Introduction: Left-digit bias is a phenomenon in which the leftmost digit of a continuous variable disproportionately influences decision-making. We examined left-digit bias towards age of patients undergoing treatment for intermediate-risk prostate cancer.
Methods: Men aged 65 to 74 years with clinically localized Gleason grade 7 prostate cancer were identified in Surveillance, Epidemiology, and End Results (SEER) and National Cancer Database (NCDB) from 2004 to 2018. Regression discontinuity analysis was used to determine if left digit age change from 69 to 70 was associated with unexpected change in treatment recommendations. Patients with Charlson score 0 were separately evaluated within the NCDB cohort.
Results: We identified 115,223 patients from SEER and 198,376 patients from NCDB. In the SEER cohort, discontinuity of -2.4% was observed from age 69 to 70 in patients recommended surgery (95% CI: -4.0 – -0.9%, p=0.02) and 2.8% in patients recommended radiation (95% CI 1.4 – 4.2%, p<0.01). Discontinuity was also observed in the NCDB cohort from age 69 to 70, with a 4.0% decrease in surgery recommendation (95% CI: -5.6 – -2.4%, p<0.01), and a 4.2% increase in radiation recommendation (95% CI: 2.7 – 5.7%, p<0.01). Significant discontinuity remained in NCDB patients with Charlson score 0 (surgery: -4.1%, 95% CI: -5.7 – -2.5%, p<0.01; radiation: 4.3%, 95% CI: 2.5% - 6.1%, p<0.01), and no significant difference in Charlson score was found for age 69 vs. 70 (mean score 0.24 vs. 0.24, p=0.44).
Conclusions: Left digit age change from 69 to 70 is associated with disproportionately increased recommendation for radiation therapy and disproportionately decreased recommendation for surgery in patients with intermediate-risk prostate cancer. Physicians treating patients for prostate cancer should be aware of this cognitive bias.
.jpg)
.jpg)
