Abstract: PD57-01
Sources of Funding: none

Introduction

To determine the incidence and predictors for malignancy in patients who undergo workup for microscopic hematuria.

Methods

We conducted a prospective cohort study of patients undergoing evaluation for asymptomatic microscopic hematuria from January 2009 to July 2016 in an integrated managed care organization in Southern California. Patients' accompanying diagnoses and baseline cohort characteristics were determined and identified using our comprehensive electronic health record system. Cancers indicated in the workup were validated by chart review. Additional cancers documented in the institutional cancer registry through July 2015 were included if diagnosed within 1 year of initial workup. Bivariate associations were assessed using the chi-square test; multivariable logistic regression was used to build a predictive risk model and create a hematuria risk index.

Results

Within a cohort of 6417 patients with microscopic hematuria, a total of 177 (2.8%) were diagnosed with a neoplasm. On multivariate analysis, age between 50-59 (OR=1.96, 2 points), age over 60 (OR=5.21, 4 points), history of gross hematuria (OR=3.15, 3 points), current or past smoking history (OR=1.51, 1 point), male gender (OR=2.57, 2 points), >25 red blood cell per high power field (OR=2.94, 2 points), Non-Hispanic Black (OR=1.73, 1 point), and Non-Hispanic White (OR=2.31, 2 points) were all significant predictors of malignancy. A modified Hematuria Risk Index (0 to 14 points) was developed from these factors, which demonstrated an improved area under the receiver operating characteristic curve of 0.841 compared to our previous model at 0.807. We observed natural breaks in the scores that grouped the patients into low (0-4 points, 41.7%), moderate (5-9 points, 49.0%), and high-risk of cancer (10-14 points, 9.2%). Malignancy was found in 0.4%, 2.5% and 15.0% of patients from the low, moderate and high-risk groups, respectively.

Conclusions

Advance age, history of gross hematuria, current or past smoking history, male gender, >25 red blood cell per high power field, and certain ethnic groups are significant predictors for malignancy in the setting of microscopic hematuria. Classification of patients into low, moderate and high-risk groups will improve patient counseling and will hopefully reduce the need for invasive endoscopy and ionizing radiation exposure for patients within the low-risk category.

Funding

none