Trial Outcomes & Findings for Patient Preferences for Leadless Pacemakers (NCT NCT05327101)

Patients were recruited from Aveir™ DR IDE trial study sites. One-hundred seventeen patients were recruited and completed the survey instrument.

Patient Preferences for Leadless Pacemakers

Ranking of six pacemaker device features from most concerning (1) to least concerning (6)

The preference weights for the RPL model. Effect-coded parameters generate log-odds preference weights representing the relative strength of preference for each attribute level versus the mean effect across levels normalized at zero. A higher weight indicates a more preferred level while a lower weight indicates a less preferred level.

The standard deviations representing the degree of variation in preference weights, with larger estimates representing preference heterogeneity.

Maximum-acceptable risk (MAR) of a complication was calculated for patients based off latent-class analysis with two groups-leadless class and transvenous class (see secondary outcome Constrained 2-class Latent-class model preference weights). The MAR represents risk level patients would be willing to accept to obtain their preferred pacemaker type, no discomfort, a device with longer battery life, and a device with more time since regulatory approval.

Maximum-acceptable risk (MAR) of an infection was calculated for patients based off latent-class analysis with two groups-leadless class and transvenous class (see secondary outcome Constrained 2-class Latent-class model preference weights). The MAR represent the risk that patients would be willing to accept to obtain their preferred pacemaker type, no discomfort, a device with longer battery life, and a device with more time since regulatory approval.

Preference weight estimates were used to calculate the predicted probabilities that patients would choose a hypothetical pacemaker profile out of three different pacemaker types- leadless pacemaker removable, leadless pacemaker non-removable, or pacemaker with leads. Attributes for each pacemaker profile were defined using historical or published values. Preference weights from the latent class model were used to compute the probability that respondents within each class preference would choose a pacemaker profile over another.

Preference weight estimates were used to calculate the predicted probabilities that patients would choose a hypothetical pacemaker profile out of three different pacemaker types- leadless pacemaker removable or leadless pacemaker non-removable. Attributes for each pacemaker profile were defined using historical or published values. Preference weights from the latent class model were used to compute the probability that respondents within each class preference would choose a pacemaker profile over another.

Preference weight estimates were used to calculate the predicted probabilities that patients would choose a hypothetical pacemaker profile out of three different pacemaker types- leadless pacemaker removable or pacemaker with leads. Attributes for each pacemaker profile were defined using historical or published values. Preference weights from the latent class model were used to compute the probability that respondents within each class preference would choose a pacemaker profile over another.

Preference weight estimates were used to calculate the predicted probabilities that patients would choose a hypothetical pacemaker profile out of three different pacemaker types- leadless pacemaker non-removable or pacemaker with leads. Attributes for each pacemaker profile were defined using historical or published values. Preference weights from the latent class model were used to compute the probability that respondents within each class preference would choose a pacemaker profile over another.

Latent-class (LC) analysis was used to identify systematically different preference patterns across respondents. LC analysis provides a unique set of estimates of preference weights for a prespecified number of preference classes. Respondents are probabilistically assigned to classes based on the similarity of their responses to the overall preference pattern identified in each class.

Number of Discrete choice experiment (DCE) questions answered by 117 respondents who answered at least the first 8 DCE questions. After respondents answered 8 DCE questions, they were asked if they would like to complete 4 additional questions.

Respondent characteristics can be associated with class membership probabilities to "profile" the classes. These results show if respondents with certain characteristics are more likely to be in one class versus the other. The odds ratios of being in the transvenous class versus the leadless class for patient characteristics. Odds ratios greater than 1 indicate a higher likelihood of being in the class preferring transvenous pacemakers and pacemakers with longer time since government approval.