In general, the social insurance institutions in Europe have well-developed statistics about benefits, payments, and quality. With the strengthening of a more functional approach to benefits, there will probably be a need for statistics looking at the functional abilities of claimants in the future, not only of their medical diagnoses. Such accumulation of statistical information necessitates a coding system for functional status, and that could be a potential use for the ICF.

To our knowledge, no such coding of functional abilities has occurred yet. However, in the USA, the Washington Group on Disability Statistics has developed a short set of questions that internationally can be used for census purposes. The idea is that the questions can be linked to statistics on school attendance or employment. The purpose is to describe how disability in various countries can hamper participation in important life arenas. The questions are based on the ICF, and one question recognizes its basic model [35].

The development of the core set was carried out in a formal decision-making process. In a first step, the members of the working group organized eight national expert meetings to discuss and suggest relevant ICF categories. The working groups could suggest ICF categories but also other functional categories not included in the ICF.

In a second step, the core set was decided upon by voting in two rounds among 20 members of the EUMASS working group on the ICF. All participants were leading experts in social security, vocational rehabilitation, and disability evaluation. The definition and use of qualifiers was not a subject in this primary stage of the core set. The first voting round was finished during a 1-day meeting. A second voting round was completed by email.

The EUMASS Core Set is generic and contains categories that always, and irrespective of diagnosis, should be assessed and reported about by physicians in social insurance medicine. It is intended for the evaluation of rights to benefits or eligibility determination. Insurance physicians might also do assessments for sickness absence, rehabilitation, and return to work. When this is the case, the person’s resources and possibilities must be addressed comprehensively.

The core set contains only 20 categories, a smaller number than the brief core sets for chronic conditions [36]. Important categories might be missing for some countries, and national administrations might add categories according to national standards and legislation if they use the EUMASS Core Set. This has occurred in Iceland and Sweden. A low number of categories will probably increase the feasibility of the core set. In the core set, there is a fair balance between categories pertaining to physical and mental functioning.

In the selection process, body structures categories were replaced by corresponding body functions categories. This probably reflects the fact that evaluation of incapacity for work in European social insurance is related to functional disability. Body structures would probably have been preferred in an impairment-oriented assessment system, such as the Guides to the Evaluation of Permanent Impairment of the American Medical Association [37].

The majority of the selected categories pertain to the activities and participation component. They include sensory functions, basic mental functions, mobility functions, and more complex functions such as d720 complex interpersonal interactions. The selection reflects the large variations in demands for functional abilities in working life and also the complexity of disability evaluation.

Contrary to expectation (see definition in Box 6.1), no environmental factor was included in the core set. To medical experts in social security, work environment evidently is a necessary element in the disability evaluation process. There might be several explanations to this paradox. There is a large difference between European countries’ procedures in how they address work issues. In addition, the environmental factors in the ICF are probably not specific enough for a useful classification of working situations. It is also possible that disability evaluation in most countries is essentially medically defined with an emphasis on the medical consequences of illness and accidents. Social insurance physicians traditionally may accentuate the medical aspects of work incapacity. Moreover, participation categories also include the impact of environmental factors on the performance of the claimant [38]. It is possible that medical advisors integrate environmental factors in a more implicit, rather than explicit, way, when they assess work ability.

Although qualifiers were not included at this primary stage of the core set, their importance was recognized. Since there were no empirical data as to which level of functioning is making work participation impossible, it was stated that the level of functioning has to be related to common work demands within the framework of the national social insurance legislation of a country. However, an attempt to use qualifiers was done in a validation study.

To investigate the strength and weakness of the EUMASS Core Set, an international study of content validation was initiated [3]. The aim was to establish if the EUMASS Core Set captures the functional abilities of claimants for disability benefits, irrespective of the underlying medical condition and national social security system. The objective was to explore if the ICF categories of the core set were relevant, useful, and sufficient to express functional ability in claimants applying for long-term disability benefits in social insurance.

The validity was tested in an exploratory, cross-sectional multicenter study. Two different instruments (one for person-encounter disability evaluation and one for paper-file disability evaluation) were constructed by adding qualifiers to each of the 20 categories of the EUMASS Core Set. The qualifiers and definitions were taken from the ICF Checklist [39]. They ranged from no impairment/no limitation to complete impairment/limitation on a 5-item ordinal scale. The participating social insurance physicians could document categories that were missing, but only when they were necessary to express the functional ability of a particular claimant. They were also asked to what extent the core set was useful and sufficient to describe a claimant’s functional capacities on a 5-item ordinal scale ranging from “totally agree” to “totally disagree.”

EUMASS members of the participating countries translated the validation forms into their own languages using the ICF. They recruited 5–10 social insurance physicians that were experienced in evaluating claims for long-term work disability. Each physician applied the validation form to a maximum of ten consecutive claimants for disability evaluation.

In total, 48 physicians from six countries evaluated 446 claimants for long-term work disability benefits. The physicians from Belgium, France, Iceland, and Romania performed the evaluation in personal encounters. In Norway, the physicians exclusively did file-based disability evaluations, while Germany did both. The majority of claimants were women, and many were in unskilled occupations such as cleaning work. Musculoskeletal and mental disorders were the most frequently reported health conditions. These characteristics indicated that the sample was fairly representative of claimants for disability benefits in Europe.

The physicians used all 20 ICF categories of the core set to describe functional ability in their reports. The most frequently listed limitations of claimants were sensation of pain (66 %), lifting and carrying objects (64 %), and handling stress and other psychological demands (63 %). Subgroup analysis showed that the physicians of all six countries applied each of the 20 ICF categories of the core set at least once to express limitations, and each of the 20 ICF categories was represented at least once as a limitation in the main disease groups.

The physicians mentioned 42 different categories as missing to describe the claimants’ functional abilities: 27 categories of body functions (17 of which were mental), 11 of activities and participation, 3 of environmental factors, and 1 of personal factors. Sixteen categories were mentioned more than once. Categories were missed most frequently for claimants with mental disorders.

In 68 % of the cases, the physicians rated the core set as useful to express functional ability (27 % “totally agree,” 41 % “partly agree”). Physicians from Norway and Belgium found the core set significantly less useful compared to physicians from Germany, Romania, France, and Iceland. No difference in usefulness was found among the main disease groups.

The physicians also rated the core set as sufficient to express functional ability in 63 % of the cases (23 % “totally agree,” 40 % “partly agree”). The physicians from Norway, Belgium, and Romania perceived the core set as significantly less sufficient, compared to physicians from Germany, France, and Iceland. They were significantly less satisfied with the sufficiency of the core set when they rated persons with circulatory disorders such chronic heart failure.

The validation study showed that the core set includes relevant categories, but it is not completely comprehensive since 42 additional categories were suggested. The physicians perceived the core set as useful and sufficient to express functional disability in the context of working life, but the judgments varied among countries and diagnostic groups. The findings suggest that the EUMASS Core Set can provide support in evaluating long-term work disability. The study demonstrated that it is possible to use an instrument to evaluate functional ability despite different national or local processes in disability evaluation. Such an instrument could promote transparency, reliability, homogenous presentation in practice, and data exchange in research. With adding more categories of mental functions such as temperament and personality functions or psychic stability, the EUMASS Core Set could also be more comprehensive.

The EUMASS Core Set has been used and studied in other countries, notably in Germany, Sweden, and Iceland. In a German study, Timner found that insurance physicians frequently rated limitations in the category “handling stress and other psychological demands” [4]. The category “pain” was used less frequently compared with the validation study [3]. In Sweden, insurance physicians have tested an instrument with 18 categories based on the EUMASS Core Set in long-term work disability claimants. The physicians also evaluated the degree of the limitation, if the recorded limitations are a consequence of disease, e.g., reduced physical endurance caused by heart failure, and if they are based on observed findings, e.g., spirometric testing [22]. Preliminary results of the Swedish instrument testing show that “handling stress and other psychological demands” was the category most frequently reported as a limitation and as a consequence of disease [40]. In a study in Iceland, the insurance physicians found that the core set did not capture mental illness well enough [30].

The current evidence calls it necessary to conduct more validation studies. They should include more European countries, particularly the Netherlands and the UK, where there is a tradition of reporting functional ability systematically. Moreover, studies should report findings according to disease groups to look at disease-specific burden and more information of where in the process of disability evaluation medical examiners specify functional disability. A larger sample would allow sensitivity analyses to explore the best cutoff for relevant thresholds in activity limitations.

Another important consideration is to also explore if social insurance officers handling disability pension claims are able to integrate information from the core set in the decision-making process. It should also be investigated if claimants find the core set appropriate to express their functional disability. It is also a frequent observation in countries which use instruments in work disability evaluation, such as the UK, the Netherlands, Sweden, and Iceland, that detailed instructions are needed to support the use of the instruments [12].

The ICF was translated to Swedish and published in 2003 by the National Board of Health and Safety. The introduction was well timed with respect for the need of reforms in Sweden, since the number of days paid for sickness allowance had doubled in the preceding 5-year period. Research showed that half of the increase was caused by an increased number of claims and the other half by prolonged periods of disability [41].