Sonographic assessment of renal length in the first year of life: the problem of "spurious nephromegaly"

Purpose. Interest in the potential diagnostic or prognostic implications of nephromegaly as evidence for compensatory renal hypertrophy has recently b een emphasized in a variety of clinical settings. This project was designed to compare the results of linear and nonlinear sonographic models in the i nterpretation of renal size and growth during the first year of life. Materials and Methods. We identified all renal and abdominal ultrasound exa minations that were performed between March 1994 and October 1997 in full-t erm infants under age 1 year during which (1) both renal lengths were measu red and (2) both kidneys appeared anatomically normal. Using three differen t computerized algorithms based on published standards for sonographic rena l length in relation to age, we calculated z-scores for the renal lengths a nd compared the results of the three methods: in method A the standards at birth, 1 week, 4 months, 8 months, and 1 year were all used; in method B th e 1-week standard was omitted; in method C the standards at 1 week, 4 month s, and 8 months were omitted. Results. We evaluated 1,234 renal measurements in 617 patients (293 boys, 3 24 girls; mean age 0.24 year). Compared with method A, z-scores were signif icantly increased when either method B or C was used (P < 0.0001). The mean increment in z-score was + 0.433 for method B and + 1.135 for method C. Th e prevalence of "nephromegaly" (z > + 2) was significantly increased when s ubannual standards were omitted (P < 0.0001): using method A, 20 (1.6%) kid neys were large for age compared with 74 (6.0%) using method B, and 214 (17 .3%) using method C. All kidneys that were large for age based on method A were also large for age using both methods B and C. The rate of false-posit ive diagnosis of nephromegaly was 73% (54/74 kidneys) with method B and 91% (194/214 kidneys) with. method C. Conclusions, Although the use of multiple subannual standards for renal len gth in infants less than 1 year of age is time consuming and mathematically more complicated, omission of these standards results in a statistically s ignificant increase in the frequency of "spurious" nephromegaly. Learning objectives. Precise application of published standards is importan t in the interpretation of sonographic measurements of renal length. Omissi on of the subannual standards for renal length in children who are less tha n 1 year of age can result in an incorrect impression of nephromegaly.