Rationale and impact
These sections briefly explain why the committee made the recommendations and how they might affect practice. They link to details of the evidence and a full description of the committee's discussion.
Evidence showed that anti-cyclic citrullinated peptide (CCP) antibodies and radiographic damage at baseline were both important prognostic factors for subsequent radiographic progression. Anti-CCP antibodies are usually measured and X-rays often taken as part of diagnosis. When this has not been done, the committee agreed that the tests should be performed as soon as possible. The results will inform discussions with the patient about how their rheumatoid arthritis (RA) might progress and reinforce the importance of active monitoring and rapidly seeking specialist care if the disease worsens.
There was limited evidence on poor function, as measured by the Health Assessment Questionnaire (HAQ), as a prognostic factor. However, the committee agreed that functional ability (measured, for example, by HAQ) should be determined at diagnosis to provide a baseline for assessing response to treatment at the annual review.
Evidence suggests that all people with RA should be offered the same management strategy; however, in the committee's experience some people may respond less well and have more progressive radiographic damage and impaired function. Because the evidence was limited as to whether people with poor prognostic markers should follow a different management strategy to improve radiographic and functional (HAQ) outcomes, the committee agreed to make a research recommendation.
Anti-CCP antibodies are usually measured so there should be no change in current practice. X-raying the hands and feet and measuring functional ability at baseline reflects current best practice, but not everyone with RA currently has these investigations. There may be an increase in the number of X-rays, especially in units without early inflammatory arthritis clinics, but this is unlikely to have a substantial resource impact.
Measuring functional ability at baseline will involve a change of practice for some providers, but the cost is low and so this is not expected to have a substantial resource impact.
Full details of the evidence and the committee's discussion are in evidence review B: Risk factors.
Ultrasound is not used widely in diagnosing RA, but use is increasing and depends on the clinic and the rheumatologist. Evidence was inconsistent and too limited for the committee to make any recommendation for or against its use in diagnosis. The committee noted that the studies generally included only people with clinically definite synovitis and agreed that ultrasound may be more useful when there is uncertainty about the diagnosis after clinical assessment. They decided to make a research recommendation to inform future guidance on who (if anyone) should have ultrasound to aid diagnosis.
Full details of the evidence and the committee's discussion are in evidence review A: Ultrasound for diagnosis.
Evidence showed that a treat-to-target strategy was more effective than usual care for managing RA and improved outcomes at no additional cost. The committee agreed that this approach was more likely to achieve rapid and sustained disease control.
No evidence was identified to indicate whether a target of remission or low disease activity was more effective. However, the committee agreed that remission (for example, a DAS28 score of less than 2.6) is the most appropriate target for most people, but for some who are unable to achieve remission despite a treat-to-target approach with appropriate escalation, low disease activity (for example, a DAS28 score of less than 3.2) is acceptable. It was agreed that for those identified as being at risk of poor prognosis, a target of remission may be more appropriate.
No studies were identified that compared different frequencies of monitoring specifically in people with active disease. The committee noted that the 2009 guideline recommended monthly monitoring and that this was used in some of the studies of a treat-to-target strategy. The committee agreed that monthly monitoring of C-reactive protein (CRP) and disease activity was most appropriate for active disease. This allows dose escalation of disease-modifying anti-rheumatic drugs (DMARDs), checking the need for short-term bridging treatment with glucocorticoids and whether people are tolerating the drug regimen, assessing side effects, providing support and encouraging adherence.
There was no evidence that people with a poor prognosis should have different management in terms of the treatment target or the frequency of monitoring. However, in the committee's experience RA often responds less well to standard management in this group. The committee agreed that the recommendations on treat-to-target with monthly monitoring should ensure that people with a poor prognosis receive effective treatment, but they decided to make a research recommendation to inform future guidance for managing RA in this group.
A treat-to-target strategy is current best practice in most NHS settings. The 2016 National Clinical Audit for Rheumatoid Arthritis and Early Inflammatory Arthritis indicated that healthcare professionals set a treatment target for about 90% of their patients. Although the 2018 recommendation specifies a target of remission or low disease activity, rather than a disease level previously agreed with the person, the committee agreed that these are the targets commonly used and so this is unlikely to involve a significant change in practice.
Monthly monitoring was recommended in the 2009 guideline, but the committee acknowledged that many clinics do not monitor active disease this often. A regional survey (Tugnet 2013) reported that about two-thirds of people with RA received monthly CRP monitoring but only a quarter had monthly monitoring of disease activity (with about 40% in dedicated early arthritis clinics) until disease control was achieved. The committee were unsure whether these rates reflected practice across England and noted that practice had improved since the survey was conducted in 2011. However, the committee agreed that monthly monitoring would likely involve a change in practice in some clinics.
Full details of the evidence and the committee's discussion are in evidence review C: Treat-to-target.
Evidence showed that starting treatment with more than 1 conventional DMARD (cDMARD) was no more effective than starting with a single cDMARD. The committee agreed that cDMARD monotherapy might have fewer side effects and recommended cDMARD monotherapy as first-line treatment. This differed from the 2009 guideline which recommended combination therapy. The difference is largely a result of inclusion of different evidence and a different approach to analysing that evidence.
Many of the studies included in the 2009 guideline used cDMARDs that are no longer commonly used in UK practice (for example, ciclosporin), and these studies were excluded from the evidence for the 2018 update. In addition, the 2018 update included new evidence published after the 2009 guideline. Further, a different approach to analysing the evidence was taken, with the 2018 update aiming to identify the most effective cDMARD strategy (monotherapy, sequential monotherapy, step-up therapy, step-down therapy or parallel combination therapy) as well as which cDMARD should be used. The 2009 guideline compared treatment strategies only, regardless of the particular cDMARDs, and combined evidence according to treatment strategy.
The evidence included in the 2018 update was therefore different to that included in 2009 and supported cDMARD monotherapy as first-line treatment.
Evidence from randomised controlled trials (RCTs) in people who had never had a DMARD showed no consistent differences in the effectiveness of methotrexate, leflunomide and sulfasalazine as monotherapies. The drugs also had similar costs. The committee agreed that any of these drugs can be used as first-line treatment.
Hydroxychloroquine was less effective, but fewer people stopped treatment because of side effects. The committee agreed that hydroxychloroquine could be considered for people with mild or palindromic disease.
Evidence for different first-line treatment in people with a poor prognosis was limited so the committee decided not to make a separate recommendation for this group. They agreed that the recommendation for dose increases and treating to target (with the aim of keeping disease activity low) should ensure adequate treatment for these people. Given the limited evidence in this area, the committee also decided that the possible benefit of managing RA with a poor prognosis with a different strategy was a priority for future research.
Evidence supported adding another cDMARD when needed (step-up strategy) rather than replacing the cDMARD with another (sequential monotherapy). The committee acknowledged that more side effects were possible with a step-up strategy, but in their experience these could be managed by drug monitoring and were outweighed by the clinical benefit of combination treatment when monotherapy was inadequate. A published cost analysis supported a step-up approach rather than sequential monotherapy.
No evidence was found for subcutaneous methotrexate, but the committee agreed that the effects may be superior and side effects fewer than with oral cDMARDs. However, because subcutaneous methotrexate is significantly more expensive than other cDMARD options, the committee was not able to recommend this without evidence of clinical benefit and cost effectiveness relative to oral cDMARDs. The committee decided to make a research recommendation to inform future guidance.
The 2009 guideline recommended a combination of cDMARDs (including methotrexate and at least 1 other cDMARD) for newly diagnosed RA and emphasised the importance of starting effective cDMARD therapy as soon as possible.
The 2009 recommendation to start with combination therapy was not widely adopted. The 2016 National Clinical Audit for Rheumatoid Arthritis and Early Inflammatory Arthritis reported that only 46% of people with RA received combination cDMARDs at any time. Currently there is variation in practice regarding the choice of cDMARD(s) and treatment strategy, with many healthcare professionals preferring to start with monotherapy and only use combination therapy when response is inadequate.
The 2018 recommendations to start with monotherapy and add drugs when the response is inadequate are unlikely to have a substantial impact on practice or resources, as they align with the current approach taken by many healthcare professionals. However, the recommendations should result in a more consistent treatment strategy and reduce the number of people prescribed combination therapy on diagnosis.
The 2009 guideline recommended methotrexate as one of the first drugs used in combination therapy. The 2018 recommendations do not specify which cDMARD should be used at any stage of treatment. Again, this will be unlikely to have a significant impact on practice, and methotrexate is likely to remain one of the most commonly prescribed drugs.
The recommendations on dose escalation and reduction have not changed substantially from the 2009 guideline and reflect current clinical practice. The committee clarified that dose reduction and the use of a step-down strategy should only be considered after a person has maintained the treatment target for at least 1 year without the use of glucocorticoids.
Full details of the evidence and the committee's discussion are in evidence review F: DMARDs.
Evidence from RCTs on the use of short-term bridging treatment with glucocorticoids to relieve symptoms while people are waiting for a new DMARD to take effect was limited. There was some evidence that fewer people withdrew from the studies due to inefficacy or adverse events when they were taking glucocorticoids, although there was no evidence that glucocorticoids were effective in terms of disease activity score, quality of life or function, as studies did not report these outcomes. In the committee's experience people with active arthritis may benefit from the anti-inflammatory effects of glucocorticoids. However, for others with less active disease this additional treatment may not be needed. The committee agreed that short-term glucocorticoids could be considered on a case-by-case basis.
Because of the lack of good quality evidence, the committee decided to make a research recommendation to determine the effectiveness of short-term glucocorticoids for adults taking a new DMARD, including the most effective regimen.
Most healthcare professionals offer short-term bridging treatment with glucocorticoids to adults starting a new DMARD. They can continue to offer this but the recommendation encourages them to consider whether this additional treatment is always needed. Therefore this is unlikely to result in additional spending for the NHS.
Full details of the evidence and the committee's discussion are in evidence review H: Glucocorticoids.
Evidence suggested that non-steroidal anti-inflammatory drugs (NSAIDs) may offer a small benefit in relieving symptoms for adults with RA (including pain and stiffness). The committee agreed that this was likely to outweigh the increase in gastrointestinal adverse events associated with NSAIDs. To minimise adverse events, the committee agreed that NSAIDs should be used at the lowest doses and for the shortest possible time, with a proton pump inhibitor, and that risk factors for adverse events should be reviewed regularly. The recommendations for analgesic treatment in this guideline replace those in the 2009 guideline.
There was limited evidence on paracetamol, opioids and tricyclic antidepressants and no evidence for nefopam, gabapentinoids or selective serotonin reuptake inhibitor (SSRI) and SSNRI antidepressants. The committee acknowledged that the 2009 guideline had recommended analgesics other than NSAIDs for pain control. However, the 2009 guideline indicated that the evidence on analgesia other than NSAIDs was 'sparse'. No further evidence on these drugs was identified since the publication of the 2009 guideline. The committee for the 2018 guideline decided to make a research recommendation rather than a practice recommendation on analgesia other than NSAIDs.
Current practice regarding the choice of analgesic is variable, with paracetamol, compound analgesics and NSAIDs all commonly used to control symptoms. Choice of analgesic tends to be based on individual effectiveness as well as the person's risk profile, tolerance, and side effects. In particular, there are some groups of people for whom NSAIDs are unsuitable because of contraindications, comorbidities or tolerability, and other people who are currently benefiting from analgesic drugs other than NSAIDs. The current approach is likely to continue but there may be an increase in prescribing of NSAIDs instead of other analgesic drugs for people with newly diagnosed RA.
Full details of the evidence and the committee's discussion are in evidence review G: Analgesics.
No evidence was identified on monitoring frequency once the treatment target has been achieved. However, the committee agreed that once people with RA had achieved the treatment target, and this was sustained at a 6-month follow-up appointment, there was no need for additional routine appointments to be scheduled other than the annual reivew. All people with RA should have an annual review.
In people with established RA (RA for at least 2 years), the evidence suggested that patient-initiated rapid access and scheduled medical review every 3 to 6 months were similarly effective. The committee agreed that all adults with RA should have rapid access to specialist care for disease flares, and ongoing drug monitoring.
Randomised controlled evidence did not support using ultrasound for routine monitoring of RA. However, in the committee's experience ultrasound can be useful for monitoring when clinical examination is inconclusive or is inconsistent with other signs of disease activity (for example, pain or markers of inflammation). The committee decided to make a research recommendation to inform future guidance about using ultrasound in these situations.
The frequency of monitoring and review appointments for people who have reached the treatment target vary around the country, with some people being seen more often than needed and others not receiving adequate follow-up. The 2018 recommendations are likely to reduce unwarranted variation.
Most people with RA currently have rapid access to specialist care when they have a flare. The 2016 National Clinical Audit for Rheumatoid Arthritis and Early Inflammatory Arthritis reported that 92% of people had access to urgent advice, with 97% of providers running a telephone advice line. Therefore the recommendation will not affect current practice.
Use and availability of ultrasound varies widely across the country and even between healthcare professionals in the same department. Some healthcare professionals use it routinely whereas others use it on a case-by-case basis. The recommendation should reduce the overall use of ultrasound while still allowing its use for selected subgroups.
Full details of the evidence and the committee's discussion are in evidence review E: Frequency of monitoring.