Unraveling the Multilevel Factors Influencing Transition to Radical Treatment in Medicare Beneficiaries Newly Diagnosed With Low-Risk and Favorable Intermediate-Risk Prostate Cancer Initiating Active Surveillance: A Fine-Gray and Cause-Specific...
Author(s)
Bernard Bright K. Davies-Teye, MD, MPH, PhD1, Eberechukwu Onukwugha, MSc, PhD1, M. Minhaj Siddiqui, MD, FACS2, Julia F. Slejko, MS, PhD1, Zafar Zafari, PhD1, C. Daniel Mullins, PhD1.
1Department of Practice, Sciences, and Health Outcomes Research, University of Maryland School of Pharmacy, Baltimore, MD, USA, 2University of Maryland School of Medicine, Baltimore, MD, USA.
1Department of Practice, Sciences, and Health Outcomes Research, University of Maryland School of Pharmacy, Baltimore, MD, USA, 2University of Maryland School of Medicine, Baltimore, MD, USA.
OBJECTIVES: Active surveillance (AS) is a guideline-recommended strategy for managing low-risk and favorable intermediate-risk localized prostate cancer (LIPCa: cT1-cT2c, cN0, cM0, grade group (GG)≤2), reducing overtreatment and complications from radical treatment (RT). However, many Medicare beneficiaries with LIPCa initiating AS transition early to RT. This study quantified the association between multilevel factors and RT transitions.
METHODS: We conducted a retrospective cohort study using SEER-Medicare data (2006-2020) linked to PolicyMap Census tract-level social determinants of health data. Incident LIPCa cases were identified using tumor-node-metastasis and American Urological Association risk stratification algorithms. AS initiation within 18 months post-diagnosis was identified using a validated algorithm. Adherence to AS protocols and transitions to RT were identified using procedure/ICD-9/10-CM codes. We estimated adjusted subdistribution hazard ratios (asdHRs; 95% CIs) using Fine-Gray subdistribution hazard models. Sensitivity analyses estimated a cause-specific hazard model with a frailty term and examined risk-stratified subgroups.
RESULTS: Among 2,826 AS initiators (median follow-up=4.2 years; 95% CI=4.0-4.3), 34.1% transitioned to RT. RT transitions were lower among the low-risk subgroup (asdHR=0.815; CI=0.696-0.955), unmarried (asdHR=0.803; CI=0.667-0.967), or those with BPH (asdHR=0.835; CI=0.726-0.961). RT transitions were higher in patients with prior AMI (asdHR=1.823; CI=1.017-3.270) or AS initiation in 2014-2018 (vs. 2007-2009: asdHR=1.453-1.590). Among low-risk patients, diabetes was protective (asdHR=0.726; CI=0.557-0.945), while hyperlipidemia increased RT transitions (asdHR=1.283; CI=1.059-1.553). Among favorable intermediate-risk patients, CHF (asdHR=2.205; CI=1.005-4.837) and obesity (asdHR=1.957; CI=1.185-3.232) increased transitions. Residing in tracts with the highest internet access (asdHR=1.283; CI=1.027-1.603) and moderate cost-burdened renters (asdHR=0.476; CI=0.248-0.912) was associated with higher and lower RT transitions, respectively. Being managed by radiation (vs. urologists: asdHR=1.955; CI=1.372-2.784) and surgical (asdHR=2.735; CI=1.005-7.442) oncologists increased RT transitions, while high-volume practices (asdHR=0.850; CI=0.738-0.980) reduced RT transitions.
CONCLUSIONS: Among AS initiators, patient-, neighborhood-, and provider-level factors were independently associated with RT transitions. Further studies should investigate the causal mechanisms linking these factors and RT transitions.
METHODS: We conducted a retrospective cohort study using SEER-Medicare data (2006-2020) linked to PolicyMap Census tract-level social determinants of health data. Incident LIPCa cases were identified using tumor-node-metastasis and American Urological Association risk stratification algorithms. AS initiation within 18 months post-diagnosis was identified using a validated algorithm. Adherence to AS protocols and transitions to RT were identified using procedure/ICD-9/10-CM codes. We estimated adjusted subdistribution hazard ratios (asdHRs; 95% CIs) using Fine-Gray subdistribution hazard models. Sensitivity analyses estimated a cause-specific hazard model with a frailty term and examined risk-stratified subgroups.
RESULTS: Among 2,826 AS initiators (median follow-up=4.2 years; 95% CI=4.0-4.3), 34.1% transitioned to RT. RT transitions were lower among the low-risk subgroup (asdHR=0.815; CI=0.696-0.955), unmarried (asdHR=0.803; CI=0.667-0.967), or those with BPH (asdHR=0.835; CI=0.726-0.961). RT transitions were higher in patients with prior AMI (asdHR=1.823; CI=1.017-3.270) or AS initiation in 2014-2018 (vs. 2007-2009: asdHR=1.453-1.590). Among low-risk patients, diabetes was protective (asdHR=0.726; CI=0.557-0.945), while hyperlipidemia increased RT transitions (asdHR=1.283; CI=1.059-1.553). Among favorable intermediate-risk patients, CHF (asdHR=2.205; CI=1.005-4.837) and obesity (asdHR=1.957; CI=1.185-3.232) increased transitions. Residing in tracts with the highest internet access (asdHR=1.283; CI=1.027-1.603) and moderate cost-burdened renters (asdHR=0.476; CI=0.248-0.912) was associated with higher and lower RT transitions, respectively. Being managed by radiation (vs. urologists: asdHR=1.955; CI=1.372-2.784) and surgical (asdHR=2.735; CI=1.005-7.442) oncologists increased RT transitions, while high-volume practices (asdHR=0.850; CI=0.738-0.980) reduced RT transitions.
CONCLUSIONS: Among AS initiators, patient-, neighborhood-, and provider-level factors were independently associated with RT transitions. Further studies should investigate the causal mechanisms linking these factors and RT transitions.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
EPH272
Topic
Epidemiology & Public Health, Health Service Delivery & Process of Care, Methodological & Statistical Research
Topic Subcategory
Safety & Pharmacoepidemiology
Disease
Oncology, Personalized & Precision Medicine, Reproductive & Sexual Health, Surgery