Arozullah AM (Jesse Brown VAMC), Gordon HS
(Jesse Brown VAMC), Yarnold P
(Northwestern University), Soltysik R
(Northwestern University), Ferreira MR
(Jesse Brown VA), Wolf MS
(Northwestern University), Molokie R
(Jesse Brown VA), Bhoopalam N
(Hines VA), Bennett CL
(Jesse Brown VA)
To determine socio-demographic, health behavior, and health status contributions to the rate of advanced stage prostate cancer.
Patients diagnosed with prostate cancer within the prior six months were enrolled at Jesse Brown and Hines VA Medical Centers through the oncology and general medicine clinics. Individuals with dementia, blindness, severely impaired vision, deafness, uncorrectable hearing problems, or too ill to participate were excluded. Each patient was interviewed to assess socio-demographics, health literacy, social support, health status, employment history, health risk behavior, prior cancer screening, health service access and utilization, trust, and satisfaction. Cancer stage and Charlson co-morbidity index were determined by medical and pathology record reviews. Optimal discriminant analysis (ODA) minimizes misclassification and was used to assess the relationship between potential predictors and cancer stage at presentation (stage 4 versus stages 1-3). Classification tree analysis utilized multivariable ODA.
We enrolled 389 patients, and 6.9% presented with stage 4 prostate cancer. There were no significant racial differences in the proportion with stage 4 presentation. Predictors of stage 4 presentation included lower self-reported health responsibility (10.5% vs. 3.2%, p < .02), Charlson index greater than 5 points (46.7% vs. 1.7%, p < .001), fair or poor health status (13.7% vs. 4.8%, p < .01), lower self-efficacy (18.7% vs. 5.4%, p < .05), and lower exercise (11.8% vs. 4.2%, p < .02). Multivariable ODA identified health responsibility as a first attribute/predictor of stage 4 presentation. Among patients with lower health responsibility, higher Charlson score was associated with stage 4 at presentation. Among patients with higher health responsibility, those with worse health status and higher tangible social support were significantly more likely to present with stage 4 disease. Race was not a significant predictor of stage 4 presentation in multivariable analysis.
Co-morbid conditions and health status, rather than socio-demographics, appear to be associated with prostate cancer stage at presentation. Specifically, higher Charlson score and worse self-reported health status predicted stage 4 presentation.
Focusing efforts to improve prostate cancer screening among patients with co-morbid conditions and worse health status may reduce the rate of stage 4 prostate cancers.