Abstract
Objectives: Differences in testosterone replacement therapy (TRT) gel products may affect patient satisfaction, quality-of-life, and treatment response and adherence. This study investigated preferences for TRT gel in terms of formulation and administration.
Methods: 525 male adults aged 45 years and over completed a discrete choice experiment. Respondents made repeated choices between two hypothetical testosterone gel treatments described according to four attributes: form, ease of use, impact of use on showering/swimming, and location/dosage of the application. Choice data were analyzed using a latent class model.
Results: Three preference classes were identified. Respondents across all classes displayed a preference for the gel being dispensed in smaller units with accurate dosing, waiting shorter times after the gel application before swimming/showering, and using 2.5 gm of gel to be applied to the inner thigh/abdomen as opposed 5 gm to shoulder/abdomen. The importance of these characteristics differed across classes, with preference class membership predicted by age and education level. For instance, younger men (aged 45–64 years) were more likely to belong to a class that prioritized reduced waiting time before being able to undertake activities. Formulation was not an important driver of choice.
Conclusions: Preferences demonstrate a predilection for TRT gel dispensed in small units allowing precise dosing, shorter waiting time after application, and application to the inner thigh/abdomen. However, the strength of importance of these characteristics differs between men. This study highlights the attributes of TRT gel considered important to patient subgroups, and which may ultimately affect treatment response, medication adherence, and patient quality-of-life.
Acknowledgments
This study was funded by Kyowa Kirin International (KKI) manufacturer of Tostran©. KKI was not involved in the drafting or review of this manuscript.
Disclosure
Dr Jenny Retzler reports grants from Kyowa Kirin, during the conduct of the study. Dr Adam B Smith reports grants from Kyowa Kirin International (KKI), during the conduct of the study. Ms Ana Sofia Oliveira Gonçalves reports grants during the conduct of the study. Professor Jennifer A Whitty reports grants from York Health Economics Consortium, during the conduct of the study. The authors report no other conflicts of interest in this work.
Supplementary materials
Figure S1 Class 1 preferences for each attribute level. For each attribute, the columns represent the marginal preference for the identified attribute level (left column) over its alternative level (right column). Error bars represent the 95% confidence intervals ().
Figure S2 Class 2 preferences for each attribute level. For each attribute, the columns represent the marginal preference for the identified attribute level (left column) over its alternative level (right column). Error bars represent the 95% confidence intervals ().
Figure S3 Class 3 preferences for each attribute level. For each attribute, the columns represent the marginal preference for the identified attribute level (left column) over its alternative level (right column). Error bars represent the 95% confidence intervals ().
