Response rates to a mailed survey targeting childhood cancer survivors: a comparison of conditional versus unconditional incentives.

OBJECTIVE Mailed surveys are widely used to collect epidemiologic and health service data. Given that nonresponse can threaten the validity of surveys, modest incentives are often used to increase response rates. A study was undertaken among childhood cancer survivors and their parents to determine if response rate to a mailed survey differed with provision of immediate versus delayed incentives. DESIGN A self-administered survey designed to ascertain health behaviors was mailed to 397 childhood cancer survivors (and their parents if the survivor was <18 years of age). Subjects were randomized into two groups based on gender, age, race, and cancer type. One group received a 10 US dollars incentive with their blank survey (unconditional incentive), whereas the other group received the incentive upon receipt of their completed survey (conditional incentive). If children were minors, both the parent and the child received incentives. RESULTS No significant differences in response rates were observed with respect to gender, age, race, or cancer type. However, significant differences in response rates were observed between incentive groups, with unconditional incentives yielding significantly higher response rates than conditional incentives for child survivors who were > or =18 years (64.4% versus 49.0%), as well as younger child survivors (62.5% versus 43.6%) and their parents (64.8% versus 41.5%; all P < 0.05). CONCLUSIONS The provision of an immediate incentive generated significantly higher response rates to this mailed health survey among childhood cancer survivors and their parents. Given that survey studies are commonly conducted across various pediatric populations, these findings may help inform the design of future pediatric survey research.