Clinical utility of gray scale renal ultrasound in acute kidney injury

We conducted a retrospective review of all RUS performed from 2004 to 2006 at Memorial Hermann Hospital-Texas Medical Center in Houston. The University of Texas Committee for the Protection of Human Subjects approved this study and granted a waiver of authorization. Patient demographics and pertinent clinical data were obtained from the medical record for the patients who underwent RUS.

All patients who underwent a RUS were identified over the three-year period through an electronic medical record. Patients were excluded from analysis if they were institutionalized, less than 18 years of age, pregnant, recipient of a renal transplant, or clinical data were missing. AKI was defined as a 50% increase in baseline creatinine or a urine output of less than 0.5 ml/kg/hr for more than 12 hours. The etiology of AKI was determined by consensus agreement of the study authors based on review of the medical record. Data collected included demographic and clinical variables. To determine the cause of AKI the presence of contributing factors such as sepsis, hypotension, recent surgery, decompensated heart failure, volume overload, use of intravenous radiocontrast media, and use of nephrotoxic medications were recorded. Diagnoses predisposing to urinary tract obstruction were ascertained for each patient from review of the medical record. Prior abdominal or pelvic malignancy, benign prostatic hypertrophy, nephrolithiasis, ectopic pregnancy, pelvic inflammatory disease, prior pelvic surgery, neurogenic bladder, anatomic genitourinary abnormality, abdominal trauma, and prior renal surgery were considered risk factors for urinary tract obstruction. Obstructive nephropathy was considered the cause of AKI if prompt renal recovery (decrease in serum creatinine within 24 hours) ensued after a urological procedure or bladder catheterization. Renal ultrasound results were obtained from the radiologists’ reports in the electronic medical record.

A total of 4443 RUS were identified over the three-year period. After applying exclusion criteria, 1471 studies remained, of which 810 studies were performed for the evaluation of AKI, and the remaining 661 were performed for other indications (e.g., evaluation of chronic kidney disease, suspected abscess, hematuria, renal colic, and resistant hypertension).

Comparison of demographic and clinical variables was performed among patients with AKI based on the presence or absence of hydronephrosis on renal ultrasound. Statistical differences between the groups with and without hydronephrosis for the tested dichotomous variables were determined using bivariate logistic regression. A multivariable logistic regression model was created to test the independent associations of clinical and demographic variables with presence of hydronephrosis on RUS. All variables which on bivariate logistic regression analysis had an association with hydronephrosis with P-value less than 0.2 were included in the multivariate model. All statistical analyses were performed using Stata, version 12.1 (StataCorp, http://​www.​stata.​com).

Characteristics of the 1471 patients who underwent renal ultrasound and did not meet exclusion criteria are listed in Table 1. Patients who underwent RUS for AKI had a mean age of 63.6 years. There were slightly more men in the group (53.2%) and more than one-third were in the intensive care unit (37%). About one-half of the patients who underwent RUS had history of CKD. Only a small proportion of patients with AKI had a history of nephrolithiasis or benign prostatic hypertrophy (1.9%) while a history of abdominal malignancy was found in 8.2%. Neurogenic bladder, retroperitoneal fibrosis, prior ectopic pregnancy, prior pelvic or renal surgery, prior pelvic inflammatory disease, and abdominal trauma were present in a small number of patients (2.8%). The etiologies of AKI determined on clinical grounds are listed in Table 2.

Table 3 summarizes the findings on RUS for patients with AKI. Hydronephrosis was found in 5.2% (42 of 810) patients with an equal number having unilateral and bilateral abnormalities. Increased renal echogenicity was reported in nearly 40% of studies. Simple cysts were the most common anatomic abnormality found on RUS and were present in about 13% (101 of 810) of studies.

Demographic and clinical characteristics for patients with AKI are presented in Table 4 according to the presence or absence of hydronephrosis. Age 65 years or older, prior diagnosis of abdominal malignancy, and a history of nephrolithiasis were each associated with a higher likelihood of finding hydronephrosis on RUS. The presence of any one of the group of multiple other medical conditions thought to predispose to urinary tract obstruction (neurogenic bladder, retroperitoneal fibrosis, prior ectopic pregnancy, prior pelvic or renal surgery, prior pelvic inflammatory disease, and abdominal trauma) was also associated with a higher likelihood of detecting hydronephrosis on ultrasound.

By multivariable logistic regression a history of abdominal cancer was associated with over 3-fold higher odds of finding hydronephrosis in AKI and reached statistical significance. Having a medical condition in the broad group encompassing other factors thought to predispose to urinary tract obstruction (neurogenic bladder, retroperitoneal fibrosis, prior ectopic pregnancy, prior pelvic or renal surgery, prior pelvic inflammatory disease, and abdominal trauma) was associated with 7-fold higher odds of hydronephrosis. Patients aged 65 years or older had over twice the odds of hydronephrosis being found by renal ultrasound in the setting of AKI (Table 5).

For the patients with AKI who were found to have any hydronephrosis, 54.8% (23 of 42) had known medical conditions predisposing to obstruction. Of the 42 patients with hydronephrosis on RUS, in only 45.2% (19 of 42) was urinary tract obstruction determined to be the cause of AKI. In the remaining patients, hydronephrosis was mild and the cause of AKI was attributed to alternative causes. Of the patients considered to have AKI from urinary tract obstruction, 74% (14 of 19) had at least one of the risk factors associated with obstruction. Thus there were only 0.6% (5 of 810) patients with AKI from urinary tract obstruction without any suggestive medical history, and 4 of these 5 patients were older than 65-years.

Of the 810 renal ultrasounds performed to evaluate for AKI, 336 (41.5%) included Doppler evaluation of renal blood flow. Five of these studies (1.5%) found evidence of renal artery stenosis (RAS), although in none of these cases were those findings thought to contribute to AKI Two studies (0.6%) showed renal artery occlusion, and in both cases this finding had been suspected prior to the study based on the clinical situation. There were no reports of renal vein thrombosis among the Doppler ultrasounds performed for AKI. Resistive indices were reported on only a small number of the studies.