"CRP-first" algorithm to guide imaging in suspected renal colic: a retrospective UK cohort study

To evaluate whether admission C-reactive protein (CRP) can triage patients with suspected renal colic to low dose non contrast CT KUB or contrast enhanced CT of the abdomen and pelvis (CTAP) at first presentation.

Retrospective single centre diagnostic accuracy study in a United Kingdom emergency department. Index test was admission CRP with a prespecified cut-off of 5 mg/L (positive if CRP ≥ 5 mg/L). Reference standard was CT classified a priori as: A normal, B simple calculus, C complicated calculus, D alternative acute diagnosis. The target condition for accuracy analyses was C or D. We constructed a 2 × 2 table and calculated sensitivity, specificity, predictive values and likelihood ratios with 95% confidence intervals.

Of 1,096 CT examinations during the study window, 233 were for suspected renal colic; 58 patients met eligibility and had admission CRP available (29 with CRP < 5 mg/L and 29 with CRP ≥ 5 mg/L). The target condition was present in 26/58 (44.8%). Using CRP ≥ 5 mg/L, sensitivity was 0.73 (95% CI 0.54-0.86), specificity 0.69 (0.51-0.82), positive predictive value 0.66 (0.47-0.80), negative predictive value 0.76 (0.58-0.88), likelihood ratio positive 2.34 (1.16-4.70) and likelihood ratio negative 0.39 (0.20-0.77).

CRP provided modest but clinically interpretable probability shifts for complicated stones or alternative acute pathology. A CRP first approach may support initial imaging selection between CTAP and CT KUB. Prospective multicentre validation is required.