Article Text

A systematic literature review of strategies promoting early referral and reducing delays in the diagnosis and management of inflammatory arthritis
  1. Edith Villeneuve1,
  2. Jackie L Nam2,
  3. Mary J Bell3,
  4. Christopher M Deighton4,
  5. David T Felson5,
  6. Johanna M Hazes6,
  7. Iain B McInnes7,
  8. Alan J Silman8,
  9. Daniel H Solomon9,
  10. Andrew E Thompson10,
  11. Patience H P White11,
  12. Vivian P Bykerk12,
  13. Paul Emery13
  1. 1Department of Rheumatology, CHUM, Montreal, Canada
  2. 2Section of Musculoskeletal Disease and NIHR Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK
  3. 3Department of Rheumatology, Sunnybrook Health Sciences Centre, Toronto, Canada
  4. 4Department of Rheumatology, Royal Derby Hospital, Derby, UK
  5. 5Clinical Epidemiology Research and Training Unit, Boston University, School of Medicine, Boston, Massachusetts, USA
  6. 6Department of Rheumatology, ErasmusMC, Rotterdam, Netherlands
  7. 7Glasgow Biomedical Research Centre, University of Glasgow, Glasgow, UK
  8. 8Arthritis Research UK, Chesterfield, Manchester, UK
  9. 9Division of Rheumatology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA
  10. 10Department of Rheumatology, University of Western Ontario, London, Canada
  11. 11Division of Rheumatology, The George Washington University, Washington DC, Washington, USA
  12. 12Division of Rheumatology and Immunology, Brigham and Women's Hospital, Boston, Massachusetts, USA
  13. 13Section of Musculoskeletal Disease and and NIHR Leeds Musculoskeletal Biomedical Research Unit, University of Leeds, Leeds, UK
  1. Correspondence to Professor Paul Emery, University of Leeds, Section of Musculoskeletal Disease and NIHR Leeds Musculoskeletal Biomedical Research Unit, Leeds, UK; p.emery{at}


Background Despite the importance of timely management of patients with inflammatory arthritis (IA), delays exist in its diagnosis and treatment.

Objective To perform a systematic literature review to identify strategies addressing these delays to inform an American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) taskforce.

Methods The authors searched literature published between January 1985 and November 2010, and ACR and EULAR abstracts between 2007–2010. Additional information was obtained through a grey literature search, a survey conducted through ACR and EULAR, and a hand search of the literature.

Results (1) From symptom onset to primary care, community case-finding strategies, including the use of a questionnaire and autoantibody testing, have been designed to identify patients with early IA. Several websites provided information on IA but were of varying quality and insufficient to aid early referral. (2) At a primary care level, education programmes and patient self-administered questionnaires identified patients with potential IA for referral to rheumatology. Many guidelines emphasised the need for early referral with one providing specific referral criteria. (3) Once referred, early arthritis clinics provided a point of early access for rheumatology assessment. Triage systems, including triage clinics, helped prioritise clinic appointments for patients with IA. Use of referral forms standardised information required, further optimising the triage process. Wait times for patients with acute IA were also reduced with development of rapid access systems.

Conclusions This review identified three main areas of delay to care for patients with IA and potential solutions for each. A co-ordinated effort will be required by the rheumatology and primary care community to address these effectively.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


There is good evidence that early disease modifying antirheumatic drug (DMARD) initiation is associated with improved clinical and radiographic outcomes.1,,5 Achieving sustained remission or low disease activity is also more easily attained with early therapy.

Evidence also suggests that early management of patients with inflammatory arthritis (IA), in particular rheumatoid arthritis (RA), be primarily undertaken by an arthritis expert, usually a rheumatologist. Studies showed that RA patients were diagnosed earlier, received DMARD therapy more frequently and achieved better clinical and radiographic outcomes when managed by rheumatologists.6,,10 An UK National Audit Office report also concluded that initial management in secondary care was more cost-effective.11

Despite this, delays occur between patients seeking medical attention and assessment by a rheumatologist for treatment initiation.5 ,12,,14 Patient-related factors13 ,15 and lag times between initial clinical assessment and diagnosis have been documented as significant causes.5 ,14 ,16

For purposes of this literature review, delays to diagnosis and starting treatment have been divided into four main steps (figure 1):

  1. From the patient at symptom onset to assessment in primary care

  2. From primary care provider (PCP) to rheumatology referral

  3. From rheumatology referral to assessment

  4. From rheumatology assessment to commencement of DMARD therapy.

Figure 1

Potential strategies to reduce delays in the diagnosis and management of inflammatory arthritis.

An American College of Rheumatology-European League Against Rheumatism (ACR-EULAR) taskforce was assembled to analyse strategies addressing these delays. As initiatives promoting earlier classification17 and treatment18 were already in place, the group focused its efforts on addressing referral delays (ie, the first three steps).


A systematic literature review was performed according to the Cochrane systematic reviews guidelines19 to identify effective strategies to reduce delays in the diagnosis and management of IA, in particular RA.

Objectives were outlined according to a PICO (patient, intervention, comparison, outcome) framework.20 Patients were defined as adults with musculoskeletal (MSK) symptoms. The intervention could be any case finding strategy aiming to improve identification or referral of patients with IA. Where present, this was compared with a group where the strategy was not used. Outcome measures for improving identification or referral were documented. Medline, Embase, Cinahl and the Cochrane Library were searched for articles published between January 1985 and November 2010 and EULAR and ACR abstracts between 2007–2010 obtained. The search was done with an experienced librarian (see details of search terms and search strategy in the online supplementary material). No language restriction was used. Additional articles were retrieved by a grey literature search using the grey matters tool.21 Since the original search, an update was also performed by hand searching the literature. To identify strategies that may not be published, a survey was conducted through the ACR and EULAR societies inviting rheumatologists and health professionals to share information on strategies that they had implemented or were aware of to assist with early identification and referral of patients with IA.

Two reviewers (EV and JLN) independently screened titles and abstracts of retrieved references, reviewed full papers and independently extracted data of selected articles. Discrepancies were resolved by discussion. The aim was to include all potential strategies; a formal quality assessment of studies was not undertaken as this may have resulted in exclusion of some.


A total of 8668 articles were retrieved after initial removal of duplicate articles. Seventeen were selected for full review. Nine articles were found by hand search (four through an update following the original search), 10 from conference abstracts, 10 from grey literature and one from the survey (figure 2). Of the 47 articles included, 32 described strategies for earlier identification or improved referral and provided some outcome measure of efficacy and 15 described systems that may reduce referral delays. There were eight articles addressing step 1 (patient to PCP), 19 for step 2 (PCP to rheumatology referral) and 20 for step 3 (referral to rheumatology assessment). These are detailed in table 1 and discussed below.

Figure 2

Literature search flow chart.

Table 1

Strategies for identification and referral of patients with inflammatory arthritis

Table 2

Summary of strategies for early identification and referral of patients with inflammatory arthritis

Strategies from patient at symptom onset to primary care

Strategies at this level included community case-finding and public awareness programmes and website information.

Community case finding strategies

A community health fair screening approach using a questionnaire and autoantibody testing has been developed to identify individuals with undiagnosed IA.22 The diagnostic accuracy of the combinations of the CSQ (the Connective Tissue Disease Screening Questionnaire) and testing for rheumatoid factor (RF) and anti-cyclic citrullinated peptide antibodies (anti-CCP) to identify IA yielded maximal sensitivity, specificity and positive and negative predictive values of 95.3%, 99.2%, 71.4% and 97.7% respectively. Further analysis of this strategy on a larger cohort is awaited to determine the optimal diagnostic accuracy and cost-effectiveness.23

A simple self-administered test comprising a firm handshake, a four-finger grip around a pencil and pincer grip of a sheet of paper to detect inflammatory hand disease has also been described.24 Although easy to perform, many patients required clinical review due to the low specificity of this tool.

Two other projects were found inviting individuals with joint symptoms for screening and counselling for arthritis but no reported measures of efficacy were documented.25 ,26

Public awareness programmes

In one public awareness initiative of MSK conditions, a large bus was adapted and positioned at easily accessible places for informal medical counselling.27 The programme was accompanied by countrywide media announcements. RA was considered for the first time in 1.1% of patients who were subsequently referred to a rheumatology unit for assessment.

Internet and website information

With the growing internet access patients increasingly use the web as a source of information or means for self-diagnosis. However, no validated internet-based strategies to identify IA were found. One internet-based self-administered questionnaire with a version specific for RA provided scores based on positive answers but no information was given regarding the likelihood of having IA and no algorithm for action in terms of seeking care.28

There were two evaluations of the internet. These suggested that reading level was too high for the average reader,29 and that the majority of the sites were commercially driven.30

Strategies from primary care to rheumatology referral

Recognising patients with possible IA at the primary care level is key to early referral. Strategies at this stage included PCP and health professional education programmes, the use of patient self-administered questionnaires, an MSK screening examination, and use of referral guidelines.

PCP and health professional education programmes

Eight references were found reporting on strategies to improve PCP knowledge of IA including workshops, joint consultation with PCPs and rheumatologists, a tele-clinic and distribution of educational material. All reported some success in terms of improved awareness, knowledge and ability to detect IA and quality of the referral process.31,,38 One described a continuing medical education (CME) initiative to educate PCPs on the benefits of screening and diagnosing patients with suspected RA and co-managing them with a specialist.35 In addition, an increase in short-term knowledge, the percentage of patients with RA referred to rheumatologists increased from 37.4% to 41.8% over a 4-month period. Another programme, based on clinical practice guidelines adapted for primary care and consisting of an accredited inter-professional workshop and 6 months of activities to reinforce the learning, showed that decisions to refer early RA to rheumatology increased from 43.2% to 54.6%.39

Joint consultations between PCPs and rheumatologists also influenced referral behaviour with a decrease in referrals of patients who may be managed in primary care potentially reducing waiting lists of patients requiring rheumatology assessment.31 ,32

Patient self-administered questionnaires

Four self-administered questionnaires were found.40,,43 These were reported to be simple and easy to perform. Their use may provide an initial guide for PCPs to identify patients with IA.

Two were developed for use in the preprimary care setting.40 ,41 One, an 11 item early IA detection tool developed and validated by Bell et al,40 included questions relating to symptoms of IA, functional ability, personal and family history of RA and a diagnosis of psoriasis.44 The other, a web-based screening tool for RA, was based on patients' self-reported pattern of joint involvement, symptom duration and the absence of symptoms typical of fibromyalgia.41 ,45 A third, using seven weighted questions and a diagram to identify painful and swollen joints, was designed primarily for use by rheumatologists.42 Evaluated to identify RA according to the 1987 RA classification criteria, it was noted to be specific in excluding patients with ankylosing spondylitis, psoriatic arthritis and osteoarthritis, but less so for patients with arthralgia some of whom may have had early IA. The authors had planned to develop a new version of the questionnaire using the 2010 RA classification criteria.17

A simple measure using the ratio of the patient pain visual analogue scale (P-VAS) and a questionnaire assessing difficulty with activities of daily living (D-ADL) has also been developed to distinguish patients with IA from those presenting with non-inflammatory causes of pain.43 This has yet to be tested in an undiagnosed population.

Other potential tools

Although primarily performed by rheumatologists and other MSK specialists, screening in primary care has been used to improve the detection of MSK conditions.46 Use of the GALS (Gait, Arms, Legs and Spine) screening examination47 by physiotherapists to detect RA48 yielded a sensitivity and specificity between 50% to 77% and 75% to 100% respectively, suggesting that this may help PCPs rule out non-IA conditions and maintain priority access for patients with IA.

A computer-based program consisting of nine items (six symptom related and three based on laboratory results –RF, anti-CCP and erythrocyte sedimentation rate) has been designed to help PCPs diagnose RA with a provisional diagnosis given based on the presence or absence and severity of each item. However, this program has not yet been evaluated in terms of improving identification and referral of patients with IA.49

Referral guidelines

Many national and international societies have published IA management guidelines, most emphasising the importance of early rheumatology referral. One reference provided an evidence-based clinical guide for referral with specific elements including number of swollen joints and duration of early morning stiffness.50 Many early arthritis clinics (EACs) have used similar criteria for inclusion in their clinics.51,,53

Strategies from rheumatology referral to rheumatology assessment

Strategies addressing delays from time of referral to rheumatology assessment included triage of referrals and use of referral forms, the development of EACs and other types of rapid access services for patients who present acutely with IA.

Triage of referrals

Five papers were identified reporting the use of triage systems for referral to a rheumatologist. These were found to help prioritise appointments for patients with IA and reduce unnecessary referrals.

Two used an A to C/D grading system where A was assigned to urgent cases, including individuals with new IA, to be seen within 2–4 weeks.54 ,55 Use of standardised forms provided a checklist of essential information including clinical findings, radiographic and blood results and response to non-steroidal anti-inflammatory drugs and corticosteroids.56

Reviewing patient records prior to making an appointment was another effective form of triage.57 However, availability of patient records would be required and may not always be readily accessible in all centres.

An audit of one of the triage systems54 showed that some patients without IA were given priority clinic appointments at PCP request when the clinical information in the referral letter would have suggested otherwise. This was often on the basis of a false-positive RF. A small prospective observational cohort study found anti-CCP testing to be of benefit to prioritise clinic access in these patients.58

Referral forms

The ability to triage referral letters appropriately is dependent on the information provided by the referent. Many rheumatology referrals have been found to lack basic details leading to inappropriate patient triage.

Several groups have designed referral forms to improve the triage process. Our search retrieved four forms that have been tested — these have helped standardise and provided important information for triage.59,,62 All included elements of history, physical examination and laboratory tests and some asked for an attempt at diagnosis. Requesting referents to rate the level of urgency of their referrals was also used to guide evaluation of the referral.60

Two groups assessed the use of referral forms together with a triage system and have shown a significant increase in the ability to detect urgent referrals61 and reduce waiting time for appointments compared with triage alone.62

Triage clinics

Our search retrieved two articles describing triage clinics in which individuals were initially assessed by a PCP or health professional before an appointment with a rheumatologist was given. In one, both rheumatology nurses and PCPs performed well with sensitivity, specificity, positive and negative predictive values ranging between 87% and 92%.63 In another, a high degree of patient satisfaction was reported.64 No data on the improvement of waiting times were described.

Rapid access services

The availability of rapid access services provided an additional means for patients with IA who require urgent appointments to be seen.

Several papers on clinic or healthcare reorganisation aiming to provide early access to newly presenting patients have shown reduction in waiting times and increase in referrals for patients with IA. These included changes in appointment scheduling and designing a protocol for PCPs for certain rheumatic conditions,65 design of an early access MSK programme combining rheumatology, orthopaedics and physiotherapy resources66 and the implementation of rapid/immediate access clinics where new patients were seen within 1 day to 2 weeks of contacting/referral to the department for a brief assessment prior to appointment scheduling or further recommendation.67,,69 These clinics may prove particularly effective for early access to rheumatology services where rheumatologists are few in number.

Two groups also described rapid access services for patients presenting with acute problems of which a proportion were newly diagnosed IA — in one staff were available at all times for telephonic discussion and arrangements made for patients to be seen, and in another urgent new referrals were seen at a treatment service set up as a designated twice-weekly session.70 ,71

Early Arthritis Clinics

Many EACs have reported on patient symptom duration at first visit and proportions of patients diagnosed with IA. Two papers, both from the Leiden EAC, compared outcomes to routine care and showed that symptom duration at first PCP visit and at first rheumatology clinic visit were shorter in patients referred to EACs (6.4 vs 17.3 weeks and 14.1 vs 37 weeks, respectively).72 A similar analysis a few years later73 showed that symptom duration was considerably shorter in both groups (median (range) 4.4 weeks (0.1–87) and 17 weeks (0.1–104) at first EAC and routine care clinic visits, respectively) suggesting an increased awareness of the clinic from referring physicians and/or improvement in clinic service provision over time.


RA is a prevalent condition with important socioeconomic implications. In England it is estimated that 580 000 adults have RA; 26 000 new cases are diagnosed annually with approximately 45% of working age.50 Twenty per cent have been reported to quit working due to RA after 5 years.74

Evidence from published literature and clinical practice clearly supports the need for early diagnosis and treatment to reduce joint damage and improve clinical outcomes,2 ,50 ,75 ,76 ideally within 3 months of symptom onset.2 ,50 ,77 Further evidence has shown benefit from early management by a rheumatologist.6 ,9 ,76 However, delays on the part of the patient in seeking medical care15 and subsequent rheumatology referral and assessment are still noted. In a recent study, the median delay across the 10 European centres from symptom onset to rheumatology assessment was 6 months, with the percentage of patients seen within 3 months of symptom onset ranging from 8% to 42%.14 In some centres (Birmingham, Heraklion) patient delay was a key component while in others (Berlin, Vienna) and in Leiden5 patients presented early. In most, referral from PCP to a rheumatologist was an important contributor to overall delay, with a median of at least 2 months in seven of the 10 centres.

Our literature review and survey results found several strategies addressing these. Some were well established and had outcome measures demonstrating efficacy in improving referral. Others were less well developed and some still relatively new in their design.

To address delays from patients at symptom onset to primary care assessment, community mass case-finding strategies have been implemented though cost effectiveness still needs to be established. Education of the target population about IA prior to screening has been suggested as a key factor to optimise diagnostic accuracy of testing.22

Public awareness is an integral part of early referral15; however, little was found in terms of education of the general population. Although the internet is widely used to disseminate information, the quality of questionnaires and website information would need improvement if these were to be used to assist referral of patients with early IA.

The role of PCPs is important for early referral as the majority of people with arthritis symptoms will seek help in the first instance from a PCP48; MSK conditions have been documented as the most common reason for recurring PCP visits, comprising up to 30% of consultations.40 A study encompassing a literature review, best practice guidance and views of expert informants suggested that high quality PCP care would need to fulfil two core aspects of care and management of patients with arthritis – first early recognition of symptoms and second a prompt referral to a specialist for diagnosis, ideally within the first or second consultation.76 However, it has been reported that <50% are referred to a specialist within 3 months5 ,49 and that people with RA may visit their PCP an average of four times before being referred. Lack of knowledge and awareness and lack of incentives for payment and quality measures have been reported as factors contributing to the delays. Insidious symptom onset and atypical clinical presentations of RA may be other contributing factors.

Our review found that many guidelines confirmed the need for early referral for persons with suspected IA33 ,34 ,50 ,78 with few providing specific guidance.50 PCP and health professional education programmes, which may take the form of CME and feedback given in medical correspondence, have shown to improve knowledge and referrals of patients with IA. Education of medical students is another important factor although not specifically addressed here.

Early referral requires both parties willing and able to communicate with each other.76 Triage has been shown to prioritise referrals but is dependent on information given. Referral forms provided practical points to guide referral. These together with the use of triage systems have demonstrated improved prioritisation of referrals and decreased time from referral to rheumatology assessment. The cost-effectiveness of other measures for example, anti- citrullinated protein antibodies (ACPA); testing by PCP may be an area for research.

The establishment of EACs79 has improved access to early rheumatology assessment. Many PCPs use rheumatology guidelines and criteria for early referral. Although they vary across the different EACs, a combination of early referral criteria and point of referral to specialised care organised in an EAC or equivalent clinic appears effective. Other rapid access services have provided means for patients presenting with acute IA to be seen early.

Our literature review has its limitations. The majority of the data for this review originate from Europe and North America. Work still needs to be done in other healthcare environments to assess strategies to reduce delays.


Early rheumatology referral and treatment is vital for patients with IA. To succeed, this will require measures addressing delays at all levels — from patient at symptom onset to primary care assessment and referral to rheumatology assessment and initiation of therapy — and will require a combination of education of all involved and an adequate practice organisation to ensure rapid access for patients with IA.


The authors would like to thank Deirdre Andre for her help with the literature search and Amy Miller, Regina Parker and Caroline Pasche for their help in conducting the survey.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Funding The study was funded by the American College of Rhematology/European League Against Rheumatism.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.