Rational Testing

Raised inflammatory markers

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e454 (Published 3 February 2012)
Cite this as: BMJ 2012;344:e454

Recent rapid responses

Rapid responses are electronic letters to the editor. They enable our users to debate issues raised in articles published on bmj.com. Although a selection of rapid responses will be included as edited readers' letters in the weekly print issue of the BMJ, their first appearance online means that they are published articles. If you need the url (web address) of an individual response, perhaps for citation purposes, simply click on the response headline and copy the url from the browser window.

Displaying 1-9 out of 9 published

In their rational testing paper the authors suggest to wait and see in case raised levels of inflammatory markers are detected incidentally (1). Based on our experience we would suggest not to wait too long.

Using 18F-FDG-PET we have evaluated 10 patients (median age 76 years, range 40-87) attending our outpatient Rheumatology Clinic, who had laboratory markers of inflammation (median ESR 99, range 44-120; median CRP 9.5, range 5-18) and definitely non-specific constitutional symptoms, including fever. In 6 of them we obtained evidence of large-vessel vasculitis: the aorta was affected in 5 and the vertebral and femoral arteries in the remaining patient. These patients underwent treatment with methyl-prednisolone (0.5-1 mg/kg) with a fast decline of ESR and CRP values and a substantial resolution of the constitutional symptoms. The dose was tapered in the following months.

We are still following the patients, who have had no further complications. These patients, after undergoing time and money consuming evaluation, would have been probably treated empirically with steroids by most physicians. Instead, the use of 18F-FDG-PET has lead to an early definite diagnosis of vasculitis, thus providing a rationale for the use of steroids and probably preventing dramatic complications (dissection).

1. Watson J, Round A, Hamilton W. Raised inflammatory markers. BMJ 2012;344:e454

Competing interests: None declared

Marco Fusconi, Rheumatologist

Magda Frisoni, Daniela Zauli

Azienda Ospedaliero-Universitaria, Via Massarenti 9, Bologna, Italy

Click to like:

We note the author’s¹ response to a previous letter discussing the role of ESR in exclusion of myeloma. We are concerned that it has been accepted that a normal ESR may be seen only rarely in myeloma. This is dangerous advice particularly with respect to light chain myeloma, one of the most devastating forms of myeloma, often presenting with renal failure.

As per the British Committee of Standards in Haematology (BCSH) guideline² we believe that all patients with suspected multiple myeloma should be investigated with screening tests including full blood count (FBC), ESR, urea, creatinine, calcium and importantly, electrophoresis of serum and concentrated urine.

Furthermore a serum free light chain assay should be performed in any patient in whom there is a strong suspicion of myeloma and serum protein electrophoresis is negative. This assay is particularly important in the diagnosis of light chain only myeloma and non-secretory myeloma in whom no monoclonal protein would be detected by routine serum electrophoresis.

A recent review³ recommended serum free light chain assay to be used in conjunction with serum protein electrophoresis when screening for the presence or absence of a monoclonal plasma cell disorder such as myeloma.

There is little published work investigating the value of inflammatory markers in the diagnosis or exclusion of myeloma. The paper quoted 4 by Watson et al examined patients with an ESR of 100mm/hr or more; grossly elevated rather than simply elevated and was published in 1979. We performed a literature search but found no more recent papers.

We therefore performed a concise study of our own patients. The last 100 patients diagnosed with myeloma at North Glasgow Hospitals were identified from our registry. ESR values at presentation were available in 89 cases. Subtypes of myeloma were recorded; IgG 49%, IgA 20%, Light chain 25%, IgE 1%, Amyloid 3%, Non-secretory 1%; to ensure the sample was representative. The local normal range for ESR is ≤12mm/hr.

We found 10% of patients to have ESR ≤12mm/hr, 48% of patients ESR ≤60mm/hr and 70% of patients ESR ≤100mm/hr at time of diagnosis. In particular, the patient with non-secretory myeloma had an ESR ≤30mm/hr and 68% of patients with light chain myeloma an ESR ≤60mm/hr.

The low contribution of light chain to the value of ESR helps explain these findings. ESR is mainly controlled by fibrinogen and other acute phase markers.5 The molecular weight of fibrinogen is more than ten times that of light chain, markedly limiting the potential effect a rise in light chain can have on ESR value.

We hope this brief study demonstrates that a normal ESR cannot be used to confidently exclude a diagnosis of myeloma and is not a rare occurence, particularly when there is no monoclonal protein present.

1. Watson J, Round A, Hamilton W. Raised inflammatory markers. BMJ. 344:1-52,7843:43-45.

2. BCSH and UKMF Guidelines on the Management and Diagnosis of Multiple Myeloma. J.Bird et al. Sept 2010.

3. Kyle RA, Rajkumar SV. Criteria for diagnosis, staging, risk stratification and response assessment of multiple myeloma. Leukemia (2009) 23:3-9

4. Ford MJ, Innes JA, Parrish FM, Allan NC, Horn DB, Munro JF. The significance of gross elevations of the erythrocyte sedimentation rate in a general medical unit. E.J. Clin Invest. 9(3):191-4,1979.

5. Lewis SM, Bain BJ, Bates I. Dacie and Lewis Practical Haematology.

Competing interests: None declared

Catherine Ogilvie, Haematology speciality trainee

Lesley Stirton, Richard Soutar

Gartnavel General Hospital, 1053 Great Western Rd, Glasgow, G12 0YN

Click to like:

We would like to thank all the correspondents for their comments, which are most helpful. The ESR calculator is useful, though our local laboratory has now ceased analysing ESRs (preferring viscosity), for the reasons given in the article. It was useful to hear of differential rises in inflammatory markers in lupus, and that this may help to differentiate between this and other rheumatological disorders.

In reply to the other comments, it is worth starting from the position that no test is perfect.

For childhood appendicitis, we gave details of a systematic review,1 which showed an association between a high ESR/CRP and the chance of appendicitis. The Stefanutti paper was published after this review, though Stefanutti concluded ‘White blood cell count or CRP values alone do not appear to provide any useful additional information to the surgeon.’2 Again this seems to be a case of associations being real but insufficiently strong to override other clinical features.

Two correspondents were concerned about using a negative inflammatory marker as a rule out test: for myeloma and for polymyalgia / temporal arteritis. We made it clear that negative tests can occur in polymyalgia: the Ellis paper, studying a hospital population, had the highest proportion at 22%, 3 though the recent BMJ review quoted 4% overall.4 Similarly, we agree negative tests can rarely occur in myeloma. These rare occurrences emphasise the importance for all clinicians of being prepared to revisit their diagnosis (or non-diagnosis) if the patient does not improve. That is the point at which additional testing or referral is often warranted.

1. Bundy DG, Byerley JS, Liles EA, Perrin EM, Katznelson J, Rice HE. Does this child have appendicitis? Jama 2007;298(4):438-51.
2. Stefanutti G, Ghirardo V, Gamba P. Inflammatory markers for acute appendicitis in children: are they helpful? Journal of Pediatric Surgery 2007;42(5):773-76.
3. Ellis M, Ralston S. The ESR in the diagnosis and management of the polymyalgia rheumatica/giant cell arteritis syndrome. Annals of the Rheumatic Diseases, 1983;42:168-72.
4. Hassan N, Dasgupta B, Barraclough K. Giant cell arteritis. BMJ 2011;342:d3019.

Competing interests: None declared

William T Hamilton, professor of primary care diagnostics

Jessica Watson, Alison Round

PCMD, Veysey Building, EX2 4SG

Click to like:

I'm grateful to the authors for a good overview of this topic.

I'm a little unhappy about one bit: ' ... this is clinically supported by the negative likelihood ratio of 0.33, meaning that bacterial infection is about a third less likely once a negative test has been reported'. I wonder if 'is' should read 'was'.

It depends on the prevalence of bacterial infection. If the initial odds (probability) of bacterial infection are 9:1 (90%), a negative likelihood ratio of 0.33 gives revised odds (probability) of 3:1 (75%) - which is a sixth less likely. If the initial odds (probability) of bacterial infection are 1:3 (25%), a negative likelihood ratio of 0.33 gives revised odds (probability) of 1:9 (10%) - which is almost two-thirds less likely.

Initial odds (probabilities) are different in different contexts so the same likelihood ratio changes the odds (probabilities) differently in different contexts. Unfortunately, likelihood ratio for apparently the same test is also different in different contexts.(1)

(1) Treasure W. Diagnosis and Risk Management in Primary Care. Radcliffe. London, 2011.

Competing interests: I've written about likelihood ratios in a book published by Radcliffe for profit - Diagnosis and Risk Management in Primary Care: words that count, numbers that speak.

Wilfrid Treasure, Salaried GP

Whalsay Health Centre, Symbister, ZE2 9AE

Click to like:

Raised inflammatory markers as a diagnostic and prognostic tool

We read with interest the article “Raised inflammatory markers” by Watson et al who made an interesting point about biochemical markers in vascular and inflammatory conditions and haematological malignancies. However, they did not address the diagnostic and prognostic value of raised inflammatory markers in common surgical conditions such as acute appendicitis and acute pancreatitis.
Recently, the interest in CRP and WCC as diagnostic markers for acute appendicitis has grown, as it is clear that when both tests are combined, sensitivity levels go up to 0.98 and normal values of WCC and CRP are extremely unlikely in pathologically confirmed appendicitis.2 Khan et al report positive predictive values for WCC and CRP as 92% and 96% in confirmed appendicitis.3

Ranson and Glasgow scores are commonly used for predicting the severity of acute pancreatitis, however, Zrnić et al report that LDH and CRP are similarly effective in diagnosing severity in the early stages. Diagnostic accuracy for both CRP and Ranson score are 50% on the third day and they concluded that LDH and CRP are simple and economical tests which help to predict complications in acute pancreatitis.4,5 High CRP levels can be used prognostically to identify patients who are more likely to develop complications, or those with severe disease which is not obvious at the outset.6 These findings were corroborated by a Finnish study which concluded that of the 1050 acute pancreatitis cases none were life threatening when leucocyte and CRP levels were normal on admission.7

1. BMJ 2012;344:e45. Raised Inflammatory Markers. Watson J, Round A, Hamilton W
2. J Pediatr Surg. 2007 May;42(5):773-6.
Inflammatory markers for acute appendicitis in children: are they helpful? Stefanutti G, Ghirardo V, Gamba P
J Ayub Med Coll Abbottabad. 2004 Jul-Sep;16(3):17-9.
3. The role of white cell count and C-reactive protein in the diagnosis of acute appendicitis.
Khan MN, Davie E, Irshad K.
4. Lijec Vjesn. 2007 Jan-Feb;129(1-2):1-4.
[C-reactive protein and lactate dehydrogenase as single prognostic factors of severity in acute pancreatitis]. Zrnić IK, Milić S, Fisić E, Radić M, Stimac D
5. Gut. 1994 Jun;35(6):822-7.
Inflammatory response in the early prediction of severity in human acute pancreatitis. Viedma JA, Pérez-Mateo M, Agulló J, Domínguez JE, Carballo F.
6. Ann R Coll Surg Engl. 1988 Jul;70(4):227-32.
Prognostic markers in acute pancreatitis: can pancreatic necrosis be predicted? Leese T, Shaw D, Holliday M.
7. Scand J Surg. 2002;91(4):353-6.
Do normal leucocyte count and C-reactive protein on admission to hospital exclude a life-threatening attack of acute pancreatitis? Hämäläinen MT, Grönroos P, Grönroos JM.

Competing interests: None declared

Rajesh K Choudhary, Associate Specialist, General Surgery (Colorectal), Darlington Memorial Hospital, Darlington

Manaswini Choudhary, Medical Student Guy's, King's and St Thomas' School of Medicine

Darlington Memorial Hospital, Hollyhurst Road, Darlington, DL3 6HX

Click to like:

Watson et al state that normal inflammatory markers make myeloma very unlikely as a diagnosis and rules out the need for further testing. This is dangerous advice, as patients with light chain myeloma, non-secretory myeloma, or simply a low paraprotein level, can have a normal plasma viscosity. There is no 'rule out test' for myeloma: bloods, urine, radiology, and bone marrow all need to be taken into consideration.

Competing interests: None declared

Gwendolen Al Wakeel, Haematology Associate Specialist

North Bristol NHS Trust, Southmead Hospital, Bristol BS10 5NB

Click to like:

In response to your article on inflammatory markers we feel there is a potential danger in suggesting that normal inflammatory markers can be used as a rule-out tool in illnesses such as Giant Cell Arteritis (GCA) and Polymyalgia Rheumatica (PMR). Whilst it may be argued that patients with PMR will come to no serious long-term harm from undiagnosed illness, the same cannot be said about GCA where a delayed or missed diagnosis could have potentially serious implications including permanent loss of vision. The rates of normal ESR in PMR/GCA vary. However a normal ESR has been shown to occur in up to 22.5% of patients [1]. Patients with a delayed diagnosis of GCA are often the patients that present with atypical features [2] and a case review of 80 patients by Ellis showed that 3 patients with GCA- associated blindness had a normal ESR [1]. Whilst normal inflammatory markers make the diagnosis more unlikely, if there is a clinical suspicion, early referral should be sought, especially in suspected GCA [3,4].

(1) Ellis ME, Ralston S. The ESR in the diagnosis and management of the polymyalgia rheumatica/giant cell arteritis syndrome. Ann Rheum Dis 1983 Apr;42(2):168-170.
(2) Ezeonyeji AN, Borg FA, Dasgupta B. Delays in recognition and management of giant cell arteritis: results from a retrospective audit. Clin Rheumatol 2011 Feb;30(2):259-262.
(3) Dasgupta B, Borg FA, Hassan N, Alexander L, Barraclough K, Bourke B, et al. BSR and BHPR guidelines for the management of giant cell arteritis. Rheumatology (Oxford) 2010 Aug;49(8):1594-1597.
(4) Dasgupta B, Borg FA, Hassan N, Barraclough K, Bourke B, Fulcher J, et al. BSR and BHPR guidelines for the management of polymyalgia rheumatica. Rheumatology (Oxford) 2010 Jan;49(1):186-190.

Competing interests: None declared

Toby Helliwell, GP

Dr Sara Muller, Dr Samantha L Hider

Arthritis Research Uk Primary Care and Health Sciences Centre, Keele, University, Arthritis Research UK Primary Care Centre, Keele University, Keele, Staffordshire, ST5 5BG

Click to like:

Regarding Watson et al’s paper on inflammatory markers, there are occasionally instances where the differential response of ESR and CRP rise can provide further information regarding aetiology as this is reported in a relatively small number of conditions having been most associated with systemic lupus erythematosus in the past where more modest rises in conventional CRP are typically noted (although rises in high-sensitivity CRP are detected)1.

In summary, whilst agreeing with the main messages of Watson et al, the pattern of differential response in CRP and ESR with suggestive clinical correlates can sometimes be helpful and should not be ignored in this situation.


1. Gaitonde S, Samols D, Kushner I. C-reactive protein and systemic lupus erythematosus. Arthritis Rheum 2008;59(12):1814-20.

Competing interests: None declared

Andrew RL Medford, Chest Physician

North Bristol Lung Centre, Southmead Hospital, Westbury-on-Trym, Bristol BS10 5NB

Click to like:

It is often forgotten that an erythrocyte sedimentation rate (ESR) can in fact be adapted for age and gender as published by Miller et al. in 1983.1

If their simple formula for maximum normal ESR (for men, age in years ÷ 2)* is used then the 72 year old man quoted in Hamilton and colleagues’ article on inflammatory markers would have been allowed to have an ESR of up to 36 without the clinician having to regard it as elevated.

This easy manoeuvre greatly increases the positive predictive value of ESR testing and makes ESR far more useful in clinical practice than plasma viscosity (PV).

* for women: (age in years + 10) ÷ 2

Miller A, Green M, Robinson D. Simple rule for calculating normal erythrocyte sedimentation rate. Br Med J (Clin Res Ed) 1983; 286: 266

Competing interests: None declared

Achim R. Nestel, Consultant Ophthalmologist

North Devon District Hospital, Raleigh Park, Barnstaple EX31 4JB

Click to like: