Background
The foot is often the first area of the body to be systematically affected by rheumatoid arthritis (RA) [
1,
2]. Upon diagnosis, approximately 16% of patients with RA have foot involvement [
3], in 15% of cases the forefoot is the first area of the body to become symptomatic [
4], and virtually 100% of patients report foot problems within 10 years of RA onset [
5]. These clinical foot problems have a significant effect on the person's functional ability which is known to lead to important emotional experiences for patients, including anger and sadness [
6].
The multidimensional consequences of foot problems can be subjectively assessed and evaluated using patient-reported outcome measures (PROMs). PROMs record patients' perspectives of their health, illnesses and the impact of any clinical interventions in a valid, reliable and feasible way [
7]. They are an objective means of recording largely subjective outcomes and represent an ideal, economical and efficient method of measuring the quality and efficacy of care provided [
8], integrating important psychosocial factors into a clinical assessment that otherwise may not be gathered.
According to Bowling [
9], PROMs can be characterised in terms of their disease specificity (generic or non-disease specific), measurement objectives (discrimination, evaluation and prediction) and what they intend to measure (quality of life, health related quality of life (HrQoL) or health status).
The development of a PROM and establishment of its measurement properties most commonly entails the use of psychometric theory, which can be divided into two main methodological approaches: Classical Test Theory (CTT) and Item Response Theory (IRT) [
10]. CTT (referred to as Traditional Psychometrics) utilises both item and sample statistics and is based upon 3 concepts, together known as True Score Theory [
11]: test/observed score, true score and error score. Item response theory (known as Modern Psychometrics), on the other hand, utilises logistic response models to apportion individual items to the constructs of interest using conditional methods, according to their individual difficulty and the ability of subjects to respond positively to the items [
12].
In selecting a PROM for use in either a clinical or research environment, the decision should be made upon its conceptual basis [
13], the appropriateness of the PROM for the intended purpose and evidence for its measurement properties [
14] (see Additional File
1, Table 1). A conceptual basis/model for a PROM is required to establish a well-defended rationale for and specify clearly the outcomes of interest for the instrument. The lack of an appropriate conceptual model for a PROM can result in a number of problems, including weak or incorrect scoring, analysis and interpretation of the data yielded [
13].
Given the impact of foot problems for the adult with RA and the recognition that foot health interventions are an important aspect of health care for this patient group [
15], a measurement of changes in foot health is vital to the monitoring of the foot and interventions for it from both the clinician's and the patient's perspective. This notion aligns with Darzi [
16] who has recommended the use of PROMs, which focus on quality health care from a patient-centred perspective.
There are systematic reviews of PROMs for the foot and ankle in general [
17] and for combined objective and subjective outcome measures, which includes a narrow selection of PROMs, with application for the foot with RA [
18]. However, currently, there is no systematic review that has appraised specifically and in detail PROMs relevant for the foot with RA in terms of their conceptual bases, quality of construction and evidence for their measurement properties. Therefore, a systematic review of PROMs with relevance to the foot with RA is both timely and appropriate.
The aim of this systematic review was to appraise the quality of PROMs that may be used for the assessment and evaluation of the foot with RA in terms of their conceptual bases, quality of their construction, measurement aims and evidence for their measurement properties.
Methods
This review was conducted using appropriate systematic review methods and is reported in accordance with the PRISMA statement [
19]. A structured and exhaustive search of Pubmed, Embase, Cinahl, Ingenta, Science Direct and the Cochrane Collaboration Library was conducted on 15/10/2008 using the search terms: 'Rheumatoid arthritis' and 'foot index,' 'foot score,' 'foot instrument' and 'foot evaluation'. The search was restricted to publications in the English language. The reference lists of journal articles of interest were also searched and no restriction on year of publication was imposed to reduce publication bias.
PROMS selected for the review fulfilled the following inclusion criteria:
-
Foot region specificity.
-
Measurement of constructs relevant to the foot with RA, such as pain or activity limitation.
-
Potential for application of the PROM in a research and/or clinical environment.
An additional search for specific evidence for the measurement properties of the PROMs was conducted by using the same search strategy, using the name of the PROM and 'measurement properties,' 'reliability,' 'validity,' 'validation,' 'responsiveness' and 'interpretability' as search terms.
All PROMs included in the review were appraised according to several pre-defined quality assessment criteria, including: The Scientific Advisory Committee of the Medical Outcomes Trust [
20], The Patient-Reported Health Instruments Group [
21] and the NHS Technology Assessment Board [
7].
Discussion
This systematic review identified 11 PROMs with potential application for the assessment and/or evaluation of the foot with RA, 9 of which are based on CTT and 3 on IRT. However, only 1 of the PROMs is RA disease-specific, 1 JIA disease specific and the other 8 are generic. For the assessment and evaluation of a specific disease condition, generic PROMs may lack sufficient levels of validity, responsiveness and sensitivity [
12].
All of the PROMs vary in terms of the conceptual bases for their development, quality of the methodological procedures used for their development (item generation, selection and appraisal) and the amount of evidence available to support their measurement properties. In terms of their conceptual bases, only 5 of the PROMs identified can be considered to evaluate or assess musculoskeletal conditions of the foot from a biopsychosocial perspective [
22,
32,
34,
43,
52], one of them is RA disease-specific [
52] and the other JIA disease-specific [
43]. The other PROMs identified are formulated on conceptual bases that are either very restricted, such as the ROFPAQ [
40] with the sole assessment of pain, or rooted in reductionist biomedical models of disease, such as the FAM [
36] with its strong emphasis on structure and function.
Biomedical models, such as the International Classification for Impairments, Disability and Handicaps [
54], assume a linear progression from a health condition to impairments and disability [
55] and do not take account of the bi-directional implications of environmental or personal factors, leading to an over-simplistic appraisal of the implications of chronic diseases for patients [
56]. PROMs that use biomedical models their conceptual bases may lack a wide enough conceptual scope to comprehensively evaluate the impact of chronic conditions on feet, such as RA. Furthermore, only 5 of the PROMs identified [
32,
34,
36,
47,
52] consider the implications of footwear for patients with musculoskeletal conditions of their feet, with only one of them RA disease-specific. Given the importance of footwear for the management of musculoskeletal conditions of the foot [
57], particularly RA [
15,
58,
59], it is an important omission.
The content generation methods employed for the development of each of the PROMs vary in terms of both approaches and quality. Rattray and Jones [
25] argue that the generation of content for PROMs should entail a variety of sources, particularly the engagement of patients, to ensure that the items generated have content validity and are as relevant to patients as possible. Although many PROMs employed heterogeneous content generation methods, only 4 PROMs identified [
28,
32,
40,
52] used interviews with patients. It is known that patient and clinician perspectives of what is important are different [
60,
61], so those PROMS relying on experts only to generate content may have restricted relevance to patients.
Levels of pre-testing of the PROMs appears highly variable and, in some cases, absent from the methodological approaches used to develop them. Pre-testing of PROMs is necessary to ensure that all of the questions are consistently easy to read and understand by all respondents. This is to reduce measurement error and non-response by achieving what Groves and colleagues [
62] suggest to be content, cognitive and usability standards. Such standards can be evaluated during the pre-testing phase via expert reviews, focus group discussions, cognitive interviews and field pre-testing [
63]. Although many of the PROMs involved pre-testing of their content in terms of their content standards and usability standards, such as the FAM [
36] and BFS [
32], none of the PROMs involved pre-testing to ensure that they satisfy cognitive standards. This is evidenced by the FAAM [
49] and LFIS [
52], which feature questions containing complex terminology and potentially distressing words, respectively.
Evidence for the measurement properties of the PROMs is highly variable. Either no attempt has been made to establish evidence for a particular measurement property or procedures have been employed to demonstrate particular measurement properties that are incorrect or inappropriate. Furthermore, when considering the evidence for the measurement properties of PROMs based upon CTT, it should be realised that the procedures involved are both sample and context-dependent. This means that the evidence for the measurement properties has restricted validity for use on populations that differ from those used to develop evidence for the PROM. As most of the generic CTT-based PROMs did not include patients with RA in the development of their measurement properties, it can be considered that they have limited clinical and research utility for the assessment and evaluation of patients with feet with RA.
All of the PROMs identified consist of fixed scales, presenting every patient with the same set of items irrespective of their RA disease duration, severity and particular lifestyle. However, patients are influenced by the symptoms of RA to varying extents, depending upon the level of disease activity and duration [
60] and environmental and patient characteristics [
64]. Thus, a PROM consisting of fixed items that assume equal importance and relevance for every patient may not be the most appropriate or patient-centred means of assessing the impact of foot problems on people with RA. Although measuring the same outcomes for groups of patients may be advantageous and necessary for research such as clinical trials, for informing decision making in clinical practice, assessment of change unique to the individual has been considered to be more beneficial [
65]. This is possible through the use of patient preference-based questionnaires, which are geared towards the assessment of the specific individual and can permit the measurement of concepts such as individual health-related quality of life [
66].
Patient preference-based questionnaires have been implemented and are currently used with considerable success [
67]. These include the Patient Generated Index [
68] and the MacMaster-Toronto Arthritis Patient Preference Questionnaire [
69]. However, this systematic review could not identify RA or generic foot-specific idiographic PROMs with either the potential for use in the assessment of the foot with RA, or with any relevance to the foot.
Although this review has attempted to present a comprehensive appraisal and review of all foot-specific PROMs that are relevant for the assessment and/or evaluation of the foot with RA in both research and clinical environments, it should be realised that new PROMs may have been developed or additional evidence for the measurement of existing PROMs presented since the literature search for this review was conducted, Further, it was not possible to present the results of the literature search to experts and special interest groups to gauge its comprehensiveness prior to conducting the review. However, the findings have been scrutinised for accuracy and relevancy by two academic podiatrists and one independent academic. Despite this slight weakness in approach, this review has attempted to present a comprehensive appraisal and review of all foot-specific PROMs that are relevant for the assessment of the foot with RA in both research and clinical environments.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SW conceived the study design, conducted the systematic review, interpreted the findings and drafted the manuscript. AEW, MR and AG reviewed the manuscript and provided academic support throughout. All authors read and approved the final manuscript.