br In the past few years there has
In the past few years there has been a dramatic expan-sion in the support for MRI in the evaluation of men with known or suspected PCa.10 Recent studies have under-scored the ability of contemporary prostate MRI to offer high resolution anatomical assessment of the prostate including improved prediction of high-grade or high-stage disease.11 Moreover, MRI facilitates the performance of targeted MRI-guided in-bore, cognitive, or MR-ultra-sound fusion biopsies, which improve detection yields for occult higher-grade cancer that would be missed on
Published by Elsevier Inc.
systematic transrectal ultrasound guided biopsy alone. On the basis of encouraging single and multi-institutional controlled studies, prostate MRI is increasingly recom-mended as an important staging tool that can identify
occult aggressive disease and facilitate enrollment in AS.12
Despite growing enthusiasm, there is a lack of popula-tion-level evidence that addresses national utilization trends, or practical impact of prostate MRI on disease management. Therefore, we assessed the use of prostate MRI in a population-based, nationally representative cohort of Medicare beneficiaires with low-risk PCa, evalu-ated patient and healthcare level factors predictive of MRI use, and examined the association between prostate MRI and initial management with observation or defini-tive therapy. We tested the hypotheses that early MRI use was variable across socio-demographic and provider-level factors, and that use of MRI was associated with greater likelihood of observation as initial management.
MATERIALS AND METHODS
We used Surveillance, Epidemiology, and End-Results (SEER) records linked with individual level Medicare claims to identify men with PCa. The SEER-Medicare (±)-Baclofen provides a high level of detail regarding healthcare delivery among older Ameri-cans including clinical and demographic characteristics, treat-ments received, and subsequent outcome.13 The most recent linkage released in April 2017 includes patients diagnosed through 2013. Patients included in this study were diagnosed with nonmetastatic, lymphnode negative, first primary, low-risk PCa at the age of ≥ 66 years in 2010-2013 according to the D’Amico classification: PSA<10 ng/mL, Gleason score ≤ 3 + 3, and clinical stage ≤ T2a.14 Patients were excluded on the basis of absence of Medicare Parts A and/or B, health maintenance organization membership in the 12 month period preceding diagnosis until December 31, 2014, or death, unknown date of diagnosis, incidental detection of PCa on death certificate or at autopsy, missing tumor grade, stage or PSA level. To assure patients received MRI or treatment, we further excluded patients who died within 1 year after diagnosis.
We identified prostate MRI obtained in the period surrounding diagnosis of PCa by common procedural terminology codes (72195, 72196, and 72197). To identify prostate MRI that was performed for the purpose of local staging or risk assessment, we included studies performed in a 19-month window including a 6-month period preceding the month of diagnosis and 12 months following the month of diagnosis. To account for the possibility that MRI studies were obtained during observation to assess for disease progression, rather than for initial staging, we performed sensitivity analyses limiting the time window of inter-est to 6-9 months after diagnosis, as well as, a window of 3 months preceding until 3 months following the diagnosis of PCa. To account for MRI studies performed for local treatment planning, we regarded prostate MRI studies performed within the 28 days before the initiation of therapy as a component of treatment, and excluded them from analysis. We also performed