Mapping of the Outcome Measures in Rheumatology Core Set for Antineutrophil Cytoplasmic Antibody–Associated Vasculitis to the International Classification of Function, Disability and Health

Abstract

Objective

The International Classification of Functioning, Disability and Health (ICF) is a framework and classification of health that describes health along 4 components: body functions, body structures, activities and participation, and contextual factors. This study examined the content of instruments that constitute the Outcome Measures in Rheumatology (OMERACT) core set of outcome measures for antineutrophil cytoplasmic antibody–associated vasculitis (AAV) by “mapping” them to the ICF.

Methods

The content of the instruments included in the AAV core set were linked to the ICF by 2 independent investigators according to previously established ICF linkage rules.

Results

The AAV core set includes 3 measures of disease activity (3 versions of the Birmingham Vasculitis Activity Score), 1 damage measure (Vasculitis Damage Index), 1 patient-reported outcome (Short Form 36 health survey), and death. Linking these instruments to the ICF revealed comprehensive coverage of the ICF components body functions and body structures, limited coverage of the ICF component activities and participation, and complete absence of coverage of contextual factors.

Conclusion

ICF was found to be useful for thematic characterization of a heterogeneous group of outcome measures for AAV, i.e., a group of complex medical conditions. Linking of the instruments selected for the OMERACT AAV core set of outcome measures to the ICF classification revealed limitations in the representation of constructs related to life impact of AAV, represented by the ICF components activities and participation and contextual factors. Further research and methods development are needed to better incorporate important aspects of functioning and health relevant to patients into clinical trials of AAV.

INTRODUCTION

The International Classification of Functioning, Disability and Health (ICF) is a general health status framework based on the bio/psycho/social model of health and disease. It was endorsed by the World Health Organization in 2001 as the international standard to describe and measure health and disability (1).

Significance & Innovations.

• The International Classification of Functioning, Disability and Health (ICF) framework was shown to be useful for analyzing a complex medical condition (antineutrophil cytoplasmic antibody–associated vasculitis [AAV]) from the global bio/psycho/social perspective.

• ICF-based analysis demonstrated that the instruments used to assess outcomes in clinical trials of AAV patients mainly focus on pathophysiologic manifestations of AAV; they display limited ability to assess the overall life impact of AAV on patients.

• A patient-reported outcome measure specific to AAV is needed to be able to assess the impact of AAV on patients’ quality of life.

The ICF views health as a broad concept that is described along 4 ICF components as follows: 1) body functions, 2) body structures, 3) activities and participation, and 4) contextual factors (1). The level of functioning and health of the individual patient is viewed as being the result of the relationship between impairments of various body functions and body structures, limitations in activities, restrictions in participation, and the influence of contextual factors. Contextual factors represent the collection of an individual’s personal, social, and attitudinal environment that influences the relationship between the other ICF components and modifies their effect on the overall health; when the overall effect of the contextual factors is positive, they are called facilitators of health, and when it is negative they are called barriers to health (1). As an example, the actions of a supporting family member to help a patient get around and preserve an acceptable level of functioning may diminish the negative effect on the patient’s well-being of needing to use a wheelchair as a result of a medical condition; family support in this case represents a contextual facilitator of health.

In addition to the framework, the ICF offers a classification system to describe functioning and health using 1,424 categories organized into a 4-level, hierarchically nested structure covering the 4 components described above. A lower-level category within a particular component represents a more detailed example (and therefore a subset) of a higher level category. ICF classification provides a methodology to comprehensively describe a patient’s general health status in a standardized approach, highlighting the relationship between an individual’s physical state of health, personal or environmental factors, and functioning. A number of applications of the ICF have been explored since its endorsement in 2001, ranging from summarizing disability statistics and facilitating development of health information systems on the broad population level, (2) to describing domains that are relevant when studying patients with specific medical conditions (3).

Vasculitides are complex conditions characterized by inflammation of blood vessels. Antineutrophil cytoplasmic antibody–associated vasculitis (AAV) is the best studied group of vasculitides. The group consists of 3 conditions: granulomatosis with polyangiitis (Wegener’s) (GPA), microscopic polyangiitis, and eosinophilic granulomatosis with polyangiitis (Churg-Strauss). With the advent of effective therapy consisting of high-dose glucocorticoids combined with other immunosuppressive medications, the clinical course of AAV has changed from nearly universal early mortality to a chronic condition with a high level of morbidity and substantial economic burden. Consequently, outcomes for measuring the impact of AAV need to capture the breadth of the impact on physical, social, and emotional aspects of health-related quality of life (HRQOL) and health resource utilization relevant to patients with AAV, in addition to measuring the physiologic consequences of the disease.

Outcome Measures in Rheumatology (OMERACT) is an international organization that strives to develop optimal outcome measures for use in clinical trials through the application of a systematic outcome measure validation process (4). The OMERACT-ICF Reference Group works to integrate the ICF into the OMERACT process (5). ICF classification provides a standardized approach to describing the existing outcome measures (mapping), allowing comparison of the content of different measures (5). Such content comparison could facilitate the selection of tools most appropriate for clinical settings and for conducting clinical research.

In 2010, the OMERACT community endorsed the core set of outcome measures for AAV (6). An OMERACT core set of outcome measures is a collection of validated, responsive feasible instruments that address a set of domains of the illness under study and are considered crucial to assess in clinical studies (6). Selection of such instruments follows a standardized agreed-upon process (4). The AAV core set includes specific measures of disease activity, disease damage, and patient-reported outcomes; these instruments contain a large number of items and, until now, no attempts have been made to understand in detail the exact aspects of functioning and health they measure. The current study addresses this question by mapping the outcome measures included in the OMERACT AAV core set to the ICF classification.

MATERIALS AND METHODS

Outcome measures included in AAV core set

The OMERACT AAV core set suggests using one of the versions of the Birmingham Vasculitis Activity Score (BVAS) (7–10) as a measure of disease activity, the Vasculitis Damage Index (VDI) (11) for assessment of disease-related damage, the Short Form 36 health survey (SF-36) (12) for measurement of life impact, and death as a measure of mortality.

The 3 versions of the BVAS that are most commonly used in clinical trials are the BVAS, version 2 (BVASv2), the BVAS for Wegener’s Granulomatosis (BVAS/WG), and BVAS, version 3 (BVASv3). The BVASv2 (10) is the version that emerged and became incorporated into clinical trials soon after creation of the original BVAS (7); it is frequently referred to as simply “BVAS.” The BVASv2 (10) was designed for use with all forms of vasculitis; it consists of 66 individual items addressing 9 organ systems. The BVAS/WG is a more concise version of BVAS (34 items) specifically designed for use in patients with GPA (9). The BVASv3 is another version created through item reduction (56 items) and simplification of the original BVAS (8). These 3 versions of the BVAS have been used in multicenter, randomized trials of AAV and produce highly correlated summary scores (13).

The VDI consists of 64 items of damage organized by 12 organ systems commonly affected by vasculitides (11). The VDI has been used in almost all trials in AAV.

The SF-36 is a widely used generic HRQOL instrument. It contains 36 items that address 8 domains of quality of life (14) that can be further grouped into physical or mental component summary scores.

Death is the final constituent of the OMERACT AAV core set. When death is recorded, it is usually accompanied by the description of its cause, such as being attributable to the vasculitis, to its treatments, or unrelated to either.

Linking AAV core set items to ICF

Three versions of the BVAS (BVASv2, BVASv3, and BVAS/WG) and the VDI were linked to the ICF according to the previously established ICF linkage rules (15,16). Briefly, this involves first identifying constructs contained in each item of the composite instrument, followed by linkage of each construct to the most precise ICF category. When a construct cannot be linked to any ICF category, it is classified as either not definable (if it is covered by the ICF but cannot be classified to a specific category) or not covered (if it is not covered by the ICF). Constructs that refer to specific medical conditions were linked to the ICF category within the body functions or body structures that represent the function or structure affected by the condition. Medical conditions that imply a specific pathophysiology were additionally classified under the ICF health condition category in order to avoid losing the specific pathophysiology-related information. For example, deafness was linked to the ICF body function category b230, i.e., hearing functions, while diabetes mellitus was linked to the ICF body function category b5401, i.e., carbohydrate metabolism, as well as to the category “hc,” i.e., health condition (diabetes mellitus). The SF-36 has been previously linked to the ICF by Cieza et al (15), and this version of linking was used for this study. Linking was performed independently by 2 investigators (NM and AB) familiar with the ICF. Any disagreements were discussed and the linking was modified until both investigators were satisfied with the final result for each of the linked outcome measure tools. To increase efficiency of the linking process, the tools were linked sequentially, so that an agreement on the previous tool was reached by the 2 investigators before they proceeded to link the next tool.

The degree of agreement between the 2 investigators on linking of the composite tools was estimated by calculating the proportion of agreement in ICF second-level categories for all of the constructs contained in the composite tool, as was suggested in previous reliability studies (17).

RESULTS

The summary of the results of linking of the instruments contained in the OMERACT core set for AAV, grouped by the ICF component, can be found in Table 1 (body functions), Table 2 (body structures), Table 3 (activities and participation), and Table 4 (constructs that were linked to health condition, not covered, and not defined categories). None of the constructs contained in instruments of the AAV core set linked to the ICF component contextual factors. The BVASv2 contains 66 items, in which 84 constructs were identified and linked to 37 second-level ICF categories (17 in body functions and 20 in body structures). The BVAS/WG contains 34 items, in which 47 constructs were identified and linked to 29 second-level ICF categories (9 body functions and 20 body structures). The BVASv3 contains 56 items, in which 71 linkable constructs were identified and linked to 40 second-level ICF categories (20 in body functions and 20 in body structures). The VDI contains 64 items in which 76 constructs were identified and linked to 35 second-level ICF categories (17 body functions and 18 body structures).

Table 1.

Coverage by the OMERACT AAV core set of ICF component “body functions” (mapped to second-level categories)^*

ICF code	ICF category description	BVAS	BVASv3	BVAS/WG	VDI	SF-36
B1	Mental functions				1
14	Orientation functions	1	1
17	Intellectual functions	1	1		1
30	Energy and drive functions					3
52	Emotional functions					9
B2	Sensory functions and pain	2
10	Seeing functions		2		4
30	Hearing functions	3	2	2	1
79	Additional sensory functions, other specified and unspecified	1	1	1	1
80	Sensation of pain	6	4	1	1	2
89	Pain in body part, other specified	1	1
B3	Voice and speech functions	1
10	Voice functions	1
B4	Functions of cardiovascular/hematologic/immunologic/respiratory systems	1
10	Heart functions	6	4		3
15	Blood vessel functions		1		2
20	Blood pressure functions	1	1		1
30	Hematologic system functions				1
35	Immunologic system functions	1	1	1
40	Respiratory functions	1	1	1	3
49	Functions of the respiratory system, other specified and unspecified		1	1
50	Additional respiratory functions	3	1
60	Sensations related to cardiovascular and respiratory functions	3	1		1
B5	Functions of the digestive, metabolic, and endocrine systems	1
25	Defecation functions	1	1
30	Weight maintenance functions	1	1
40	General metabolic functions				1
50	Thermoregulatory functions	2	1	1
B6	Genitourinary and reproductive functions
10	Urinary excretory functions	6	6	3	3
79	Genital/reproductive functions, specified and unspecified				1
B7	Neuromusculoskeletal and movement-related functions
29	Functions of joints and bones, other specified and unspecified				1
30	Muscle power functions				1
80	Sensations related to muscles and movement functions				1
98	Neuromusculoskeletal and movement-related functions, other specified	1	1	1

^*Values are the number of items in the instrument that mapped to the corresponding International Classification of Functioning, Disability and Health (ICF) category. Categories from the third and fourth levels were carried up to the corresponding second-level categories.

OMERACT = Outcome Measures in Rheumatology; AAV = antineutrophil cytoplasmic antibody–associated vasculitis; BVAS = Birmingham Vasculitis Activity Score; v3 = version 3; BVAS/WG = Birmingham Vasculitis Activity Score for granulomatosis with polyangiitis (Wegener’s); VDI = Vasculitis Damage Index; SF-36 = Short Form 36 health survey.

Table 2.

Coverage by the OMERACT AAV core set of ICF component “body structures” (mapped to second-level categories)^*

ICF code	ICF category description	BVAS	BVASv3	BVAS/WG	VDI	SF-36
S1	Structures of the nervous system
10	Structure of brain	4	4	4	3
20	Spinal cord and related structures	1	1	1	1
98	Structure of the nervous system, other specified	2	3	3	1
S2	Eye, ear and related structures	2
10	Structure of eye socket	1	1	1	1
20	Structure of eyeball	7	10	7	3
30	Structures around eye		2	1
40	Structure of external ear	1	1	1
50	Structure of middle ear	1	1	1
60	Structure of inner ear	1	1	1
S3	Structures involved in voice and speech	1
10	Structure of nose	3	4	3	5
20	Structure of mouth	1	1	1	1
30	Structure of pharynx	1	1	1	2
S4	Structures of cardiovascular/immunologic/respiratory systems
10	Structure of cardiovascular system	11	7	1	6
100	Heart	4	2		5
101–3	Arteries, veins, capillaries	6	4		1
30	Structure of respiratory system	8	7	6	3
S5	Structures related to digestive/metabolic/endocrine systems	1
10	Structure of salivary glands			1
20	Structure of esophagus				1
40	Structure of intestine	2		1	3
50	Structure of pancreas	1			1
98	Structures related to digestive/metabolic/endocrine systems, other specified		1		1
S6	Structures related to genitourinary/reproductive systems
10	Structure of urinary system				1
30	Structure of reproductive system	1	1
S7	Structures related to movement
30	Structure of upper extremity	2	1	1
50	Structure of lower extremity	2	1	1
60	Structure of trunk				1
70	Additional musculoskeletal structures related to movement	1	1	1	5
S8	Skin and related structures
10	Structure of areas of skin	5	5	2	1
40	Structure of hair				1

^*Values are the number of items in the instrument that mapped to the corresponding International Classification of Functioning, Disability and Health (ICF) category. Categories from the third and fourth levels were carried up to the corresponding second-level categories.

OMERACT = Outcome Measures in Rheumatology; AAV = antineutrophil cytoplasmic antibody–associated vasculitis; BVAS = Birmingham Vasculitis Activity Score; v3 = version 3; BVAS/WG = Birmingham Vasculitis Activity Score for granulomatosis with polyangiitis (Wegener’s); VDI = Vasculitis Damage Index; SF-36 = Short Form 36 health survey.

Table 3.

Coverage by the OMERACT AAV core set of ICF component “activities and participation” (mapped to second-level categories)^*

ICF code	ICF category description	BVAS	BVASv3	BVAS/WG	VDI	SF-36
D2	General tasks and demands
30	Carrying out daily routine					2
D4	Mobility
10	Changing basic body position					3
30	Lifting and carrying objects					2
45	Hand and arm use					2
50	Walking					3
55	Moving around					3
D5	Self-care
10	Washing oneself					1
40	Dressing					1
D6	Domestic life
49	Household tasks, other specified and unspecified					1
D8	Major life areas
50	Remunerative employment
59	Work and employment, other specified and unspecified					3
D9	Community, social and civic life
20	Recreation and leisure					5

^*Values are the number of items in the instrument that mapped to the corresponding International Classification of Functioning, Disability and Health (ICF) category. Categories from the third and fourth levels were carried up to the corresponding second-level categories.

OMERACT = Outcome Measures in Rheumatology; AAV = antineutrophil cytoplasmic antibody–associated vasculitis; BVAS = Birmingham Vasculitis Activity Score; v3 = version 3; BVAS/WG = Birmingham Vasculitis Activity Score for granulomatosis with polyangiitis (Wegener’s); VDI = Vasculitis Damage Index; SF-36 = Short Form 36 health survey.

Table 4.

Concepts from the AAV core set that could not be linked to specific ICF categories^*

ICF code/category description	BVAS	BVASv3	BVAS/WG	VDI	SF-36	Death
HC (health condition)	0	0	0	4: osteoporosis, asthma, diabetes mellitus, malignancy
NC-BF (not covered [body functions])	3: sinus involvement, brain function, and cranial nerve palsy	3: sinus involvement, brain function, and cranial nerve palsy	2: sinus involvement, cranial nerve palsy	1: brain function
NC-BS (not covered [body structures])	2: sinus involvement (x2)	1: sinus involvement	1: sinus involvement	2: sinus involvement (x2)
ND (not definable)	3: persistent disease, new/worse disease	1: persistent disease	6: severe/limited disease, severe/limited flare, persistent disease remission	1: complication	12: general health, physical health, other	1
ND-BF (not definable [body functions])	1: malaise			1: malignancy
ND-BS (not definable [body structures])				4: malignancy minor/major (x2) tissue loss

^*Values are the number of items in the instrument that mapped to the corresponding International Classification of Functioning, Disability and Health (ICF) category. Categories from the third and fourth levels were carried up to the corresponding second-level categories.

AAV = antineutrophil cytoplasmic antibody–associated vasculitis; BVAS = Birmingham Vasculitis Activity Score; v3 = version 3; BVAS/WG = Birmingham Vasculitis Activity Score for granulomatosis with polyangiitis (Wegener’s); VDI = Vasculitis Damage Index; SF-36 = Short Form 36 health survey.

Four concepts from the VDI were linked to the health condition category, in addition to being linked to the closest category in the body functions or body structures component: osteoporosis, asthma, diabetes mellitus, and malignancy.

Several concepts in each of the instruments could not be linked to a specific ICF category and were considered as not covered (5 in BVASv2, 4 in BVAS/WG, 3 in BVASv3, and 3 in VDI) or as not definable (4 in BVASv2, 6 in BVAS/WG, 1 in BVASv3, and 6 in VDI). The not covered items included constructs related to body functions (sinus dysfunction, brain and cranial nerve dysfunction) and body structures (sinus damage); the not definable constructs are death and various disease states (limited, severe, new or worse, persistent disease, flare, and remission), malaise, malignancy, and tissue loss (Table 4).

The SF-36 (previously linked to the ICF by Cieza et al [15]) contains 11 questions that contain 36 items and 52 linkable constructs. These constructs linked to 14 second-level ICF categories, 3 in the body functions component and 11 in the activities and participation component. Twelve constructs were assigned to the not definable category; these were general health, physical health, and levels of intensity of activity (vigorous, moderate, etc.).

Overall, the OMERACT core set for AAV covers 29 second-level categories within ICF body functions component, 27 in body structures, 11 in activities and participation, and none in environmental factors; 4 constructs were additionally linked to the health condition category, 29 constructs were considered as not definable, and 16 constructs were not covered by the ICF classification.

The proportion of agreement between the 2 investigators (NM and AB) who performed the linking in ICF second-level categories for the first of the linked tools, the BVASv2, was 83%. Because the BVAS/WG and BVASv3 are very similar to the BVASv2, and their linking was performed after the consensus on the final linking of BVASv2 was reached by the 2 investigators, the proportion of agreement on BVAS/WG and BVASv3 was extremely high (98% and 99%, respectively). The proportion of agreement for VDI was 83%.

DISCUSSION

Since its endorsement in 2001 (1), the ICF has steadily become incorporated into the medical field as an interface to describe and compare coverage of different health domains by existing outcome measure tools, including disease-specific (osteoarthritis [17] and ankylosing spondylitis [AS] [18]) instruments and generic (19) tools.

The ICF classification was found to be useful for gaining insight into the exact content of the OMERACT core set of outcome measures for AAV. Overall, the AAV core set covers 29 second-level categories within the ICF body functions component, 27 second-level categories in body structures, only 11 second-level categories in activities or participation, and none in environmental factors. In addition, the AAV core set contains a number of concepts that were linked to the health condition, not specified, or not definable categories.

OMERACT recently developed a framework (OMERACT Filter 2.0 framework) and guidelines to improve selection of domains that should be considered when assessing outcomes in clinical studies (20,21). This framework recommends that every outcome assessment should address each of the 2 broad categories of outcomes: pathophysiologic manifestations and impact of health condition. The latter broad category consists of 3 areas: death, life impact, and resource use/economic impact. The first 2 areas are to be measured in every clinical trial, while measurement of the third area is encouraged where appropriate. In addition, adverse events (toxicity or harm) are required to be addressed separately for items not already covered in death and life impact. The selection of specific domains in each area can depend upon the setting that should be made explicit at the start of the study. Identification of contextual factors that can be confounders or mediators when interpreting the outcome is encouraged. An OMERACT core set of outcome measure tools for a specific medical condition (such as the OMERACT core set for AAV discussed in this manuscript) is meant to contain at least 1 validated instrument to address each of the relevant domains, as well as the minimal set of relevant contextual factors. When an instrument that sufficiently addresses an important domain is lacking, it will need to be developed. Researchers conducting clinical trials will then be able to choose among the instruments contained in the OMER-ACT core set for the specific medical condition in order to address all of the domains relevant to the purpose of their study.

Within the new framework, the ICF component activities and participation represents mainly the OMERACT area of life impact, the components of body functions and body structures represent pathophysiologic manifestations, and the ICF contextual factors could serve as a starting point to define the “context” that might influence the outcomes.

Therefore, the 3 measures of disease activity (BVASv2, BVAS/WG, and BVASv3) and the measure of damage (VDI) (6) represent the area of pathophysiologic manifestations in the new OMERACT framework. This study confirmed that all constructs of the items from each of these instruments linked to the ICF components body functions and body structures (Tables 1 and 2). Comprehensive coverage of these components is demonstrated in this study, further validating the different versions of the BVAS and VDI as tools for measuring disease activity and disease-related damage in patients with AAV. Life impact, on the other hand, is assessed in the AAV core set by the SF-36, a generic measure of HRQOL. The majority of its items are indeed linked to the ICF component activities and participation. Being a generic instrument, the SF-36 is likely limited in its ability to address the various activity limitations and restrictions of participation that are specifically relevant to patients with a complex medical condition like AAV, resulting in suboptimal coverage of life impact by the SF-36 specifically and by the AAV core set overall. In order to improve the ability to assess life impact for patients with AAV, measures that take into account patients’ views need to be developed and included into the OMERACT AAV core set. The OMERACT Vasculitis Working Group is currently working to develop and validate both generic and disease-specific patient-reported outcome tools (22).

The SF-36 can be used for economic evaluation through its derivative form, the SF-6D (6 dimensions) (23). Based on extrapolation of the preference values on the responses to the different items of the SF-36 obtained from the general population, a single index measure can be estimated from any completed SF-36 questionnaire that allows for the calculation of quality-adjusted life years for use in cost-utility analysis (23). Although this application of the SF-36 has not, to date, been employed specifically in the setting of vasculitis, the generic nature of the instrument and its SF-6D adaptation implies that this approach should be useful for economic evaluations of treatments or disease impact of AAV.

In contrast, the OMERACT context (or “contextual factors” in the ICF language) is not addressed by the AAV core set to any extent. The concept of “contextual factors” has only recently entered the medical field, after the AAV core set was developed. Generally, the contextual modifiers are mainly relevant to various aspects of impact of a medical condition, including life impact, economic impact, and potentially death; contextual modifiers are unlikely to have a significant effect on pathophysiologic manifestations. Most interventional trials in AAV that make use of the instruments included in the AAV core set focus on specific physiologic outcomes, such as renal function or the rate of complete remission; for these trials, personal and environmental contextual factors are, indeed, largely irrelevant. As acceptance of the OMERACT filter 2.0 framework (21) increases, the range of outcomes measured in clinical trials will likely expand and include more detailed and consistent assessment of life impact; then contextual factors will become more relevant and their incorporation into the design and analysis of clinical trials will become necessary.

It is striking that the area life impact seems to be under-represented compared to pathophysiologic manifestations in almost all OMERACT core sets. While the core sets of outcome domains for psoriatic arthritis, rheumatoid arthritis (RA), and AS comprise between 3 (psoriatic arthritis) and 8 (AS) domains that cover pathophysiologic manifestations (24), the core sets for RA and AS contain only 1 domain (physical function) from the area life impact. The OMERACT core set for psoriatic arthritis also contains the domains of participation and HRQOL, resulting in broader coverage of life impact.

The ICF offers valuable opportunities as a framework to better define the outcomes to measure that are relevant to patients and society, as well as to health professionals. Disease-specific ICF core sets represent selections of ICF categories that are typical and relevant to patients with individual medical conditions; they are analogous to the OMERACT core sets of domains, with the ICF categories replacing the description of domains. We scrutinized several of the existing ICF core sets of rheumatologic conditions that have been developed to date. The ratio of the number of ICF categories, from the ICF component body structures, to body functions, to activities and participation, to environmental factors is 1:3.8:5.8:5.0 for low back pain, 1:2.2:3.2:2.8 for osteoarthritis, 1:1.9:3.0:3.7 for osteoporosis, and 1:1.9:4.0:2.6 for RA (25) for the ICF core sets for the specified conditions, compared to the corresponding ratio of 2.6:2.5:1:0 for the non–ICF-based OMERACT AAV core set reported in this study. Such broader coverage of life impact domains by the ICF core sets relative to the existing non–ICF-based OMERACT core sets is likely a function of both the use of the bio/psycho/social framework of health (which is the basis of the ICF) and of the more structured approach to ICF-based core set development that prioritizes input of all end-users, including patients. As the similarly structured OMERACT Filter 2.0 framework gradually becomes the standard approach to development of the new outcome measures, both newly developed and revisions to previously developed, the disease-specific OMERACT core sets should display broader coverage of life impact and the context.

Despite multiple advantages of using the ICF, the classification has limitations. First, although the established linking rules maximize precision of the linking process, some subjectivity to the process is inevitable. The agreement between the 2 investigators who did the linking in this study was between 83% and 99%. The agreement in other ICF-related studies is closer to the lower end of the agreement in this study (17), likely because the current study used a sequential approach to linking of highly similar instruments, with achievement of full agreement on previous instruments before proceeding to the subsequent ones. Disagreements occurred mainly in regards to concepts for which no perfect fit to a specific ICF category could be found. Second, despite its comprehensiveness, the ICF classification is not perfectly exhaustive. Some concepts contained in the AAV core set instruments could not be linked to a specific ICF category and as a consequence had to be labeled as not definable or not covered (see Table 4). While some of such concepts are likely too broad to be linkable to a specific ICF category (for example “malaise” and “physical health”), others such as structure and function of paranasal sinuses, and function of brain and peripheral nerves, are fairly specific, and as such their absence from the ICF classification likely represents a mere oversight at the time of its development. Such limitations are recognized and the ICF will likely be updated in the near future. Until such updates are available, the ICF classification will need to be supplemented with existing outcome measures and composite outcome measurement tools when studying and describing complex systemic conditions such as AAV. Third, the ICF does not consider the concept of “time” nor the inherently time-dependent concept of “process,” which is the basis of distinction between disease activity and damage. As a result, Tables 1 and 2 demonstrate significant overlap of coverage of various ICF categories between the measures of disease activity and the measure of damage. Conceptually, disease activity measures focus on changes in body functions within specific body structures (and therefore the constructs link to both components), while the measures of damage focus on permanent disturbances of both functions and structures of the same organs. While clinically this distinction can (and should) be made, although with some difficulty (26), in the ICF it is not possible to indicate whether an impairment of a structure is temporary or permanent, resulting in such overlap.

This study demonstrates that the ICF classification is a useful framework for analyzing a complex medical condition from the global bio/psycho/social perspective. Linking the content of the OMERACT core set of measures for AAV revealed a detailed coverage of body functions and body structures (OMERACT pathophysiologic manifestations), but suggested important gaps in the coverage of the OMERACT area of life impact and complete lack of coverage of contextual factors. These results suggest the need for expanding the range of the domains that need to be addressed when assessing patients with AAV. It has been increasingly recognized that identifying the relevant domains necessitates directly engaging patients in the process of outcome measure development (27). Patients have become an indispensable part of nearly every part of the OMERACT process, among which is the work of the OMERACT Vasculitis Working Group to develop a vasculitis-specific, patient-reported outcome measure (22).