Taking the Long View: Patients Perceive Benefits and Risks of Treatment as Multidimensional

As a chronic and systemic inflammatory condition, rheumatoid arthritis (RA) affects people’s quality of life (QOL), with symptoms ranging from pain and fatigue to stiffness and restricted physical mobility. The availability of a number of longstanding and newer RA therapeutic options has helped combat troublesome aspects of the disease, including progressive joint erosion and damage. However, QOL is also affected by medication side effects that differ in frequency, severity, and duration. Lifelong therapy requires patients with RA to continuously assess and reassess the benefits vs risks of each treatment. The study by Hazlewood and colleagues, “Frequency of Symptomatic Adverse Events in Rheumatoid Arthritis: An Exploratory Online Survey,”¹ is an important contribution in advancing scientific understanding of the risk side of this balancing act. The study identifies and quantifies how patients perceive symptomatic adverse events (AEs)—a term the authors use interchangeably with “side effects”—when participating in clinical trials to develop new therapeutic options. As the authors point out, current approaches to capture AEs rely on grading systems informed by physician or research personnel inputs rather than by direct patient reporting. The study by Hazlewood et al¹ records a key step in the journey of patient input in this area, but there is a long way to go. Future work in this space should heed additional considerations.

Chief among these considerations is the reality that patients perceive side effects as multidimensional. Hazlewood et al have parsed 3 dimensions: which side effect(s), how frequently side effects are experienced, and which medications are potentially implicated.¹ Patients take note of other dimensions as well; temporality and severity of side effects are important factors not addressed by the study.¹ It is important to distinguish between long-term effects that may emerge after years of therapy vs more immediate side effects. Further, some side effects may subside between doses,² while others may be ongoing or recurring. Still other side effects may be subject to patient tolerability over time as one learns to cope with a specific side effect. Reporting side effects in one moment in time may fail to capture how time, adaptability, and changing disease status could interfere with patient reporting of symptomatic AEs.

For these reasons, future research of patient experience of side effects would benefit from a longitudinal perspective. Hazlewood et al¹ captured patients’ perspectives of AEs in one moment in time, meaning that the survey respondent might be in a period of remission, have stable disease, or be experiencing a flare or acute illness. However, current disease status may influence patient perception and experience of side effects. At CreakyJoints (creakyjoints.org), an international, digital arthritis organization, our ArthritisPower Research Registry conducts human subjects research that captures the patient perspective through both unique surveys and validated patient-reported outcome (PRO) measures on our proprietary smartphone and desktop applications. These apps allow patients to track and share their symptoms and treatments while also participating in voluntary research studies in a secure and accessible manner. The ArthritisPower infrastructure (arthritispower.org) is designed to connect with patients directly, regularly, and remotely through the digital platform, and can also be leveraged via community-based physician networks that help recruit study participants. In a proof-of-concept study to explore the temporary side effects of nausea and fatigue experienced each week by patients with RA when taking their methotrexate, we found that “digital remote patient monitoring presents an opportunity to detect and address medication tolerability in real time,” in this case, quantifying changes in fatigue and nausea PRO scores proximal to weekly dose.² Study designs such as this are needed to characterize patient experience of medication side effects to assist in grading and, ultimately, to help patients anticipate and cope with them once a medication is on the market.

The severity of side effects matters to patients and, importantly, patients and their physicians make treatment decisions in the context of risk-benefit trade-offs in the real world. Our research, in collaboration with Fraenkel et al, explored the development of a PRO measure meant to capture patients’ global experience of an RA treatment, including both benefits and harms.³ As a first step in the study, we developed an AE hierarchy to directly compare AEs to each other, essentially ranking AEs by patients’ perceptions of their severity. Because some AEs may not be directly comparable, we constructed a partial ordering of AEs into “equivalence classes” or groups of side effects that could be ranked relative to one another, even if the AEs themselves were not always directly comparable within groups. We then paired this hierarchy with the anticipated benefits of a therapy. Ultimately, the goal of the measure we created, which combines risk and benefit into a single instrument, is to “enable patients and their physicians to compare the percentage of patients experiencing each level of outcome, from most to least desirable, across treatments” and by doing so, to help patients to make decisions about medications based on their preferences, experiences, and knowledge.³ However, the measure we developed, OPEX (Overall Patient Experience),³ can be applied only to treatments on the market with a known side-effect profile; therefore, Hazlewood and team’s efforts¹ to incorporate patient perspective into drug development is welcome and necessary. Hazlewood et al chose a different approach by focusing specifically on symptomatic AEs (side effects) and drawing from the preexisting list of such side effects from cancer medications from the Patient-Reported Outcomes of the Common Terminology Criteria For Adverse Events (PRO-CTCAE).¹ Future efforts to generate data on the frequency, severity, and duration of side effects from a patient perspective must take into account the perceived and experienced heterogeneity of their impact on QOL within a hierarchical ordering of side effects. Grading systems need to take this into account because people living with RA make benefit-risk assessments based on these nuances. In practice, trade-offs are not determined by a medication’s side effect alone but also by its expected benefit, and ultimately, its overall effect on patient QOL.

Another important dimension to explore in future research is the likelihood that a side effect is attributable to the medication, rather than the RA. Patient decision making in this area is complicated because disease symptoms and experienced side effects from a medication may overlap. For example, fatigue may be experienced from RA, from other comorbidities, or as a symptomatic AE. In the study by Hazlewood et al,¹ the authors explored the association between each of the 80 symptomatic AEs reported by patients in the last 7 days and current RA medication use across 8 different medication categories, one of which was nonsteroidal antiinflammatory drugs (NSAIDs). It is worth noting that in our ArthritisPower studies,^2,3 patients with RA taking only NSAIDs (and even prednisone) are often excluded from analysis because these medications are not specifically designed to directly target RA the way disease-modifying antirheumatic drugs are. Even though some classes of medications, such as NSAIDs and glucocorticoids, may help with symptomatic relief, it could be confusing for a patient taking only NSAIDs to discern whether a perceived side effect is caused by treatment or by suboptimally treated disease when the medication is not designed to address the underlying autoimmune condition. This may be why the authors flag that, “Some of these side effects [reported by patients]…may be related to active RA, as patients taking NSAIDs or prednisone are more likely to have active (or recently active) disease.”¹ In this study,¹ respondents taking NSAIDs reported side effects such as pain, aching muscles, aching joints, headache, sleep disturbance, and others, all of which are also known symptoms of RA. Future research or analysis should consider removing NSAID-only patients to determine the impact on side-effect reporting or compare patients who take NSAIDs with those who do not.

Further, there is an urgent need to diversify our clinical trials to be more inclusive in reach. This will help us better understand how a specific RA therapy works, and is received, across different demographics and to understand the potential heterogeneity in values and preferences that may emerge across different demographics including race, ethnicity, and gender. Such an understanding can help tailor communications uniquely to different demographics and build patient trust. Like many studies, including our ArthritisPower research,^2,3 Hazlewood et al analyzed data from a fairly homogenous study population, which they acknowledge as a study limitation.¹ Patient registries like ArthritisPower are uniquely positioned to be able to recruit diverse populations into clinical trials, collect validated information on patient-generated responses to therapeutics, and establish trust within communities, because registries are repositories of real people who document their lived experience through validated measures that can then be shared with others who share similar demographics.

The Hazlewood et al study¹ is a good, first building block to explore an important phenomenon that we often hear about anecdotally as a patient organization that has a patient registry that is actively collecting longitudinal data on RA patients’ experiences. It is well known that many medications carry a side-effect profile. As new drugs are developed, a formal lexicon and a validated tool for patients to use, in combination with physician observations, to describe their experience in treatment will be extremely valuable. In addition, what remains is the need for more data to demonstrate with clear evidence the frequency and effect of symptomatic AEs on QOL, as well as which medications are associated with those AEs. Understanding how side effects affect QOL, which side effects patients are willing to tolerate, and what type of communication and support they need means ensuring that the patient perspective is incorporated at the drug development phase. Ultimately, this optimizes adherence to therapy in the real world, leading to better outcomes. In research, patient respondents quickly become numbers and data points. As we all collectively work to better understand the patient experience of RA, it is important to share what we have learned with the RA patient community. At CreakyJoints, we are committed to disseminating research findings, whether through educational articles, videos, or social media events, turning our key learnings into actionable strategies to improve the day-to-day management of RA and long-term patient outcomes.

• Copyright © 2022 by the Journal of Rheumatology