NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Collaborating Centre for Chronic Conditions (UK). Rheumatoid Arthritis: National Clinical Guideline for Management and Treatment in Adults. London: Royal College of Physicians (UK); 2009 Feb. (NICE Clinical Guidelines, No. 79.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Rheumatoid Arthritis

Rheumatoid Arthritis: National Clinical Guideline for Management and Treatment in Adults.

Show details

4Referral, diagnosis and investigations

4.1. Referral for specialist services

4.1.1. Clinical introduction

The key to the early identification of recent-onset RA is the identification of synovitis. This is inflammation of the membrane that lines the inside of synovial joints (most of the joints in the body). Inflammation manifests itself in pain, swelling, heat and loss of function of the affected joint. The joints will also be stiff, as if trying to move them through resistance, especially in the morning, and sometimes in the evening also. Occasionally the joints can also be red, but this is more unusual with RA, and a single red hot swollen joint should always be treated as septic until proven otherwise. Widespread synovitis may also lead to systemic symptoms of inflammation, with malaise, fever, sweats, fatigue and weight loss. All of these symptoms and signs may be present to a lesser or greater extent in a person with recent-onset RA.

The initial trigger for rheumatoid arthritis is not known. Whatever initiates the inflammation leads to a big increase in blood flow to the joint with resultant heat (and occasionally overlying redness). Subsequently the number of blood vessels in the synovial lining proliferate markedly, perpetuating the heat of the joint. The inner lining of the joint is composed of cells called synoviocytes, and these are normally only two to three cells deep. Half of these cells produce the lubricating and nutrition-giving synovial fluid, which forms a thin viscous layer in a normal joint. In inflammation, the synoviocytes increase in size and number and the production of synovial fluid increases markedly. This combination of increase in synovial fluid and proliferation of the synovial membrane causes the joint to become swollen, so that it has a ‘boggy’ feel on palpation. A number of pain receptors exist in the soft tissues and bone around the joint, and their stimulation by the swelling and other nerve irritants leads to pain. It is the combination of pain and swelling that results in surrounding muscle weakness and loss of function.

The main priority for non-specialists is to recognise synovitis as soon as possible. This means a low threshold for detecting heat, swelling, pain, loss of function, morning stiffness, and systemic features of inflammation. In the context of rheumatoid arthritis, if some of these signs and symptoms are present in small joints, then the threshold for considering RA needs to be lowered considerably. Some authorities have suggested if squeezing metacarpophalangeal or metatarsophalangeal joints is tender, this is sufficient to trigger concern.7,14

Colleagues in primary care face a number of challenges in recognising recent-onset RA, especially when it can initially present in such a variety of ways. Many patients present with musculoskeletal aches and pains, and even evidence of synovitis, but only a small minority will progress to rheumatoid arthritis. With the incidence of idiopathic inflammatory arthritis being around 2 per 10,000 population,5,7 a GP with a list size of 2000 patients will see one new case of RA approximately every 2 years. Other challenges for general practitioners include:

  • being asked to identify relatively rare persistent synovitis in patients that need prompt interventions, from all of the other very common causes of musculoskeletal aches and pains
  • the lack of sensitive and specific diagnostic tests. Where patients develop a polyarthritis very rapidly, the clinical diagnosis and abnormal results, as well as the pain and disability, will usually result in a prompt referral to secondary care. This type of onset is in the minority, with most patients having a much more insidious onset, or a palindromic onset where the disease might wax and wane in attacks before becoming established. The majority of patients therefore present a diagnostic dilemma in the early stages of disease.
  • Laboratory results and x-rays may all be normal in the early stages, even in patients with definite RA, particularly if this is just affecting small joints. Much of the time the key skills for identification are the clinical skills of detecting synovitis, and this is not always straight forward, even for specialists.
  • Some patients with a recent onset of persistent synovitis may gain useful benefit from NSAIDs or analgesics, but this may lull both the patient and general practitioner into a false sense of security with regard to the impact the drugs are having on the disease.

A further challenge in managing patients with recent onsets of synovitis is the delay in the person with the symptoms presenting to their GP in the first place. Recent studies have suggested that this accounts for a greater part of the delay in people with RA seeing a specialist for the first time, when compared with delays once the GP has seen the person.15 People with synovitis may have accumulated damage in their joints before presenting to any member of the medical profession, and it may be that a public health awareness campaign is required to address this. However, GPs may still delay referral, because symptoms may be vague, or signs of synovitis difficult to identify, or positive responses to analgesics and NSAIDs lead to a false sense of security. Elsewhere in these guidelines, reference will be made to the need to initiate DMARDs in the 3-month window of opportunity that makes a huge difference to long-term outcomes (See section 7.3.13). This highlights the need for:

  • the general public to be aware of the symptoms and signs of synovitis so that they can attend their GP early, and
  • ongoing attempts to reinforce to GPs (and other doctors who might see recent-onset RA, such as casualty officers) the need for prompt referral.

These challenges lead to the following questions:

  • Are there any clinical patterns of symptoms or signs of inflammatory arthritis where prognostic concerns need to register and the patient be referred to secondary services promptly?
  • If such signs and symptoms are evident, is there any evidence to suggest a time-frame for such a referral?

4.1.2. Clinical methodological introduction

We looked for studies that investigated the clinical features a non-specialist should recognise in order to refer to specialist services and how quickly the referral should be made with respect to impact on symptoms, joint damage, function and quality of life in patients with a recent onset of undifferentiated inflammatory arthritis. All types of study with a UK-relevant population were selected.

Thirteen studies15–27 were found that fulfilled the criteria. All studies were methodologically sound. Some of the studies were diagnostic studies and these had an element of prognostic design; they assessed which clinical features were able to predict patients who went on to develop RA (and fulfilled the ACR criteria) at least one year after the test was performed.

Clinical features

One cohort study16 and 12 case-series (prospective)17–25,28,29 were found which fulfilled the criteria for which clinical features a non-specialist should recognise in order to refer. All studies were methodologically sound.

The cohort study16 looked at the clinical features of N=474 patients with arthritic symptoms who attended an early arthritis clinic vs a routine clinic and these patients were followed for 1 year.

The 12 case-series looked at the clinical features of patients with early inflammatory arthritis, and in some studies, patients were followed up to look at the features of those that went on to develop RA. Studies differed with respect to:

  • sample size (range: N=41 to N=903)
  • study length (range: time not mentioned, 1–8 years).

Timing of referral

Three retrospective case-series15,26,27 were found which looked at the optimum timing of referral in patients with RA. The first case-series26 looked at the effects of early vs late referral in N=200 patients. The second case-series27 looked at the effects of delay in referral and starting DMARD therapy in N=198 patients. The third case-series15 looked at delay in referral times and reasons for delay in N=169 patients.

4.1.3. Health economic methodological introduction

No health economic papers were identified.

4.1.4. Clinical evidence statements

Timing of referral (all evidence level 3)

Table 4.7. Function

One case-series15 found that patient-dependent factors (such as delay from the onset of symptoms to the assessment in primary care), leading to a delay in consulting primary care physicians, are the principal reasons for delay in patients with RA being seen by rheumatologists.

4.1.5. Summary of evidence statements

  • The ARA criteria identify patients who are likely to have persistent synovitis and a poor prognosis, but do not perform well as diagnostic criteria in recent-onset RA.17,18,20–22 The key clinical features to facilitate identification of patients who are likely to have persistent synovitis and a poor prognosis include:

    the number of joints affected (the more joints the worse the prognosis)18,22

    the presence of both swelling and tenderness in affected joints (particularly small joints)18,19

    a positive MCP squeeze test23

    the involvement of PIPs and MCPs,23 and symmetry of joints affected

  • An inability to make a fist or flex the fingers was associated with an ability to identify RA from other diagnoses in one study.19 Ever having prolonged morning stiffness is more helpful than currently having morning stiffness for early RA.23
  • Rheumatoid factor detects less than half of eventual RA patients at presentation18
  • Acute phase markers are no different in patients with an inflammatory arthritis that evolves in RA than those that evolve into non-RA23
  • Delays in referral are associated with worse function at presentation26 and if delayed by 1 year an increase in erosive changes on x-ray27
  • The most significant factor in delaying start of DMARD was delay in referral to a rheumatologist27
  • There is evidence that the greatest delay in patients presenting to specialist care is in the patients attending their GP with symptoms in the first place, rather than the delay in patients being referred by the GP to specialist care.15

4.1.6. From evidence to recommendations

The GDG felt that a diagnosis of RA should be based on clinical findings such as history and examination, with investigations sometimes being helpful. The diagnosis should not be constrained by the ACR classification criteria, which are unhelpful in recent-onset disease. The ARA criteria were considered to be better prognostic guides than diagnostic guides in early inflammatory arthritis. In other words, if a person presents with small joint disease, then this carries a poor prognosis, and should lead to urgent referral to specialist care. The same considerations apply to the number of joints affected, where evidence suggests that the greater the number, the worse the prognosis, and the greater need for early identification and referral. The most important feature of recent-onset RA is the clinical detection of synovitis. This is a clinical skill, and in early RA all blood tests may be normal despite significant disabling disease. It was therefore considered important that this message was emphasised, and that normal investigations should not put a GP off referring patients urgently to secondary care. It was acknowledged that there is evidence that much of the delay in referral to specialist care is often beyond the control of the GP, and relates to the person themselves not presenting promptly at the start of their symptoms. Given the evidence that this delay increases the risk of damage to joints, and delays introduction of DMARDs, people who present to their GP with well-established disease should be referred urgently to specialist care to try to minimise any further damage, especially if symptoms have already been present for more than three months.



Refer for specialist opinion any person with suspected persistent synovitis of undetermined cause. Refer urgently if any of the following apply:

  • the small joints of the hands or feet are affected
  • more than one joint is affected
  • there has been a delay of 3 months or longer between onset of symptoms and seeking medical advice.

Do not avoid referring urgently any person with suspected persistent synovitis of undetermined cause whose blood tests show a normal acute-phase response or negative rheumatoid factor.

4.2. Presenting symptoms and signs

4.2.1. Clinical introduction

The term RA covers a very broad spectrum of disease. For recent-onset RA some patients will have rapid onset of disease, and quick evolution to a polyarthritis. Others will have an insidious onset and perhaps months of a mono- or oligoarthritis before gradually evolving into a symmetrical peripheral pattern. Patients with recent-onset and with established disease may also have extra-articular features that impact on the severity and disability of the disease, such as interstitial lung disease and vasculitis. Others may have no obvious extra-articular disease, or relatively mild but nevertheless irritating symptoms such as dry eyes. It would be useful to be able to identify patients who are likely to have a poor prognosis early in the course of the disease. These patients could then be monitored more closely so that a lower threshold could operate for intensive intervention to modify the course of their aggressive disease. Conversely, in patients lacking poor prognostic markers, or possessing good prognostic markers, a less intensive follow-up and treatment strategy might be pursued. Are there any markers that might help to make management more targeted, depending on disease prognosis?

4.2.2. Clinical methodological introduction

We looked for studies that investigated which clinical features of RA patients (recent-onset and established disease) can be used to identify those with a good or poor prognosis. Due to the large volume of evidence on prognostic features, studies were selected which were of a UK-relevant population; if the population was mixed arthritis there had to be >75% RA or RA subgroup analysis, and a sample size N>200. We also looked for studies that assessed which treatments are the best for patients with a poor prognosis; no limits were set for selection criteria except for UK-relevant population.

Thirty-three case-series were found that fulfilled the inclusion criteria. All studies were methodologically sound and assessed the clinical features of RA patients who had either a good or poor prognostic outcome (patients were either followed prospectively or data was gathered retrospectively).

Recent-onset RA

23 case-series30–52 were found which fulfilled the criteria. They differed with respect to:

  • sample size (range: N=211 to N=1387)
  • study length (range: 1 year to 43 years)

Established RA

10 case-series 53–62 were found which fulfilled the criteria. They differed with respect to:

  • sample size (range: N=263 to N=2448)
  • study length (range: 6 months to 50 years)

Treatment of poor prognosis patients

Three RCTs63–65 and 1 cohort study66,67 were found which fulfilled the criteria for which was the best treatment for RA patients with a poor prognosis. All trials were performed on patients with UA or a recent onset of RA. The cohort study66,67 was published as two separated papers with different follow-up times and therefore the study has only been counted once. However, results from both papers are reported and referenced here.

The three RCTs63–65 were parallel group studies. The first two RCTs63,64 looked at DMARD treatment (single vs combination) in patients with a recent onset of RA who had a poor prognosis (N=82 and N=20 respectively). The first RCT63 compared two different treatment arms: sulphasalazine (SSZ) 500 mg/day vs cyclosporine A (CSA) 1.5 mg/kg/day + methotrexate (MTX) 7.5 mg/week + corticosteroid (CS) methylprednisolone in a 48-week treatment phase. The second RCT64 compared two different treatment arms: MTX 7.5 mg/week + placebo vs MTX 7.5 mg/week + infliximab (IFX) 3 mg/kg/day in a one-year treatment phase with follow-up at one year post-treatment. The third RCT65 looked at DMARD treatment in N=110 patients with UA of which N=51 went on to develop RA. The trial compared two different treatment arms: MTX vs placebo in a 30-month treatment phase and performed a subgroup analysis of the outcome of patients in each group who had a poor prognosis (anti-CCP+ or RF+). The methodological limitations of the RCTs were as follows: those graded 1+ were either unblinded, single blind or did not have an ITT (intention to treat) analysis. The trial graded 1++ was double blinded and the authors performed an ITT analysis.

The cohort study66,67 looked at which was the best treatment in N=206 patients with a recent onset of RA who had a poor prognosis. The trial compared two different treatment arms: early treatment (DMARDs + NSAIDs) vs delayed treatment (NSAIDs then DMARDs) in a four-year treatment phase.

4.2.3. Health economic methodological introduction

The two questions were combined as the search terms were very similar. 144 abstracts were found, of which 142 were not specific for rheumatoid arthritis. The remaining two were ordered as full papers but subsequently did not meet the inclusion criteria (they were not economic evaluations). One paper68 was found by a GDG member which had been published online after the searches were run. It met the inclusion criteria and so it was appraised.

4.2.5. Health economic evidence statements

No economic evaluations of a UK population were found. Konnopka et al.68 is a Dutch study that evaluates the use of anti-CCP antibody testing compared to diagnosing RA using the ACR criteria. The model is populated with data not sourced through any formal evidence review, and the Markov model used is generally poorly explained. The assumption of the impact of a late diagnosis on HAQ progression is uncertain, and when varied from 0.01–0.15 causes the ICER to vary from dominance to over €153k per QALY. The baseline results estimate an ICER of €930 per QALY, although the differences in costs between aCCP strategy and ACR strategy (€15,010 and €14,995 respectively), and the differences in QALYs between the aCCP strategy and ACR strategy (7.1237 QALYs and 7.1073 QALYs) are relatively small.

4.2.6. Summary of evidence statements

In recent-onset disease most of the studies looked at radiological outcomes over variable follow-up periods. The themes that emerge are as follows:

  • RF titre stands out as repeatedly being a good predictor of prognosis (both for radiology and function) in most studies.31,33,36,41–44,47,48,51,52
  • Anti-CCP positivity is a predictor of prognosis.44,48,52 Interactions with RF occur, so that the worst prognosis is seen for patients positive for both RF and anti-CCP, the best prognosis in those patients negative for both antibodies, and intermediate prognoses for those positive for one antibody only.46 Other variables appearing to predict prognosis in more than one study include:

    baseline radiological score,31,42,44,45,47 nodules41,42

    acute phase markers31,33,41,42,45,48,50

    HAQ score43,45,49

    grip strength33,49

    swollen joint count.33,41,48

In established disease studies looked at a mixture of functional and radiological outcomes. Recurring themes are as follows:

  • Disability is predicted by:

    baseline disability score55,57,61

    older age57–59

    longer disease duration.57–59,61

  • Women tend to do worse than men.57,58
  • No study examined whether poor prognosis patients should be treated differently.

4.2.7. From evidence to recommendations

The GDG noted that currently there is not universal availability for testing of anti-CCP antibodies. The evidence does show that anti-CCP appears to add information over and above testing for rheumatoid factor as far as prognosis is concerned, with testing positive for a combination of both rheumatoid factor and anti-CCP being associated with a particularly poor prognosis. However, until there is evidence that a poor prognosis group identified in this way might need different management, this would not justify a recommendation for the routine use of anti-CCP testing in patients who were rheumatoid factor positive. It was also noted that an anti-CCP test could be useful in people who were positive for rheumatoid factor but in whom the clinical picture was not suggestive of rheumatoid arthritis. A positive anti-CCP test in these circumstances might suggest that the individual was at risk of developing RA subsequently and therefore merited close follow-up. However, there is currently only limited data available in this group of people, and the GDG agreed that a recommendation would be premature.

The GDG agreed that the evidence suggested that the principal strength of anti-CCP testing appeared to be in people who were seronegative for rheumatoid factor and in whom intensive combination therapy would be the initial treatment (see recommendation 16). In this group of people who might be reluctant to start intensive therapy for a recent onset of persistent synovitis without further specific tests relating to diagnosis and prognosis, being positive for anti-CCP might be used to inform their decision about taking such medication. The GDG concluded that a specific consensus recommendation for this group of patients would be appropriate, but that the introduction of more widespread testing for anti-CCP in other groups would need good evidence of cost-effectiveness and that this should form the basis of a research recommendation.



Consider measuring anti-cyclic citrullinated peptide (CCP) antibodies in people with suspected RA if:

  • they are negative for rheumatoid factor, and
  • there is a need to inform decision-making about starting combination therapy (see recommendation 16).

4.3. Investigations

4.3.1. Clinical introduction

The identification of persistent synovitis is largely a clinical skill. However, there are investigations that can help to demonstrate that there are abnormalities that require intervention. Some of these may be non-specific, such as evidence of inflammation taking place (eg elevated C-reactive protein (CRP), anaemia of chronic disease), whereas others may be more helpful in pointing towards a diagnosis (eg rheumatoid factor present in high titres, erosive change on x-rays). Some investigations help to rule out other causes of polyarthritis or polyarthralgias, such as thyroid function tests. Other investigations are useful baselines that help in management, such as renal and liver function tests prior to commencing NSAIDs or DMARDs. In this section the focus is on those tests that help with early recognition of the disease, but it needs to be borne in mind that these do not replace the need for a careful history and examination, and even if all tests are normal, this should not prevent appropriate interventions from taking place in a patient with persistent synovitis. Antibodies against cyclic citrullinated peptides (anti-CCP) have emerged in recent years as being as sensitive, but more specific than rheumatoid factors in the diagnosis of RA. These are covered in section 4.3 on presenting signs and symptoms. There is evidence to suggest that ultrasound and magnetic resonance imaging (MRI) scans are superior to clinical examination in the detection of synovitis, and that they are more sensitive to the presence of erosions and other early inflammatory and damage signs than conventional x-rays.69,70 However, the long-term significance of their findings, and the limited availability to many clinicians, restricts their utility, leaving the detection of synovitis on clinician examination as the gold standard.

4.3.2. Clinical methodological introduction

We looked for studies that assessed the ability of investigative procedures to identify patients with undifferentiated inflammatory arthritis who would go on to develop RA. Due to the large volume of evidence, studies were selected which were of a UK-relevant population, patients were pre-RA/had UA, if the population was mixed arthritis there had to be >75% RA or RA subgroup analysis, and had a sample size of N>50 (except for MRI or ultrasound studies). The studies found were diagnostic studies but with an element of prognostic design; they assessed the ability of investigative tests to predict patients who went on to develop RA (fulfilled the ACR criteria) at least 1 year after the test was performed.

Two MA,71,72 3 case-control studies73–76 and 12 case-series29,35,77–86 were found that fulfilled the criteria. One of the case-control studies was published as two separate papers74,75 reporting different outcomes and so the study has only been counted once. However results from both papers are reported and referenced here. No studies were found on ultrasound. The case-series and case-control studies were included in addition to the MAs because these studies did not appear in the MAs or were published after the MA search cut-off date. Three of these studies did appear in the MAs and were included because they reported outcomes of interest that were not included in the MAs.


The first MA71 focused on all trials looking at anti-CCP tests for the diagnosis of RA and included N=68 trials with data. Of these, N=14 trials looked at investigative procedures predicting the development of RA (N=11 involved UA patients and N=3 involved RA patients who had given blood before developing RA). The MA itself was fairly well conducted, however no tests for heterogeneity were performed. Studies included in the analysis all used the same method for detecting anti-CCP antibodies (ELISA); however, they differed with respect to:

  • Type of investigative test used (N=5 studies used anti-CCP1, N=10 trials used anti-CCP2).
  • Cut-off for anti-CCP+ (anti-CCP1 range 21.4 IU to 1000 IU, anti-CCP2 3.8IU to 50 IU).
  • Study size (UA patients for anti-CCP1: N=1,327; UA patients for anti-CCP2: N=2,017; RA patients given blood before RA development for anti-CCP1: N=79 and for anti-CCP2: N=142)
  • Study duration – length of follow-up (UA patients: range 5–36 months; RA patients given blood before RA development: range <1.5 years to 9 years).

The second MA72 looked at the diagnostic accuracy of anti-CCP tests and RF tests in patients with a recent onset of RA (<1 year duration) and included N=86 trials. Of these, N=37 studies looked at anti-CCP tests and N=50 at RF tests. Trials differed in terms of:

  • Study size (range not mentioned)
  • Study design (prospective in N=18/37 anti-CCP; N=25/50 RF)
  • Study quality – maximum score of 5 (N=1 very good quality; N=22, 30% reasonable quality; N=9, 10% poorer quality)
  • Study duration – length of follow-up (range not mentioned)
  • Comparison group (mainly patients with UA; healthy patients; other diseases; other rheumatic diseases)
  • Intervention – type of anti-CCP test (anti-CCP1 N=8; anti-CCP2, N=29)
  • Intervention – type of RF test (IgM, IgA, IgG)

Case-control studies

The 3 case-control studies73–76 all looked at investigative tests which could be used to predict the development of RA in pre-RA/UA patients. The first two case-control studies73–75 looked at investigative tests (RF and antifilaggrin antibodies(AFA), anti-CCP, RF and collagen types respectively) in people who were at risk of developing RA (both N=19,072). The studies compared cases (those who went on to develop RA) with matched controls (who did not develop RA) and had follow-up times of 22 years or 12 to 16 years respectively. The third case-control study76 looked at investigative tests (RF and AFA) in N=330 patients with UA. The study compared cases (those who went on to develop RA) with controls (patients who already had RA) and had a 1-year follow-up time.


The 12 case-series29,35,77–86 which investigative tests and procedures could be used predict the development of RA in pre-RA/UA patients. Studies differed with respect to:

  • Sample size (range: N=30 to N=1,003)
  • Study length (range: 1 year to 6.9 years mean)
  • Investigative procedure/test used [RF, anti-CCP, ACPA (Anti-citrullinated protein/peptide antibodies), CRP, APF (antiperinuclear factor), ESR (erythrocyte sedimentation rate), MRI, symptoms, radiographs, histopathology, ACR criteria and others)

4.3.3. Health economic methodological introduction

No health economic papers were identified.

4.3.4. Evidence statements

Table 4.25. Rheumatoid Factor (RF)

Table 4.26. Anti-CCP

  • One MA/SR71 reported three studies which looked at patients with RA who had donated blood samples before development of RA. Level III

    One study found that anti-CCP2 predicted RA development with low sensitivity (4%, 25% and 52% at 9 years, >1.5 years and <1.5 years before symptoms) and had a high specificity (98%). OR 28 (95% CI 8 to 95).

    One study did further analysis of the same patients and found that anti-CCP2 had highest predictive value compared to RF; OR 15.9 for anti-CCP2 and 6.8 for RF.

    One study found that 5 years before symptom onset, anti-CCP1 had a low sensitivity and high specificity for predicting RA (29% and 99.5% respectively; OR 64.5 (95% CI 8.5 to 48.9).

Table 4.27. AFA (antifilaggrin antibodies)

Table 4.28. APF (antiperinuclear factor)

Table 4.29. CRP

Table 4.30. Radiographs and MRI

Table 4.31. Other

4.3.5. Summary of evidence statements

  • RF in most studies is a useful predictor of RA development.29,76–78,82,85
  • Anti-CCP positivity is a useful predictor of RA development,29,71,74,75,84 and in comparison to RF appears to have a higher specificity, but similar sensitivity.71,72
  • Baseline CRP is a poor predictor of who will go on to develop RA.74–76,83
  • AFA had a moderate sensitivity and specificity for the development of RA and was better at predicting the development of RF+ RA than RF− RA.73,77,78,82
  • APF had a moderate sensitivity and specificity for the development of RA.77,78,81,82
  • Baseline erosions on hand x-rays show high specificity but low sensitivity for the development of RA.79
  • Erosions on MCP MRI scans have a fairly high sensitivity and specificity for the development of RA.83

4.3.6. From evidence to recommendations

Although anti-CCP antibodies are more specific than rheumatoid factor, this difference is not great, and sensitivities seem very similar. Recommendations on anti-CCP also need to be informed by health economic analysis to determine whether the extra cost and increased specificity render this test cost-effective (please refer to section 4.2.7 and 4.2.8), either for all early inflammatory arthritis, or for sub-groups (eg RA suspected, but rheumatoid factor negative). Rheumatoid factor remains a relatively cheap and useful test in undifferentiated synovitis that is helpful both diagnostically and prognostically.

After much deliberation, it was decided that x-rays of the hands and feet in early synovitis are worthwhile, because although this is a blunt instrument in detecting joint inflammation, there are occasions when erosive damage will be detected when all other tests are normal, and it also acts as a readily accessible base-line for future determinations of disease progression. As ultrasound and small joint MRI become more widely available, the long-term significance of some of the early inflammatory and erosive changes that have been described using these imaging modalities should become apparent and they may replace x-rays.



Offer to carry out a blood test for rheumatoid factor in people with suspected RA who are found to have synovitis on clinical examination.


X-ray the hands and feet early in the course of the disease in people with persistent synovitis in these joints.

Copyright © 2009, Royal College of Physicians of London.

All rights reserved. No part of this publication may be reproduced in any form (including photocopying or storing it in any medium by electronic means and whether or not transiently or incidentally to some other use of this publication) without the written permission of the copyright owner. Applications for the copyright owner’s written permission to reproduce any part of this publication should be addressed to the publisher.

Bookshelf ID: NBK51814


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...