Development of a novel RNA biomarker for prostate cancer detection using seminal fluid and mismatched opioid prescriptions following urologic surgery are among the topics being presented by UCSF clinicians and researchers at the Western Section of the American Urological Association’s SurgeWest 2025. The 101st SurgeWest meeting, taking place from Nov. 2-6, in Napa, California, is the largest regional gathering of urology professionals in the Western U.S.
This year’s program features innovative research and discussions by experts from the UCSF Department of Urology.
Sample USF Presentations:
Sunday, Nov. 2, from 10:15 a.m. to 11:45 a.m.
Walter Hsiang, MD, MBA, resident physician in the UCSF Department of Urology, presentation: “Mismatched Opioid Prescription to Patients After Urologic Surgery: A Retrospective Cohort Study” (#87) during the “Health Policy-Data Quality” session. This study evaluated the extent of mismatched opioid prescriptions (over-prescribing and under-prescribing) at the time of hospital discharge among patients undergoing inpatient urologic surgery and assessed the association between prescription mismatch and postoperative opioid refills. The study found that decreasing total dose of opioid alone is not sufficient to effectively decrease opioid refills. Hsiang reports that strategies which individualize discharge opioid dosing based on inpatient consumption may help reduce prescription mismatch, improve postoperative outcomes and enhance opioid stewardship.
Hsiang also presentation: “Use of Commercial Large Language Models to Expand Literacy Level Concordant Materials for Kidney Stone Patient Education” (#187) during the “Health Policy-Data Quality” session. In the United States, nearly 9 out of 10 patients have sub-proficient health literacy, which is associated with increased patient morbidity. In this study, Hsiang and his colleagues evaluated the accuracy and completeness of three commercial large language models (LLMs) in converting standard kidney stone education materials to the sixth-grade reading level as an initial assessment of these tools’ potential to expand access to health information for urologic patients with low literacy. Hsiang reports on LLM’s accuracy when converting kidney stone patient education materials into lower reading grade levels.
Lynn Leng, BS, medical student in the UC Berkeley–UCSF Joint Medical Program, presentation: “Rethinking Morbidity and Mortality Conferences: A Novel, Multi-Institutional Study of the M-PROVE Model” (#49) during the “Health Policy-Data Quality” session. Morbidity and mortality conferences (MMCs) serve as critical platforms to evaluate patient adverse events and promote continuous improvement in clinical practice and health systems. In 2022, UCSF piloted the Morbidity and Mortality Process Redesign to Optimize Value and Education (M-PROVE) model as a standardized approach to improve educational value and increase opportunities for systems change in patient care in an inclusive, nonjudgmental manner. This study sought to implement the M-PROVE model at three additional academic urology institutions to evaluate its effectiveness and to identify any potential barriers. Leng reports on how the M-PROVE model can be implemented at other institutions and can effectively improve attendee attitudes and enhance the perceived value of MMCs across multiple domains. Leng additional presentation: “Social Predictors of No Shows at an Urban, Safety-Net Urology Clinic: A Mixed Effects Model Study” (#211) during the “Health Policy-Data Quality” session. Clinic non-attendance negatively affects patient outcomes and disrupts clinic workflow. This study aims to identify predictors of patient no-shows at an urban, safety-net urology clinic. Predictors of urologic clinic no-shows were a history of being unhoused, historical no-show rate, inactive patient portal and increased time from appointment scheduling, of which the latter two are intervenable. This study offers upstream solutions for improving attendance for a safety-net clinic and healthcare access for urologic patients.
Amy Showen, MD, MSc, resident physician in the UCSF Department of Urology, presentation: “Characterizing 7-Day Hospital Revisits Following Outpatient Urologic Procedures: A Single-Institution Analysis Using Vizient Data” (#134) during the “Health Policy-Data Quality” session. The seven-day revisit rate is a metric gaining traction as a measure of healthcare quality and performance. Vizient, a leading healthcare performance improvement company, supports this metric’s adoption for internal benchmarking and national ranking. This study aims to characterize seven-day revisit rates after outpatient urologic procedures, identify procedure-related versus unrelated revisits, and compare revisit patterns between clinic-based and operating room (OR)-based procedures. Showen reports on the nature of seven-day hospital revisits and examines the opportunities for targeted interventions to reduce unnecessary emergency and inpatient utilization.
Marvin Carlisle, BA, UCSF medical student, presentation: “Development and Validation of a Generative Artificial Intelligence-Based Pipeline for Automated Clinical Data Extraction from Electronic Health Records: Technical Implementation Study” (#206) during the “Public Policy-Data Quality” session. Manual abstraction of unstructured clinical data for clinical research is time consuming and can be of variable quality. Large language models (LLMs) show promise in medical data extraction, yet integrating them into research workflows remains challenging and poorly described. The objective of this study was to develop and integrate an LLM-based system for automated data extraction from unstructured electronic health record (EHR) text reports within an established clinical outcomes database. Carlisle reports on the study’s successful integration of an LLM-based system for automated report extraction within an existing outcomes database and how this approach could significantly accelerate research timelines and expand feasible clinical studies for large-scale projects.
Monday, Nov. 3, from 2:15 p.m. to 3:30 p.m.
Kevin Shee, MD, PhD, resident physician in the UCSF Department of Urology, presentation: “Development of a novel RNA biomarker for prostate cancer detection using seminal fluid: results from a large multicenter clinical study” (#139) during the “Prostate Cancer 1” session. Prostate-specific antigen (PSA) testing remains the standard for prostate cancer (PCa) screening, but its limited specificity often leads to unnecessary biopsies and overdiagnosis. Given that the prostate contributes to approximately one-third of seminal fluid, this accessible biofluid represents a promising, noninvasive source for biomarker discovery. In this multicenter study, Shee and his colleagues developed a novel biomarker combining seminal fluid RNA expression with clinical data to detect clinically significant PCa (csPCa). Combining seminal fluid RNA with PSA and age improved csPCa detection.
Shee additional presentation: “External validation of a pathology-based multimodal artificial intelligence biomarker for predicting prostate cancer outcomes after prostatectomy” (#132) during the “Prostate Cancer 1” session. Radical prostatectomy (RP) improves survival in localized prostate cancer (PCa), but 20%-40% of patients experience biochemical recurrence (BCR) within 10 years, with one-third progressing to metastasis. Predictive tools for stratifying post-RP risk remain limited. Shee and his colleagues previously developed and validated a digital pathology-based multimodal AI (MMAI) model (RP MMAI v1.1) using H&E images and clinical data to predict outcomes in BCR patients. In this study, Shee presents its first external validation in both BCR and non-BCR post-RP patients.