RACIAL DISPARITIES IN CONTEXT: A MULTILEVEL ANALYSIS OF CONSERVATIVE MANAGEMENT IN MEN WITH LOW RISK PROSTATE CANCER

Purpose:

To examine the impact of county-level characteristics on African American (AA)/White (W)  disparities in the receipt of conservative management among older men with low risk prostate cancer (PCa).

Method:

We analyzed AA and W men aged 66 years or older, diagnosed with low risk PCa (AJCC classification: stage≤T2a,PSA≤10 ng ml^-1,Gleason score≤6) between 2000 and 2005 from the linked SEER-Medicare dataset. The data was enriched with US census (2000) and county business pattern data (2000-2005). Early treatment during the first 12 months of diagnosis was categorized as, 1) conservative management (CM) and, 2) any treatment receipt including radical prostatectomy, radiation, chemotherapy and hormone therapy. County-level measures included percent below poverty-level, education-level and number of health care facilities and services available per-capita.  The variation in patterns of conservative management across US counties were examined using caterpillar plots. Variance partition coefficients (VPC) were estimated to quantify the variation in CM patterns attributable to differences across the counties. Two-level random intercept/slope logit models were used to examine the cluster-specific effect of race on the receipt of CM, controlling for individual and county-level characteristics. 

Result:

Application of the inclusion criteria resulted in 22,519 low risk PCa patients. Majority (90%) of the patients were W and 20% of the patients received CM. The average age was 74 (SD:6) years. AAs were more likely to receive CM than Ws (AA: 24%, W: 19%, p<0.01). The caterpillar plot illustrates the variation across counties in the receipt of CM (Figure:1). VPC showed that 4.4% of the overall variation in receipt of CM was due to differences in contextual factors. After adjusting for individual and county-level characteristics, an AA patient was more likely to receive CM compared to a W patient living in the same county (OR: 1.48, 95% CI: 1.32- 1.67, p<0.01). The effect of race on the receipt of CM varied across counties (p<0.01). Among AAs 15.4% and, among Ws 4.6% of the remaining variance in the likelihood of receiving CM was due to between-county variation. 

Conclusion:

Counties vary in receipt of CM among AA and W low-risk PCa patients. More research is needed to characterize the role of county-level characteristics in the receipt of CM.

 

 

 

 

Figure 1: Caterpillar plot for full sample