Skip to main content
BMJ Case Reports logoLink to BMJ Case Reports
. 2018 Sep 23;2018:bcr2018225618. doi: 10.1136/bcr-2018-225618

Expand the differential…think beyond rheumatoid arthritis

Cody Schwartz 1, Altelisha Taylor 1, Zareen Zaidi 1
PMCID: PMC6157583  PMID: 30249729

Abstract

A 31-year-old male patient with severe, migratory arthralgias presented to our academic medical centre after being erroneously diagnosed and treated for rheumatoid arthritis for over 1 year. Multiple immunomodulatory therapies for rheumatoid arthritis were attempted with no relief of symptoms. Eventually, the pain was so bothersome that the patient became bedridden for 1 month prior to presenting to our facility. Our assessment revealed that the patient met the diagnostic criteria, known as the Yamaguchi criteria, needed to diagnose adult-onset Still’s disease. Yamaguchi criteria include migratory inflammatory arthritis, quotidian fevers, leucocytosis and a salmon-coloured maculopapular rash. These signs and symptoms may go unnoticed or overlooked if adult-onset Still’s disease is not considered. The patient was treated with anakinra (a recombinant human IL-1 receptor antagonist) and had rapid improvement in his symptoms, with the restoration of mobility.

Keywords: biological agents, drugs: musculoskeletal and joint diseases, connective tissue disease, rheumatoid arthritis, musculoskeletal syndromes

Background

Adult-onset Still’s disease is defined as a multigenic autoinflammatory disease characterised by daily fevers, arthritis and a salmon-coloured maculopapular rash.1 2 The condition was first evaluated, in 1897, by George Still after examining 22 children with symptoms of arthritis for which he named the disease systemic juvenile idiopathic arthritis. The name ‘Still’s disease’ became an eponym for systemic juvenile idiopathic arthritis.3 It was not until 1971 the name ‘adult-onset Still’s disease’ was coined by Bywaters after he observed 14 adults with similar symptoms as the children with systemic juvenile idiopathic arthritis (rash, arthralgia and fevers).4 The exact aetiology of adult-onset Still’s disease is still unknown. It is believed that genetic, infectious and environmental factors all play a role in the development of the disease causing a dysregulation in the innate immune system which leads to unprovoked inflammation without autoreactive T lymphocytes or autoantibodies, therefore making it different from the classical autoimmune diseases.5 The estimated annual incidence is 0.16 cases per 100 000 people in France, with equal distribution between men and women,6 0.22 cases per 100 000 people in Japan7 and 0.4 cases per 100 000 people in Northern Norway.8 It has been shown to affect both genders with predominance among younger adults aged 16–35.6 Often adult-onset Still’s disease symptoms closely resemble that of rheumatoid arthritis and may be misdiagnosed as such. When a patient is under 50 years old and have symmetric, bilateral arthralgic pain, we may make the mistake of prematurely closing in on a diagnosis of rheumatoid arthritis or seronegative spondyloarthropathy. While these two diagnoses are much more common, the prevalence of rheumatoid arthritis is around 1%,9 and the prevalence of seronegative spondyloarthropathy is around 1%–1.4%,10 it is important to keep the differential diagnosis broad initially and, in this case, consider the rare diagnosis of adult-onset Still’s disease. As a failure to recognise and diagnose, adult-onset Still’s disease could cause significant morbidity if treatment is delayed as seen in our case.

Case presentation

A 31-year-old man presented to our hospital after 1 year of unrelenting, asymmetric, migratory arthralgias manifesting in bilateral hands, wrists, knees and right shoulder at one time or another. At first, it was difficult to obtain a thorough history from the patient given a language barrier as he was not born in the USA and English was a second language. Using interpreters, we were able to get the full history after talking to him and to his sister.

The patient first developed arthritis 1 year ago in his bilateral knees. This progressed to involve other joints at different times, mostly his ankles, wrists, shoulders and knees. The joints would be swollen, warm and tender but he never had significant stiffness. He then developed fever, erythematous (non-pruritic) rash on his trunk and arms, and a severely sore throat. He was evaluated twice at an outside community hospital around November/December 2016. The first time he was treated with antibiotics without improvement. The second time he was discharged on prednisone 40 mg daily. The prednisone helped his pain significantly but some of the joint swelling remained. Lower doses of prednisone 30 mg did not seem to help him at all. He then developed daily fevers of 39°C or higher for 2 months. His rash did not always correlate with the fever and he continued to get the rash on a daily basis even on admission. His right ankle had been swollen and painful for over 6 weeks prior to admission to our facility but his other joint pain seemed to come and go over the course of a few days. He still developed the occasional sore throat but not to the degree as his first episode.

In February 2017, the patient had multiple labs performed. His serologies were unremarkable including negative rheumatoid factor (RF), anticyclic citrullinated peptide (anti-CCP) antibody, antinuclear antibody (ANA), double-stranded DNA, Smith antibody, antineutrophil cytoplasmic antibody, myeloperoxidase, proteinase 3, anti-Ro (SS-A), anti-La (SS-B), antitopoisomerase 1 (Scl-70), anticentromere antibody, Jo-1 antibody, antiphospholipid panel, antiribonucleoprotein and human leucocyte antigen-B27 (HLA-B27). His complement levels were normal (C2, C3 and C4), erythrocyte sedimentation rate (ESR) was 29 and C reactive protein (CRP) was 9.7. A ferritin level was not documented. He had a leucocytosis of 24 000 and a normal creatine kinase of 10. His care was transferred to a rheumatologist who was convinced that the patient had rheumatoid arthritis and treated him as such. The patient was started on methotrexate and adalimumab but did not find any relief of symptoms. According to the patient’s sister, after the failure of adalimumab and methotrexate, the patient was started on abatacept, followed by tofacitinib citrate and finally he was taking tocilizumab prior to presenting to the hospital (the last dose of tocilizumab was 2 weeks before admission). He continued to use prednisone but was tapered to 4 mg daily. The patient stated that no medications completely alleviated his pain, although 40 mg of prednisone daily did make the pain tolerable.

During this time, the patient’s arthralgias continue to worsen and he started developing clinical signs of steroid-induced Cushing syndrome with abdominal striae. He presented to our hospital in 10/10 pain, bedbound for 1 month, with arthralgias in bilateral wrists and shoulders, swelling and pain in his right foot, a non-pruritic pink rash on both forearms and a sore throat. The arthralgic pain was exacerbated when he attempted to move his body or walk.

The review of systems was negative for chest pain, shortness of breath, night sweats, dysuria, haematuria, diarrhoea, penile discharge, constipation, melaena, abdominal pain, fevers, chills, weight loss and changes in vision. He did not have any sick contacts and had no recent travel outside of Florida in over 4 years. He was born in Venezuela, lived in the Dominican Republic from age 1 until he was 16, then moved to the USA at the age of 16. He denied the use of any oral supplements or new soaps, had no allergies and denied any significant tick or mosquito bites. He also denied any current sexual activity but did state he was sexually active with only women in the past. He reported last having intercourse 1 year ago. He had sexually transmitted disease testing done 6 months ago including HIV, syphilis, gonorrhoea and chlamydia which were all negative.

The patient has a significant medical history composed of several disorders such as haemolytic disease of the newborn, unexplained jaundice at age 5, splenectomy at age 11, portal vein thrombosis with jugular vein anastomoses at age 16 and a bowel obstruction requiring partial colectomy at age 19. It is not clear if these multiple comorbidities were because of a genetic factor related to adult-onset Still’s disease or the fact the patient has a predilection for inflammation. It is known that adult-onset Still’s disease is related to disorders associated with macrophage activation syndrome (a disorder characterised by pancytopenia, liver insufficiency, coagulopathy and neurological symptoms) like systemic lupus erythematous and Kawasaki disease.11 It is estimated that 12% of patients with adult-onset Still’s disease will develop macrophage activation syndrome.12 Presence of lymphadenopathy and abdominal pain is associated with macrophage activation syndrome occurrence,13 but given the current lack of these symptoms and any consistent association with HLA antigens and autoantibodies,14 no clear association or genetic connection can be made to the many issues he has experienced in the past simply because of the limited information at hand. There has been one other case reported where the patient had portal vein thrombosis along with adult-onset Still’s disease but this patient did not have any of the other complications that our patient experienced.15 His family history is significant for a mom with hypertension and maternal grandparents that each died of unknown cancer in their 70s. Regarding his social history, he denies any alcohol abuse, he has never smoked or used tobacco and he does not use any illicit drugs. He is a hairdresser for a living, but given his pain and symptoms, he has been unable to do his job for the last year.

Physical examination was significant for a young man writhing with pain, unable to get out of bed, with a light pink, salmon coloured rash most prominent left forearm (figure 1A) but also on the lateral aspect of the left thigh and the right forearm. He had synovitis with effusion of the left wrist and right ankle (figure 1B), both with a limited range of motion. The bilateral shoulders had limited range of motion but there was no effusion or tenderness to palpation of the shoulders. The elbows, metacarpophalangeal joints, proximal interphalangeal joints and distal interphalangeal joints had no synovitis but there was fusiform oedema of some of the fingers without nodules, deformities, warmth or erythema. The knees and left ankle were normal. Abdominal examination showed the presence of abdominal striae, moon face secondary to iatrogenic (steroid-induced) Cushing syndrome.

Figure 1.

Figure 1

Panel A: 31-year-old patient with salmon-coloured rash seen on left forearm. Panel B: 31-year-old patient with noticeable swelling of right ankle and foot as compared with the left.

Investigations

On admission to our hospital, we obtained multiple labs and imaging modalities. Laboratory investigation was remarkable for a CRP of 5.4, ESR of 36, ferritin of 123.6 and a white cell count (WCC) of 17.3 x 10^9/L. He had a negative RF with a negative anti-CCP antibody, ANA, cytoplasmic antineutrophil cytoplasmic antibody (C-ANCA), perinuclear antineutrophil cytoplasmic antibody (P-ANCA), HLA-B27 and blood cultures had no growth after 5 days. Basic metabolic panel was normal; liver function tests (LFTs) were significant for a low albumin of 2.9 and total protein low of 5.4. Right ankle X-ray showed prominent soft-tissue swelling about the right ankle with right ankle joint effusion and patchy bony demineralisation about the right ankle and hind foot. Right-hand and left-hand X-rays both showed periarticular osteopenia with loss of joint spaces in the intercarpal joints but no erosive changes. MRI of the right ankle showed inflammatory changes centred on the tibiotalar articulation with severe cartilage loss and patchy subchondral marrow oedema along the articular surfaces of the tibial plafond, talar dome and lateral malleolus. A mild ankle joint effusion was also appreciated on the right. Chest X-ray showed no abnormalities.

Differential diagnosis

Differential diagnoses for the arthralgias and rash include many possibilities including infections such as hepatitis, endocarditis, bacteraemia, rheumatic fever and tuberculosis (TB); rheumatological/immunological causes such as rheumatoid arthritis and seronegative spondyloarthropathies; vasculitidies like polyarteritis nodosa; and malignancy such as leukaemia and lymphoma. Hepatitis was ruled out with the negative hepatitis panel including hepatitis B and C. Endocarditis was unlikely given the patient’s lack of risk factors like intravenous drug use and lack of congenital heart disease, normal chest X-ray and a normal cardiovascular examination. Bacteraemia was a possibility but was ruled out given blood cultures were done and revealed no growth. Rheumatic fever was a possibility given the patients arthralgias, non-pruritic rash and sore throat, but he did not have other elements of the Jones criteria (no carditis, Sydenham chorea or subcutaneous nodules),16 and he denied a prior strep infection. TB was ruled out with a negative T-spot TB test. Rheumatoid arthritis was less likely considering the patient’s negative RF antibody, negative ANA and negative anti-CCP antibody. Also, his symptoms had not improved despite being treated for the past 9 months with various modalities including methotrexate (most common initial treatment in which 30%–40% do not have a good response to methotrexate),17 glucocorticoids and other immunomodulatory agents. Seronegative spondyloarthropathies was also a possibility given the patient’s elevated ESR and CRP but his HLA-B27 antigen was negative and he lacked other common symptoms like psoriasis, uveitis and arthralgias mostly related to the vertebral column. Other rheumatological and immunological diagnoses were considered but his P-ANCA and C-ANCA were shown to be negative. Vasculitic illnesses, like polyarteritis nodosa, were also a possibility as they can present with arthralgias, rash and fever. However, the patient had no evidence of ischaemic organ damage or renal disease given the normal blood urea nitrogen and creatinine. Malignancy such as lymphoma and leukaemia were possibilities in a young patient but this was unlikely given the patient’s lack of diffuse lymphadenopathy and mostly unremarkable complete blood count differential. However, further workup for lymphoma or leukaemia, with a lymph node or bone marrow biopsy, would have been considered if other, more probable, diagnoses had been ruled out.

Treatment

Initial treatment for adult-onset Still’s disease is based on severity of disease and extent of damage. Patients with mild disease can be treated with non-steroidal anti-inflammatory drugs (NSAIDs) (ibuprofen 800 mg three to four times daily or naproxen 500 mg two times a day). However, NSAIDs alone are only 20% effective in adult-onset Still’s disease.18 Those who do not respond to NSAIDs or have moderate disease should be treated with glucocorticoids. Approximately 70% of patient’s will respond to glucocorticoids alone or will respond to glucocorticoids after a trial of NSAIDs.19 If the patient does not respond to glucocorticoids, they should be started on a trial of methotrexate with or without glucocorticoids. One study showed a response of 88% to methotrexate in those people who were resistant to glucocorticoids.20 If the patient continues to be refractory, as in our patient, then immunomodulatory methods should be pursued. Biological agents such as tumor necrosis factor inhibitors (etanercept, infliximab and adalimumab), IL-1 inhibitors (anakinra, canakinumab and rilonacept) and IL-6 inhibitors (tocilizumab) have all been studied and used for the treatment in glucocorticoid-resistant adult-onset Still’s disease. TNF and IL-6 inhibitors target more of the pure arthritic symptoms while IL-1 inhibitors, like anakinra, target systemic symptoms.2 The patient was first given adalimumab for which he only saw brief relief of his arthralgic symptoms. Next, tocilizumab was tried but once again the patient showed only improvement of his arthralgic symptoms. Given our patient’s systemic involvement (fevers, rash and sore throat), therapy with TNF inhibitors (adalimumab) and IL-6 inhibitors (tocilizumab) did not fully treat his adult-onset Still’s disease as they target more arthralgic symptoms and less systemic symptoms. With this information, we started our patient on anakinra (a recombinant human IL-1 receptor antagonist) 100 mg administered subcutaneously daily. Prior to discharge, the patient received four doses of anakinra. Due to the patient’s insurance issues, he was unable to afford anakinra as outpatient but was able to obtain canakinumab (an anti-IL-1 beta monoclonal antibody) 150 mg administered subcutaneously every 30 days provided by the rheumatology clinic.

Outcome and follow-up

When the patient received just one dose of anakinra in the hospital, he had rapid improvement of his arthralgias, rashes and fevers. His pain was reduced from 10 out of 10 to 5 out of 10. After 3 days of treatment, and a total of four doses, the patient reported complete resolution of arthralgic pain, complete disappearance of the rash and relief of his sore throat. Even though the patient was unable to continue anakinra as outpatient, he had a great response to canakinumab. On a recent follow-up with rheumatology clinic, and after contacting the patient and his sister over the phone, the patient seems to be tolerating canakinumab very well. Currently, he reports that his symptoms are well controlled and he has not experienced any flare-ups. He does state some mild, residual right ankle pain which is greatly improved from previous. The improvements in his symptoms can be compared with a relatively recent study performed by Colafrancesco et al which investigated treatment with IL-1 blocking agents in 140 patients with adult-onset Still’s disease. The results showed that after just 3 months of therapy with the IL-1 blocking agents, anakinra and canakinumab, patients achieved almost complete remission. Proving that IL-1 inhibitors appear to be a highly effective treatment of adult-onset Still’s disease.21

Discussion

This case demonstrates that even with the patient’s young age of 31 and the triad of arthralgias, rash and fevers making up most of the Yamaguchi criteria, many community physicians repeatedly missed the diagnosis of adult-onset Still’s disease. Intuitive clinical reasoning (a rapid process used frequently by experts) led them to focus on the more common diagnosis of rheumatoid arthritis (given the patient’s arthralgias). This guided treatment for nearly a year, despite the patient’s lack of response to steroids. While other arthralgias, like rheumatoid arthritis or spondyloarthropathies or autoimmune pathologies, should be included on the differential, adult-onset Still’s disease should also be considered. With the above clinical signs and symptoms, along with negative ANA, RF and other autoimmune antibodies, this indicates that adult-onset Still’s disease should be higher on the differential. Laboratory findings such as an elevated ESR, CRP, WCC and serum ferritin may also be found. Worsening of the condition and/or negative response to high-dose steroids are additional factors to consider.

Patients with adult-onset Still’s disease should be screened and stratified accordingly based on a validated prognostic tool called the systemic score. The systemic score assigns 1 point to each of 12 manifestations: fever, typical rash, pleuritis, pneumonia, pericarditis, hepatomegaly or abnormal LFTs, splenomegaly, lymphadenopathy, leucocytosis >15 000/mm3, sore throat, myalgia and abdominal pain (maximum score: 12 points). A score of ≥7 has a strong prognostic impact identifying patients at risk for adult-onset Still’s disease-related death.22 Our patient had a systemic score of 5 which places him at a lower risk of adult-onset Still’s disease-related death. Also, macrophage activation syndrome occurrence significantly reduced the survival rate in patients with adult-onset Still’s disease,13 but it is unclear if his history is consistent with macrophage activation syndrome or an unrelated disorder. While this type of case will likely be worked up by a rheumatologist, a delay in diagnosis may be likely. In general, is important to keep adult-onset Still’s disease on the differential in people who present into the emergency room or outpatient clinic with the marquee findings of the relapsing arthralgias, fever and rash.

Learning points.

  • A thorough history and comprehensive physical examination are the first steps in assessing a patient with suspected adult-onset Still’s disease.

  • If a patient does not respond to multiple different treatment modalities for rheumatoid arthritis, then another diagnosis should be considered.

  • Other complications such as infection, malignancy and other rheumatological diseases must be ruled out prior to making the diagnosis of adult-onset Still’s disease.

  • Multiple treatment modalities including non-steroidal anti-inflammatory drugs, steroids, disease-modifying antirheumatic drugs and immunomodulatory therapies should be considered until complete remission of adult-onset Still’s disease is achieved.

Footnotes

Contributors: CS was the main author of the case report. He planned, wrote and edited the article. AT was the second author of the case report. She developed an abstract and helped write the case report. ZZ reviewed the case report offering editing and updates to be added.

Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests: None declared.

Patient consent: Obtained.

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

References

  • 1.Mandl LA, 2017. Clinical manifestations and diagnosis of adult Still’s disease [Internet] https://www.uptodate.com/contents/clinical-manifestations-and-diagnosis-of-adult-stills-disease (Cited 31 Jan 2018).
  • 2.Gerfaud-Valentin M, Jamilloux Y, Iwaz J, et al. Adult-onset Still’s disease. Autoimmun Rev 2014;13:708–22. 10.1016/j.autrev.2014.01.058 [DOI] [PubMed] [Google Scholar]
  • 3.Middlemiss JH. Juvenile rheumatoid arthritis (Still’s disease). Proc R Soc Med 1951;44:805–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bywaters EG. Still’s disease in the adult. Ann Rheum Dis 1971;30:121–33. 10.1136/ard.30.2.121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.McDermott MF, Aksentijevich I, Galon J, et al. Germline mutations in the extracellular domains of the 55 kDa TNF receptor, TNFR1, define a family of dominantly inherited autoinflammatory syndromes. Cell 1999;97:133–44. 10.1016/S0092-8674(00)80721-7 [DOI] [PubMed] [Google Scholar]
  • 6.Magadur-Joly G, Billaud E, Barrier JH, et al. Epidemiology of adult Still’s disease: estimate of the incidence by a retrospective study in west France. Ann Rheum Dis 1995;54:587–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Wakai K, Ohta A, Tamakoshi A, et al. Estimated prevalence and incidence of adult Still’s disease: findings by a nationwide epidemiological survey in Japan. J Epidemiol 1997;7:221–5. 10.2188/jea.7.221 [DOI] [PubMed] [Google Scholar]
  • 8.Evensen KJ, Nossent HC. Epidemiology and outcome of adult-onset Still’s disease in [42] Behrens EM. Macrophage activation syndrome in rheumatic disease: what is the Northern Norway. Scand J Rheumatol 2006;35:48–51. [DOI] [PubMed] [Google Scholar]
  • 9.Spector TD. Rheumatoid arthritis. Rheum Dis Clin North Am 1990;16:513. [PubMed] [Google Scholar]
  • 10.Reveille JD, Weisman MH. The epidemiology of back pain, axial spondyloarthritis and HLA-B27 in the United States. Am J Med Sci 2013;345:431–6. 10.1097/MAJ.0b013e318294457f [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Grom AA, Mellins ED. Macrophage activation syndrome: advances towards understanding pathogenesis. Curr Opin Rheumatol 2010;22:561–6. 10.1097/01.bor.0000381996.69261.71 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Arlet JB, Le TH, Marinho A, et al. Reactive haemophagocytic syndrome in adult-onset Still’s disease: a report of six patients and a review of the literature. Ann Rheum Dis 2006;65:1596–601. 10.1136/ard.2005.046904 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Ruscitti P, Iacono D, Ciccia F, et al. Macrophage activation syndrome in patients affected by adult-onset still disease: Analysis of survival rates and predictive factors in the gruppo italiano di ricerca in reumatologia clinica e sperimentale cohort. J Rheumatol 2018;45:864–72. 10.3899/jrheum.170955 [DOI] [PubMed] [Google Scholar]
  • 14.Rossi-Semerano L, Koné-Paut I. Is Still’s Disease an Autoinflammatory Syndrome? Int J Inflam 2012;2012:1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Morita H, Nishiwaki H, Nagayama Y, et al. Portal vein thrombosis in adult-onset Still’s disease: a case report and literature review. Rheumatol Int 2009;29:1515–8. 10.1007/s00296-009-0894-7 [DOI] [PubMed] [Google Scholar]
  • 16.Gewitz MH, Baltimore RS, Tani LY, et al. Revision of the Jones Criteria for the diagnosis of acute rheumatic fever in the era of Doppler Echocardiography: A scientific statement from the American Heart Association. Circulation 2015;131:1806–18. 10.1161/CIR.0000000000000205 [DOI] [PubMed] [Google Scholar]
  • 17.Weinblatt ME, Kaplan H, Germain BF, et al. Methotrexate in rheumatoid arthritis. A five-year prospective multicenter study. Arthritis Rheum 1994;37(10):1492–8. 10.1002/art.1780371013 [DOI] [PubMed] [Google Scholar]
  • 18.Mandl LA, 2017. Treatment of adult Still’s disease [Internet]. https://www.uptodate.com/contents/treatment-of-adult-stills-disease (Cited 31 Jan 2018).
  • 19.Franchini S, Dagna L, Salvo F, et al. Efficacy of traditional and biologic agents in different clinical phenotypes of adult-onset Still’s disease. Arthritis Rheum 2010;62:2530–5. 10.1002/art.27532 [DOI] [PubMed] [Google Scholar]
  • 20.Fautrel B, Borget C, Rozenberg S, et al. Corticosteroid sparing effect of low dose methotrexate treatment in adult Still’s disease. J Rheumatol 1999;26:373. [PubMed] [Google Scholar]
  • 21.Colafrancesco S, Priori R, Valesini G, et al. Response to Interleukin-1 Inhibitors in 140 Italian Patients with Adult-Onset Still’s Disease: A Multicentre Retrospective Observational Study. Front Pharmacol 2017;8 10.3389/fphar.2017.00369 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Ruscitti P, Cipriani P, Masedu F, et al. Adult-onset Still’s disease: evaluation of prognostic tools and validation of the systemic score by analysis of 100 cases from three centers. BMC Med 201614:194 10.1186/s12916-016-0738-8 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from BMJ Case Reports are provided here courtesy of BMJ Publishing Group

RESOURCES