Assessing Treatment Preference in Pediatric Growth Hormone Deficiency: Challenges and Proposed Solutions

OBJECTIVES: Patient preference information (PPI) of the desirability or acceptability to patients and/or caregivers of alternatives among health treatments is an increasingly important type of patient experience data. Developing PPI measures can be challenging and no single methodology is the gold standard. Challenges include balancing stated vs. revealed preferences, applicability to study designs where treatment options are not experienced, interpretation of findings, as well as decisions about the best respondent to provide the data, especially when patients are children.

The purpose of this presentation is to suggest a process for developing PPI measures for growth hormone deficiency (GHD) which helps address these challenges. This process draws from best practices for patient-reported outcome measure development and discrete choice methodologies.

METHODS: Concept elicitation interviews were conducted with 5 clinical experts, 15 children with GHD, and 15 caregivers of children with GHD regarding the attributes and concepts which underpin treatment preferences. Qualitative data were analyzed using adapted grounded theory methodology, and items were generated and cognitively assessed.

RESULTS: Child and caregiver versions of treatment preference measures were developed: the GHD-Preference Measure, which is applicable when the respondent has experienced different treatment options, assesses preference as well as the “why” behind the preference, and the GHD-Attribute Measure, which is applicable when only one treatment option has been experienced, assesses the strength or “presence” of preference attributes.

CONCLUSIONS: By leveraging rigorously collected, patient-centered data, both a preference and an attribute measure can be developed simultaneously. These measures provide PPI that is easily interpreted and may be completed by caregivers when the patient is not the primary preference decision maker. Use of these measures and the process for creating them will provide real world data useful to researchers assessing PPI, clinicians helping patients make treatment choices, and regulators assessing risk/benefit during the drug development process.