Osteoporosis (OP) is a chronic disease with a high burden of illness characterized by low bone mass and microarchitectural deterioration in bone tissue. These bone changes lead to enhanced bone fragility and increased fracture risk.1,2 The number of hip fractures related to OP is increasing by 1% to 3% per year in most world regions.3–6 In general, OP and bone fractures can result in limitations of physical activities and restrictions in daily activities and societal participation.7–9
Physical therapy is a frequent component of prevention services, acute and post–acute health care, and rehabilitation of individuals with OP.10–19 Physical therapists conduct both standard physical assessment and OP-specific assessments to evaluate spinal mobility, strength and muscle endurance, aerobic capacity of the cardiovascular system, balance, functional ability, and pain.20–29 The interventions used by physical therapists for the management of OP primarily focus on patient education, pain management, reducing fall risk, remodeling bone tissue, and reducing fear of falling.30–36 Recently, a number of systematic reviews and guidelines have summarized physical therapy assessments and interventions related to OP.37,38
To optimize interventions aimed at maintaining function and minimizing disability, the physical therapist should have an in-depth understanding of the patients' functioning and health status as well as a conceptual framework to guide communication among health professionals, clinical research, and patient care. Addressing this need, the World Health Organization (WHO) has introduced the International Classification of Impairments, Disabilities, and Handicaps (ICIDH) in 1980.39 The ICIDH has stimulated discussions of disability concepts and has received both positive and negative reviews in the literature.40,41 Taking into account the criticisms of existing disablement frameworks, such as the ICIDH or the Nagi model,42,43 the WHO released a major revision of the ICIDH in 2001, the International Classification of Functioning, Disability and Health (ICF). The ICF is based on the integrative biopsychosocial model of functioning and disability and provides a comprehensive understanding of the patients' functioning and health.44
According to the ICF's biopsychosocial model of functioning and disability, an individual's functioning is a complex interaction among his or her health conditions and the personal and environmental contextual factors influencing Body Functions and Body Structures and Activities and Participation (Figure 1). There is a dynamic interaction among these entities. Functioning denotes the positive aspects, whereas disability denotes the negative aspects of the interaction between an individual with a health condition and the contextual factors (environmental and personal factors).
The structure of the ICF model of functioning and disability uses a classification system made up of increasingly more specific descriptors-labeled parts, components, chapters (level 1), and categories (levels 2, 3, and 4) (see Figure 2). Body Functions are the physiological functions of body systems including psychological functions. Body Structures are the anatomical parts of the body such as organs, limbs, and their components. Loss or abnormalities in body structure or physical function is called “impairment.” The component Activities and Participation describes domains of functioning from both an individual and societal perspective. Activity limitations are difficulties an individual may have in executing activities, whereas participation restrictions are problems an individual may experience in involvement in life situation. In contrast to other disability models, the ICF classifies Contextual Factors that may either facilitate or hinder functioning and, therefore, influence disability. The 2 components of Contextual Factors are (1) Environmental Factors, which include factors in the physical, social, or attitudinal world and (2) Personal Factors, which include concepts such as gender, age, habits, and coping style.44
Within each component, except personal factors, there is an exhaustive list of ICF categories, which are the units of the classification. ICF categories are coded by the component letter and a suffix of 1 to 5 digits. The letters b, s, d, and e refer to the components Body functions (b), Body structures (s), Activities and Participation (d), and Environmental factors (e) (see Figure 2). This letter is followed by a 1-digit number indicating the chapter, followed by a 2-digit code for the second level, a 1-digit code for the third level, and a 1-digit code for the fourth level. Within each component, the categories are arranged in a stem/branch/leaf scheme. Consequently, a higher-level (more detailed) category shares the lower-level categories of which it is a member, in addition to an increased level of specificity. That means the use of a higher-level category implies that the lower-level category is applicable, but not vice versa.44 To give an example, the third-level ICF category b7303 Power of muscles in lower half of the body is one element of the second-level category b730 Muscle power functions with the specificity of in lower half of the body indicated by the last digit “3.” The second-level category, b730 Muscle power functions, is an element of chapter b7 “Neuromusculoskeletal and Movement-Related Functions” with the specificity of muscle power functions indicated by the digits “30.” Finally, chapter b7 “Neuromusculoskeletal and Movement-Related Functions” is part of the ICF component b Body Functions.
The content and the structure of the ICF both have potential value for the health professions involved in OP care.45 The ICF as a whole is composed of more than 1400 categories, making it a highly comprehensive classification. This comprehensiveness is a major strength of the ICF on one hand, but also the major challenge to its feasibility on the other hand. To facilitate the implementation of the ICF into clinical practice, “ICF Core Sets” were developed for a number of health conditions including OP. The development of the ICF Core Sets followed a standard approach that included a formal decision-making and consensus process, integrating evidence gathered from preparatory studies by appropriate experts in the field.46 Preparatory studies included a worldwide Delphi study with 21 physicians and physical therapists, a systematic review on outcome measures used in 107 OP clinical trials, and an empiric data collection on 32 German patients with OP.47 Based on the results of these studies, a panel of 15 OP experts (10 physicians, 5 physical therapists) from 7 different countries decided on the composition of the “Comprehensive ICF Core Set for OP” in a formal consensus process. The Comprehensive ICF Core Set for OP, which emerged from this evidence-based decision process, includes 72 ICF categories covering a spectrum of problems of functioning in patients with OP.
On the basis of the Comprehensive ICF Core Set for OP, the impairments in Body Functions and Structures, limitations in Activities, restrictions in Participation, and the influential Environmental factors of a patient with OP can be described. A functional profile could then be created and used as a reference for follow-up. Such a profile could also serve as the basis for the assignment of the core competencies within the therapeutic team. The responsibilities for the treatment of a specific problem can be assigned to the team members according to the corresponding ICF categories. Thereby, role overlaps in the therapeutic team, and treatment gaps can be detected.48 Compared with the use of profession-specific assessment and documentation systems, the major advantage of the “ICF Core Set for OP” is its common language that can be understood by all health professions.
Currently, the ICF Core Set for OP is undergoing worldwide testing and validation using a number of approaches. One key aspect is content validation from the perspective of health professionals to determine whether the problems in functioning, which are substantial targets of the specific interventions applied by health professionals, are represented in the Comprehensive ICF Core Set for OP. To give an example, if joint mobility is an essential component of a physical therapy intervention program for individuals with OP, the ICF Core Set for OP should include joint mobility categories; thus reinforcing the necessity of joint mobility assessment and providing a structure for documentation of changes in joint mobility during the episode of care. Consequently, if the corresponding ICF category for joint mobility is not included in the current version of the Comprehensive ICF Core Set for OP, the Core Set is lacking content validity from the perspective of physical therapists.
The objective of this study was to examine the content validity of the Comprehensive ICF Core Set for OP from the perspective of physical therapists. The specific aims were to identify the patients' problems, resources (eg, personal resources like self-esteem, health behavior, or social support), and environmental factors that impact physical therapist management of patients with OP and to examine how these patient problems, resources, and environmental factors are represented by the Comprehensive ICF Core Set for OP.
A 3-round electronic-mail survey of physical therapists was conducted using the Delphi technique.49,50 The principal aim of the Delphi technique is to gain consensus from a panel of individuals with knowledge of the topic being investigated.51 These informed persons are commonly titled “experts.”52 The Delphi technique is a multistage process, where each stage builds on the results of the previous one, and a series of rounds are used both to gather and provide information about a particular subject. The technique is characterized by its anonymity to avoid dominance of single individuals in the group, iteration that allows panel members to change their opinions in subsequent rounds, and controlled feedback showing the distribution of the group's responses, as well as each individual's previous responses.49–55
Recruitment of Participants
In the preparatory phase of the study, physical therapy associations, as well as the International Osteoporosis Foundation and related associations, were identified by World Wide Web Internet search and contacted. In addition, an Internet search in PubMed, MEDLINE, CINAHL, PEDro, AMED, and The Cochrane Library was performed for authors of publications on physical therapy and OP. Moreover, professional recommendations were used to identify individual physical therapists experienced in the treatment of OP. University hospitals and physical therapy schools were contacted as well.
The criterion to be included in this study was participant-confirmed professional experience in the treatment of OP. The initial letter that was sent to the potential participants specified that participants should be “physical therapists experienced in the treatment of OP.” The first contact included an invitation to participate and a detailed description of the project's aims, the Delphi process, and the timeline.
The study was conducted in English language. The process and verbatim questions of the electronic-mail survey using the Delphi technique are presented in Figure 3. The participants were given 3 weeks to submit their responses for each round. Reminders were sent electronically 1 week and 2 days before deadline, and 1 week after deadline.
In the first round of the Delphi procedure, an information letter including instructions and an Excel file containing an open-ended questionnaire were sent to all the participants. In the questionnaire, the participants were requested to list all patients' problems, resources, and aspects of environment, which are treated by physical therapists in patients with OP. In addition, the participants were asked to complete questions on demographic characteristics and professional experience.
The responses received in round 1 were then tied to the ICF by a linking process (see the following). Problems, resources, and environmental factors not represented in the ICF were placed in the Not-covered category. Aspects related to the characteristics of a person were listed as Personal factors.
In the second Delphi round, the participants received the list of the ICF categories linked to the responses of the first round. The instruction letter included a brief introduction into the ICF terminology and a link to an Internet Web site, where further information on the ICF terminology and codes was provided in different languages. The participants were asked whether these ICF categories represent the problems, resources, and environmental factors of patients with OP, who were treated by physical therapists. The percentage of participants who agreed was calculated.
In the third Delphi round, the participants were requested to answer the same questions taking into account the answers of the group, as well as their previous responses.
Each response of the first Delphi round was linked to the most precise of the 1400 ICF categories by 2 physical therapists (B.K. and A.G.), who were trained in the ICF model and the 10 linking rules established in former studies.56 To give an example, from the reported problem “decreased muscle power,” the concept “muscle power” was extracted and then assigned to the ICF category b730 Muscle power functions. Concepts not addressed in the ICF were coded as Not-covered and concepts addressing personal factors according to the ICF definition were coded as Personal factors. First, the responses were linked by the 2 physical therapists independently. The agreement between the 2 persons was quantified using κ statistics. Second, these same physical therapists compared the links made for each specific response, discussed each disagreement, and came to agreement about the most appropriate ICF category to assign each response.
Statistical analysis was performed using SAS for Windows V8 (SAS Institute Inc, Cary, North Carolina). Descriptive statistics were used to characterize the sample and frequencies of answers. K statistics with bootstrapped confidence intervals were used to describe the agreement between the physical therapists who performed the linking.57,58
Recruitment and Participants
A total of 107 physical therapy associations, such as the World Confederation of Physiotherapy (WCPT), the European Confederation of Physiotherapy (ER-WCPT), and the Chartered Society for Physiotherapy (specialists in women's health) were contacted. Seventeen associations named 31 persons for potential participation, and 10 physical therapists agreed to participate. Thirty-three associations for OP, women's health, and geriatrics were invited to inform their members. Six of these associations named 8 physical therapists and 4 agreed to participate. More than 500 University hospitals, rehabilitation centers, and schools for physical therapy were contacted. Twelve of these facilities named 13 physical therapists, 12 of whom agreed to participate. On the basis of personal recommendations of other participants (“snowball system”) and personal contacts, 140 physical therapists were contacted and 13 agreed to participate. A total of 402 authors of publications on physical therapy and OP were contacted. Thirty-four of these agreed to participate. Two additional physical therapists were identified by Internet search, both of whom agreed to participate.
Although the letter of inquiry noted that participants should be “physical therapists with experience in the treatment of patients with OP,” 2 scientists involved in OP research but with low practical knowledge were recommended by another participant. We decided to include them because of their outstanding research expertise in the field of OP.
The Delphi procedure was conducted between June 2007 and September 2007. Fifty-nine of the 75 physical therapists (78.6%) agreed to participate in the study filled in the first round questionnaire; 2 of the questionnaires could not be analyzed because they were not filled in or it was not possible to read the answers contained in the file. The demographic and professional characteristics of the participants are shown in Table 1.
In the first Delphi round, 816 concepts were named and linked to 160 ICF categories. Fifty-two of the 57 participants (91.2%) returned the second-round questionnaire. The third-round questionnaire was completed by 52 of the 52 (100.0%) participants.
Linking of the Concepts to the International Classification of Functioning, Disability and Health
The 160 linked ICF categories covered all ICF components. Thirty second-level categories, 23 third-level categories, and 5 fourth-level categories were linked to the component Body functions. The component Body structures was represented by 6 second-level categories, 7 third-level categories, and 6 fourth-level categories. Twenty-eight second-level categories and 27 third-level categories of the component Activities and participation and 20 second-level categories and 8 third-level categories of the component Environmental factors were linked. Seventeen concepts were linked to the not yet developed ICF component Personal factors. Nine concepts were found not to be covered by the ICF.
The Kappa statistic for the agreement between the physical therapists who performed the linking was 0.42 with a 95% bootstrapped confidence interval of 0.41 to 0.42.
Representation of the Participants' Responses in the Comprehensive ICF Core Set for OP
In total, from the 160 ICF categories linked to the participants' responses, 95 reached an agreement more than 75% in the final round and were considered for comparison with the current Comprehensive ICF Core Set for OP.
Seven of the ICF categories under the component Body Functions were linked to the participants' responses represented in the Comprehensive ICF Core Set for OP at the same level of classification. Seventeen responses were linked to more detailed third-level ICF categories, for example, b2800 Generalized pain, which is represented in the Comprehensive ICF Core Set for OP by the second-level category b280 Sensation of pain (see Table 2, categories marked by “xb”). Thirteen ICF categories from the component Body functions are found not to be represented in the Comprehensive ICF Core Set for OP, 3 covering mental functions, 3 covering sensory functions and pain, and 7 related to neuromusculoskeletal and movement functions (see Table 2, categories in bold face type).
Of the component Body structures, 17 ICF categories could be linked to the participants' responses (see Table 3). Among these, 4 categories are included in the Comprehensive ICF Core Set for OP at the same level of classification, whereas 12 third- and fourth-level categories, for example, s4302 Thoracic cage, are represented by their corresponding second-level categories (see Table 3, categories marked by “xb”). The ICF category “s710 Structure of head and neck region” was the only category with an agreement more than 75%, which is not represented in the Comprehensive ICF Core Set for OP.
Most of the ICF categories linked could be assigned to the ICF component Activities and Participation. Of the 34 categories linked, 10 are included at the same level of the classification and 13 at the more detailed third-level category rather than corresponding second-level classification in the Comprehensive ICF Core Set for OP (see Table 4, categories marked by “xb”). Eleven categories are not represented in the Comprehensive ICF Core Set for OP, for example, d420 Transferring oneself, but reached an agreement more than 75% (see Table 4, bold face type).
Of the component Environmental factors, 10 categories were linked to the participants' responses. All of them are represented in the Comprehensive ICF Core Set for OP, either on the same level or at a different level of the classification (see Table 5, categories marked by “xb”).
Fifteen responses were linked to the not-yet-developed ICF component Personal factors and reached an agreement more than 75% (see Table 6). All of these issues addressed either health conditions, which are often associated with OP, such as fitness level, or factors that describe the patients' way to deal with their condition like independence and knowledge about the disease. Nine responses of the participants were found not to be covered by the ICF (see Table 7).
This study examined the content validity of the Comprehensive ICF Core Set for OP from the perspective of physical therapists. In this study, content validity refers to the extent to which the patients' problems, resources, and environmental factors identified by physical therapists as relevant to their management of patients with OP are represented in the previously published Comprehensive ICF Core Set for OP. Factors with a participant agreement more than 75% in the final Delphi round were considered relevant. Consequently, ICF categories with an agreement more than 75%, which are not represented in the Comprehensive ICF Core Set for OP, may indicate missing factors and will be the main focus of the following discussion.
In general, the factors identified by physical therapists related to the components Body Structures and Environmental Factors were also represented in the Comprehensive ICF Core Set for OP, whereas shortcomings could be identified regarding the factors identified for the components Body functions and Activities and Participation.
The body function categories b1266 Confidence, b130 Energy and drive functions, and b1301 Motivation are global mental functions identified as relevant by physical therapists in this Delphi process but not included in the Comprehensive ICF Core Set for OP. These aspects of mental function may influence patients' compliance with physical therapy interventions. In fact, patients with OP require high levels of attention and motivation when being treated and educated by physical therapists, often over a long period.59 However, it is questionable whether these mental functions actually are treated by physical therapists as requested in our study. We suppose it was sometimes difficult for the participants to differentiate between patient problems that are directly treated by physical therapists and problems that may affect physical therapy.
Several ICF categories with high agreement among the participants (b235 Vestibular functions [88%], b2401 Dizziness [78%], and b260 Proprioceptive functions [94%]) were not represented in the Comprehensive ICF Core Set for OP. There are numerous reports in the literature of the importance of these factors in influencing the risk of falls and fractures of individuals with OP.15,16,21–27,30,60,61
Moreover, neuromusculoskeletal and movement functions such as b735 Muscle tone functions, b760 Control of voluntary movement functions, and b7602 Coordination of voluntary movements are not represented in the Comprehensive ICF Core Set for OP. These functions are relevant for both bone-remodelling and fall-prevention. In physical therapy practice, reduced strength and flexibility of specific muscle groups is frequently seen in patients with OP. The restoration of muscle balance using exercise therapy is a central task of physical therapists.7–13,15,16,20–31,61
Further attention should be directed toward the category b720 Mobility of bone functions, which was identified by the participants in our study with an agreement of 98%. In contrast, b720 Mobility of bone functions was excluded in the consensus conference, which led to the current version of the Comprehensive ICF Core Set for OP. This inconsistency might be related to the lower level of ICF knowledge and training of the participants in this study compared to those involved in the consensus conference. Participants in this study may have focused on the title of the category and concluded that this category covers the elasticity of the cancellous bone, which is indeed reduced in patients with OP.62 However, having a closer look at its higher-level categories, namely mobility of scapula, pelvis, carpal bones and tarsal bones, it becomes clear that the elasticity of the cancellous bone is not covered by b720 Mobility of bone functions. Thus, it could be useful to improve the clarity of this category title in future revisions of the ICF.
From the component Activities and Participation, the second-level categories—“d230 Carrying out daily routine,” “d420 Transferring oneself,” “d435 Moving objects with lower extremities,” “d460 Moving around in different locations,” “d570 Looking after one's health,” and “d920 Recreation and leisure”—or the corresponding higher-level categories were identified by the participants as being treatment goals of physical therapy but are not included in the Comprehensive ICF Core Set for OP. This finding emphasizes the broad spectrum of activities and participation treated by physical therapists in patients with OP and indicates a need to add these ICF categories to the Comprehensive ICF Core Set for OP.
A number of concepts named by the participants were linked to the yet undeveloped ICF component Personal factors. Most of them address attitudes supporting the patients' independence in managing his or her disease and may be influenced by physical therapists using patient education. Various studies and systematic reviews support the positive effects of patient education on self-management, self-efficacy, and coping with disease.7–12,17,20,32,33,37,38 The considerable number of Personal factors mentioned by the participants in this study emphasizes the need to develop this ICF component to provide a comprehensive and complete description of aspects that influence a patient's functioning and health.
A few concepts were retrieved in this study, for example, balance and posture, which are not covered by the ICF. In the consensus conference leading to the Comprehensive ICF Core Set for OP, b755 Involuntary movement reaction functions was extensively discussed and finally accepted to cover balance, balance reactions, and posture. However, in this study, balance and posture were considered “not classified” by the 2 persons who have linked the answers of the participants. They agreed that the definition of this ICF category, namely functions of postural reactions, righting reactions, body adjustment reactions, balance reactions, supporting reactions, defensive reactions, does not precisely enough cover balance and posture, which are commonly used concepts in physical therapists' daily practice.60,61
The Delphi technique turned out to be an appropriate method for this study objective. The response rates were more than 78% in the single rounds, which exceeded the response rates of approximately 50% reported in the literature.63,64
However, some limitations regarding the reliability and the external validity of the study are included. First, the level of agreement between the persons who independently performed the linking assessed by κ coefficient was only moderate and lower compared to other studies that used similar methods.65,66 Because many detailed third- and fourth-level categories were linked and the κ coefficient was calculated for agreement on the most specific level, it cannot be excluded that the agreement was higher on a lower level of precision. Contrary to the moderate initial agreement, the persons were able to reach consensus after having discussed the ICF categories they had initially disagreed on. This demonstrates that the linking procedure is complex and a further refinement of the linking rules could be useful to increase reliability in future studies.
The second limitation refers to external validity. Participants from all 6 world regions defined by the WHO were recruited, representing a wide range of expert opinion. The recruitment of OP physical therapists in nonindustrial countries was difficult. This led to an underrepresentation of the African Region and Eastern Mediterranean Region in our study. This unbalanced representation may reflect a generally lower number of physical therapists or a low priority of this health condition in these world regions, which may also be due to a relatively low prevalence of OP.67,68 Furthermore, it is not clear whether a different population with physical therapists from smaller hospitals and smaller cities would have shown other results. Finally, a selection bias cannot be excluded because only a small number of the persons who were invited finally participated in the study.
In summary, in our study, we found the categories from the ICF components “Body Structures” and “Environmental Factors” included in the current version of the Comprehensive ICF Core Set for OP were largely representing what physical therapists agreed upon as important to their interventions for individuals with OP. However, a number of ICF categories addressing body functions, activities, and participation, which were found to be relevant for physical therapists were not yet included in the Comprehensive ICF Core Set for OP and should be investigated further. Because the Comprehensive ICF Core Set for OP is a tool for multidisciplinary use, it is essential to consider the perspective of other health professions when deciding about a future inclusion of these categories. On the basis of these results, a revised version of the Comprehensive ICF Core Set for OP could be created. In the revision process, it should be decided whether the ICF categories identified by the different health professions as not being represented in the current version should be added or whether these additional profession-specific categories should make up a profession-specific list of ICF categories that may be applied in addition to the Comprehensive ICF Core Set for OP. This decision should also consider data on the feasibility, because the feasibility in clinical practice could be jeopardized by a considerable enlargement of the number of ICF categories included in the Core Set. Furthermore, this study identified some relevant concepts in physical therapy, which are not covered precisely by the ICF and should be considered in future revisions. In addition, the personal factors that were identified could contribute to the future development of this ICF component.
The resulting final version of the Comprehensive ICF Core Set for OP will have a number of potential areas of application and great potential for further developments. It defines which areas are relevant in relation to functioning in patients with OP and consequently what to measure. Therefore, it can be used as a starting point in the assessment of patients with OP providing a common standardized language for all health professions. A functioning profile can be created on the basis of OP assessment and subsequently used to document interventions and also used as a reference for follow-up.47 Being used in a multidisciplinary team, the Comprehensive ICF Core Set for OP can be seen as a common platform from which the different professionals start their assessments and plan interventions and to which they return when discussing the treatment of the patient and the patients' problems, resources, and environmental factors.69,70 In addition, scores that combine the information of all single ICF categories into a few numbers are anticipated to be useful in clinical practice. Data collected within an ongoing international validation study will be used to develop such scores as recently demonstrated for the “Comprehensive ICF Core Set for Osteoarthritis.”71 Moreover, because the Comprehensive ICF Core Set for OP defines what to measure but not how, future studies could focus on the operationalization of the ICF categories and identify appropriate instruments to assess the single ICF categories. Finally, all these developments will be summarized in a user manual for the “ICF Core Sets for OP” to facilitate its application in clinical practice and research.
The authors thank all participants of the Delphi survey for their valuable contribution to the study and their time in responding to the demanding questionnaires.