Introducing effective management early in the course of inflammatory arthritis is integral to the successful treatment of people with rheumatoid arthritis and other conditions that lead to persistent synovitis, such as psoriatic arthritis.1 Early diagnosis and treatment with disease-modifying anti-rheumatic drugs (DMARDs) has been shown to lead to improvements in disease remission, response to treatment, disability and radiographic progression.2–4

The National Institute for Health and Care Excellence (NICE) has published quality standards for the management of early inflammatory arthritis (EIA).5 These quality statements are reflected in the 2013 UK best practice tariffs for people with EIA, which were introduced by the UK Department of Health and developed in association with the British Society for Rheumatology (BSR) and Arthritis Research UK. Rheumatology services received financial incentives to achieve the following targets for people referred with EIA during the first year of their care:6

1. Patients who have suspected EIA who are seen within 3 weeks of referral, are diagnosed as not having EIA and discharged within 6 weeks of referral.
2. Patients with suspected EIA who are seen within 3 weeks of referral, who have DMARD initiated within 6 weeks of referral and receive regular follow-up and monitoring (which would usually involve a minimum of 4 consultant-led follow-ups over the first year of treatment) with evidence of appropriate titration of therapy.
3. Patients who meet NICE eligibility criteria for biologics and are commenced on biologics in the first year of monitoring.

These tariffs have subsequently been updated, with the most recent criteria published for 2019/20.7 The latest iteration still recommends seeing people with suspected EIA within three weeks of referral and commencing DMARD therapy within six weeks of referral, but the criteria relating to the first year of their care is no longer included. In the absence of New Zealand standards of practice for people with suspected EIA, and given the similarities between rheumatology in the UK and New Zealand, we thought it was reasonable to assess our practice against the BSR standard.

The NICE quality standards have also formed the basis for national rheumatology audit in England and Wales, which evolved to the National Early Inflammatory Arthritis Audit (NEIAA) that has reported on data collected during 2018 to 2019. Some key findings of this audit were that the median time to first assessment of referred patients with suspected EIA was 28 days and 38% were seen within three weeks; DMARD therapy was initiated within six weeks of referral in only 54% of patients with confirmed EIA; nearly a third of patients reported significant work impairment at presentation.8

A previous audit of 128 new cases of rheumatoid arthritis from the Wellington region that were diagnosed between 2005 and 2008 found that 22% of cases were given an urgent triage allocation, and for these patients the median time from referral to the first clinic visit was 33 days (IQR 15–52). Seventy-five percent were given a semi-urgent triage allocation, with a median time from referral to clinic of 42 days (IQR 31–77).9 The Wellington Regional Rheumatology Unit (WRRU) currently aims to see patients triaged as ‘urgent’ within four weeks and ‘semi-urgent’ within four to six months. The main outcome of that audit was to encourage triage of likely rheumatoid arthritis as urgent.

Aim

The aim of this audit was to describe the care of patients with suspected early inflammatory arthritis at the Wellington Regional Rheumatology Unit, and to compare this against the 2013/2014 BSR best practice tariffs.

Data collection and methods

Patients seen in rheumatology clinics as new patients from 1 January 2015 were eligible for inclusion in the audit. Referral letters for patients were reviewed to determine whether the patient was referred with suspected EIA. For some clinics, online referrals were visible. Where this was not the case, a hard copy of the case notes was reviewed. Consecutive rheumatology clinics from 1 January 2015 were examined, with a goal of obtaining 100 patients referred because of suspected EIA. Patients seen in any of the regional rheumatology clinics were included (Hutt Hospital, Wellington Hospital, Kenepuru Hospital, Greytown Medical Centre).

Patients were included in the audit if the referrer specifically questioned whether a new diagnosis of an inflammatory arthritis was present, including undifferentiated inflammatory arthritis, rheumatoid arthritis, spondyloarthropathy or connective tissue disease with associated inflammatory arthritis. Patients were also included if the referrer described any of the following cardinal features of joint inflammation: swollen joints, red joints, early morning stiffness, symptoms relieved by activity, raised inflammatory makers or positive rheumatoid serology (rheumatoid factor or anti-CCP antibodies) in the context of joint pain.

Patients were excluded if they were seen as an inpatient consultation, if they carried a pre-existing diagnosis of an inflammatory arthritis, if the referrer was seeking advice for a condition not forming part of the inclusion criteria (including osteoarthritis, crystal arthritis, fibromyalgia or connective tissue disease without joint symptoms) or if the patient failed to attend the first appointment offered to them.

Clinical notes, including the electronic health record, from the patients referred with suspected inflammatory arthritis were examined. The key quality targets were:

• the percentage of patients seen within three weeks of referral
• the percentage of patients with confirmed early inflammatory arthritis commenced on disease-modifying therapy within six weeks of referral
• the percentage of patients with confirmed early inflammatory arthritis who received four or more rheumatology clinic appointments within the first year of contact involvement with the service.

Duration of symptoms were estimated from the information available in the clinical record and categorised as more than two years, between three months and two years and less than three months. Although the current audit standard recommends that the time between patients’ presentations to primary care with symptoms of EIA and their referrals to specialist services should be three days, we were not able to identify the date of presentation to primary care from the referral letter, so this information was not included. In any case, this is a performance standard for primary care rather than the rheumatology services.

Results

There was a total of 334 new patient referrals between 1 January 2015 and 30 April 2015. Of the 275 referrals examined, 117 met the inclusion criteria for suspected EIA. The characteristics of these patients are shown in Table 1. Referral information for an additional 59 patients was unable to be obtained (43 sets of notes were not able to be retrieved from clinical records, and 16 sets of notes were retrieved but were missing a referral letter). The reasons for excluding 158 patients who were not referred with suspected EIA are shown in Table 2.

Data relating to performance indicators for the people referred with suspected early inflammatory arthritis are summarised in Table 3. Of the 117 patients with suspected EIA, the median time from referral to clinic appointment was 11.4 weeks (IQR 6.6–13.3 weeks) (also see Appendix Figure 1). There were 61 patients (52%) who received a confirmed diagnosis of inflammatory arthritis. Most of these patients (44/61, 72%) were commenced on DMARD therapy, almost always at their first appointment (39/44, 89%); but only 16 (26%) had a DMARD commenced within six weeks of referral. Of the patients with inflammatory arthritis, 29 (48%) had four or more follow-up visits within the first year. The median number of clinics in the first year was four (IQR 3–4). A separate analysis was also carried out for patients with confirmed rheumatoid arthritis (also shown in Table 3).

Table 1: Characteristics of patients referred with suspected inflammatory arthritis (n=117).

Table 2: Reasons for excluding patients who did not meet inclusion criteria.

Table 3: Performance indicators.

With respect to the BSR best practice tariffs:

• There were four patients with suspected inflammatory arthritis who were seen within three weeks and discharged within six weeks after being diagnosed as not having early inflammatory arthritis (3.4%).
• There were eight patients seen within three weeks who were diagnosed with early inflammatory arthritis (6.8%). Six out of these eight were commenced on DMARD therapy within six weeks; for the other two, DMARD therapy was not indicated. Two out of the six patients who commenced DMARD therapy had four or more clinic visits within the first year. Therefore, according to the BSR best practice tariff, only 2/117 (1.7%) patients were seen within three weeks, commenced on DMARD and followed appropriately over the first year.
• There was one patient who was commenced on biologic therapy within the first year.

With respect to the current WRRU triage process, 25 of the 117 patients referred for suspected EIA were given an urgent triage category and were seen at a median of 3.4 weeks (IQR 2.9–4.9); 89 were given a semi-urgent triage category and were seen at a median of 12.4 weeks (IQR 10.57–13.7); two were given a low priority and were seen at 16.9 and 6.9 weeks; and for one patient the triage category was unclear (seen at 16.7 weeks). Of those triaged as urgent, 6/25 (24%) did not have EIA, and of those triaged as non-urgent, 41/92 (45%) did have EIA. When assessing against current WRRU standards, 64% of urgent referrals were seen by four weeks and 100% of semi-urgent referrals were seen within six months.

Patients with shorter symptom duration tended to be seen slightly more quickly (Figure 1, Kruskal–Wallis p=0.04). Fewer patients with longer symptom duration (>2 years, 39%) were diagnosed with inflammatory arthritis compared with patients with shorter symptoms duration (≤3 months, 61%; three months to two years, 59%; Chi square 6.07, 2 df, p=0.05).

Figure 1: Box-plot showing the distribution of time between referral and clinic appointment by duration of symptoms.

Discussion

Although timely assessment and management of patients referred for suspected early inflammatory arthritis is acknowledged to be an indicator of high-quality rheumatology care, it has been difficult to establish systems and capacity to facilitate this. In this audit of a regional rheumatology service, we found that the average time to first specialist appointment for people referred with suspected EIA was 11.4 weeks, and only 8/61 (13%) patients who were subsequently confirmed to have EIA were seen within the recommended three weeks. Furthermore, only 26% of patients with confirmed EIA were commenced on DMARD therapy within six weeks of referral.

A previous audit suggested that shorter waiting times could be achieved for referrals of people with suspected rheumatoid arthritis (RA) if they were triaged as urgent.9 We also observed this, as referrals triaged as urgent waited a median of 3.4 weeks and 64% were seen within four weeks. However, only 25/117 (21%) of referrals in the current audit were triaged as urgent, and 19/61 (31%) of those who were ultimately confirmed with EIA. Without an increase in capacity to see urgent cases, it is likely that simply changing the triage category would only increase waiting times for this group.

Furthermore, in this audit, the accuracy of the triage category to identify those ultimately confirmed to have EIA was poor, with 25% of urgent triages not actually having EIA and 45% of non-urgent triages having EIA. Better identification of EIA at the triage stage could potentially be achieved by encouraging referrers to include relevant information or to create a standardised referral template that mandates critical information.

In comparison to the 2018/19 UK national audit, the time to first specialist appointment for suspected EIA in this audit was much worse (median of four weeks versus 11.4 weeks).8 This might reflect more service capacity in the UK National Health Service, since the ratio of publicly funded rheumatology FTE to population in the Wellington region is 0.55 per 100,000,10 which is considerably less than the ratio of 0.87 per 100,000 in the 2009 UK rheumatologist workforce survey.11 In fact, the ratio of the median time to clinic between Wellington and the UK (2.85) is similar to, but not the same as, the ratio of rheumatologist FTE per 100,000 between Wellington and the 2009 UK FTE per 100,000 (1.58). This might suggest that, in addition to capacity, efficient systems (such as dedicated EIA pathways) may be important.

What is the direct effect on patient outcomes of waiting an additional seven weeks to be seen by a rheumatologist? Although this audit cannot answer that question directly, a systematic review12 described studies showing that patients with inflammatory arthritis who waited for less than 12 weeks to be seen at specialist clinics were less likely to require orthopaedic surgery (absolute risk reduction 6.5%)13 and were more likely to achieve sustained drug-free remission from symptoms (absolute risk reduction -7.9%).14 An additional seven weeks of waiting is over half of this 12-week window. Another study in rheumatoid arthritis, which attempted to further define the ‘window of opportunity’ when treatment is more likely to be successful, found that, in order increase the chances of sustained drug-free remission, the optimal time from symptom onset to first visit,  was within 11.4 weeks (95% CI 7.7–79) for anticitrullinated peptide antibody (ACPA) positive RA and 15 weeks (95% CI 9.7–48.7) for ACPA negative RA.15

Possibly the best way of reducing waiting times for people with EIA is the development of dedicated early arthritis clinics. In the 2018/19 UK national audit, patients referred via an EIA pathway were much more likely to meet the quality performance indicator of assessment within three weeks of referral (46% compared to 26%). Some additional resources and accurate triage processes are likely to be necessary to implement this.

We acknowledge some limitations that prevent strong conclusions. Some eligible patients were excluded, because of absent referral communications from the medical record (59/334, 18% of referrals could not be located). One possible reason for this is that, for the time period being studied, not all referrals were archived electronically. A significant number of the paper referrals could not be located even after requesting them from clinical records. This will be less of an issue if our audit is repeated, now that all referrals to our department are archived electronically. The sample size was small and more precise estimates of service performance would require more systematic and regular audit exercises. We did not take into account the duration of symptoms prior to presentation at a general practitioner. Nevertheless, the overall findings are similar to our previous audit, which helps confirm their veracity.

A regular national audit of rheumatology service performance, similar to in the UK, would be very helpful to confirm these findings, determine regional variations and improve the quality of rheumatology services in New Zealand.

Appendix

Appendix Figure 1: The cumulative probability plot of waiting time between referral and rheumatology clinic appointment.