Patient-reported outcomes for arthritis: Time to focus on personal life impact measures?


Only patients themselves can report their perspective on the outcomes of illness and its treatment. For some treatment effects, the patient may be the only source of information. Therefore, patient-reported outcomes (PROs) are standard measures of pain and have become standards for functional disability and global health/disease activity due to their inclusion in American College of Rheumatology (ACR) criteria and other disease activity measures (1, 2). They have been shown to be as effective as the traditional physician- or laboratory-reported outcomes in reflecting changes in disease activity over time and predicting long-term morbidity and mortality (3). In addition, there are a range of health-related quality of life instruments to measure the broader impact of conditions (4). However, these currently used PROs are not necessarily synonymous with the patient perspective. The US Food and Drug Administration, recognizing the need to incorporate this wider patient perspective in drug evaluation programs, has called for (and issued draft guidance on) standardization of methods for developing PROs that incorporate the patient perspective (5). We believe that the development of measures across rheumatic conditions and a wide range of interventions will be beneficial.

Examining the patient's experience, as reported in new data from the BeST study in this issue of Arthritis Care & Research (including patient-reported functional difficulties) (6), may be particularly crucial in the context of chronic conditions that have a major impact on patients but are only partially treatable, which is a common situation in rheumatology. The multidisciplinary team intervenes not only to change the disease process itself, but also to ameliorate the short- and long-term consequences of the condition. This can be achieved through other therapeutic measures (such as compensating for loss of function by increasing general fitness) or through changes in personal or social circumstances (such as education about appropriate domestic arrangements or altering the living environment). PROs have increasingly been used in rheumatology to test treatment strategies in clinical trials where interventions may be close to the disease pathology, but PROs have a much wider use in health-related domains such as informing individual patient care in a routine practice setting, in health policy and resource allocation, in the measurement of health care quality (7), and capturing participation in society (8). The patient's perspective of the impact of rheumatic conditions is broad (9), whereas the clinician tends to focus on pathology and functional disability.

The growing amount of literature on the patient perspective has revealed 2 important issues: clinicians and patients have different perspectives on outcomes (10), and patients prioritize treatment outcomes that are not routinely measured, such as well-being, normality, and sleep (11–14). Differences between what clinicians and patients believe is important are sometimes due to prioritizing treatment of disease over its consequences, and/or a lack of research about particular aspects of the disease. For example, the significance or multidimensionality of fatigue in rheumatoid arthritis (RA) was not recognized by health professionals (15), nor was dyscognition in fibromyalgia (“fibro-fog”) (16).

Mapping the experience of symptoms in detail, the impact of conditions on different domains of life, and patients' important treatment outcomes may enhance rheumatology research at many levels, from the underlying mechanisms of the disease to the broad social impact. First, understanding the experience of symptoms in detail may provide new insights into the mechanisms of the disease and/or provide evidence for the cohesion (or otherwise) of symptoms caused by different pathologic processes. For example, morning stiffness was improved by time-released glucocorticoids when other symptoms remained unchanged (17). Second, discrepancies between the patient's experience and apparent disease control may be identified. Treatment may result in a clinically significant reduction in the Disease Activity Score (DAS), but not change in a patient's important outcomes, such as being able to climb stairs. Third, patient-clinician communication may be improved by establishing the different ways in which common terms, such as flare, are used (18), or the treatment outcomes that make sense to patients. For example, in the BeST study reported in this issue of Arthritis Care & Research, the percentage of less difficulty in doing housework was measured as part of the McMaster Toronto Arthritis Patient Preference Disability Questionnaire (MACTAR) (6). This is potentially a more comprehensible way of describing the effectiveness of a treatment strategy to a patient than a DAS score. As van der Kooij et al suggest, patient perceptions of health can influence patient compliance to therapy and the consequent costs of RA to society (6).

Fourth, research on patient-important outcomes may help to establish which interventions may have a significant impact on self-management of symptoms and which outcomes should be measured. For example, a cognitive–behavioral therapy intervention for fatigue in RA is underway to improve patients' management of the symptom (19). Although traditional measures may be included, such as the Health Assessment Questionnaire (HAQ) and the DAS, different types of PROs are required for psychological and educational interventions. Fifth, the consequences of rheumatic conditions affect not only the individuals with the diagnosis, but society as a whole. For example, a work productivity questionnaire is currently being constructed to enable the effect of work presenteeism (at-work productivity loss) and absenteeism to be more accurately measured (20). These broader impacts of disease are represented in frameworks such as the International Classification of Functioning, Disability and Health (ICF), which covers both the individual and social implications of rheumatic conditions (8).

In the identification of new areas of interest to the clinician and the patient, it is clear that there are not always instruments available that are fit for purpose, either because the measure was developed for a different condition or does not cover all of the dimensions emerging from qualitative research. There are the usual methodologic issues with regard to PROs, such as validity and reliability (5, 21–23), but there are some bigger issues to be addressed. First, it is important that the process of mapping patient-important domains across rheumatic conditions is standardized. This would include the collaboration with patient research partners (24). A review of methods used to date should be performed and best practice endorsed by authorities such as the ACR and the Outcome Measures in Rheumatology Clinical Trials (OMERACT) group. Second, guidelines to determine which measures to include in clinical trials of different types of pharmacologic and nonpharmacologic interventions should be developed. The guidelines might use the ICF as a framework to ensure a comprehensive perspective. Therefore, in addition to the ACR core set of variables that would capture body functions and structures, and a set of patients' most important outcomes to capture activities and participation, validated “bolt-on” outcomes would be added specific to the intervention (25).

We suggest that this view of PROs does not relate just to the source of the data (the patient), but encompasses a more comprehensive concept, better captured in the term personal life impact measures (PLIMs). This takes into account the broader impact of living with the disease and its consequences. It also emphasizes the individual importance of outcomes to a patient, i.e., their personal impact. For example, measuring a patient's difficulty in bathing may not be appropriate if the individual showers out of choice rather than necessity. Tools such as the MACTAR (26) or the Personal Impact HAQ (27) incorporate this idea of valued functions and activities and may give a more accurate reflection of changes in disease activity or of its consequences.

Before guidelines for standardization can be constructed and agreed upon, there are several issues researchers may consider. First, researchers should collaborate with patient research partners from the outset, both in the identification of a need for a new measure and their development. Second, more research is required around patients' important outcomes for different rheumatic conditions, and particularly their measurement in different cultural groups (28). Third, sufficient information should be available in the methods of publications to enable replication of measures. For example, the exact wording of visual analog scales and their anchors should be reported. Fourth, where studies use both traditional measures of disease activity and PROs (or PLIMs), statistical analysis of their relationship would allow us to see whether clinical improvements are reflected in patients' assessments of change.

It appears that there is still considerable work to do in the field of PROs in rheumatology. There is potential for the patient perspective to inform all aspects of rheumatology research and for PROs to take a broader and more significant role as different types of interventions are trialed. Discussions at OMERACT have pointed to an important direction for development: a synthesis between the ICF, which could act as a guide to the breadth of domains that may be included in PRO measurement, and the concept of PLIMs, which would provide appropriate adjuncts to the ACR core set to ensure that the patient perspective is included in a wide variety of intervention trials (25). It is hoped that by refocusing on the patient perspective, new insights into the conditions and their consequences will be revealed, and the true impact on patients' lives can be captured.