FOAM4GP Map – My ESR is bigger than your CRP. Or do we care not?


So which one should I order and when? We all know ESR and CRP can be useful as non-specific markers of inflammation but over the years, with changing trends, trying to know which one to order has confused me. I thought today I’d look at both and their pros, cons, and uses.

I’ve made a mistake. I had a patient with possible pneumonia and I was uncertain if they clinically needed to go to hospital. At a weak moment I thought – “I know, a CRP will help me decide” – if it’s really high I will send them. Then it came back at 76. Oh. Now what? That’s not high or low. And I’ve already sent her to hospital anyway…. Why did I do that for!? Do we often really need a CRP or ESR at all!?

CRP has been getting interesting press lately suggesting doctors use this test in the clinic to determine whether a patient needs antibiotics for a specific infection – though even this is controversial – The Guardian Science 2014 Link.

What should we really be doing? And what about the ol’ ESR? Anyone remember that test? Is it gone?

Westergren pipet array.JPG

Westergren Pipet Array

Q1. What is an ESR?

ESR stands for Erythrocyte sedimentation rate. The ESR is the rate (mm/hr) at which erythrocytes settle in plasma when in a vertical tube. This speed is affected by many factors but mainly fibrinogen. It will be innaccurate in patients with abnormal red cells (eg.sickle cells) or haematocrits. Age, sex, menstrual cycle, pregnancy and drugs will also affect its result.

If you like watching paint dry (or red cells fall), here is a video – Link

Q2. What is a CRP?

C reactive protein (CRP) is a protein produced by the liver in response to triggers from macrophages and adipose cells, which binds to the surface of dead or dying cells to activate the complement system. Initially it was thought to be a pathogen as it was elevated in patients who had disease; but it was quickly determined that it was released from the liver and a native protein. (Interesting fact – if you have significant liver disease CRP will be inaccurate!)

File:C-reactive protein.png

CRP Molecule

CRP will rise within 2 hours of inflammation onset and peak at 48hrs. A normal level of CRP is not truly known! A large study of 21000 people (National Health and Nutrition Evaluation Survey) found it varied with age, sex, and race.

Australian Prescriber article (2007) on CRP and its use can be found here – Link (Great table here on appropriate indications for CRP)

Q2a. What is a “high sensitivity” CRP?

These are more refined tests which can pick up on much smaller CRP rises (including < 3 – which is the cut off for a normal CRP). Conventional use for this test does not seem clear however further research is investigating its use in helping to risk stratify cardiovascular risk.

Q3. What are the differences between the ESR and CRP?

Best Practice AC (NZealand) – Has a great summary document here listing some of the major differences – Link

The key differences are:

CRP rises within hours, ESR takes longer.

CRP has less variation with age and sex

CRP is more sensitive to subtle changes in the acute phase response.

Q4. Are there times I should order both?

Royal College of Pathologists of Australasia article on ESR and CRP ordering (June 2014) – Link

ESR and CRP are often ordered together for monitoring Rheumatoid Arthritis (RA), Giant Cell Arteritis (GCA), and Polymyalgia Rheumatica (PMR). I continue to search as to why – please let me know in the comments if you know the reason for this. This article in the Mayo medical laboratories discusses some indications for testing both but is far from conclusive – Link. ESR has been historically used for PMR and GCA and therefore the majority of studies are based on ESR values.

Q5. Bacterial or Viral? Can inflammatory markers help?E._coli_Bacteria_(16598492368)

Australian Prescriber: “As an adjunct to clinical assessment, a C-reactive protein test may be useful in differentiating between bacterial and viral infections. A very high C-reactive protein (greater than 100 mg/L) is more likely to occur in bacterial rather than viral infection, and a normal C-reactive protein is unlikely in the presence of significant bacterial infection. However, intermediate C-reactive protein concentrations (10-50 mg/L) may be seen in both bacterial and viral conditions.

Measurement of another acute-phase reactant, procalcitonin, has been advocated as an alternative marker in these circumstances, but data are too preliminary to recommend its universal adoption.” Link.

Q6. Can we have false positives/negatives?

ESR: Anaemia – The ESR is only an indirect measurement of serum acute phase protein concentrations, particularly fibrinogen. It is also influenced greatly by the size, shape, and number of red blood cells (RBC), as well as by other constituents in the blood such as immunoglobulins. Anemia will also increase the sedimentation rate. The sedimentation of RBC is presumably impeded by other RBC; sedimentation is thus more rapid in anemia, in which this retardation is lessened, thus increasing the ESR.

ESR: Increased proteins/globulins – Fibrinogen, β-globulins, α-globulins, albumin levels can all alter the outcome of an ESR (such as MGUS).

ESR: Other:There are many other small factors which can influence ESR result. A great table is here from the AAFP (Table 2) – Link

CRP: Systemic Lupus Erythematosus (SLE) – Systemic Lupus Erythematosus patients seem to have lower CRP levels. It is thought that the type 1 interferons in SLE can suppress hepatocyte creation of CRP and give low readings. However, if CRP significantly elevated in SLE this is more specific for bacterial infection.

CRP: Drugs – OCP and the presence of an intrauterine device (which apparently can cause ‘inflammation’) can cause false positives. It is reported that false negatives for CRP can result from NSAID, steroids and salicylates use. (Though my argument here is – is this a false negative or have they just got less inflammation?)

Q7. IMPORTANT question – when should I not order EITHER!?

So inflammatory markers have been used as a surrogate for ‘how sick’ a patient is. European family practitioners have been using point-of-care CRP for some time to reduce antibiotic prescribing with mixed results (BMJ). It has also be used to help guide success of treatment such as improvement in cellulitis through antibiotic use. However is this ever really required? Can we not see the cellulitis improving clinically? Here are a few studies looking at the use of inflammatory in different scenarios (nearly all ED though unfortunately! Come on GP research!)


Possible Bacteraemia, acute abdomen or osteomyelitis / septic arthritis – CRP has been shown not to be helpful in any of these cases – MJA 2013

Possible Bacteraemia – Not a useful tool in diagnosing severe bacterial infection – Emerg Med J.

Possible bacterial infections in children – Impact of CRP use is very small leading to unnecessary expenditure – BMC Pediatrics 2012

Acute otitis media – CRP Does not help define who needs antibiotics – Pediatr Infect Dis. Link

Surgical vs non surgical causes for Non specific abdominal pain – CRP did not help distinguish surgical vs non surgical causes – Ann R Coll Surg Engl.

Appendicitis in children – C-reactive protein concentration was a poor marker (63.7% sensitivity ; 52.174.3% specificity, ROC curve analysis) – Croat Med J. 2007

Bacterial vs viral causes of pneumonia in children – Unlikely to be helpful unless very high – Pediatr Infect Dis J. 2000

UTI in infants – CRP was an inaccurate tool for discriminating for UTI – Pediatrics. 2000


Acute otitis media – Very high levels of CRP were correlated with bacterial ear infections – Pediatrics 1995

Pyelonephritis in infants – If DMSA was not available it is reasonable to treat a febrile UTI if CRP >66 if >2 days fever or CRP >27 if <2 days – J Microbiol Immunol Infect. 2007

Common Sense use – The final word

File:Sullivan to MacKenzie.png

8. Can CRP be used to risk stratify cardiac disease risk?

Inflammation has been established as playing a role in atherosclerosis formation and hence, cardiovascular disease. High sensitivity CRP detects levels of inflammation. So obviously if you have a high CRP you are likely developing atherosclerosis. Doesn’t it?

Well no. The evidence is being continually debated and definitive randomized evidence for its role is lacking. There are no clear consensus guidelines on the use of CRP to help risk stratify intermediate risk cardiovascular disease patients into low or high risk categories. Watch this space however – further large trials are ongoing currently. Here is a great article from  Journal of the American College of Cardiology 2013 which talks about the possible pathophysiology and highlights the controversial aspects of the debate.

Final word:

CRP and ESR are useful tests used in the correct context. The RCPA article “making sense of inflammatory markers” 2014 has a great list of chronic diseases where ongoing monitoring of disease activity can be useful e.g. Sarcoidosis, pancreatitis, vasculidities, crohn’s disease etc.

In acute diagnoses it has limited use. I think the common sense blow torch should be applied when you order it – “Will it change my management?” If you are uncertain and think a high CRP may sway you to consider sending a patient to hospital then this could be appropriate use. If you are doing it ‘just to be safe’ then I would rethink your diagnosis and management and be more certain of your own skills – rather than relying on a possibly unhelpful test. Also, finally before ordering the test think to yourself “What if the result is NOT what I am expecting. What will I do then?”

Important to have these little mental reminders prior to ordering inflammatory markers so you don’t shoot yourself in the foot (I can tell you YOUR inflammatory markers will be up if that happens…)

I’m off to have a Friday drink (Apparently this lowers levels of CRP – Hurrah! – J Amer Board Fam Med Link)

Cheers! Rob.

PS. When getting the article peer reviewed, a couple of my colleagues had some very ‘sage’ like advice. 

1. “My current teaching in ED is to tell my RMOS to assume it will come back as 42 (answer to life, universe and everything) not bother ordering it and carry on as usual as if nothing happened”

2. “As an ED colleague once quipped to a registrar who ordered a lot of CRP, “you can die with a normal CRP!”


8 thoughts on “FOAM4GP Map – My ESR is bigger than your CRP. Or do we care not?

  1. Great piece Rob! This was a nice reminder of that bit of medicine that had become blurry in my memory over time (not sure it was ever really clear) and I like the common sense reminders at the end. Thanks for sharing.

  2. Nice balanced approach Rob. We all wish there was a simple answer but in reality the simplest answer is to first go back to basics and do the proper history and exam and use the tests as they are meant to be used, to refine when needed, not in place of good clinical assessment.

  3. Pingback: LITFL Review 192 | LITFL: Life in the Fast Lane Medical Blog

  4. Great article Rob! Glad you’ve brought up this issue. I’ve been trying to get my med students and junior doctors away from using “inflammatory markers” indiscriminately. Rather than being too disparaging of the tests, however, I think it needs to be said that the usefulness of a test always depends on the pre-test probability and the clinical question you want that test to answer.

    CRP is probably inappropriate to answer the question: should my pneumonia patient go to hospital? That’s a decision involving the patient’s condition, comorbidities and circumstances. But it may be an entirely appropriate test in a patient already assessed by the ED doctor as possible appendicitis and referred to a surgeon. In this population, a combination of WCC and CRP has good NPV and PPV for appendicitis (Andersson. Br J Surg. 2004 Jan;91(1):28-37) – useful if there is some reason you wish to avoid CT scanning or unnecessary surgery.

    Personally, it’s been a long time since I’ve ordered a CRP. It’s hard to recall a time when ordering it has changed my management. But it’s because I’m me and I work where I work (rurally). When I was a junior doctor, my bosses asked me to order a CRP probably appropriately (e.g. monitoring response of chronic osteomyelitis to antibiotics). Thankfully, I trained in Emergency Departments where ordering a CRP required senior approval.

    On the other hand, I have never understood how to use an ESR. Thanks for getting us thinking.

  5. Pingback: Weekly Resources #1 | GOFOAM

  6. Pingback: Do you know what you don’t know? Want to find out? | FOAM4GP

  7. Pingback: LITFL Review 192 – FOAM Ed

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s