A critical appraisal of the radiological evaluation of nephrocalcinosis

Nephron Clin Pract. 2007;106(3):c119-24. doi: 10.1159/000102999. Epub 2007 May 22.

Abstract

Background/aim: The level of agreement concerning the diagnosis of nephrocalcinosis (NC) based on ultrasonography (US), computed tomography (CT) or kidney/ureter/bladder (KUB) X-ray was assessed.

Methods: Sequences of KUB+US+CT from 62 patients, 48 with at least one exam suggesting NC and 14 with pelvicalyceal calcifications (nephrolithiasis) were reviewed twice by 3 radiologists (firstly randomized and secondly presenting KUB+US+CT of each patient together).

Results: The intraobserver concordance varied from 76 to 90% for KUB, 77 to 85% for US and 82 to 89% for CT. There was a significant change in diagnosis between the 1st and 2nd reviews for observer 1 in KUB and CT, for observer 2 in US, but not for observer 3. Evaluating patients' exams together did not provide a better agreement. The highest sensitivity and specificity (92 and 89%, respectively) were only attained when 2 exams suggested NC diagnosis, being CT one of them. These enabled us to suggest that 33 out of 48 (62.5%) patients had NC (evidenced in US+CT+KUB (81.8%), US+CT (12.1%) or KUB+CT (6.1%).

Conclusion: The low level of concordance renders the radiological diagnosis of NC difficult. Nephrocalcinosis should be confirmed by CT combined with either US or KUB.

Publication types

  • Comparative Study
  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Child
  • Child, Preschool
  • Female
  • Humans
  • Male
  • Metabolic Diseases / diagnosis
  • Metabolic Diseases / etiology
  • Middle Aged
  • Nephrocalcinosis / complications
  • Nephrocalcinosis / diagnostic imaging*
  • Observer Variation
  • Radiography
  • Sensitivity and Specificity
  • Ultrasonography