3046 — Identifying Provider Factors associated with PSA screening among Elderly Veterans with Limited Life Expectancy.

Tang VL, San Francisco VA Medical Center; Shi Y, San Francisco VA Medical Center; Fung K, San Francisco VA Medical Center; Espaldon R, San Francisco VA Medical Center; Tan J, San Francisco VA Medical Center; Sudore R, San Francisco VA Medical Center; Walter L, San Francisco VA Medical Center;

Objectives:
Despite guidelines recommending against prostate-specific antigen (PSA) screening in elderly men with limited life expectancy, screening is common. The decision whether to screen rests on the health care provider, however little is known about the provider-level determinants of PSA screening. We sought to identify provider characteristics that are associated with PSA screening rates in elderly men with limited life expectancy.

Methods:
We conducted a cross-sectional study of 826,286 men who were age > = 65 years, eligible for PSA screening, and had laboratory tests performed at 1 of 130 VAs in 2011. Limited life expectancy was defined as age > = 85 with Charlson score > = 1 or age > = 65 with Charlson score > = 4 in 2011. Providers who ordered the index screening PSA or ordered the majority of the patient's lab tests in 2011 were identified. The primary outcome was the percentage of men who had a screening PSA at a VA in 2011 and the percentage with limited life expectancy. The primary predictors were provider characteristics, including provider type, training level, gender, and specialty, collected from the VA Corporate Data Warehouse. We performed log-Poisson regression models for the association between each provider characteristic and PSA screening among men with limited life expectancy adjusting for patient characteristics (age, race, marital status, income, and education) and provider clustering.

Results:
Among the 203,717 (25%) men with limited life expectancy, 39% received PSA screening. Of providers who ordered the majority of labs in men with a screening PSA, 69% ordered the PSA. After adjusting for patient characteristics, higher PSA screening in men with limited life expectancy was associated with provider training level (42% for attendings/nurse practitioner/physician assistant versus 26% for resident/fellow trainees, P < 0.001) and specialty (82% Urology versus 41% Medicine, P < 0.001). Provider type and gender were not predictive.

Implications:
Over a third of men with limited life expectancy are receiving PSA screening. Physician trainees have substantially lower PSA screening rates than physician attendings, nurse practitioners or physician assistants as do generalist providers versus urologists in men with limited life expectancy.

Impacts:
Interventions to reduce PSA screening in elderly men with limited life expectancy should be targeted to high screening providers.