External Validation of a Preoperative Renal Stone Grading System: Reproducibility and Inter-rater Concordance of the Guy's Stone Score Using Preoperative Computed Tomography and Rigorous Postoperative Stone-free Criteria

Abstract

To validate the Guy's stone score (GSS) using preoperative computed tomography (CT) and to assess its inter-rater concordance and association with rigorous definitions of stone clearance. The preoperative CT scans of 166 consecutive percutaneous nephrolithotomy (PCNL) patients treated by a single surgeon were independently reviewed by 2 urology residents and graded according to GSS. Concordance was calculated using Cohen's kappa score. Residual fragments (RFs) were evaluated on CT or plain radiography on postoperative day 1. GSS was correlated with 3 different outcomes; RFs<4mm, RFs<2mm, or no RFs. Higher GSS was associated with decreased stone clearance by any metric on a CT scan: RFs<4mm (P= .03), RFs<2mm (P= .02), or no RFs (P= .02). On plain radiography, higher GSS was only associated with lower likelihood of no RFs (P<.005). Inter-rater concordance was good (κ= 0.72), with 78% of cases categorized the same by both raters. Twelve of 36 cases (33%) of disagreement were between categories II and III and 20 of 36 cases (56%) pertained to unclear definitions of "partial staghorn stone" and "abnormal anatomy." The GSS is a straightforward grading system of the complexity of renal stones. When applied to preoperative CT scans, it offers good inter-rater concordance and is associated with rigorous endpoints of stone clearance. The inter-rater concordance could be further improved by explicit definitions of abnormal anatomy, partial vs complete staghorn stones, and the size of a calculus that constitutes a separate stone.