Dupless: Toward a patient-friendly approach for erectile dysfunction nature differentiation – a study of 291 penile duplex Doppler ultrasound assessments

Abstract

Background: Erectile dysfunction (ED) is a condition commonly classified as either psychogenic or organic. Traditional age-based categorizations are considered overly simplistic, yet many clinicians continue to rely on initial evaluation—patient symptoms and history, physical examination, blood tests, and questionnaires—for diagnosis due to limited modern tools. Objectives: This study aims to evaluate the predictive value of patient characteristics in individuals with “ED of indeterminate origin” following initial evaluation. Identifying these variables could enhance early diagnosis and reduce reliance on invasive procedures. Materials and methods: A retrospective cohort study was conducted on patients who underwent penile duplex Doppler ultrasound between January 2018 and January 2024 due to “ED of indeterminate origin”. Patient data, including demographics, lifestyle factors, and medical history, were collected and analyzed using unpaired t-tests, chi-squared tests, Fisher's exact tests, and multivariate logistic regression to assess their predictive value. Results: Among the 291 patients in the cohort, 165 (56.7%) were diagnosed with organic ED and 113 (38.8%) with psychogenic ED. Significant differences in age, history of diabetes mellitus, and drug use were noted. Logistic regression revealed multicollinearity among the variables and explained only 5.8% of the variance in ED etiology. Subgroup analysis revealed that diabetes mellitus predicts organic ED in patients aged 40 years and older, while psychopathology is linked to psychogenic ED. No significant predictors were identified for patients under 40 years. Discussion and conclusion: The findings of this “Dupless” study highlight the limitations of relying solely on initial evaluation to differentiate ED etiology, stressing the need for additional diagnostic tools. While some predictive factors were identified, they proved insufficient for clinical use. Thus, an urgent need exists for the development of modern, noninvasive diagnostic tools to enhance ED classification. Future research could explore machine learning models to uncover complex patterns not evident in traditional statistical methods.