I can still remember my first weekend on call as a surgical house officer - rushing around on the morning post-take ward round when a radiology request card was thrust in my hand, and I was sent like a lamb to the slaughter to the on-call radiologist to request an ultrasound for a patient with suspected cholecystitis.
This particular radiologist was well known to have a penchant for making new house officers crumple if we dared request out of hours radiology without the most water-tight argument. ‘HOW WOULD THIS INVESTIGATION CHANGE YOUR MANAGEMENT?’ came the well-rehearsed mantra. Well, frankly I had no idea; I’d only been a doctor for a week and had no clue if the surgeons were planning to operate over the weekend or not, so I was duly sent packing back to my senior to get more information.
Brutal though that first encounter was, it did teach me an important lesson - knowing what test to use, when and for what conditions is a crucial skill in medicine, particularly in General Practice. A really excellent study in the July BJGP brought into sharp focus an area where there is wide variation in practice, namely the use of inflammatory markers for diagnosis.
Inflammatory marker testing in primary care is rising every year, but little is known about whether this is helping diagnostic accuracy or simply generating a cycle of further testing for little gain. Increasingly it appears we are doing CRP and ESR to try to rule out serious disease and reassure us and our patients, but is this helping? Well, this was one of the questions posed by an excellent cohort study in the July 2019 BJGP. The authors looked at a large cohort from the CPRD from 2014, including just under 160,000 patients that had inflammatory marker testing (ESR, CRP or plasma viscosity) and just under 40,000 matched controls who did not have testing. Those with known cancer, auto-immune disease or chronic infection were excluded and the authors then reviewed the numbers of participants who developed serious disease (including cancer, auto-immune disease and infections) in those with raised or normal inflammatory markers. Much of the testing was for non-specific symptoms (with tiredness being one of the commonest reasons for testing) and the results showed that of the ~28% who had raised inflammatory markers, 85% of these had no significant disease (i.e. 85% were false positives).
The other issue the study confirmed is that inflammatory marker testing showed very poor sensitivity for serious disease, meaning they are generally not suitable as rule-out tests either. I was particularly interested in those with a cancer diagnosis - based on the results 21/1000 tested had cancer with 10 having positive inflammatory markers and 11 negative results i.e. doing a CRP/ESR added no value to diagnosis of cancer, and in those with negative results there is a real risk of delayed diagnosis by giving us and our patient false reassurance. The one exception to this would be suspected Myeloma - previous research in the BJGP in 2018 showed that a combination of a normal Hb, ESR and calcium was sufficient to rule out myeloma in the vast majority of the cases, but this is clearly the exception not the rule with inflammatory marker testing.
The message from this research seems quite clear - injudicious use of inflammatory marker testing for non-specific symptoms is likely to cause more harm than good. It will generate huge numbers of false positives, and subsequent unnecessary onward testing and appointments, as well as false reassurance in some patients with normal CRP/ESR who do have serious underlying disease, and a risk of delayed diagnosis. ‘HOW WOULD THIS INVESTIGATION CHANGE YOUR MANAGEMENT?’ - a question we should all be asking when considering use of inflammatory marker testing in primary care.
This article was originally publish on the NB Medical Education Hot Topics Blog on 1 August 2019.