Sonography as a predictor of human immunodeficiency virus-associated nephropathy

Objective. To determine whether renal sonography can be used to predict the pathologic diagnosis of human immunodeficiency virus-associated nephropathy. Methods. This cross-sectional study evaluated 87 human immunodeficiency virus-positive patients who underwent both kidney biopsy and renal sonography after referral to the Johns Hopkins Renal Clinic from January 1995 to July 2002. Using a standardized measure of echogenicity, an independent blinded radiologist reviewed the original sonographic images. Sensitivity, specificity, positive and negative predictive values, receiver operating characteristic curves, and likelihood ratios were determined with the use of the biopsy pathologic report as the criterion standard. Results. Thirty-four patients (39%) had biopsy-proved human immunodeficiency virus-associated nephropathy. A higher serum creatinine level, greater proteinuria, and black race were associated with human immunodeficiency virus-associated nephropathy, whereas age, sex, hypertension, and diabetes were not. Sensitivity and specificity for the highest 2 levels of echogenicity were 96% and 51%, respectively. Sensitivity and specificity for the highest level of echogenicity were 40% and 95%. The likelihood ratio for the diagnosis of human immunodeficiency virus-associated nephropathy on the basis of the highest echogenicity score was 7.4 (95% confidence interval, 1.3-73.0; P = .006). The likelihood ratio for the lowest 2 echogenicity scores was 0.08 (95% confidence interval, 0.002-0.57; P = 0.003). Kidney size was not associated with human immunodeficiency virus-associated nephropathy status. Conclusions. This study provides evidence that, among patients with human immunodeficiency virus and kidney disease, the highest and lowest levels of sonographic echogenicity have diagnostic value in respectively establishing or excluding human immunodeficiency virus-associated nephropathy.