Thanks to everyone who contributed to the discussions about Jennifer Price, a 46 year old legal secretary who had widespread joint and muscle pain.
You can see the highlights of the discussion on Twitter in the storify.
This week’s question setter was Dr Eleana Ntatsaki, Rheumatology Registrar and Clinical Teaching Fellow at UCLMS. Our expert was Dr Laura Congi, Rheumatology Consultant, Whittington Health.
Information available at the start of the week:
Miss Jennifer Price, a 46 year old legal secretary, has had widespread joint and muscle pain associated with poor quality of sleep and excessive tiredness for one year. She is worried she has an autoimmune disease and comes to see you in a GP clinic to request some investigations.
Question 1: What further information do you need to help you decide whether to investigate further?
Further information released during the week:
Miss Price has had aches and pains in her joints and muscles but hasn’t noticed any stiffness or any swollen joints. She has been stressed lately due to work-related issues and feels that this has made her irritable bowel syndrome worse. She has not lost any significant weight. She is not on any regular medication and has no known allergies. She has been taking over the counter paracetamol and ibuprofen which do not help much. She does not smoke and drinks alcohol rarely. Her maternal aunt has rheumatoid arthritis which is well controlled on immunosuppressive drugs (etanercept, a TNF inhibitor, also known as a ‘biologic’).
Her observations are:
- Temperature: 37.0 C, HR 71bpm, RR 13/min, GCS 15/15, BP 110/75mmHg, BM 5.2mmol/L, Sats 99% on air.
- A: airway clear
- B: air entry bilaterally, no crepitations or wheeze
- C: normal pulse, warm peripheries, capillary refill < 2 seconds, heart sounds normal
- D: orientated, alert, pupils equal and reactive, fundoscopy no evidence of retinopathy or retinal vasculitis
- E: No skin rashes, mouth ulcers or lymphadenopathy. Abdomen soft, non-tender, no organomegaly, bowel sounds present. No synovitis, multiple tender trigger points (see resource for details)
- Urinalysis: no blood or protein
- Bloods: Hb 120 g/L, WCC 4×109/L, Neutrophils 2.8×109/L, Lymphocytes 2.2×109/L, platelets 200×109/L, ESR 11 mm/hr, CRP 8 mg/L, Corrected calcium 2.4 mmol/L, ALT 32 IU/L, creatinine 72 µmol/L, TSH 3 mU/L, Rheumatoid Factor negative, ANA positive in low titre (1:80 titre)
Further questions discussed on Twitter:
Question 2: What would you look for on examination?
Question 3: What tests would you request?
Question 4: How would you interpret these results?
- Only request a test if you know what you will do with the result. Think about pre-test probability.
- A full history must be taken and an examination completed to look for evidence of autoimmune disease before considering blood tests
- Most people with a positive ANA test do not have a connective tissue disease
Selecting tests and considering health economics
- Postgraduate Medical Journal article: ‘Introduction to health economics for the medical practitioner’ http://pmj.bmj.com/content/79/929/147.full
- Site explaining Bayes theorem and pre-test probability: http://www.thennt.com/diagnostics-and-likelihood-ratios-explained/)
ANA testing and interpretation
- American College of Rheumatology information on ANA tests: http://www.rheumatology.org/practice/clinical/patients/diseases_and_conditions/ana.asp
- GP notebook page on interpreting ANA test results: http://www.gpnotebook.co.uk/simplepage.cfm?ID=x20030808022432739580
- eMedicine article on ANA testing: http://emedicine.medscape.com/article/2086616-overview#showall
- Clinical and Experimental Rheumatology journal article: ‘Appropriateness in anti-nuclear antibody testing: from clinical request to strategic laboratory practice’ http://www.ncbi.nlm.nih.gov/pubmed/15144133
- Annals of the rheumatic diseases article on ANA testing guidelines: ‘ANA screening: an old test with new recommendations’ http://www.ncbi.nlm.nih.gov/pubmed/?term=ANA+screening%3A+an+old+test+with+new+recommendations
Dr Eleana Ntatsaki, Rheumatology Registrar, and Dr Laura Congi, Rheumatology Consultant, say:
This was an interesting scenario which brought out some important principles about selecting the right investigations and explaining results, particularly when these results are ‘borderline’. It also led us to consider how to make good use of resources which requires clinical judgement and application of knowledge based on clinical experience.
Clinical reasoning for choice of investigations
A good principle is that you should not order a test if you don’t know what you will do with the result! Before you order a test, think why you are requesting it for this particular patient and what your next steps will be if it comes back either abnormal or normal. To help you think about whether a test will help you answer a question remember that its usefulness is affected by the pre-test probability and the positive and negative predictive values of the tests. In clinical practice a useful concept is the likelihood ratio, which combines sensitivity and specificity. The likelihood ratio for a test result is defined as the ratio between the likelihood of observing that result in a patient with the disorder in question, and the likelihood of that result in a patient without the disorder.
Our first concern is of course the patient in front of us: if they want a test and that test will be useful for them as an individual then we would perform the test. We may think that even if there is not a clear benefit, if there is no harm then why not order a test? However, we should not forget the bigger picture and the health economics of ordering unnecessary tests.
Appropriate use of autoantibody testing (ANA)
ANA is often requested in patients with nonspecific joint pains and when positive frequently prompts referral to secondary care. Before requesting ANA remember:
- Most patients with +ve ANA do not have SLE/connective tissue disease
- ANA is increasingly found in ageing patients making results difficult to interpret It is always best to only request ANA when your pre-test suspicion is high. Things that may lead to a high clinical suspicion of connective tissue disease include: frequent mouth ulcers, malar rash/photosensitive skin rash, pleuritic chest pain/breathlessness, joint swelling, hair loss, persistent dry eyes or mouth, a past history of DVT/PE or recurrent miscarriage (suggesting the possibility of antiphospholipid syndrome), Raynaud’s phenomenon
- If you think there is a high pre-test probability and then test for ANA and it is positive then it is important to look for additional context with which to interpret the result: is there any evidence of neutropaenia, lymphopaenia, thrombocytopaenia or unexplained anaemia (could there be haemolysis?); is the ESR elevated (in active SLE the ESR is often raised in the setting of a normal CRP, providing there is no intercurrent infection, so these are helpful tests requested together).
- Sometimes further investigations are helpful. A positive dsDNA is much more specific for SLE and should prompt referral. Similarly low C3/C4 could indicate active connective tissue disease. ENA can also be helpful eg Ro and La may be positive in SLE and Sjogren’s
- It is important to remember that sometimes tests are useful for diagnosis but not monitoring. ANA is not a useful tool for monitoring SLE so once a diagnosis has been confirmed it does not need to be repeated.
- Since connective tissues diseases are complex to diagnose and manage, you should always refer or discuss patients where the above findings raise your suspicion of CTD.
- If you do suspect SLE always check a urine dip as the presence of blood and protein would require urgent assessment.
- Don’t forget a positive test does not confirm a diagnosis of connective tissue disease. ANA can be raised in a number of other conditions, and in a proportion of the normal population
Managing patient expectations – what to discuss with Miss Price
Miss Price has no clinical features to suggest she has a CTD. Her symptoms point towards a diagnosis of chronic pain /fibromyalgia. In this context, there is no indication to check her ANAs as there is no clinical suspicion (and therefore a very low pre-test probability) that this may be a CTD. In her case, a borderline positive ANA might be relevant to a family history of autoimmunity, but would not necessarily mean that she herself has a CTD, in the absence of clinical symptoms to suggest this.
As a result of rapid changes in laboratory technology and patients’ expectations clinicians may these days instigate tests in order to meet patient’s expectations without a clear clinical indication. A battery of blood tests and investigations cannot replace a detailed history and comprehensive examination, that remain the cornerstones of good clinical care. In an attempt to exclude an underlying CTD, we may have caused more confusion to our patient by ordering an ANA test. Despite reassuring Ms Price about the absence of clinical symptoms, it would be understandable for her to focus on the borderline positive ANA result. She may see this as evidence that she has a CTD. Having a single positive ANA test is often the sole reason for many unnecessary referrals to the specialist clinic.
Possible harms for the Ms Price in this case are increased anxiety about the possibility of a serious condition and a reduced likelihood of pursuing and committing to treatments aimed at treating her chronic pain/fibromyalgia. This may lead to worsening or prolonged symptoms. Our relationship with Ms Price may also be damaged if she feels we are not acting on a result by not referring her to secondary care. Possible harms overall also include the cost of the test itself and the referral to secondary care. There may be an opportunity cost associated with each of these as well as a direct financial cost.
Moving forward we would want to support Miss Price in pursuing treatments for fibromyalgia. However, if she remained concerned about the possibility of autoimmune disease we may need to refer her to a specialist autoimmune/lupus clinic for a second opinion.