Related Experiment Video
Updated: Jun 30, 2026
A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Physician, affective, and cognitive variables differentially predict 'initiation' versus 'maintenance' PSA screening
Nathan S Consedine1, Michael A Christie, Alfred I Neugut
1Department of Psychology, Long Island University (Brooklyn Campus), Brooklyn, New York, USA. nconsedi@liu.edu
Objectives:
To evaluate the relevance of demographic, physician, and psychological characteristics to PSA screening in ethnic subpopulations and ascertain whether the same characteristics distinguish men who have never had a PSA from those who screen infrequently and those who screen yearly (adhere).
Design And Methods:
Stratified cluster-sampling was used to recruit 533 men (45-70 years) from four ethnic groups: African-American; European-American; immigrant Jamaican; and immigrant men from Trinidad and Tobago. Men provided demographic and structural (insurance, regular physician, annual exam, and physician recommendation), cognitive (risk and efficacy perceptions, knowledge), and emotional variables (cancer worry and embarrassment), and reported on PSA screening history. Multinomial logistic regression used these variables to predict three screening classifications (never screened, partially adherent, and adherent).
Results:
Multinomial logistic regression showed that minority men were less likely to report either never screening or yearly screening, while younger men were more likely. Lack of a regular physician (OR=2.87, 95% CI 1.39-5.84), an annual exam (OR=1.73, 95% CI 0.91-3.28), and low recommendation (OR=3.76, 95% CI 2.13-6.66) were associated with being categorized as a never (vs. partially adherent) screener, but only annual exam (OR=0.26, 95% CI 0.10-0.63) was associated with yearly screening. Lower cancer worry was marginally associated with never screening (OR=0.59, 95% CI 0.38-1.04), while knowledge was associated with screening yearly over time (OR=0.46, 95% CI 0.28-0.77).
Conclusions:
Demographic, physician, and psychological variables are differentially associated with never, less than yearly, and yearly screening classifications. Minority men were unlikely to have never screened, but were also less likely to screen yearly. Physician variables were associated with the difference between not screening and partially adherent, but not between partially adherent and yearly screening suggesting that the role of physicians in PSA behaviour over time would benefit from further study.
