Introduction: Low-value PSA testing is responsible for substantial waste and potential harm to patients, yet it accounts for nearly half of prostate cancer (PCa) screening. Decision fatigue, the progressive decline in consistency and quality of choices with repetitive decision-making, has been observed in breast and colorectal cancer care. Our aim was to determine whether low-value PSA testing patterns by outpatient providers are consistent with decision fatigue.
Methods: Men without PCa at a large academic health system from July 2011–June 2018 were identified. Outpatient encounters across various specialties were stratified by clinical guidelines as whether a PSA test order would be appropriate or low-value. The primary endpoint was whether a PSA test was ordered. Logistic generalized estimating equations were used to analyze PSA test likelihood by appropriateness, with spline functions representing trends by hour. Models were adjusted for patient-, provider-, and appointment-level factors and stratified by specialty.
Results: Of 1,761,815 outpatient encounters, a PSA test was ordered during 3.7% when it would be considered appropriate and 1.8% when it was low-value. The overall likelihood of an encounter resulting in any PSA test was greatest at 8:00am, tapering off by 12:00pm (OR=0.66; 0.58–0.75) and persisting through 4:00pm (OR=0.65; 0.56–0.77). Testing patterns differed between specialties (Figure), with non-urologists exhibiting a proportionately greater decline in the likelihood of an appropriate test as the day progressed. Urologists showed a different pattern, with appropriate decisions relatively preserved even as inappropriate testing declined in the middle of the day.
Conclusions: In a framework where PSA testing decisions are considered complex and the default position is to not test, PSA testing among non-urologists is consistent with decision fatigue. That is, testing is most likely early in the day when complex decisions are easiest. The pattern of PSA testing among urologists is different and may not be explained by decision fatigue. This may be due to differences in the default position for testing or greater PSA testing expertise, which lightens the cognitive load of decision-making. Source of