Assessment of rheumatoid arthritis disease activity and physical function

INTRODUCTION — The creation of instruments for the structured assessment of disease activity has facilitated the effective use of disease-modifying antirheumatic drugs (DMARDs) to control disease activity in patients with rheumatoid arthritis (RA) and the development of new DMARDs for treatment of this disorder; these instruments include composite measures incorporating several key clinical and laboratory variables that allow for a standardized quantitative assessment of disease activity and the response to therapy.

These assessment tools for measuring disease activity have been of particular benefit in helping to achieve therapeutic goals in patients with RA, including amelioration of the symptoms and signs of active disease, resultant prevention or mitigation of structural joint damage, and the consequent avoidance of functional impairment and restoration of physical function [1,2].

Clinical indicators employed in the assessment of RA disease activity are discussed here, including key individual variables (eg, swollen joint counts and acute phase reactant measurements) and the use of composite indices for disease activity assessment; indices of functional capacity and disability are also presented here. The clinical features of RA and biologic markers of disease activity are discussed separately. (See "Clinical manifestations of rheumatoid arthritis" and "Biologic markers in the assessment of rheumatoid arthritis" and "Investigational biologic markers in the diagnosis and assessment of rheumatoid arthritis".)

ASSESSMENT OF DISEASE ACTIVITY

Key individual variables used in composite measures — Core sets of clinical and laboratory variables for the assessment of disease activity in patients with rheumatoid arthritis (RA) were defined by several expert groups beginning in the 1990s [3-5]; these variables, in different combinations and relative weights, have been used to create the composite disease activity measures that are used in clinical practice and research. (See 'Composite indices for disease activity assessment' below.)

Most major sets of variables that were identified for this purpose included swollen and tender joint counts, patient assessment of pain, patient global assessment (PGA) of disease activity, evaluator global assessment of disease activity (EGA), and a measure of the acute phase response [3-5]. The methods used to identify these variables were driven by clinical data and met several criteria for validity [6-8].

As an example, the American College of Rheumatology (ACR) core dataset includes a total of seven measures: three performed by the evaluator (swollen joint count, tender joint count, and global assessment of disease activity); three collected by patient self-report (functional status, pain, and global assessment of disease activity); and a laboratory measure (either an erythrocyte sedimentation rate [ESR] or C-reactive protein [CRP]) [4].

The ACR and World Health Organization (WHO)/International League of Associations for Rheumatology (ILAR) core sets both include EGAs of disease activity and physical function [4,5]. By contrast, in the European Alliance of Associations for Rheumatology (EULAR; formerly known as European League Against Rheumatism) core set, physical function is regarded as an outcome variable rather than a process variable and, therefore, is not included in the EULAR core set [3].

Specific measures of RA disease activity used in core sets include the following [9]:

●Swollen and tender joint counts – Joints are typically assessed according to two characteristics:

•Soft tissue swelling and effusion (the swollen joint count)

•Pain on pressure or motion (the tender joint count)

In general, neither weighting joints by their size (ie, the cartilage surface area) nor grading them by degree of swelling or tenderness confers greater validity or reliability than swollen and tender joint counts alone [10,11].

The 28 joint count has become a standard for use in both clinical practice and clinical trials [11-14]. The 28 joint count excludes assessments of the foot and ankle joints because the interpretation of swelling and tenderness in these joints is confounded frequently by disorders other than RA [11,13,14]. However, employing a 28 joint count for determining levels of disease activity and response does not obviate the need to also assess ankles and feet in clinical practice. Although the 28 joint count has been criticized for leaving out assessment of the feet, it has been thoroughly validated and employed reliably in clinical trials and other analyses [15-17]. It is the basis of the most frequently used composite measures of disease activity for both clinical practice and clinical trials, such as the Clinical Disease Activity Index (CDAI), the Disease Activity Score using 28 joints (DAS28), and the Simplified Disease Activity Index (SDAI). (See 'Composite indices for disease activity assessment' below.)

●Pain – Pain is usually measured by visual analog scales [14-16], which most often use horizontal 100 mm lines. Patients indicate their degree of pain (typically over the preceding week) by placing a mark between "no pain" (left end, 0 mm) and excruciating pain (right end, 100 mm). Alternatives to visual analog scales include numerical rating scales (ranging from 0 to 10 in steps of 1 or 0.5) and categorical scales (eg, 5-point Likert scales), both of which are also reliable and sensitive to change [18-20].

●Patient and evaluator global assessments – Global disease activity rated by the patient, the evaluator, or both (and termed, respectively, the patient global assessment [PGA] and evaluator global assessment [EGA]) is assessed in a similar manner to pain using visual analog scales, numerical rating scales, or Likert scales. Measuring the PGA acknowledges the importance of patient-reported outcomes. Whereas the EGA typically integrates information from both subjective and objective variables, the PGA is considered a subjective measure, which is strongly weighted for pain [21]. PGAs are usually scored at higher levels than are EGAs because patients are more likely than clinically trained evaluators to construe functional disability as a manifestation of active disease and because medically trained evaluators have greater context for comparison with other patients with RA, while patients necessarily can only compare their current state with their own previous status.

The typical question asked of the patient for a PGA is: "Considering all the ways your arthritis has affected you, how is your arthritis today?" This assessment is presumed to also account for aspects of disease activity that might be less well appreciated by the clinician (eg, fatigue or sleep disturbances).

●Acute phase response measures – Levels of the acute phase response, as indicated particularly by measurement of the ESR and CRP, constitute the most objective measures of disease activity. Acute phase reactant levels correlate well with both clinical disease activity measurements and radiographic progression of joint damage [22-27]. The ESR and CRP are the two biomarkers used most widely to assess disease activity. These tests are widely available, relatively inexpensive, and reflective of the cascade of inflammatory events associated with active RA; the CRP, in particular, is well standardized. CRP and ESR generally increase and decrease in line with disease activity. However, neither CRP nor ESR may be elevated in approximately half of patients with active RA [28]. (See "Biologic markers in the assessment of rheumatoid arthritis", section on 'Disease activity and prognosis: Acute phase reactants' and "Acute phase reactants".)

Measurements of serum proinflammatory cytokines that drive the acute phase response, such as levels of interleukin (IL) 1 beta, IL-6, or tumor necrosis factor (TNF)-alpha, are not more useful in assessing RA activity than levels of acute phase reactants and are prone to greater variation from laboratory to laboratory.

Drug effects on the acute phase reactants may influence how well their levels reflect overall disease activity since some agents, especially those that directly inhibit proinflammatory cytokines or their signal transduction, such as IL-6 receptor (IL-6R) inhibitors or Janus kinase (JAK) inhibitors, and thus interfere with the hepatic production of acute phase reactants, may reduce or normalize the acute phase response independently of clinical improvement and as a result, may confound the assessment of disease activity and usual laboratory signs of infections. (See 'Disease Activity Score using 28 joints (DAS28)' below.)

CRP and ESR also usually increase with infections. Certain drugs, especially agents that target TNF, T-cell costimulation, B cells, IL-6 receptors, and the JAK/signal transducer and activator of transcription (STAT) pathway, may blunt the infection-induced response of acute phase reactants. Therefore, greater emphasis should be placed on the clinical evaluation rather than acute phase reactants for the assessment of infection.

●Other variables – Other variables associated with disease activity include the duration of morning stiffness, degree of fatigue, and the reduction in functional capacity. Morning stiffness is not included among the core set of variables because of its greater variability and lower sensitivity to change compared with other measures.

A variety of patient-reported instruments are available to measure levels of fatigue, including the vitality/fatigue scales that constitute part of the Short Form-36 (SF-36) (see 'The Short Form-36 (SF-36)' below). A specific measure for fatigue is the Functional Assessment of Chronic Illness Therapy (FACIT) fatigue measure [29]. More commonly, however, fatigue is assessed simply through the use of a visual analog scale.

With some preparation and minor modifications of the usual work flow in clinical settings, all of the variables comprising core datasets for RA activity assessment can be collected in a standard fashion: (1) Tender and swollen joint counts can be obtained by any health professional with appropriate training; (2) visual analog scales for PGAs and EGAs of disease activity require less than one minute to complete (figure 1); (3) laboratory assessment of the ESR, CRP, or both is inexpensive and widely available; (4) Health Assessment Questionnaire Disability Index (HAQ-DI) assessments may be completed by patients before the visit.

Role of structured composite measures for assessment — Structured composite measures are used to facilitate regular monitoring of disease activity in clinical practice and to assess disease activity in clinical trials; such measures are more reliable than individual measures. It is important that such measures capture several variables related to a single dimension, such as arthritis, and thus are unidimensional and do not mix the assessment of one disease characteristic with that of another disease dimension (such as rheumatoid nodules) into a combined score (multidimensional instruments) [30].

●Role of structured composite measures as part of monitoring – Structured assessments of disease activity and of functional status are important parts of the monitoring of patients with RA. The initial evaluation and subsequent periodic monitoring should include a quantitative composite measure of disease activity, in addition to clinical and laboratory monitoring of disease and for screening of drug toxicities. The periodic use of structured quantitative assessments of disease activity (eg, the CDAI) is complementary to the ongoing regular monitoring of disease manifestations, including extraarticular manifestations and complications of the disease; of disease progression and joint damage, including use of selected imaging studies; and of functional status and disability. (See "General principles and overview of management of rheumatoid arthritis in adults", section on 'Assessment and monitoring'.)

●Composite measures versus individual variables in assessment – Limitations inherent in the use of individual variables alone for assessment of disease activity have led to the development of the composite indices that each combine several core set of variables (table 1). The individual variables of disease activity (see 'Key individual variables used in composite measures' above) reflect major characteristics of the disease within the population of patients with RA as a whole. However, given the heterogeneity of RA, the predominance of individual indicators may be highly variable across individual patients and may even vary with time within individual patients.

As an example, some patients have little tenderness despite the presence of many swollen joints, while in others, tenderness may be more prominent; additionally, within a single patient, the extent of swelling or tenderness may fluctuate during a short time period without any major change in the overall state of disease activity. As a result, the evaluation of a single core variable (eg, the tender joint count or ESR) may not accurately reflect the full spectrum of the disease. In addition, the individual evaluation of all variables within core sets often leads to heterogeneous responses and substantial methodologic problems in clinical trials. Bringing together several variables related to the same manifestation (arthritis) has substantially improved the way disease activity is measured in RA by increasing accuracy, reliability, and robustness of the resulting composite measurement [6-8,31]. The formulas for calculating the indices vary in complexity; one disadvantage of these indices is that some (eg, the DAS28) are challenging to compute without tools designed for this purpose [12,23,32-35].

●Implications for disease management – Given their advantages as reliable measures over the disease course, baseline disease activity estimated by these composite scores, and especially disease activity after three months of therapy with disease-modifying antirheumatic drugs (DMARDs), is highly predictive of disease activity at later points in time (six months or one year) [36,37], facilitating the adoption of treatment strategies that allow rapid, dynamic changes of therapeutic agents in patients who continue to have high (or sometimes moderate) disease activity by three to six months from start of therapy.

The ability to reliably measure disease activity facilitates use of treat-to-target strategies. With few exceptions [38], clinically active RA and the processes leading to joint destruction are linked, such that damage usually progresses in the presence of active disease [38-45]. Monitoring disease activity at regular, short-term intervals and appropriate modifications of DMARD therapy to establish and maintain control of disease result in improved radiographic and functional outcomes in patients with RA. (See "General principles and overview of management of rheumatoid arthritis in adults", section on 'Tight control'.)

●Response criteria – Structured measures of disease activity and response criteria both allow for an assessment of the degree of disease activity that is present, the extent of improvement that has occurred, and whether a state of remission has been achieved. Response criteria (eg, the ACR response criteria for 20 percent improvement in disease activity [ACR20]) are more widely used for clinical trials than in daily practice. (See 'Response criteria' below.)

●Functional assessment – Additional measures, which partly reflect disease activity and partly reflect disease outcome, include measures of physical function (eg, the HAQ) and health status measures (eg, the SF-36). (See 'Assessment of physical function' below.)

COMPOSITE INDICES FOR DISEASE ACTIVITY ASSESSMENT

Choosing a composite measure — We suggest use of a structured composite measure rather than an individual variable or general clinical impression to help guide treatment decisions by clinicians caring for patients with rheumatoid arthritis (RA). We prefer the Clinical Disease Activity Index (CDAI), which is calculated as a simple numerical sum of swollen and tender joints, the patient global assessment (PGA), and the evaluator global assessment (EGA) (see 'Clinical Disease Activity Index (CDAI)' below). It can be used for clinical decision-making purposes even in the absence of information about acute phase reactants. It is easy to obtain and calculate despite the inclusion of formal joint counts (the 28 joint count). The erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP), if available, may be useful in validating the clinician's impressions based on the CDAI or other composite index of choice. Reasonable alternatives include the other indices suggested by the American College of Rheumatology (ACR) described just below.

●For patient care – Several disease activity measures have been identified by the ACR as being preferred options to choose from for this purpose in clinical practice [46]:

•Clinical Disease Activity Index (CDAI) (calculator 1) (see 'Clinical Disease Activity Index (CDAI)' below)

•Disease Activity Score using 28 joints (DAS28) with the ESR (calculator 2) and DAS28 with the CRP (calculator 3) (see 'Disease Activity Score using 28 joints (DAS28)' below)

•Simplified Disease Activity Index (SDAI) (calculator 4) (see 'Simplified Disease Activity Index (SDAI)' below)

•Routine Assessment of Patient Index Data 3 (RAPID3) (see 'Routine Assessment of Patient Index Data 3 (RAPID3)' below)

•Patient Activity Scale (PAS) II (see 'Patient Activity Scale (PAS) II' below)

The choice between these measures is based upon clinician preference. The RAPID3 and PAS II require only patient-reported data, while the CDAI includes patient and provider data. All of the measures accurately reflect disease activity; are sensitive to change; discriminate well between low, moderate, and high disease activity; have remission criteria; and are feasible to perform in clinical settings [46,47].

The authors prefer the CDAI over other measures because it is appropriate for defining all disease states, including remission; is not substantially impacted by preexisting joint damage; and it is not affected differentially by certain agents:

•Inclusion of joint counts – The use of swollen and tender joint counts in the CDAI is consistent with the recommendation of the European Alliance of Associations for Rheumatology (EULAR) and the treat-to-target task force to use measures that include joint counts [1,2]. Swollen joint counts (and also acute phase reactants) predict progression of joint damage, while patient-reported outcomes do not.

•Limitations of patient-reported data for distinguishing disease activity and damage – Patient-reported data reflect disease activity in the early disease stage but reflect, often to a large extent, the damage accrued rather than disease activity. Thus, such reports can be misleading for the assessment of patients with longstanding, destructive disease who constitute a substantial proportion of patients with RA. (See 'Health Assessment Questionnaire (HAQ)' below.)

•Lack of dependence upon acute phase response measures – The CDAI does not include an acute phase reactant, which is an advantage when using drugs that directly interfere with the acute phase response. (See 'Key individual variables used in composite measures' above and 'Disease Activity Score using 28 joints (DAS28)' below.)

•Utility for defining remission – The CDAI has been endorsed for use by the ACR and EULAR joint effort to define remission. Given that remission is the main therapeutic target in the treatment of RA, it is important to note that ACR and EULAR do not recommend using the DAS28 to define remission [48], and the US Food and Drug Administration (FDA) regards the remission definition by DAS28 as only consistent with low disease activity [49]. Additionally, with respect to patient-reported data, as noted above, remission may be difficult to classify in patients with longstanding and destructive disease [50].

●For clinical research – For clinical trials, our preferred composite measure is the CDAI given the limitations regarding the DAS28 described above and below. Use of the CDAI for clinical trials outcome measurement has several advantages: it allows determination of disease activity throughout the duration of a trial; it has validated state criteria, such as high disease activity (eg, as an inclusion criterion for trials) or remission or low disease activity (as outcomes criteria); and validated response criteria also exist with a similarity to the ACR response criteria. Because it lacks an acute phase reactant, its use in comparative trials would not lead to false results for drugs that directly interfere with the acute phase response and therefore would not lead to exaggerated remission rates with such agents [51,52]. Finally, flare criteria have been defined using the CDAI; these criteria streamline the use of the CDAI for adjudicating disease activity as discussed below.

Composites involving clinician assessment

Clinical Disease Activity Index (CDAI) — The CDAI, a simplification of the SDAI, does not require the measurement of an acute phase reactant, but otherwise uses the same measures as the SDAI (table 1 and figure 2). (See 'Simplified Disease Activity Index (SDAI)' below.)

The CDAI, which ranges from 0-76, correlates well with other disease activity scores and response criteria, as well as with progression of joint damage and functional impairment [23,35,53]. The CDAI is interpreted as follows:

●<2.8 – Remission

●2.8 to 10 – Low disease activity

●11 to 22 – Moderate disease activity

●>22 – High disease activity

The advantage of the CDAI is that it facilitates immediate treatment decisions based entirely on clinical criteria and includes assessment of the joints, the principal "target organ" in RA. This attribute is useful in clinical trials and practice, since it circumvents the potential problem of laboratory to laboratory variation in the measurement of acute phase reactants. A calculator for the CDAI is available (calculator 1).

Since the CDAI does not incorporate measures of acute phase reactants, it presumably constitutes the most valid measure across all agents, including drugs that interfere directly with the acute phase reactants (see 'Disease Activity Score using 28 joints (DAS28)' below). However, the contribution of CRP to the SDAI is only minimal [23], and therefore, these two scores give very similar results irrespective of the type of drug used.

Disease Activity Score using 28 joints (DAS28)

●DAS28 and the original DAS – The Disease Activity Score (DAS) and its widely used modification, the DAS28, which utilizes a 28 joint count, provide a global summative and continuous score for disease activity assessment [32]. The DAS28 differs from the original DAS in that it excludes the grading of joints and reduces the number of joints evaluated to 28 [12,35]; it has been widely used in practice and in clinical trials. Calculators for easily computing the DAS28 using the ESR (calculator 2) and the DAS28 using the CRP (calculator 3) are available.

The basis for the original DAS was the clinician's decision to raise or lower disease-modifying antirheumatic drug (DMARD) doses based on a largely qualitative assessment of disease activity and reflected clinical practice in approximately 1990 [32]. Several features made the DAS challenging to use in clinical practice and trial settings, leading to the subsequent development of the DAS28. First, the DAS employs the Ritchie Articular Index [33], a measure with major shortcomings, in terms of feasibility and reliability, to evaluate joint tenderness. Second, the DAS employs a 44 joint count to record the number of swollen joints. Additionally, the formula for calculating the score is quite complex, using different weights for each of the variables as well as square roots or logarithmic transformations for some variables (table 1).

●Definition of activity levels by DAS28 scores – Ranges of DAS28 scores that correspond to high, moderate, and low disease activity have been proposed (table 1 and figure 2). High disease activity relates to DAS28 >5.1, moderate disease activity to DAS28 of >3.2 to 5.1, low disease activity is regarded in the range of 2.6 to 3.2. A cutoff point for "remission" has also been proposed (DAS28 <2.6). However, depending on the treatment used, between 15 and 50 percent of patients with DAS28 scores of 2.6 (the cutpoint for remission) continue to have 2 or more swollen joints; some may even have more than 10 swollen joints yet still be categorized as in "remission" using this composite index [54-57]. (See 'Criteria for remission' below.)

●Role and impact of acute phase responses in scoring – The DAS28 can be calculated using either the ESR or CRP. Originally, these two variants were thought to be interchangeable, but this is not the case, and the cutpoints for the various disease activity states have consistently been shown to be lower for DAS28-CRP than DAS28-ESR [58,59]. Indeed, to match the remission rates found using the SDAI (see 'Simplified Disease Activity Index (SDAI)' below), the current cutpoints would need to be lowered to 2.2 for the DAS28-ESR and to 1.9 for the DAS28-CRP [58,59]. At the same time, however, the issue of residual joint activity at these adapted, lower levels that indicate remission was only mitigated but is still substantial [60].

The greater relative weight the DAS28 gives to acute phase measures can result in a disproportionately lower estimate of disease activity compared with other measures, such as the SDAI or CDAI (see 'Simplified Disease Activity Index (SDAI)' below and 'Clinical Disease Activity Index (CDAI)' above) when used to assess disease activity in patients receiving agents such as interleukin (IL) 6 inhibitors or Janus kinase (JAK) inhibitors. These agents have more pronounced effects on acute phase reactants relative to their overall clinical benefit compared with other drugs [51,61,62]. As revealed by its formula, the DAS28 also gives half as much weight to a swollen joint as it does to a tender joint, although swollen joint count reflects the level of inflammation and risk of damage progression more properly. Regarding JAK inhibitors, it is of particular interest that DAS28-CRP and DAS28-ESR become substantially dissociated; indeed, rates of DAS28 <2.6 in patients taking JAK inhibitors are three times higher when DAS28-CRP is used compared with DAS28-ESR [62] owing to the fact that CRP levels are generally normalized with the available JAK inhibitors, while the ESR is not.

Simplified Disease Activity Index (SDAI) — The SDAI, which employs five of the core set of variables (table 1), is calculated using a linear sum of unweighted, untransformed variables [35].

The SDAI is defined as the simple sum of the following:

●Tender joint count (using 28 joints)

●Swollen joint count (using 28 joints)

●PGA (0 to 10 scale)

●EGA (0 to 10 scale)

●CRP level (mg/dL)

The SDAI has been validated for use in both clinical practice and clinical trials [54,63,64]. Moreover, it was shown in 2005 to have the highest sensitivity and specificity for predicting clinicians' decisions to change DMARD therapy when compared with other composite scores [64,65] and to correlate best with sonographic outcomes. A calculator for the SDAI is available (calculator 4), although the SDAI (and the CDAI (see 'Clinical Disease Activity Index (CDAI)' above)) can readily be performed without the use of a calculator in daily practice.

Cutpoints have been established for the various activity states (table 1 and figure 2). A cutpoint of 15 for the SDAI had the best combination of sensitivity and specificity (90 and 86 percent, respectively) when compared with the treating clinicians' decisions to change DMARDs because of active disease [64]. The cutpoint for remission, an SDAI of ≤3.3, does not allow the presence of more than two joints that are swollen or tender; because of its stringency, it also constitutes the main index-based ACR/EULAR remission definition [48,66]. The established cutpoints of 11 and 26 separate low disease activity (≤11) from moderate (≤26) and high (>26) disease activity [53,54]. The SDAI ranges from 0 to about 90.

Patient-reported instruments

Role of patient-reported instruments — Several self (patient)-reported instruments are available for the assessment of RA activity. These questionnaires can be completed by the patient with minimal instruction or assistance. They do not involve a physical examination by a clinician and do not include laboratory data. These features make them appear to be more efficient in the clinic. However, there are several disadvantages to these measures. The patient-reported instrument cannot distinguish between physical limitations due to RA and those due to secondary osteoarthritis (OA) or other disorders. Moreover, they will show worse results with increasing joint damage independent of actual disease activity. Because of this raised floor in the score of patients with longstanding disease, the reversibility of patient-reported measures will be limited in these patients, who comprise most patients with RA in clinical practice, despite some improvement with newer medications and treatment strategies. Since they lack a formal joint count, their predictive value for progression of joint damage is also limited.

The RAPID3 [67] and the PAS II [68] are both among the instruments with sufficient reliability and demonstrated validity in clinical practice to be considered as options for assessment of disease activity in clinical practice, when other alternatives, including the DAS28, CDAI, and SDAI are not feasible [46]. (See 'Routine Assessment of Patient Index Data 3 (RAPID3)' below and 'Patient Activity Scale (PAS) II' below.)

Other measures used in the past have included the Rheumatoid Arthritis Disease Activity Index (RADAI) [69] and the Rapid Assessment of Disease Activity in Rheumatology (RADAR) [70]. Although the use of patient-reported indices of disease activity such as these are appealing in RA, a condition associated with large personal dimensions, these instruments are rarely used in clinical trials and almost never in clinical practice. The patient-reported instruments used most often in clinical trials are global assessments of disease activity (eg, visual analog scales), Health Assessment Questionnaire (HAQ) scores, and the Short Form-36 (SF-36). (See 'Key individual variables used in composite measures' above and 'Health Assessment Questionnaire (HAQ)' below and 'The Short Form-36 (SF-36)' below.)

Routine Assessment of Patient Index Data 3 (RAPID3) — The RAPID3, as an expansion of the RAPID, constitutes an index that is based upon patient self-reported outcomes only and is simple for the patient to complete either in clinic or at home. It includes the HAQ, physical function, pain, and PGA, all normalized to 0 to 10, counted together, and divided by 3 to yield a score on a scale of 0 to 10 [67,71].

The RAPID3 does not require a formal joint count, although swollen joint counts correlate with progression of joint damage [72,73]; nonetheless, it does provide similar quantitative information to the DAS28 or the CDAI regarding disease activity [67], especially in early disease. However, in established disease where functional impairment, to a large extent, is irreversible due to the accrued joint damage (see 'Health Assessment Questionnaire (HAQ)' below), the RAPID3 likely does not correlate well with the actual degree of clinical improvement. This view was supported by the findings in a post-hoc analysis of two clinical trials, one in patients with early RA and the other in patients with established disease [74,75]. RAPID3 remission and low disease activity states discriminated between responses to tofacitinib and methotrexate in early RA, but in the trial enrolling patients with established RA, the difference in change between tofacitinib and placebo in the RAPID3 score failed to reach statistical significance despite the statistical superiority of tofacitinib over placebo using ACR response criteria, including the ACR response criteria for 70 percent improvement in disease activity (ACR70) [74,75]. Moreover, other comorbidities, such as OA and fibromyalgia, may inflate the score apart from RA disease activity and confound its interpretation.

Patient Activity Scale (PAS) II — The PAS II is a composite index composed of a visual analog scale for pain, scored from 0 to 10; a PGA, also expressed on a visual analog scale from 0 to 10; and the HAQ or the HAQ-II [68]. The PAS II uses fewer questions than the original PAS. It has the same advantages and disadvantages as other patient self-report instruments. (See 'Role of patient-reported instruments' above and 'Routine Assessment of Patient Index Data 3 (RAPID3)' above.)

RESPONSE CRITERIA

Types of criteria — Response criteria, defined for both moderate and major changes in disease activity, have been developed for all the major rheumatoid arthritis (RA) activity assessment indices. Response criteria may involve either the comparison of a patient's activity score to a baseline value for that same patient (eg, American College of Rheumatology [ACR] response criteria) or the achievement of a certain disease activity state, such as remission or low disease activity (eg, Clinical Disease Activity Index [CDAI] or Simplified Disease Activity Index [SDAI]). Such criteria can be applied to large numbers of patients in the context of clinical trials, and also in gauging the treatment responses of individual patients over time. (See 'ACR response criteria' below.)

The SDAI and CDAI are continuous scores for which validated response criteria and disease activity state definitions exist that are useful in clinical practice and are also widely used in the reporting of results for clinical trials (see 'SDAI and CDAI response criteria' below). The SDAI and CDAI allow for monitoring of overall disease activity, as well as determination of states such as remission or low disease activity. Also, SDAI and CDAI responses are a simple calculation of change, can be done easily in clinical practice, and are even simpler than the European Alliance of Associations for Rheumatology (EULAR) response criteria (see 'EULAR response criteria' below), which require a combination of achieving a minimum change and attaining a disease state.

ACR response criteria — The categorical response criteria embodied in the American College of Rheumatology (ACR) response criteria for 20, 50, and 70 percent improvement in disease activity (ACR20, ACR50, and ACR70, respectively) measures the frequency of benefit (ie, the proportion of patients achieving a defined response from a treatment intervention). The ACR20 response, a standard measure for most clinical trials performed since 1995, is defined as improvement of at least 20 percent in the number of both swollen and tender joints, as well as at least 20 percent improvement in three or more of the five remaining core set of variables (table 2).

An example of an early attempt to define minimal response requirements was the Paulus criteria [76]. These criteria formed the basis for the ACR response criteria [77], which (in their initial iteration) derived the variables required to realize 20 percent improvement in disease activity.

●ACR20 response – The ACR20 response discriminates accurately between the effects of active medications and those of placebo. However, because of improvements in RA treatment, achievement of an ACR20 is no longer considered a clinically meaningful improvement in disease activity. Patients who achieve only ACR20 responses may still have substantial disease burdens from active RA. Moreover, the ACR20 has other potential weaknesses in some applications: (1) it does not directly measure any responses >20 percent in magnitude over baseline and therefore may underestimate the extent of improvement; (2) it characterizes the percentage of patients who meet this cutoff, rather than the response of the average patient; and (3) it measures the change in patients' disease activity compared with baseline, but does not quantify disease activity at the end of the period of interest. However, the ACR20 is still the best discriminating response criterion for clinical trials, since it peaks at 12 weeks, while the ACR50 and ACR70 require longer (ie, 20 to 24 weeks) to peak, especially for patients with early disease [78].

●ACR50 and ACR70 responses – More ambitious criteria for improvements in disease activity have been proposed; namely, the ACR50 and ACR70 responses [79], corresponding to 50 and 70 percent improvements, respectively. In contrast to ACR20 responses, patients notice dramatic differences following the achievement of ACR50 and ACR70 responses. In clinical trials of biologic agents and disease-modifying antirheumatic drugs (DMARDs), either alone or in combination, the percentage of patients achieving ACR20, 50, and 70 responses with the study agent (or combination of agents) have been on the order of 40 to 80 percent, 25 to 60 percent, and 10 to 40 percent, respectively. An ACR70 response corresponds well with achievement of a low disease activity state [80]. However, the ACR50 and 70 responses take longer to peak compared with the ACR20. Furthermore, the ACR50 and ACR70 are no better than the ACR20 at discriminating active treatment from placebo, particularly at early time points [78].

The ACR-N response measure expresses changes in the ACR core set of variables as a continuous rather than a categorical measure [35]. Although it had been used in several trials, it has not become a standard response metric for clinical trials and is no longer being used. The ACR-N judges changes in the following three variables: swollen joint count, tender joint count, and the median of the five remaining core set of variables, using the 0 to 100 percent improvement that is the smallest among these three measures. Thus, the ACR-N quantifies the patient response as a continuous measure in a single number.

Another attempt to expand the ACR response criteria was the Hybrid-ACR measure, which also is no longer used in clinical practice or research [65].

EULAR response criteria — The European Alliance of Associations for Rheumatology (EULAR) response criteria are based upon the Disease Activity Score using 28 joints (DAS28) (see 'Disease Activity Score using 28 joints (DAS28)' above). These criteria categorize improvement into either good or moderate responses (table 1) [81,82].

●EULAR good response – For a good response, the decline in score must exceed 1.2 and result in the achievement of low disease activity (ie, DAS28 <3.2).

●EULAR moderate response – A moderate response, on the other hand, may be achieved by a decline in the DAS28 by >1.2 (without reaching low disease activity); or by a decline of 0.6 to 1.2, plus reaching at least moderate disease activity (ie, DAS28 <5.1).

Comparisons of the ACR and EULAR Response Criteria indicate that moderate EULAR responses are achieved more often than ACR20 responses in most studies. Good EULAR responses are observed more frequently than are ACR70 responses. (See 'ACR response criteria' above.)

SDAI and CDAI response criteria — The Simplified Disease Activity Index (SDAI) and the Clinical Disease Activity Index (CDAI) are continuous scores for which validated response criteria exist that are useful in clinical practice and are also used increasingly in clinical trials reporting. Response cutoffs have been defined for the CDAI and SDAI that correspond with the traditional ACR responses; these definitions of minor, moderate, and major response have been defined as relative improvements of SDAI or CDAI of 50, 70, or 85 percent, respectively [83]. (See 'Simplified Disease Activity Index (SDAI)' above and 'Clinical Disease Activity Index (CDAI)' above.)

The SDAI and CDAI thus allow for monitoring of overall disease activity as well as determination of states such as remission. Also, SDAI and CDAI responses are a simple calculation of change, can be done easily in clinical practice, and are even simpler than the EULAR response criteria (see 'EULAR response criteria' above), which require a combination of achieving a minimum change and attaining a disease state.

Criteria for remission — Definitions of remission were developed jointly by the ACR and EULAR for use in clinical trials and in clinical practice [48,66] and have undergone validation in additional studies and in comparisons with other criteria [57,84-86]. In these studies, the ACR/EULAR definition has been widely validated as a better predictor of good radiographic outcomes than other measures, such as the DAS28.

These remission criteria have also been adopted by the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA) for use by pharmaceutical companies in clinical investigation for drug development.

●Remission criteria for clinical trials – For clinical trials, the ACR/EULAR remission criteria recommend use of either of the following definitions as the primary and the other as a secondary outcome measure:

•SDAI ≤3.3 (calculator 4) (see 'Simplified Disease Activity Index (SDAI)' above)

or

•All of the following (Boolean-based criteria):

-Swollen and tender joint counts (using 28 joint count) each ≤1

-Patient global assessment (PGA; on a 0 to 10 scale) ≤2

-C-reactive protein (CRP; expressed in mg/dL) ≤1

Of note, the PGA had originally been limited by a value of 1. However, as seen in subsequent analyses, among patients who missed the Boolean criteria by just 1 item, the PGA of 1 was the reason for missing remission in almost two-thirds of the patients [21]. Studies have now revealed that a threshold of 2 for the PGA was better aligned with SDAI and CDAI remission and with the ACR/EULAR index-based remission definitions, and still allowed maintenance of structural and functional integrity of remission, while deleting the PGA would result in a poorer physical function state in such a definition of remission [87]. These data were independently confirmed and, consequently, the ACR and EULAR redefined the remission definition as stated above [88].

The CDAI is also widely used for defining remission in clinical trials because it does not include an acute phase reactant; it therefore does not lead to misleading results regarding remission if agents that interfere with cytokines or their signaling independently of clinical improvement are used. (See 'Clinical Disease Activity Index (CDAI)' above.)

In a related fashion, the ACR/EULAR task force dismissed the use of the DAS28 to determine remission because of its inclusion of the acute phase response in its calculation and a large number of residual swollen joints in many patients defined as in remission by this score (see above) [48].

●Remission criteria for clinical practice – For clinical practice, either of the clinical trial criteria, modified by exclusion of the CRP, may be used:

•CDAI ≤2.8 (calculator 1) (see 'Clinical Disease Activity Index (CDAI)' above)

or

•All of the following:

-Swollen and tender joint counts (using 28 joint count) each ≤1

-PGA (on a 0 to 10 scale) ≤2

●Remission with residual disease activity – In concept, the state of remission constitutes a clinical condition in which no active disease is present. However, as noted, definitions of disease remission technically permit some degree of active disease within the calculated ranges defined by their variables as consistent with remission (table 1) [48,54,64,66,89,90]. The identification of remission is hampered practically by the presence of irreversible joint damage, which may lead to abnormalities confused with residual disease activity. Similarly, comorbidities such as fibromyalgia or osteoarthritis (OA) may confound the designation of remission. Importantly, however, scores should be used to assess the disease of interest and not applied when the scores may be significantly confounded by a comorbid condition. In such situations, alternatives such as swollen joint counts or acute phase reactant levels may have to be used, but this approach should be avoided, if possible, since these variables do not sufficiently correlate with the important outcome of physical disability and quality of life.

Because progressive joint destruction may occur and function may decline, albeit at relatively slow rates, when disease activity persists even at a low level, stringent criteria are important as a means of distinguishing remission from low disease activity [54,91-93]. In this regard, as briefly noted above, remission criteria by SDAI and CDAI, which constitute the index-based ACR/EULAR provisional definitions of remission, as well as Boolean-based criteria, are not only more stringent but also more reliable than DAS28 remission criteria, especially since they do not lead to different results with different types of therapies [48,51,66].

Criteria for flare — A definition of flare using patient-defined anchors for worsening has been proposed for the SDAI and CDAI. This proposal defines flares as an SDAI increase of 4.7 points or a CDAI increase of 4.5 points. Using this definition, fares are associated with worsening tender and/or swollen joints, joint damage, and physical function. Both the SDAI and CDAI definition of flares can be used in clinical practice and trials [78].

COMPOSITE BIOMARKER SCORING OF DISEASE ACTIVITY — A multibiomarker disease activity (MBDA) assay has been developed to measure disease activity based on a scoring system derived from blood levels of a variety of biomarkers [94]. However, it is not clear when and how to optimally use this test, which is described in greater detail separately. (See "Biologic markers in the assessment of rheumatoid arthritis", section on 'Multi-protein biomarker algorithms'.)

While the MBDA assay reflects clinical disease activity, it may be misleading in patients with infections or other comorbidities. Moreover, according to one study, the MBDA may respond differently depending upon the drug used, despite similar levels of clinical improvement; with the use of IL-6 receptor inhibitors and presumably also Janus kinase (JAK) inhibitors, MBDA testing may result in highly exaggerated responses [95]. Thus, further study needs to be done in these regards before it can be adopted in clinical practice. Since clinical assessment shows good correlation with disease activity, it remains the gold standard for clinical practice. Such biomarker tests, if sufficiently validated in various populations, may be useful for research purposes, but they do not replace clinical assessment.

ASSESSMENT OF PHYSICAL FUNCTION

Choosing a functional measure — Functional status can be quantitatively assessed and followed by use of patient-reported functional status assessment measures as a complement to the medical history and inferences drawn from clinical assessment. A number of these measures are available and have been used in both clinical practice and for research. Several versions of the Health Assessment Questionnaire (HAQ) and the Physical Function 10-item Short Form are all acceptable options:

●ACR-preferred options – The American College of Rheumatology (ACR) has identified three measures as those options most appropriate to choose from for routine clinical use. These include [96]:

•HAQ-II (see 'Health Assessment Questionnaire (HAQ)' below)

•Multidimensional HAQ (MD-HAQ) (see 'Health Assessment Questionnaire (HAQ)' below)

•Physical Function 10-item Short Form (Patient-Reported Outcomes Measurement Information System Physical Function 10-item Short Form [PROMIS PF10a]) (see 'Physical Function 10-item Short Form (PROMIS PF10a)' below)

●Options preferred by the authors – The authors' preference is the traditional HAQ (using 20 questions). It has been used in almost all clinical trials of the last decades and has proven reliable across therapies and sensitive to change. It can be filled out by patients before their visit, either online or while sitting in the waiting room. The authors also consider the HAQ-II as a useful alternative instrument. (See 'Health Assessment Questionnaire (HAQ)' below.)

Functional impairment may relate to both active rheumatoid arthritis (RA), manifested by symptoms such as pain, swelling, and stiffness of the joints, and to structural joint damage occurring as the consequence of previously active disease [50,97-100]. Disease outcome and functional outcome in RA are discussed in more detail separately. (See "Disease outcome and functional capacity in rheumatoid arthritis", section on 'Functional capacity'.)

Because active RA exerts a substantial effect on physical function, instruments designed to measure physical function are also useful indicators of disease activity. Such instruments, developed for generic (ie, not disease-specific) use, are employed commonly in clinical trials, but used less often in clinical practice.

The two physical function assessment tools used most frequently are the HAQ and the Medical Outcomes Study Short Form-36 (SF-36) [101-104]; in clinical practice, we prefer the HAQ over the SF-36, while in clinical trials, both scales are usually employed. While the SF-36 is more complex (36 items), generic, and an instrument of health status, the HAQ Disability Index (HAQ-DI) was originally developed for RA, is shorter (20 items), and is much easier to calculate.

Health Assessment Questionnaire (HAQ)

●Original (traditional) HAQ and HAQ Disability Index – The original complete HAQ is a comprehensive instrument designed to assess patient disability, discomfort, medication side effects, costs, and mortality. The HAQ-DI is the component of the HAQ that is used frequently in clinical trials and clinical practice. The HAQ-DI, a component of the ACR core dataset for the evaluation of RA disease activity and outcome, evaluates patients' ability to perform activities of daily living through their answers to 20 questions designed to assess upper and lower extremity use. These questions are organized into 8 categories: dressing, rising, eating, walking, hygiene, reach, grip, and usual activities. Each question is answered on a 4-level scale of impairment ranging from 0 to 3: 0 = no difficulty; 1 = some difficulty; 2 = much difficulty; and 3 = inability to do.

The final HAQ-DI, which ranges from 0 to 3, is the mean of scores from all 8 categories. HAQ-DI scores <0.3 are considered normal; however, the mean HAQ of the population rises with age [105]. Higher HAQ-DI scores indicate increasing disability. The minimal clinically important difference in serial HAQ-DI scores has been suggested to be 0.22 on the group level for patients with RA. Several modifications of the HAQ-DI have been employed in clinical trials and practice (see below) [103].

Although disease activity influences the HAQ-DI [98,102], and mainly so especially in early disease, the score reflects both joint damage and disease activity, with an increase of the influence of damage over disease activity with increasing disease duration and/or damage. Therefore, results of the HAQ-DI require careful interpretation when used for disease activity assessment, as the irreversible, damage-related component of disability as assessed by the HAQ-DI increases with the degree of joint damage and with disease duration [50,100]. Thus, it has been proposed that the activity-related HAQ (ACT-HAQ) should be distinguished from its damage-related component (DAM-HAQ) if one wants to distinguish between the consequences of damage and activity on physical function. Thus, the HAQ-DI level that can be reached by appropriate therapy ("floor") can be quite high due to preexisting joint damage, even if disease activity has been controlled (clinical remission).

●Modified HAQ and Multidimensional HAQ – A simplification of the HAQ-DI, the modified HAQ (M-HAQ), was undertaken because the HAQ-DI comprises many questions, and it can take several minutes for patients to complete the form. In the M-HAQ, the questions from the original HAQ-DI were reduced to one or two per category, also deriving its total score by taking the average of the eight categories [106]. A further effort led to the MD-HAQ, amending the 8-item M-HAQ to include 14 items. These additional items improved the "floor effect" that had been noticed with the use of the M-HAQ, by which patients could have normal scores despite experiencing meaningful functional limitations [103].

●HAQ-II – A modification of the HAQ, the HAQ-II was developed because the HAQ is lengthy (34 questions) and has certain psychometric issues regarding linearity of the score and clarity of the questions [107]. The HAQ-II uses 10 questions with 4 levels of difficulty each. Half of the questions come directly from the HAQ and, like the HAQ, it is transformed into a 0 to 3 scale. The HAQ-II is simpler to answer and to score than the HAQ. However, although it is available in Dutch and Spanish, it has not yet been translated and validated in other languages [108-110].

The forms for the HAQ-DI, the HAQ-II, and the MD-HAQ can be obtained from online sources, such as rheuminfo.com/physician-tools/health-assessment-questionnaires-haq-haq-ii-mdhaq/.

Physical Function 10-item Short Form (PROMIS PF10a) — The Physical Function 10-item Short Form (Patient-Reported Outcomes Measurement Information System Physical Function 10-item Short Form [PROMIS PF10a]) is a short patient questionnaire, which is an additional patient-reported option that can be used to assess functional capacity. It consists of 10 questions related to physical function that are categorized into 5 states of difficulty (not at all, very little, somewhat, quite a lot, cannot do). It can be obtained from the National Institutes of Health (NIH) website (https://www.healthmeasures.net/explore-measurement-systems/promis).

The Short Form-36 (SF-36) — The SF-36 is a patient-reported instrument designed to assess overall health status; it is usually utilized to measure patients' quality of life and used primarily in research [104]. The SF-36 is not disease specific, although it has been validated in numerous diseases, including RA, and is often measured in phase 3 clinical trials. It is very rarely used in clinical practice due to its complexity.

The instrument consists of 36 questions organized into 8 domains: physical function, physical role, general health, bodily pain, mental health, social function, vitality/fatigue, and emotional role. Data from these 8 domains can be summarized into 2 categories: a physical component score and a mental component score. SF-36 results are particularly useful for comparing quality of life across cohorts of patients with different diseases, but also correlate well with other measures of RA activity. Increasing SF-36 scores are indicative of improving health status.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Rheumatoid arthritis".)

SUMMARY AND RECOMMENDATIONS

●Disease activity assessment – Key components to the assessment of disease activity in patients with rheumatoid arthritis (RA) include the following (see 'Key individual variables used in composite measures' above):

•Swollen (SJC) and tender joint counts (TJC)

•Patient assessment of pain

•Patient global assessment (PGA) of disease activity

•Evaluator global assessment of disease activity (EGA)

•Acute phase reactants (eg, C-reactive protein [CRP], erythrocyte sedimentation rate [ESR])

●Composite measures assist clinical decision-making – We suggest that clinicians use a structured composite measure (many of which incorporate the key components listed above) to guide treatment decisions for patients with RA.

Structured composite measures of disease activity are more reliable than individual measures or general clinical impressions for assessing disease activity. Composite measures are also predictive of future disease activity and allow the early identification of patients who require treatment intensification. (See 'Role of structured composite measures for assessment' above.)

●Selection of a composite measures of disease activity – We prefer the Clinical Disease Activity Index (CDAI) (calculator 1), which is calculated as a simple numerical sum of swollen and tender joints, the PGA, and the EGA. (See 'Choosing a composite measure' above and 'Clinical Disease Activity Index (CDAI)' above.)

However, there are several reasonable alternatives to the CDAI, including:

•Disease Activity Score using 28 joints (DAS28) with the ESR (calculator 2) and DAS28 with the CRP (calculator 3) (see 'Disease Activity Score using 28 joints (DAS28)' above)

•Simplified Disease Activity Index (SDAI) (calculator 4) (see 'Simplified Disease Activity Index (SDAI)' above)

•Routine Assessment of Patient Index Data 3 (RAPID3) (see 'Routine Assessment of Patient Index Data 3 (RAPID3)' above)

•Patient Activity Scale (PAS) II (see 'Patient Activity Scale (PAS) II' above)

The CDAI, SDAI, and DAS28 involve clinician assessments; the DAS28 and SDAI require a laboratory measurement; and the RAPID3 and PAS II are patient-reported instruments. (See 'Composites involving clinician assessment' above and 'Patient-reported instruments' above.)

●Definition of remission – In clinical practice, remission is often defined as a CDAI ≤2.8. Alternatively, a patient who meets all of the following criteria may be said to be in remission:

•Swollen joint count ≤1

•Tender joint count ≤1

•Patient global assessment (on a 0 to 10 scale) ≤2

In clinical trials, remission has been defined using the American College of Rheumatology (ACR) response criteria, the European Alliance of Associations for Rheumatology (EULAR; formerly known as European League Against Rheumatism) response criteria, and the SDAI response criteria, all of which use specific cut-offs to grade the patient’s therapeutic response. (See 'Response criteria' above.)

●Assessment of functional status – Patient-reported functional status assessment measures complement the medical history and clinical assessment and provide a numeric measure of functional status that can be tracked over time. Examples of patient-reported functional status assessment measures include:

•Health Assessment Questionnaire Disability Index (HAQ-DI)

•HAQ-II

•Multidimensional HAQ (MD-HAQ)

•Patient-Reported Outcomes Measurement Information System Physical Function 10-item Short Form (PROMIS PF10a)

The last three options have been endorsed for routine clinical use by the ACR. (See 'Assessment of physical function' above.)