Consumer Preferences for Attributes of Influenza Vaccines in the United States: Results from a Discrete-Choice Experiment Study

OBJECTIVES: Quantify preferences for influenza vaccine attributes among consumers in the United States (US).

METHODS: Using a discrete-choice experiment, US adults were presented with a series of experimentally designed pairs of hypothetical influenza vaccines comprising five attributes with varying levels: absolute vaccine efficacy (AVE) (15%-60% of influenza infections prevented), hospitalization prevention (same as most other vaccines, more than most other vaccines), durability (protection fades within 6 months, constant protection for at least 6 months), and side effects (risk of moderate-to-severe injection site reactions [ISR] [0%-50%] and risk of flu-like symptoms [FLS] due to the vaccine [0%-60%]). In each choice question, respondents selected the preferred vaccine or “no flu vaccine.”

Random-parameters logit analysis results were used to calculate conditional relative attribute importance (CRI) out of 100% and maximum acceptable risks of vaccine side effects in exchange for improvements in vaccine efficacy.

RESULTS: The sample included 400 respondents aged 18-64 years and 201 respondents aged ≥ 65 years. Half had at least one non-age risk factor for severe influenza.

On average, consumers preferred influenza vaccination to opting out of vaccination. Consumers placed greatest importance on avoiding risk of FLS (CRI, 40%), followed by increasing AVE (CRI, 36%) and avoiding ISR risk (CRI, 21%). On average, hospitalization prevention and durability had no influence on vaccine choice.

Consumers were less tolerant of FLS than ISR. For an improvement in AVE from 15% to 25%, consumers would be willing to accept a 34% risk of ISR or a 19% risk of FLS.

For an improvement in AVE from 25% to 50%, consumers would be willing to accept a 31% risk of ISR or a 16% risk of FLS.

CONCLUSIONS: Adults in the US were willing to accept increases in vaccine-related risks for improved AVE, and they were more tolerant of ISR risks than FLS risks.