Advertisement

Regional variation in the diagnosis and treatment of prostate cancer in a Medicare population

Login to Access Video or Poster Abstract: MP47-15
Sources of Funding: The Hitchcock Foundation

Introduction

The Dartmouth Atlas has reported wide variation among geographic regions in the diagnosis and treatment of prostate cancer. The relationship between these events, however, is not well understood, and merits attention because of uncertainty regarding appropriate levels of diagnosis and treatment in older patients. We sought to delineate regional associations between diagnosis and treatment of prostate cancer in a Medicare population._x000D_ _x000D_

Methods

We performed a cross-sectional analysis of a 100% Medicare fee-for-service sample from 2012, including men aged 68 years or older. The unit of analysis was hospital referral region (HRR) (n=306). Rates of prostate cancer diagnosis and definitive treatment (surgery or radiation therapy) were calculated, adjusting for age and race. Correlation coefficients were calculated between these events. Regions were stratified by quartile for each practice, and those with concordant or discordant practices were identified.

Results

There was wide regional variation in both the diagnosis and treatment of prostate cancer. Rates of diagnosis ranged from 0.4-2.2% (1.0%),and rates of treatment ranged from 0.0-0.3% (0.11%). There was moderate correlation between events (r=0.36; p<0.0001), which persisted within age-stratified samples. However, there was a wide range of both concordant and discordant practices in regional prostate cancer care, e.g. low diagnosis + high treatment, high diagnosis + low treatment, etc.

Conclusions

There is wide variation in regional rates of prostate cancer diagnosis and treatment among older Americans. Diagnosis and treatment correlated overall, highlighting risk of a “cascade” of events post-diagnosis and underscoring concerns for over-detection in the elderly. Interestingly, some regions evidenced “counterintuitive” practice patterns, e.g. low rates of diagnosis and high rates of treatment, and vice versa. Further study is needed to understand the impact of variation on prostate cancer care.

Funding

The Hitchcock Foundation

Authors
Rachel A. Moses
Andrea M. Austin
Donald Carmichael
Elias S. Hyams
back to top