This non-comparative, retrospective study aimed to characterize the profiles of erectile disfunction (ED) patients prescribed with tadalafil by clustering the homogenous subpopulations with different ED presentations and risks.
Men with ED, aged ³18 years (y) and prescribed with tadalafil therapy in the Optum database (2012-2019) were included. The study endpoint was to characterize patients based on the presence or absence of risk factors and ED complications, and segmentize into phenotypically similar clusters using machine learning techniques.
Among the 218,047 patients (mean age, 57.1 y) included 81% were Caucasian, 83% had cardiovascular (CV) risk factors, and 5.3 mean symptoms/patient. Five clusters were identified among which 3 were homogeneous with clear distinction from each other: (a) younger patients with few ED or other CV risk factors (n=46,526; mean age, 48.6 y, and 2.1 mean symptoms/patient); (b) benign prostate hyperplasia and/or prostate cancer (n=39,428; mean age, 65.3 y, and 4.4 mean symptoms/patient); (c) middle-aged ED with comorbidities (n=68,418; mean age, 55.7 y, and 5 mean symptoms/patient). Remaining 2 clusters were less homogenous with overlapping characteristics of other clusters (b and c) reflecting the natural evolution of ED from one cluster to another: (d) elderly patients with ED and CV history (n=29,963; mean age, 65.5 y, and 8.8 mean symptoms/patient) and (e) psychological disorders (n=43,712; mean age, 54.5 y, and 7.9 mean symptoms/patient).
This novel study was able to segmentize the heterogenous population of tadalafil-prescribed ED patients into 5 distinct phenotype clusters with specific trajectories in ED and overhealth progression which offers an opportunity for targeted interventions in those with various phenotypes. These findings may provide a better perspective on the association between ED, use of tadalafil, concomitant medications, and CV risk.