Microhematuria findings contribute to evidence base

The following is in response to an editorial by J. Brantley Thrasher, MD (“In support of AUA guidelines,” February 2013, page 7). The editorial concerned the article, “Microhematuria a poor predictor of urinary tract cancer” (February 2013, page 8).

My co-authors and I would like to thank Dr. J. Brantley Thrasher for his interest in our work. As Dr. Thrasher appropriately indicated, the AUA asymptomatic microhematuria guidelines panel, led by Dr. Rodney Davis, performed a thorough analysis of the available evidence and are to be commended for their work. They concluded the guideline recommendation with a request for researchers to further investigate and report more evidence for this important topic. We feel that our findings contribute to that knowledge base.

We would like to clarify a few points regarding Dr. Thrasher’s observations of our report:

First, nearly two-thirds of the entire cohort had two positive urinalyses (73% original, 52% validation cohorts), which gave us enough variability in referrals to validate the AUA microhematuria panel’s findings that two of three urinalyses are no more accurate than a single positive UA, as reflected in the new 2012 AUA guideline. In our study, there was no difference in outcome between patients referred with two positive UAs versus one positive UA.

Second, 100% of the cohort’s 4,414 patients underwent cystoscopy, and all pathologic diagnoses of malignancy (bladder, renal cancer) were confirmed. As we described, while 151 of our urologists in four independent geographic regions participated in the study, some did not. To evaluate the potential for selection bias, we performed a manual chart review of a random sample of those hematuria evaluations that were not included in the study sample. We found that every case was fully worked up, including cystoscopy, and the cancer detection rate was identical to the study group.

Third, this study was not simply a chart review of an EMR system. All data were prospectively collected directly from our 151 urologists utilizing an innovative data collection tool called a SmartList that was embedded into our Epic EHR. This tool collected information in a way that allowed us to electronically extract all relevant findings that were input directly from our urologists in coded fields. Historically, these data have been only possible to acquire through a manual chart review. Thus, we were able to electronically discern a history of gross hematuria in the past 6 months, the urologist’s interpretation of imaging, the urologist’s finding at the time of cystoscopy, and the urologist’s diagnosis.

Through other data systems linked to our EMR, we were able to link urinalysis findings, gender, age, race, ethnicity, and smoking history for 100% of the cohort. All pathologic findings were validated and confirmed electronically.

Again, my co-authors and I thank the editor for his interest in our work and thank the AUA for their leadership and ongoing support to improve urologic health worldwide. We fully support the work of the AUA guidelines process, and we agree with Dr. Thrasher that the recommendations will only be as good as the evidence on which they are based. The AUA Health Policy Council has undertaken an initiative to capitalize on the attributes of “meaningful use” and the potential that the electronic health record can deliver if it can be harnessed to answer the most important outstanding questions that face us today.

Finally, we understand that work is already under way elsewhere to validate our findings, and we plan to report additional findings from our institution in the near future.

Ronald Loo, MD

Downey, CA