close friends or relatives?’, which was linked to the ICF category d750 Informal social relationships . Talking on the telephone to friends was found to be an important issue for patients with OA in the CMCI joints in a qualitative study because cell phones are considerably smaller than traditional phones and require increased fine motor skills of the user [ 35 ]. HAQ and SACRAH include items that reflect both a performance and capacity perspective. However, the HAQ reflects performance only from two perspectives, namely ‘walk outdoors on flat ground’ and ‘run errands and shop’. The SACRAH assesses a ‘regular daily work’ and ‘intensive work’ from the perspective of whether the patient experiences joint pain during work. Depending on the understanding of the patient who fills out this questionnaire, this item may assess both performance and capacity or capacity only in relation to the joint pain.

The third aim was to determine which aspects of the ICF Core Set for OA are covered by the questionnaires. The ICF Core Set for OA includes substantially more ICF categories than were linked to the concepts in all questionnaires. However, this ICF Core Set has been developed for hand, hip and knee OA. Therefore, the hand-related aspects of the ICF Core Set for OA that were linked to the questionnaires, and are shown in Table 1 , could thus be a first step towards a standard on what to measure in hand OA trials. On the basis of this content comparison, researchers and clinicians could thus select an appropriate questionnaire that covers the relevant study endpoints.

Conclusion

The content of the questionnaires assessing functioning in patients with hand OA differs considerably. On the basis of this content comparison, researchers and clinicians are able to select an appropriate questionnaire that covers the relevant study endpoints. When selecting an instrument for comprehensive measurement of functioning, researchers and clinicians may wish to include both one instrument with a low diversity ratio (measuring disease-specific aspects) and another instrument with a high diversity ratio (measuring broader aspects of functioning including some aspects of participation). In addition, the hand-related aspects of the ICF Core Set for OA that were linked to the questionnaires could be a first step towards a standard on what to measure in hand OA trials.

Acknowledgement

The authors would like to thank Dr Valerie Nell and Dr Lucila Stange Rezende for reviewing the manuscript.