PDE5is-naïve versus non-naïve patients at first investigation for erectile dysfunction—findings from a long-term, real-life cross-sectional study

Background: Phosphodiesterase 5 inhibitors (PDE5is) represent a first-line pharmacological therapy for erectile dysfunction (ED). Men could obtain PDE5is for recreational purposes without any proper medical prescription. Objective: We aimed to analyze clinical characteristics of patients who already used any PDE5i for ED without previous formal medical prescription. Materials and methods: Data from 2012 heterosexual, sexually active men seeking first medical help for ED at our outpatient clinic between 2005 and 2022 were analyzed. All patients were assessed with a comprehensive sexual and medical history and completed the International Index of Erectile Function (IIEF) at baseline. Comorbidities were scored with the Charlson comorbidity index (CCI). Thereof, according to exposure to any PDE5i before their first visit, patients were subdivided into: PDE5i-naïve and non-PDE5i-naïve patients. Descriptive statistics tested the sociodemographic and clinical characteristics of both groups. A logistic regression model predicted the likelihood of being PDE5i-naïve at the baseline. Linear regression analysis (LRA) estimated the likelihood of being PDE5i-naïve versus non-PDE5i-naïve over the analyzed timeframe. Lastly, local polynomial regression models graphically explored the likelihood of being PDE5i-naïve at the first clinical assessment over the analyzed timeframe, and the sensitivity analyses tested the probability of being PDE5i-naïve at baseline. Results: Overall, 1,491 (70.9%) patients were PDE5i-naïve and 611 (29.1%) were non-PDE5i-naïve at the first assessment. PDE5is-naïve patients were younger, with a lower prevalence of CCI ≥ 1 and of normal erectile function (EF) than non-PDE5i-naïve men (all p < 0.05). Multivariable logistic regression found that patients with lower BMI (OR: 0.99), higher IIEF-EF scores (OR: 1.02), lower rates of severe ED (OR: 0.94), and who had been assessed earlier throughout the study timeframe (OR: 1.27) were less likely to be PDE5i-naïve at baseline. Univariate LRA revealed that younger patients (Coeff: −0.02), with lower CCI (Coeff: −0.29) and higher alcohol intake per week (Coeff: 0.52) were more likely to be PDE5i-naïve over the analyzed timeframe. Moreover, for the same IIEF-EF score, patients with higher CCI had lower probability of being PDE5i-naïve. Conclusions: Self-prescription of PDE5is is an attitude presents in the general population, despite this phenomenon has decreased overtime. Current data outline the importance to keep promoting educational campaigns to promote PDE5is as effective and safe medicinal products, while avoiding their improper use.