Grant Number:	1R01CA300038-01A1 Interpret this number
Primary Investigator:	Daskivich, Timothy
Organization:	Cedars-Sinai Medical Center
Project Title:	Integration of Life Expectancy Estimates Into Electronic Health Record Decision Support for Men with Prostate Cancer
Fiscal Year:	2026

PROJECT SUMMARY Life expectancy (LE) is a critical factor in prognosis and treatment decision making for men with prostate cancer (PC). Limited LE is associated with lower likelihood of sufficient longevity to benefit from treatment, higher morbidity after treatment, and decreased treatment effectiveness. As a result, LE is the first triage point for all PC risk subtypes in PC guidelines, identifying men who are best served with observation rather than aggressive treatment. The population of men with localized PC and limited LE who fit these criteria is sizeable; we recently found that a third of men over 65 years and one half of men over 75 diagnosed with PC have LE<10 years. Yet despite the prominent role of LE in guidelines, patients with limited LE are still often overtreated. Of men with PC over age 65 with LE<10 years in the 2000s era, we previously reported that over half were treated aggressively with surgery or radiation. More recently in the active surveillance era, we found that overtreatment of men with limited LE has continued and even increased for high-grade cancers. One reason for this mismanagement is that providers don’t have accurate, patient-specific LE estimates available at the point of care. To improve the uptake of LE, we developed a PC-specific LE-prediction tool based on a weighted scale of age and comorbidities. This scale—the PC-specific comorbidity index (PCCI)—was externally validated across 181,000 men in a national VA sample. Its simple computation and operationalization of comorbidity data using ICD codes makes the PCCI ideally suited for integration into EHR clinical decision support pathways. We have created PCCI-based LE tools in EPIC and in the national Veterans Health Administration (VA) note template but have delayed clinical deployment of these tools awaiting funding to examine their impact on decision making. We herein propose an implementation study to evaluate the impact of patient-specific, automated LE estimates in EHR clinical decision support pathways in both the national VA and four academic centers. We will examine the effectiveness of PCCI-based LE estimates at reducing over- and undertreatment, as well as implementation outcomes focusing on facilitators and barriers of uptake, including physician risk communication, patient and physician risk perception, and patient decisional conflict. We will also use patient and physician input to optimize data visualization and identify barriers prior to implementation. We hypothesize that integration of LE estimates into the EHR will allow physicians to more easily implement guidelines for management based on LE. Having a quantitative, patient-specific reference at the point of care will improve quality of physician communication of LE, which will increase patient knowledge about LE and reduce decisional conflict, ultimately leading to reduction of over- and undertreatment at a population level.





None