Renal ultrasonography in the evaluation of acute kidney injury: Developing a risk stratification framework

Background: In adult inpatients with acute kidney injury (AKI), clinicians routinely order a renal ultrasonography (RUS) study. It is unclear how often this test provides clinically useful information. Methods: Cross-sectional study, including derivation and validation samples, of 997 US adults admitted to Yale-New Haven Hospital from January 2005 to May 2009, who were diagnosed as having AKI and who underwent RUS to evaluate elevated creatinine level. Pregnant women, renal transplant recipients, and patients with recently diagnosed hydronephrosis (HN) were excluded. Demographic and clinical characteristics were abstracted from the medical records. A multivariable logistic regression model was developed to create risk strata for HN and HN requiring an intervention (HNRI); a separate sample was used for validation. The frequency of incidental findings on RUS was assessed for each stratum. Results: In a derivation sample of 200 patients, 7 factors were found to be associated with HN: history of HN; recurrent urinary tract infections; diagnosis consistent with obstruction; nonblack race; and absence of the following: exposure to nephrotoxic medications, congestive heart failure, or prerenal AKI. Among 797 patients in the validation sample (mean age, 65.6 years), 10.6% had HN and 3.3% had HNRI. Of 223 patients in the low-risk group, 7 (3.1%) had HN and 1 (0.4%) had HNRI (223 patients needed to be screened to find 1 case of HNRI). In this group, there were 0 incidental findings on RUS unknown to the clinical team. In the higher-risk group, 15.7% had HN and 4.7% had HNRI. Conclusion: In adult inpatients with AKI, specific factors can identify patients unlikely to have HN or HNRI on RUS.