1073 — The Effect of Healthcare System on Appropriateness of Prostate Cancer Imaging: Do Patients Get Better Care in VA or Fee-for-Service Medicare?
Makarov DV, VA NY Harbor Healthcare System; Gold HT, NYU School of Medicine; Walter D, NYU School of Medicine; Eugenio E, VA Puget Sound Healthcare System; Sherman S, VA NY Harbor Healthcare System; Gross CP, Yale School of Medicine; Zeliadt SB, VA Puget Sound Healthcare System;
To compare the appropriateness of prostate cancer imaging among patients treated in the VA (an integrated healthcare system) versus those treated in a Fee-for-Service Medicare system.
We performed a retrospective cohort study of patients with incident prostate cancer. In order to compare appropriateness of prostate cancer imaging between VA and Medicare patients, we created a novel dataset by merging data for patients with incident prostate cancer diagnosed from 2004-2008 in the VA Central Cancer Registry and the SEER-Medicare linked database. We used a logistic regression model clustered by VA region to test the association between appropriateness of prostate cancer imaging and healthcare system. Our dependent variable was appropriateness of prostate cancer imaging, as defined by the National Comprehensive Cancer Network; we stratified our analysis based on guidelines that a patient should have imaging ("high risk") or should not have imaging ("low risk"). Our independent variable of interest was healthcare system (VA, Fee-for-Service Medicare, or VA with utilization of Medicare services). We adjusted for clinicopathologic and sociodemographic characteristics that could confound this relationship.
We identified 27,811 VA only, 14,385 VA with some Medicare, and 56,671 Medicare only patients. Among low risk patients in whom imaging is not guideline-indicated, patients treated in the VA (39%) and VA patients utilizing some Medicare services (44%) had less inappropriate prostate cancer imaging than did their Medicare (47%) peers (RR:0.79; 95%CI: 0.67-0.92 vs RR: 0.87; 95%CI 0.76-0.98, respectively where Medicare is the reference). However, among high risk patients where imaging is guideline-indicated, there was no difference in appropriate utilization of imaging between all three groups: VA (69%), VA with some Medicare (72%) and Medicare only (67%); RR 1.00; 95%CI 0.95-1.06 vs RR 1.04; 95%CI 0.98-1.09, respectively where Medicare is the reference.
Using a novel, merged dataset,we determined that patients with incident prostate cancer treated in the VA received the same amount of appropriate imaging as compared to their peers in Medicare but received 21% less inappropriate imaging.
Our results show that an integrated healthcare system with a lack of physician financial incentives to provide more care can deliver optimal imaging for prostate cancer patients.