|Values are valid only on day of printing.|
Erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) are nonspecific tests for acute inflammatory processes. Although they largely generate the same information, both tests are often requested. Studies have demonstrated equivalent performance of ESR and CRP in a majority of patient populations, with the exception of patients with temporal arteritis and polymyalgia rheumatica. At Mayo Clinic, we identified and evaluated patient populations where ESR and CRP were ordered concurrently to determine whether there was any added value to ordering both tests.
Retrospective ESR and CRP results from tests that were ordered alone (ESR n=26,346, CRP n=11,333) and simultaneously (n=29,702) on the same patient on the same day were evaluated over a 1- year period (11/1/08 to 10/31/09). The concordance between positive and negative ESR and CRP results was evaluated based on ESR cutoffs for positive results of >22 millimeters/hour for males or >29 millimeters/hour for females and CRP concentrations >8 mg/L. The diagnosis ICD-9 codes associated with each laboratory test order were obtained from the electronic medical record for the simultaneous ESR and CRP results and were analyzed for agreement. The diagnosis codes were then categorized by indication for testing and the groups were sorted to associate each set of patient results with a single indication for testing when multiple indications were found.
In 95% of cases, ESR and CRP were ordered for the same medical indication. ESR was measured alone for 48% of total requests, while CRP was measured alone for 27%. The overall concordance between ESR and CRP results was 81% (16% positive, 65% negative). The indications for testing with the greatest concordance included inflammation (86%, n=3315) and neoplasm (83%, n=571). The categories with the least concordance included polymyalgia rheumatica (73%, n=368) and systemic lupus erythematosus (74%, n=410). Prosthetic joint/device infection (n=5572) and a variety of rheumatic diseases including rheumatoid arthritis (n=2215) had approximately 80% concordance.
We concluded that ESR and CRP are being ordered for the same purpose. ESR was ordered alone twice as frequently as CRP alone. When CRP was ordered, ESR was ordered concurrently 73% of the time despite an overall analytical concordance of 81%. These data demonstrate that antiquated tests such as ESR continue to be routinely utilized despite emergence of newer, more automated assays such as CRP. Furthermore, these findings suggest that there is redundancy in test orders for assessing acute inflammation for most clinical indications and identifies opportunities for the clinical laboratory to provide education to health care providers and subsequently guide proper test utilization.
Authored by: Amy Saenger, PhD and Darci Block, PhD