Outcome assessment and limitations

The widely heard quote 'you can't manage what you don't measure', likely dates back to Lord Kelvin, who in his 1883 lecture delivered at the Institution of Civil Engineers, London said: …when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind…[1] In trying to solve the challenges of managing haemophilia – whether for the individual patient or in studies of patients in general – we are further ahead if we can accurately and precisely measure the outcomes we are interested in. Much of haemophilia care is related to preventing damaging arthropathy and its resulting impact on quality of life – so musculoskeletal outcomes are of prime importance in clinical care and in research. Some have proposed that we develop an accepted 'core set' of measures that can define health in the context of haemophilia [2,3]. The World Health Organization (WHO) International Classification of Functioning and Health (ICF) is a framework that we will use to help structure our discussion of elements that may play a part in that core set (see figure) [4]. Until recently, range of motion (ROM) was the most commonly used physical outcome measure for evaluating the effects of intervention on joint health [5,6]. As it became necessary to develop an instrument that could assess a wider spectrum of physical changes that occur as a result of joint damage, the World Federation of Hemophilia (WFH) endorsed the Physical Examination (PE) Scale in 1985 [7,8]. However, with the advent of primary prophylaxis, it became evident that the score was not sensitive to early change, as many joints scored zero (normal) on the WFH PE Scale [9,10]. In addition, it did not take into account the normal physiological changes that occur in children [10]. These observations provided the impetus to develop new scoring systems. The Colorado PE instruments (full and half point), described by Manco-Johnson et al. [10], the Young Child Scale [10], and the PedNet (Stockholm) instrument were developed in an attempt to increase the sensitivity of physical assessment [11]. Subsequently, these scores were combined by the International Prophylaxis Study Group (IPSG) to produce the Hemophilia Joint Health Score (HJHS). The aim was to produce a score that would be sensitive to early change, account for normal development in children and be reliable, valid and practical to administer. The validity and reliability in the earlier versions have been found to be good [12]. Following a multi-centre validation study in 2011, a version 2.1 was developed by removing or modifying redundant or less sensitive items [13] – with a total possible score of 20 for each joint, in addition to a maximum score of 4 for assessment of global gait. The HJHS will need additional evaluation in other patient populations, and in other centres not involved in its design, to assess its applicability and usefulness in patient care and research. Global assessment of joint health should not only include changes in joint architecture and joint function, but should also evaluate how these changes affect both the patient's ability to perform activities and participate in social activities [14]. In 2004, van Genderen et al. developed the Haemophilia Activities List (HAL), a haemophilia-specific, self-assessment questionnaire to assess functional abilities in haemophilia [15]. It consists of 42 activity items, divided among seven domains: 'Lying down/sitting/kneeling/standing', 'Functions of the legs', 'Functions of the arms', 'Use of transportation', 'Self Care', 'Household tasks' and 'Leisure activities and sports' [15,16]. The total score is normalized to 100. The HAL was validated in 2006 [16] and has good convergent validity (r = 0.47–0.84) and internal consistency (Cronbach's α = 0.61–0.97). The HAL is a self-reported questionnaire, and is language and culture specific. In a cohort of patients from India, it was found that several questions, like those relating to household tasks and leisure, were not attempted by most patients; only 10% of the patients completed all 42 items [17]. The HAL also requires the subject to be literate, as it is a self-administered questionnaire. While the HAL was developed in close collaboration with, and validated for use in adults [15,16], the paediatric version of the HAL (called PedHAL) was developed for use in children [18]. The scores of the domains of the PedHAL, version 0.1 correlated significantly with the subscale 'physical functioning' of the CHQ-50, ranging from 0.48 to 0.74. However, profound ceiling effects were present in all PedHAL subscales, with the median score for all domains being 100 [18]. In addition to the self-reported HAL, the Functional Independence Score in Haemophilia (FISH) was developed as a performance-based assessment tool, to objectively measure the patient's functional ability [17]. The final assessment included eight activities (eating, grooming, dressing, chair transfer, squatting, walking, step climbing and running) that were graded from 1 to 4 according to the amount of assistance required to perform them. It has good internal consistency (Cronbach's alpha of 0.85) and is highly reliable, with a pooled intra class correlation coefficient of 0.98. It has good construct validity, with a good correlation with other self-rated assessment tools, including the HAL. The FISH had a good correlation with the clinical score (r = –0.61) and the radiological score (r = –0.38) [17]. The FISH was originally designed to compare a patient's basic functional ability with that of normal healthy individuals, and was not designed to assess challenging activities in individuals. Therefore, like the PedHAL, it may have a ceiling effect when applied to those with minimal musculoskeletal changes [19]. It has, however, been used effectively in studies from varied cultural backgrounds [20–23]. While the assessment of Activities involves the ability to execute tasks or actions, Participation is defined as involvement in a life situation, such as sport, leisure, work or social events [4]. The ICF provides a single list of activities and participation in nine domains. According to their needs and purposes, investigators designate some domains as activities and others as participation [4]. Although there are several generic instruments used to assess participation, only a few have been used in haemophilia [2]. While the items in FISH are primarily in the domain of 'activities', the HAL has several questions that involve the subject's interaction with others and with the environment. Assessing 'participation' across cultures is challenging, as several items/questions may not be equally relevant. In a study from India several items related to participation in the HAL had poor cross-cultural validation [17]. The Canadian Occupational Performance Measure (COPM) is an open-ended questionnaire that allows patients to prioritize their main concerns – both in the domains of functional activities and in participation. It has been shown to be useful in making individualized management plans for patients with haemophilia [24]. Its ability to assess different intervention programmes is, however, limited. The feasibility of developing a tool to assess participation, which is contextually relevant and universally applicable, needs to be explored. It has been a long felt need to develop a core set of disease-specific tools to assess the different domains of musculoskeletal outcome as defined by the ICF. The WFH has taken the first step by identifying a core set of tools, and making them available on the World Wide Web (http://www.wfh.org/2/7/7_0_Compendium_Assessment_Tools.htm). These tools need to be used more widely in centres not involved in their development, to judge their acceptability across different countries. Long-term studies are necessary to determine their efficacy in assessing the severity of joint arthropathy. Radiological assessment has been one of the oldest clinimetric tools used to measure progression of joint arthropathy. With newer imaging modalities, it has been possible to detect changes in joints before they are clinically apparent. It is therefore an important tool for evaluating the effectiveness of different prophylaxis regimens. Ultrasound, CT and MRI are useful in evaluating complications like pseudotumour which still exist in parts of the world where there is no prophylaxis. Radiography is the baseline imaging for haemophilic arthropathy. There were many early radiological descriptions and classifications over the years [25,26]. In 1977, Arnold & Hilgartner refined the classification into five stages [27]. In 1980, Pettersson proposed a scoring system which assesses the radiological abnormalities in six commonly affected joints (knees, elbows and ankles) [28]. This was incorporated into the first joint-scoring system endorsed by the WFH, and is still being used in the measurement of long-term outcomes in haemophilia. Although plain radiography mainly assesses osteochondral changes, which are late in the natural history of haemophilic arthropathy, it remains the modality of choice for baseline clinical assessment, as well as for comparing outcome of differing prophylaxis regimens. Magnetic resonance imaging (MRI) has had a major impact in understanding early joint arthropathy, and also in detecting changes much before they are apparent on plain radiographs. MRI has advantages over radiography, providing the best detail for soft tissue and cartilage changes with no ionizing radiation [29,30], and is considered the 'gold standard' among the imaging modalities currently available. Changes observed on MRI were first described in 1986 by Kulkarni et al. [31], and several other reports of MRI use soon followed. Many scales (Denver, European) were developed by different workers, to assess joint damage and thus monitor and compare prophylaxis regimens [32,33]. To facilitate international comparison of data and enhance the accumulation of experience with MRI scoring, the international MRI expert subgroup of the International Prophylaxis Study Group (IPSG) has developed a consensus scale assessment of haemophilic arthropathy [29,34,35]. Whereas MRI imaging picks up several early changes before they are seen on plain radiographs, the implications of these minor changes in terms of individual joint function remain to be determined. MRI is also expensive, time consuming, sometimes requires sedation and may involve very long waiting periods. Ultrasonography (US) has been used for quite some time to assess effusions and haematomas in persons with haemophilia. The modality is ubiquitous, and an examination is relatively inexpensive, quick, does not have ionizing radiation and can be repeated as needed even at the bedside. US detects soft tissue changes like joint effusion and synovial thickening. Doppler interrogation easily shows hyperaemia in acutely inflamed soft tissues. Osteochondral changes – like erosions – can be detected along the periphery of the articular margins. There are imaging protocols available [36], as well as scoring systems. In the assessment of arthropathy, US is still in an early stage and it may be important to correlate the findings with MRI, which is well established. Compared to other modalities, US requires additional expertise in interpretation, and is user dependent. Although imaging has helped in understanding the integrity of the haemophilic joint reasonably well, scoring systems for MRI and US are likely still evolving. The challenges in the use of radiological tools include availability, cost, expertise, long durations required for the studies and sometimes long waiting periods. Better understanding of how the radiological changes correlate with progression of joint arthropathy, clinical significance of Doppler and early MRI findings and the physician's need (question to be answered) are essential, and will determine the use and future directions of imaging. Quality of life (QoL) is a ubiquitous term that has been used for centuries, initially describing a country's standard of living. More recently: the term quality of life is used to evaluate the general well-being of individuals and societies [37]. The World Health Association has defined quality of life in this way: Quality of life is defined as individuals' perceptions of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns [38]. As such, QoL is necessarily subjective and very broad in its construction. For many individuals, including persons with haemophilia, physical health makes only a very small contribution to their assessment of QoL [39]. Many other factors, like romantic and social relationships, wealth, material possessions, achievements in school or work and so on, are included when an individual assesses his or her own QoL. The most important goals in treating persons with haemophilia are arguably preventing mortality and improving QoL; however, as health practitioners we may not be able to have an impact on aspects of QoL other than health. For that reason, many groups have developed ways of measuring health impact or health status. Questionnaires assessing health or health status, and reported by patients themselves, have often been labelled 'health related quality of life' (HRQL) measures [40]. The World Health Organization definition of health is a: state of complete physical, mental and social well-being and not merely the absence of disease or infirmity [41]. Their ICF model provides an organized way to conceive of health in the context of a disease like haemophilia. It posits that the disease has a direct impact on body structures and functions, activities and participation. Furthermore, these domains are influenced by environmental and personal factors. Given the descriptions above of health and QoL, it follows that a good measure of HRQL would measure the impact of disease(s) at the physical, mental and social levels. It would do this using the domains of structure and function (anatomy, physiology, etc.), activity limitation (where activities might be defined as instrumental activities of daily living) and social participation. A good HRQL measure would investigate environmental and personal aspects that influence function. Finally, HRQL should be measured from a subjective point of view (i.e., from the patient's perspective), take into account the individual patient's values, desires, expectations and autonomy of choice (i.e., measure function against what the patient wants to achieve, rather than what the questionnaire developers expect or feel to be normal) [39,40,42]. Health status and HRQL measures are sometimes divided into generic and disease-specific. Generic measures have the advantage of being applicable across the whole spectrum of diseases, and therefore allow standardization and comparison between persons with haemophilia and patients with other diseases. However, the effects of interventions directed specifically towards haemophilia (e.g. factor prophylaxis to reduce bleeding frequency) may not be measurable with generic measures (that may not, for example, have specific questions about bleeding). Several generic HRQL measures have been used in haemophilia studies, namely the Short Form 36 (SF-36), SF-12, Sickness Impact Profile (SIP) and the Quality of Well-Being Index (QWB) [43]. Bullinger, et al., representing the International Prophylaxis Study Group (IPSG) have reviewed the haemophilia-specific HRQL questionnaires [43]. For adults, these are the Haemo-Qol-A [44], Haem-A-Qol [45], the Hemofilia-QoL [46], Hemolatin-QoL [47] and QUAL-HEMO. For children, these are the Haemo-QoL [48], CHO-KLAT [49] and a proxy measure for very young patients [50]. Most of these tools were studied and shown to have good to excellent reliability and construct validity; i.e., most of these tools have demonstrated measurement properties that make them suitable for use in studies and in the clinic. Furthermore, in general, these tools address five domains of health – physical, emotional/social, functional, mental and treatment-related – and can therefore be thought of as addressing some of the main areas defined by the ICF model. There are some ways, though, in which these measures fall short of the ideal detailed above. Although persons with haemophilia (or in some cases their parents) answer these questions from their own experience – the questions asked, the scoring options listed and the values attached to those scores reflect the values and expectations of the questionnaire developers,1 rather than the values and expectations of the individual patient answering the questionnaire. That is, these measures are not fully subjective, and do not take into account individual autonomy of choice. They are missing key elements important for the assessment of QoL. A better name for these tools, then, might be 'self-reported measures of health status'. These questionnaires are all fine tools that have solid measurement properties and are useful in the evaluation of haemophilia health status, but they do not fulfil the construct of QoL (if we accept the definitions put forth by the WHO). If we truly want to measure QoL, the WHO definition is so broad that it probably does not make sense to consider disease-specific measurements at all. (Disease-specific health status measures, in contrast, make eminent sense.) To satisfy the requirements for autonomy, a QoL measure should allow patients to pick those domains and items of life that have the most meaning to them. To be truly subjective, a QoL tool should allow patients to define their own values, expectations, hopes and realizations for each of these items [42]. Alternatively, when such a precise understanding is not necessary, we may simply use global measures (like simple visual analogue scales [51] or global utility measures [52,53]) that allow patients to make these subjective and autonomous assessments internally. The WHO ICF is a useful framework for identifying the important domains of health that can make up a core set of assessments for persons with haemophilia. We have good tools for many, but not all the domains of health. For assessing the domain of structure and function, we have the Hemophilia Joint Health Score, Pettersson radiograph score and the IPSG MRI consensus scale; US scales are being developed. For assessing the domain of activity/activity limitation we have the Haemophilia Activities List (and PedHAL), and the Functional Independence Score in Haemophilia. Tools for measuring participation have been less well studied. The overall construct of health may be measured by a variety of disease-specific, and generic, so-called 'health related quality of life' questionnaires – but additional work must be done to identify ways of incorporating autonomy of choice and subjective meaning into the measurement of quality of life. Dr. Feldman holds peer-review funding from Bayer and Baxter; he is a member of DSMBs for Novartis and Pfizer.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.