SEARCH

SEARCH BY CITATION

Keywords:

  • Juvenile idiopathic arthritis;
  • Access to care;
  • Management

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Objective

To document pathways of care, management, and interval from onset of symptoms to first pediatric rheumatology multidisciplinary team (PRhMDT) assessment for children with incident juvenile idiopathic arthritis (JIA).

Methods

We conducted a retrospective observational study of children with incident JIA over a 3-year period.

Results

The study included 152 patients with JIA (87 females). The median interval from symptom onset to first PRhMDT assessment was 20 weeks (range 0–416), with significant variation between JIA subtypes (P = 0.0097); children with extended oligoarticular JIA had the longest interval (median 60 weeks, range 12–320). Prior to pediatric rheumatology assessment, many children had referrals to multiple secondary care specialties and had been subjected to multiple and often invasive procedures including arthroscopy/synovial biopsy (18 [11.8%] of 152), but none were referred for ophthalmologic screening, physical therapy, or nursing input and a diagnosis of JIA was rarely made (98%). At first PRhMDT assessment, most patients had untreated active disease with active joint count ≥1 (135 [89%] of 152), restricted joint count ≥1 (135 [89%] of 152), and functional disability by Child Health Assessment Questionnaire score >0 (53 [68%] of 118); 1 child had undetected uveitis. Following PRhMDT assessment, interventions were invariably indicated, including joint injections (69 [45%] of 152), methotrexate (49 [32%] of 152), and physical therapy programs (all patients).

Conclusion

Delay in access to pediatric rheumatology assessment is common with complex pathways of referral. Many children were subjected to inappropriate invasive investigations and many had prolonged untreated active disease at the initial PRhMDT assessment. This delay is likely to affect long-term outcome and warrants further exploration.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Optimal management of juvenile idiopathic arthritis (JIA) requires specialist pediatric rheumatology multidisciplinary team (PRhMDT) care. Referral to an experienced team is advocated as soon as the diagnosis is suspected (1) and a recent UK standard of care for adults and children states that patients are to be assessed by 4 weeks from the referral date (2). The rationale for early referral is based on studies demonstrating that long-term outcome is often poor (3–7), there is a window of opportunity early in the disease course to alter the natural history of the disease process (8, 9), evidence supporting earlier and more aggressive medical intervention (10–12), and the benefits of regular eye screening (13, 14).

JIA constitutes a heterogeneous group of diseases with variable clinical presentations and outcome. Children may present initially to hospital services (i.e., secondary care) in various guises and to different specialties including general pediatrics, accident and emergency, and orthopedics. Delay in access to PRhMDT care has been reported in Australia, with children referred from orthopedics having the longest interval in presentation (15). A similar phenomenon of delay in access to appropriate care has been reported in adults with rheumatoid arthritis (16, 17), despite evidence to support better outcome with earlier intervention (16–18).

The Regional Paediatric Rheumatology (PRh) service in Newcastle, UK operates a “fast track” service for children with suspected inflammatory joint disease, with most patients being assessed within 2 weeks of the referral letter being received (19). Anecdotally, however, we have observed prolonged intervals from onset of symptoms to receiving a PRh referral. The goal of this study was to document the pathways of care and treatment of children with incident JIA, from the onset of their symptoms to first assessment by the PRhMDT, with the objective of identifying where delays in access to PRh care may occur.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

All consecutive patients with incident JIA presenting to the Regional PRh service in Newcastle, UK over a 3-year period from 2000 were included. All case notes and referral letters were reviewed retrospectively to record clinical features at initial presentation, investigations, management, health care contacts and the specialties involved, dates of onset of musculoskeletal (MSK) symptoms, documentation of a diagnosis of JIA, and the source and date of the referral to PRh. At the first PRh appointment with the pediatric rheumatologist, routine clinical assessment documented JIA core outcome variables (20) including active and limited joint counts, functional disability (using the Child Health Assessment Questionnaire [C-HAQ]), acute-phase reactants, physician global assessment score of severity (using a 0–100-mm visual analog scale [VAS]), pain scores, and parent or patient global assessment scores (VAS). These data were recorded on a pseudoanonymized Access database, compliant with good clinical practice (21), and were analyzed using Graphpad Prism software (GraphPad, San Diego, CA). This study had audit registration with the Newcastle Hospitals National Health Service (NHS) Trust and was exempt from ethical approval because it was a case note audit of current clinical practice in comparison with standards of care for inflammatory arthritis (2), and because there was no contact with families or health care professionals involved in the pathways of care.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

The Regional PRh service covers a childhood population of 500,000 and receives >500 new patient referrals per year, of whom ∼50 have a diagnosis of JIA. Of the 152 children in the study, 87 (57%) were female, the median age at the time of diagnosis was 9 years (range 1–18 years), and the median age at onset of symptoms was 8 years (range 0.5–14 years) (Table 1). The suspected diagnosis of inflammatory arthritis or JIA was documented in the referral letter to the PRh service for 3 (2%) of 152 patients; in all other cases JIA was confirmed at the time of the initial PRh assessment (150 [98%] of 152 patients). Children were classified at the time of presentation to PRh services using International League of Associations for Rheumatology (ILAR) criteria (22), and the majority of children had a diagnosis of oligoarticular JIA (51%), with the least common subtypes being systemic-onset JIA, psoriatic arthritis, and enthesitis-related arthritis (Table 2). Several children (10.5%) were classified as undifferentiated because they did not satisfy criteria for given subtypes or because they had features of >1 subtype.

Table 1. Demographic and referral details of incident cases of JIA*
DetailsValue
  • *

    Values are the median (range) unless otherwise indicated. JIA = juvenile idiopathic arthritis; PRh = pediatric rheumatology.

Number of patients152
Female sex, no. (%)87 (57)
Age at diagnosis, years9 (1–18)
Age at onset of symptoms, years8 (0.5–14)
No. of health care contacts (primary or secondary care) prior to PRh first assessment3 (0–5)
Interval from onset of symptoms to first PRh assessment, weeks 
 Source of referral to PRh 
  General pediatrics (n = 85 [56%])16 (0–416)
  Orthopedics (n = 36 [24%])24 (4–260)
  Primary care (n = 17 [11%])12 (0–320)
  Adult rheumatology (n = 10 [7%])30 (0–260)
  Other (n = 4 [3 plastic surgery, 1 clinical genetics] [3%])60 (36–364)
Table 2. Interval from symptom onset to first pediatric rheumatology assessment, by subtype in patients with juvenile idiopathic arthritis*
JIA subtype (by ILAR classification [22])Median interval from onset to first PRh assessment weeks (range)Patients seen >10 weeks from symptom onset n (%)
  • *

    Statistical comparison of subtype and delay >10 weeks (Kruskal-Wallis) P = not significant. JIA = juvenile idiopathic arthritis; ILAR = International League of Associations for Rheumatology; PRh = pediatric rheumatology; RF = rheumatoid factor; PsA = psoriatic arthritis; ERA = enthesitis-related arthritis.

Total patient group (n = 152)20 (0–416)112 (74)
Systemic-onset (n = 8 [5.2%])6 (2–36)2 (25)
Oligoarticular (n = 78 [51%])20 (0–416)60 (77)
Extended oligoarticular (n = 7 [4.6%])60 (12–320)7 (100)
RF positive polyarticular (n = 3 [2%])16 (13–28)2 (67)
RF negative polyarticular (n = 21 [13.8%])32 (4–364)19 (90)
Juvenile PsA (n = 8 [5.2%])53 (24–260)6 (75)
ERA (n = 6 [3.9%])10 (8–70)4 (75)
Undifferentiated (n = 16 [10.5%])24 (4–364)10 (63)

Interval from onset of symptoms to pediatric rheumatology care.

The median time from onset of symptoms to the first PRh appointment was 20 weeks (range 0–416 weeks) (Table 1). The waiting time to the first PRh clinic appointment (from the date of the letter of referral being received by the PRh office) was 10 days (range 0–24 days), demonstrating compliance with the UK standard of care (2). The total interval from onset of symptoms to the first PRh assessment was significantly different between the JIA subtypes (Kruskal-Wallis P = 0.0097), with patients with extended oligoarticular JIA having the longest interval and patients with systemic-onset JIA having the shortest interval (Table 2 and Figure 1). Notably, there was a striking variation within subtypes, exemplified by the oligoarticular-onset subtype (median interval 20 weeks, range 0–416 weeks). The majority of children had a delay >10 weeks from onset to PRhMDT care, although the difference in interval between the subtypes was not statistically significant (Table 2).

thumbnail image

Figure 1. The interval from onset of symptoms to first pediatric rheumatology assessment (given as median with ranges). Subtypes of JIA were by ILAR criteria (22). JIA = juvenile idiopathic arthritis; SOJIA = systemic-onset JIA; oligo = oligoarticular JIA; ext oligo = extended oligoarticular JIA; poly RF+ = rheumatoid factor positive polyarticular JIA; poly RF− = rheumatoid factor negative polyarticular JIA; PsA = juvenile psoriatic arthritis; ERA = enthesitis-related arthritis.

Download figure to PowerPoint

Presenting musculoskeletal features.

The MSK features at presentation were joint swelling (70%), pain (64%), joint stiffness (33%), limp or non–weight bearing (26%), and fever (5%). Many children had multiple MSK symptoms at presentation (Figure 2), with pain and joint swelling being the most common (49%). There was no documentation of rash in this cohort. All patients with fever (n = 7) presented to secondary care rather than to primary care services. The most common documented initial joint involvement was the knee (60% of patients), followed by foot and ankle (21%), hand and/or fingers (15%), wrist (7%), hip (3%), elbow (1%), and neck (1%).

thumbnail image

Figure 2. Documented presenting musculoskeletal features in incident cases of patients with juvenile idiopathic arthritis.

Download figure to PowerPoint

Referral pathways.

All patients were assessed within the UK NHS, with no referrals within the private health care sector. A median of 3 (range 0–5) different specialties were involved prior to the first PRh assessment (Table 1), mostly general pediatrics (56%) and orthopedics (24%), and many patients were referred back and forth between primary and secondary care specialties (Figure 3). The first documented health care contact was primary care (59%) with the remainder presenting to secondary care specialties (namely, accident and emergency, orthopedics, general pediatrics, or adult rheumatology). The majority of children were referred to PRh by secondary care specialties (89%) and predominantly by general pediatrics or orthopedics. The interval from onset of symptoms to PRh initial assessment varied by source of referral (Kruskal-Wallis P = 0.0046), with the median interval for patients referred from orthopedics (36 of 152) being 24 weeks (range 4–260). A few patients were referred from other specialties and had even longer intervals (plastic surgery: n = 3, median interval 60 weeks, range 36–364, and clinical genetics: n = 1, interval 320 weeks).

thumbnail image

Figure 3. Care pathways from onset to pediatric rheumatology. GP = general practitioner; A&E = accident and emergency; peds = general pediatrics; ortho = orthopedics; rheum = adult rheumatology; other = hospital specialties other than general pediatrics, orthopedics, and adult rheumatology; ped rheum = pediatric rheumatology.

Download figure to PowerPoint

Management prior to initial pediatric rheumatology assessment.

Documented investigations and interventions prior to initial PRh assessment are shown in Table 3. Most patients (72%) had documentation of blood tests, often repeated on different occasions. Six (3.9%) children had an autoantibody screen (including antinuclear factor and rheumatoid factor); the median interval from onset to PRh assessment for these patients was 28 weeks (range 12–208 weeks). Serum ferritin was assessed in 4 (3%) of 152 patients, all of whom had fever and MSK symptoms. The majority of children had MSK imaging, 28 (18%) of 152 had arthrocentesis, and 18 (12%) of 152 had arthroscopy/synovial biopsy (invariably of a knee). None of the patients received intraarticular corticosteroid injections. Notably, only 22 (14%) of 152 patients were documented to have ever received analgesia or nonsteroidal antiinflammatory drugs (NSAIDs). Five patients were started on methotrexate (MTX), all of whom had presented initially to adult rheumatology. None of the children had been referred for ophthalmologic screening or physical therapy input.

Table 3. Case notes documentation of investigations and interventions prior to pediatric rheumatology assessment*
 No. (%)
  • *

    ANA = antinuclear antibody; RF = rheumatoid factor; CT = computed tomography; MRI = magnetic resonance imaging; NSAIDS = nonsteroidal antiinflammatory drugs; DMARDs = disease-modifying antirheumatic drugs.

Blood tests (including full blood count, acute-phase reactants)109 (72)
Autoantibody screen (ANA/RF)6 (4)
Ferritin4 (3)
Radiographs92 (61)
Bone scan6 (4)
CT scan2 (1)
MRI scan26 (17)
Ultrasound of joint(s)33 (22)
Joint aspiration and arthroscopy/biopsy18 (12)
Joint aspiration alone10 (7)
Referred for ophthalmologic screening0 (0)
Referred for assessment by physiotherapist/occupational therapist0 (0)
Pescribed analgesics/NSAIDs22 (14)
Prescribed DMARds5 (3)

Baseline assessment at initial pediatric rheumatology assessment.

At baseline assessment by the pediatric rheumatologist, most children (89%) had evidence of active joint disease or limited joint movement in ≥1 joint (including 2 with micrognathia) with functional limitation and significant pain experience (illustrated by C-HAQ and pain scores, respectively) (Table 4). All patients were then referred for ophthalmologic screening (1 patient was found to have asymptomatic uveitis) and complete assessment by the physical and occupational therapists; the majority (87.5%) were started on an outpatient rehabilitation program. All families received education and support from the clinical nurse specialists. Most patients required medical treatments: 113 (74%) of 152 were prescribed NSAIDs, 69 (45%) of 152 were scheduled for intraarticular steroids, 49 (32%) of 152 were started on MTX (oral in 33, parenteral in 16), and 17 (33%) of 152 were prescribed corticosteroids (parenteral in 6). The median interval from onset of symptoms to starting MTX was 32 weeks (range 2–212).

Table 4. Clinical features at first pediatric rheumatology assessment*
Core outcome variables (20)Patients with data available, no. (%)Median score (range)No./total no. (%)
  • *

    C-HAQ = Child Health Assessment Questionnaire; ESR = erythrocyte sedimentation rate.

  • Recorded on 0–3 scale where 3 = severe functional disability and 0 = no disability.

  • Global assessment scores and pain score recorded on 0–100-mm visual analog scale where 100 = severe disease and severe pain, respectively.

Active joint count152 (100)2 (0–55) 
 Active joint count ≥ 1  135/152 (89)
Restricted joint count152 (100)2 (0–55) 
 Restricted joint count ≥ 1  135/152 (89)
C-HAQ score118 (78)0.5 (0–2.625) 
 C-HAQ score >0  53/118 (68)
ESR, mm/hour108 (71)25 (0–140) 
 ESR >10  55/108 (51)
Pain score (0–100) as recorded by parent/child114 (75)25 (3–100) 
 Pain score >0  41/114 (55)
Parent/patient global assessment score (0–100)114 (75)20 (0–92) 
 parent/patient global assessment score >0  89/114 (78)
Physican global assessment score (0–100)114 (75)20 (0–90) 
 Physician global assessment score >0  86/114 (76)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Despite the availability of a Regional PRh service in the northeast of England for many years (23), this study demonstrates that delay in access to PRh services remains common and for many children results in a prolonged period of untreated active disease, which is likely to have an adverse impact on prognosis. Similar delays are reported elsewhere including Australia and North America (15, 24).

Our methodology included all incident cases of JIA and, with complete case notes retrieval, indicates an accurate reflection of current clinical practice. However, the study was observational and retrospective, relying on accuracy of case note and referral letter documentation; we did not have ethical approval to interview families or obtain medical records from other hospitals or primary care services.

The case mix of this cohort was typical for a Regional PRh service in the UK, Europe, and North America (25–27), with the majority of children having oligoarticular-onset JIA. The relatively high number of children with undifferentiated JIA may be explained by the fact that subtypes were assigned at presentation to PRh services rather than following the recommended minimum of 6 months of observed disease duration (22).

The UK standards of care for inflammatory arthritis (2) propose that patients are assessed by a specialist team within 4 weeks of referral. This interval is primarily a reflection of waiting times to first clinic visit and our cohort was compliant with this standard, although many were referred from specialties other than primary care. However, for the majority of children the interval from onset of MSK symptoms to PRh assessment was >10 weeks (median 20 weeks), which we regard as unacceptable and which suggests that the problem lies in the process leading up to instigation of the referral.

There was a significant difference in delay between JIA subtypes, with children with systemic-onset JIA having the shortest interval, a predictable observation given that such children are profoundly unwell and the differential diagnosis includes malignancy and infection. Children with extended oligoarticular JIA, juvenile psoriatic arthritis, and enthesitis-related arthritis had the longest interval; in these cases the child may not be systemically unwell and the course may be indolent with physical signs that are subtle (e.g., small joint involvement of the hand or foot) or unrecognized (e.g., enthesopathy), resulting in JIA not being suspected and the delay contributing to the reported adverse prognosis (12, 28, 29). Joint swelling (with or without joint stiffness or limp) was the most common documented presentation and the knee was the most commonly affected joint. Pain is not a good predictor of inflammatory joint disease in children (30) and more than one-third of children in this study had no documented pain, an observation that is reported elsewhere for JIA (31). Furthermore, there was no documentation of rash in this cohort; it is well recognized that the rash of systemic-onset JIA is often overlooked (32) and the rash of psoriasis may develop after the onset of arthritis (33).

In most cases, JIA was diagnosed at the initial PRh assessment and was rarely documented in the referral letter to the PRh. The nature of this study does not allow us to know if JIA was suspected albeit not documented. We could assume that autoantibody assessment might imply suspicion of an autoimmune phenomenon, but in this cohort autoantibodies were rarely requested and there was no impact on the interval to PRh assessment. We suggest therefore that JIA was not considered by the health care professionals to whom these children presented, and referral to PRh was made in the absence of an alternative definitive diagnosis.

The longest interval from onset of symptoms to PRh care was noted in those children referred by orthopedics, which has been noted previously in Australia (15). The reasons for this are unclear. Many children were subjected to multiple referrals, and even among patients ultimately referred from primary care, many had been previously assessed by different secondary care specialties with referrals back and forth (Figure 3), contributing to the delay. The care pathways prior to PRh resulted in many children having multiple investigations including invasive and potentially harmful procedures. In the absence of trauma or sepsis, JIA is the most likely diagnosis in a child presenting with a monarthritis, and arthroscopy and synovial biopsy are rarely warranted. Imaging such as computerized tomography and isotope bone scans incur considerable radiation exposure; magnetic resonance imaging and ultrasound are often used but the diagnosis of JIA can usually be made without these tests. Blood tests must be justified clinically, particularly because the volume of blood obtained may be small. Serum ferritin is useful in the differential diagnosis of the unwell child (34, 35) and hyperferritinemia suggests a diagnosis of systemic-onset JIA.

The diagnosis of JIA rests on a competent MSK history and clinical examination, the persistence of physical signs for at least 6 weeks, and exclusion of other diseases by specific investigations. Referral for PRh assessment may help with the diagnostic process, investigation, and management (36). We recommend early referral to PRh for children with unexplained MSK signs and symptoms, in the absence of trauma, sepsis, or malignancy, that have not settled by 2–3 weeks.

The current UK standards of care recommend early involvement of the PRhMDT to provide a coordinated program of disease education, support, and rehabilitation (2). Notably, none of these children had had any documented involvement of nurse specialists or therapists prior to the PRh referral, despite the majority of children having evidence of joint restriction and functional disability. Many children required medical interventions including joint injections and disease-modifying antirheumatic drug (DMARD) therapy. Such delay in appropriate treatment, particularly DMARD therapy, is likely to have a detrimental effect on long-term prognosis (10–12, 29). The finding of micrognathia in 2 patients emphasizes chronicity prior to PRh assessment and is likely to contribute to further morbidity (37, 38). None of the children with JIA in this cohort had been referred for eye screening prior to the PRh referral. Oligoarticular onset is the most common presentation of JIA (typified by this cohort) and carries the highest risk of chronic anterior uveitis with a worse prognosis if detected late, especially if the child presents with visual symptoms (14). Therefore, all children with suspected JIA should be referred promptly to PRh care to confirm the diagnosis, access the PRhMDT, and facilitate eye screening.

The reasons for the delay to PRh care are likely to be multifactorial and include health care organizational, disease-specific, and social patterning factors (39). Evidence suggests that poor pediatric MSK clinical skills are a contributing barrier to care. Trainees in pediatrics and primary care report poor self-rated confidence in MSK clinical skills (40, 41) and PRh training in primary care is deemed inadequate (42). Adult MSK clinical skills are taught as part of core teaching in UK medical schools including an MSK screening assessment (43, 44) and a regional examination (45). It is unclear how much teaching exposure is given to PRh in medical schools, but notably children are not small adults, and the clinical presentations of JIA differ from those of adults with inflammatory joint disease (31). It is hoped that a new pediatric MSK screening tool (46) will improve pediatric clinical MSK skills of medical students and will be a foundation for improved clinical performance in their future as practicing clinicians. However, there is also need to improve knowledge and raise awareness that early referral to a skilled PRhMDT does make a difference in JIA outcome, and to clarify the pathways of care for children with suspected rheumatic disease.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Dr. Foster had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Foster, Eltringham, Kay.

Acquisition of data. Eltringham.

Analysis and interpretation of data. Foster, Eltringham, Kay, Friswell, Abinun, Myers.

Manuscript preparation. Foster, Eltringham, Kay, Friswell, Abinun, Myers.

Statistical analysis. Eltringham, Myers.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

We gratefully acknowledge the help of Mrs. Terry Bell for her secretarial assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES