Physician Adherence to Treat-to-Target and Practice Guidelines in Rheumatoid Arthritis

Principles of treat-to-target (T2T) have been widely adopted in both multinational and regional guidelines for rheumatoid arthritis (RA). Several questionnaire studies among physicians and real-world data have suggested that an evidence–practice gap exists in RA management. Investigating physician adherence to T2T, which requires a process measure, is difficult. Different practice patterns among physicians are observed, while adherence to protocolized treatment declines over time. Rheumatologist awareness, agreement, and claims of adherence to T2T guidelines are not always consistent with medical records. Comorbidities, a difficult disease course, communication barriers, and individual preferences may hinder an intensive, proactive treatment stance. Interpreting deviations from protocolized treatment/T2T guidelines requires sufficient clinical context, though higher adherence seems to improve clinical outcomes. Nonmedical constraints in routine care may consist of barriers in healthcare structure and socioeconomic factors. Therefore, strategies to improve the institution of T2T should be tailored to local healthcare. Educational interventions to improve T2T adherence among physicians may show a moderate, although beneficial effect. Meanwhile, a proportion of patients with inadequately controlled RA exists, while management decisions may not be in accordance with T2T. Physicians tend to be aware of current guidelines, but their institution in routine practice seems challenging, which warrants attention and further study.


Treatment of Rheumatoid Arthritis to Target-What Are the Benefits?
Treat-to-target (T2T) refers to a set of principles that guide physician decision making by aiming to achieve prompt and effective control of the inflammatory process in rheumatoid arthritis (RA) [1], which has led to its endorsement in international recommendations [2]. The T2T strategy includes: Defining a target of therapy (most often remission); close and frequent assessment of disease activity with validated composite measures; regular adjustment of treatment plan if said target is not reached; consideration of patients' clinical characteristics; and finally, their involvement in treatment decisions and planning [2]. The rationale can be drawn from evidence supporting intensive steering and medication over conventional strategies, which is substantiated by the benefits in clinical outcomes, including better physical function and reduced structural changes [3][4][5]. Systematic reviews have previously reported on the benefits of aiming for remission, which also includes a favorable effect on other elements of T2T, i.e., comorbidity, cardiovascular risk, and productivity [6]. Recent studies have shown that T2T yields higher remission rates and improves quality of life over routine care [7]. Other reports also indicate that utilizing a T2T approach holds high potential also in patients with inadequate response to conventional synthetic disease modifying anti-rheumatic drugs (DMARDs), which remain fundamentals in the current treatment armamentarium [8,9]. These studies underscore the importance of applying T2T in a personalized approach of care, for which there may be signs of suboptimal application.
In this report, we provide a narrative review of current literature examining the multiplanar aspects of T2T with perspectives on physician adherence and optimal management of RA.

Management of RA May Be Suboptimal
Observational data have previously suggested a substantial "evidence to practice gap", i.e., in the institution of T2T, with the prevalence of inadequate disease control reported in a significant proportion of RA patients [10][11][12]. A recent systematic report on real-world remission rates under treat-to-target strategy shows an improvement over recent years, although sustained remission is still regarded as rare [13]. Taylor et al. provided an analysis of multinational data on RA management, reporting that no T2T approach was present in approximately half of patients, and only close to one third had remission as a chosen target [14]. When examining the subset of patients with RA diagnosis stated within two years, more patients lacked a treatment target, despite 38% of them suffering from moderate to severe RA. Meanwhile, the considerable majority of physicians in this subgroup (70%) were satisfied with RA control. Another recent study reported that among inadequate TNF inhibitor responders, suboptimal treatment decisions, such as lack of intensification, may be widely prevalent despite moderate to severe RA activity [15]. Although causal attribution cannot be strictly determined across these studies, it can be observed that the major goal of T2T, which is achieving remission, has not been reached for many patients. T2T has been adapted in numerous guidelines [12], which would imply that its application and core principles are widely adhered to, or at least aimed for. Meanwhile, observational studies have indicated that guideline adherence (including T2T elements) may be suboptimal in real-life, though it varies among particular guideline components [16]. While the choice of DMARDs in routine care is largely consistent with international recommendations [17,18], other studies indicate that nearly half of newly diagnosed RA patients are not initiated with a DMARD and receive symptomatic treatment [19]. This suggests that the choice of therapeutic agents is often concordant with recommendations, as compared to patterns of practice, which are more difficult to change. Understanding the practice patterns of rheumatologists may provide insights into unsatisfactory statistics of treatment in the real world, which is often discordant with efficacy reported in trials, and may relate to insufficient institution of T2T.

Improving Physician Adherence to Treat-to-Target May Aid in Achieving Remission
Many factors affect the achievement of remission in RA, with high importance of early treatment with conventional synthetic DMARDs, or in some instances biologic DMARDs, are well established [9]. However, management of RA should also follow the T2T strategy; with an ongoing process of decisions aiming for remission shaping the treatment plan. Importantly, outside of protocolized trails these decisions are routinely at the treating rheumatologist's discretion, though in theory dictated by practice guidelines. Considering studies among rheumatologists have shown that awareness and/or agreement with guidelines is not synonymous with their actual practice [18,20,21], investigating physician adherence is of particular interest. Kuusalo et al. performed a retrospective analysis of the new Finnish RA Combination Therapy (NEO-RACo) trial and reported that good physician adherence to remission-driven protocol was related to lower RA activity and even remission. When examining short and long-term predictors of remission, physician adherence was the most significant factor at three months, and the only determinant at a two year follow-up [22]. Notably, when considering these findings and that only four patients had a null (absolute) score for adherence, achieving optimal outcomes may require only a high rather than absolute degree of adherence, which is in line with other studies that have attempted to determine an adherence cut-off (respectively, 80% and 70% for remission and low disease activity using 28 joint count disease activity score (DAS28)) [23]. When examining the above and below optimal cut-off point for provider compliance and clinical outcomes, the odds of attaining DAS28 remission were near eight-fold greater. Although this illustrates the potential benefits of improving physician adherence, a cause-effect relationship cannot be easily determined.

Barriers to Guideline Implementation and Treat-to-Target
In the clinic, the treating physician often has to deal with limited time, communication difficulties, and healthcare regulations, which may all pose barriers to T2T implementation. In a discrete choice experiment for a model patient with RA, drug efficacy, economic aspects, and patient preferences were demonstrated to influence decisions among rheumatologists [24]. Studies have also indicated that both patient and physician preferences, as well as provider hesitation to therapy intensification are obstacles to T2T implementation [25]. These studies highlight the role of the payer and individual preference in the routine setting, which may also be constrained by a rheumatologist's hesitation over therapy side-effects. Patient-related factors, comorbidities, and toxicities have all been previously associated with persistent or recurrence of T2T protocol deviation [26]. However, it should be noted that T2T strategy emphasizes the importance of patient outcomes and comorbidity. A personalized approach, which takes into account the patients clinical characteristics and wellbeing, while targeting remission, is in line with T2T.
Electronic unavailability and time shortages have previously justified infrequent use of quantitative measures in routine care [27]. Assaying acute phase reactants also seems to be a substitute for composite indices [21], while remission is variably defined, frequently without validated scores [18]. Meanwhile, close and frequent monitoring of disease activity with established indices remains a staple of T2T. Other studies report that despite agreement with T2T, physicians may still not discuss treatment decisions with patients from their lack of understanding or inability to share decisions [28]. These findings show that certain suboptimal practice patterns may persist, despite an awareness of T2T.
Vermeer et al. analyzed a random RA patient sample from the Dutch Rheumatoid Arthritis Monitoring registry (DREAM) cohort, which included several hospitals adopting a T2T strategy (for details, see Table 1) [29]. Adherence to treatment advice was reported as lower when remission was not present, though the justification was often thought to be valid. On a side note, analyses of Behandel Strategieen (BeSt) and IMPROVED revealed that targeting remission may lead to lower rates of adherence to DAS-steered protocol [30]. These studies may suggest that rheumatologists are less inclined to follow treatment guidelines when RA is not well-controlled, or the measures of disease activity do not reflect the physician's view. Indeed, variability between personal judgment and composite indices, e.g., providers view of remission and presence of DAS28 ≤ 2.6, may contribute to findings of nonadherence [29].
In some countries, studies alarmingly indicate that treatment choices can be unjustified, patients are insufficiently monitored, and there is a lack of concordance with the current recommendations for practice [31]. Prior reports of the cross-country inequities in DMARD access provide insights into different local economic regulations [32]. These differences imply that rheumatologists may not always have access to the full treatment armamentarium, which limits the treatment adjustments for patients with active, unresponsive disease. Multinational surveys have indicated a wide variability in referral and early RA assessment pathways, while the impact and practice of guidelines may also be particularly country specific [33]. It seems that any interventions to improve T2T adherence will have to be revised by national expert societies, and tailored to the structure of local healthcare, which may also entail identifying different areas of "unmet needs" and assessing their priority.

Feasibility of Treat-to-Target and Decline in Adherence over Time
Versteeg et al. showed that a protocolized T2T strategy was successfully applied in the DREAM cohort, with favorable clinical outcomes reported in long term follow-up [34]. Importantly, contraindications and comorbidities did not lead to exclusion, with deviations from protocol allowed. This indicates that in order to follow T2T strategy, adjustments of treatment plan may be necessary to account for comorbidity and other patient-related factors. In a report from the Danish Registry for Biologic Therapies in Rheumatology (DANBIO), which was designed around data from clinical trials and incorporates automatic disease activity score (DAS) calculation, systematic monitoring was concluded to be feasible for routine care [35].
Although studies with a pre-defined protocol may not adequately reflect implementation of T2T/guidelines in routine practice, they lend credible evidence on the relationship between measures of T2T adherence and the providers' behavior. Wabe et al. retrospectively studied an early RA population, where protocol deviation occurred in one fourth of visits (of which 43% justified clinically), was lowest during the first six months and considerably increased in subsequent timepoints of a three-year follow-up [26]. Patient-related factors, comorbidities, and toxicities were all significantly associated with persistent or recurrence of protocol deviation. Other studies have indicated that patient and physician preferences, as well as provider hesitation to therapy intensification are challenges to T2T implementation [25]. It has also been shown that the rate of treatment accelerations per visit decreased over time in both usual care and dedicated T2T centers [25]. Markusse et al. performed a post hoc analysis of the multicenter BeSt trial data and observed that the adherence rates declined substantially over a 10-year follow-up, down to approximately 60% [36]. Disagreement with subsequent treatment steps, reported by physicians in a questionnaire, was associated with a higher chance of protocol nonadherence. However, when considering the latter responses overall, only every fourth physician actually violated protocol. This highlights the difference between a standardized, controlled clinical trial and routine care scenarios, where the individual opinion is likely to take priority.

Evaluating Physician Adherence is Difficult
Defining T2T adherence is difficult with it being a multifaceted strategy and process. Studies tend to focus on an aspect of guidelines, which may obscure the view on the general strategy. Escalas et al. reported prospective data from an early RA cohort where following analysis and use of a propensity model, adherence to three 2007 European League Against Rheumatism (EULAR) recommendations was determined. Although substantial for individual recommendations (>50%), the treatment adherence rate to all components was only 23% [37]. Other studies reported guideline adherence variability between 21 and 72%, with an approach to treat highest scoring indicators as a relative norm [16]. Not many studies have broadly examined physician adherence to several components of T2T at once, however, Yu et al. performed an analysis of medical records from the TRACTION trial and observed that physician adherence was sub-optimal when considering a score of T2T implementation, observing a wide variability in T2T component across visits (details, Table 2) [39]. Therefore, the discrimination into what is considered compliant with the "general strategy of T2T" and not "usual care" is problematic.
Harrold et al. reported findings from a multicenter trial comparing T2T and usual care, which unexpectedly showed treatment adjustments and clinical outcomes were similar across groups (details in Table 2) [25]. This particular study illustrates other confounds of investigating provider adherence; the inclusion criteria of centers having to "agree" to implement T2T, and the provider awareness of the ongoing study itself.
Adherence to T2T also remains a process to which the associated clinical benefits are related to the effectiveness of a treatment regimen in a particular demographic. Unfortunately, the current drugs are not universally effective, nor without adversity, therefore T2T adherence will not always yield satisfactory results. It has previously been emphasized that an intuitive association between T2T protocol compliance and beneficial clinical outcomes does not imply causality and requires consideration of other factors, e.g., a more manageable disease course, medication side effects, and patient characteristics [23,26]. This in turn is what makes investigating and delineating the raw effect of adherence to T2T among providers difficult.

Lag Time in Real-World Institution of Practice Guidelines
Some studies have described that institution and adherence to T2T guidelines occurred shortly following their appearance, attributing favorable control of disease activity and remission rates to this strategy, which was substantiated by an annual overview prior to and shortly after publication [38]. A retrospective chart review study has previously shown that disease activity and functional assessments increased in number from 2010 to 2012, with excellent treatment and management adherence to quality indicators [41]. In a recent systematic review of real-world T2T evidence [13], it has been shown that over time the rates of achieving remission have gradually improved, though this cannot be attributed to T2T alone. It should be noted that alongside T2T, novel drugs and trends in therapy are being introduced, which may also account for the improvement in disease control.
Considering some reports of suboptimal RA management (see Section 2), it may be argued that a delay in the institution of "T2T philosophy" will require a few years to significantly change office-based practice, and therefore a lag time will naturally occur before T2T is prevalent. However, studies from a large United States registry have reported that care in concordance with American College of Rheumatology guidelines is not associated with time since their publication [42]. Other reports examining recommendation adherence based on prescription decisions, prior to and after publication of guidelines for disease activity-driven therapy, revealed that providers may still be more inclined to apply less aggressive therapy, even in patients with uncontrolled RA [43]. Meanwhile, adequate control of disease activity is a principle concern in RA care.

Clinical Inertia and Comorbidity
Studies show that patients with active RA may not be managed in consistency with recommendations, which advocate intensification of therapy in uncontrolled disease [16]. This phenomenon has been termed as "clinical inertia", and is increasingly recognized in rheumatology [42], though it has been reported for a variety of chronic conditions, e.g., hypertension, dyslipidemia, or diabetes [44]. The recently updated T2T guidelines have underscored the importance of individual patient-level outcomes, including comorbidity, when undertaking clinical decisions [2]. The T2T strategy recommends a comprehensive approach to the patient, though reducing the inflammatory burden of RA should be a priority. International studies have shown cardiovascular (CV) comorbidity is the most prevalent in RA [45], while other have indicated that it may also shape a difficult-to-treat patient profile [46]. Meanwhile, observational data indicate that in RA patients screening for CV disease prevention is similar to the general population despite a substantially higher vascular risk [47,48]. It may be that hesitation or lower familiarity with comorbidity may lead rheumatologists to "inertia" in refraining from screening and CV prevention. Although, some reports show that CV risk factor recognition and prevention is high [49]. The phenomenon of clinical inertia requires further study in RA. Particular attention should be given to comorbidities such as interstitial lung disease, which is second to CV disease with regard to mortality, while no optimal treatment has been determined, and awareness may be lower [50].

Communication and Personalized Care
Planning and enacting the strategy for treatment requires a degree of mutual understanding, which is difficult to build when there are discordant views, particularly on disease activity and therapy. Nonadherence to T2T owing to patient preferences and differences in patient-provider assessment of disease impact were outlined in a TRACTION trial report [51], which is consistent with other studies [30,52,53]. Although patient choice is undeniably important, it should not pose a significant barrier to optimal treatment strategies. Communicating the importance of abrogating inflammation, while prioritizing health-related quality of life and patient outcomes is what makes up a personalized approach, which will likely make the patient more amenable to the "stringency" of T2T.
Current evidence supports that discordance between patients' and physicians' evaluation of RA course often occurs, e.g., in assessments of global activity [54]. There are a variety of factors, such as pain, fibromyalgia, and depression that may well contribute, even with a lack of synovitis in ultrasound [55,56]. This side to chronic disease may often be underappreciated by physicians, while patients may be less inclined to follow through with more intensive treatment regimen if they feel their concerns are not addressed. Studies of patient and physician-based surveys have underscored the importance of communication, to which there are varying perceptions from both sides, with patients fears and concerns remaining [57]. Apart from being a key aspect of T2T, the mutual decision-making process may be crucial to consistent and effective T2T implementation. There is evidence to suggest that themes such as patient anxiety over therapy [15], pain and impaired mental function [58,59], psychosociological needs [60], and other nonspecific symptoms, e.g., fatigue [61], are important to individual-oriented management.
It has been noticed that while disease control has improved over recent years, many of these patient-centered themes have not improved, with poor indicators of quality of life [62]. Addressing the patients' needs in these "challenging" areas is significant for the physician building a working relationship, which may ultimately improve the adherence to more rigorous treatment, benefiting the patient in the long-term. Patient-recorded outcome measures pose a useful tool to determine the impact of RA on patient-centered outcomes of high importance, e.g., pain or fatigue, which may then on be more easily evaluated, quantified, tracked over time and addressed in the treatment plan [62]. Patient assessments offer time and cost-efficiency, location independence, alongside a friendly mode of following RA course, which may also be more actively engaging [63]. Moreover, they are less prone to bias than the often-lengthy recall period at consultation and may more accurately provide an overview of trends in disease. There is a varying extent of validation for some of these measures, which has been outlined elsewhere [63].

Interventions to Improve Adherence and Future Perspectives
Physicians themselves have to be aware, understand, and agree with T2T in order to remain pro-active in this strategy. Studies have shown some benefit from educational interventions in improving physician adherence to guideline elements in rheumatology, though the benefits are moderate [64,65]. TRACTION trial sub-analyses showed a significant difference in T2T adherence between physicians and participating centers [39]. Rheumatologist experience was also noted as a significant factor influencing adherence, which is confirmed by data from observational studies [16]. Considering variability in practice occurs on an individual level, some practitioners may only need to re-familiarize themselves with the strategy, while others will have to acquire more information on the evidence, benefits, and core principles. A randomized trial has shown that a learning collaborative offers the potential to improve T2T adherence among rheumatologists (details, see Table 2) [40]. Lesuis et al. have also discussed the usefulness of registries with real-time feedback on quality care, which may also provide benchmarks for internal reference in healthcare [16].
The T2T strategy advocates a patient-centered approach that may more easily be achieved by an interdisciplinary team. Rheumatologists could acquire the aid of other medical staff in more time-consuming assessments, at the same time incorporating the input from patient questionnaires, while reaching out for specialist advice in treating comorbidity. Pilot studies of nurse calculations of composite disease measures, i.e., DAS28, indicate that although potentially helpful, there is no substantial evidence to support this intervention alone [66]. However, there are several lessons to be learned from the DREAM cohort, where consensus over recommendations was achieved by all practicing rheumatologists, and strategy of care pre-planned to limit overtime during clinical visits [29]. This indicates that alongside a learning collaborative, centers could benefit from an organizational framework to adopting T2T.
Over two decades ago, practice patterns in RA care were observed among rheumatologists, with substantial individual differences and a tendency to overuse certain clinical measures [67].
It seems that despite the wide endorsement of T2T reported in multinational surveys [68], individual practice patterns that may be discordant from the recommended approach remain a valid concern. We are inclined to think that changes to optimize practice will be difficult to implement as adherence to structured advice (i.e., T2T protocol) has been shown to decline over time, even in the setting of a controlled experiment. Education on the importance of T2T and patient-oriented care will likely have to be stratified on an individual physician-level, in which actual practice and consistent feedback may enable assimilation through personal experience. Digitalized tools may be appropriate not only for determining adherence to T2T, but also for automatized calculation of composite indices, and their incorporation into office-level, electronic records that may provide real-time input for the physician. Designing interventions to improve guideline implementation into practice, rather than awareness and agreement alone, are important to address in future research. Moreover, clarification of real-world cost-efficiency is another crucial aspect. Healthcare centers could provide an internal registry with quality control, with an appropriate structure of care to facilitate frequent inter-physician collaboration.

Conclusions
Real-world data and sub-analyses of trials in RA indicate physician adherence to treat-to-target recommendations or relevant elements of protocol is not universal, differs between particular components, decreases over time, and in some cases may be sub-optimal. Evaluating barriers to T2T and/or guideline adherence in routine care is difficult; remembering that individual preferences, clinical characteristics, and the patient-provider relationship play an important role. Understanding of and justification why T2T is not upheld requires not only on a patient's or physician's view, but also a healthcare system perspective. Focusing on the patient throughout management requires adequate communication and mutual understanding, which may be improved through utilizing patient-reported outcome measures. Finally, different modes of education may facilitate an improvement in adherence to T2T among providers.