Search for a command to run...
Physician questionnaires are commonly used in health services research; however, many survey studies are limited by low response rate. We describe the effectiveness of a method to maximize survey response without using incentives, the effectiveness of survey reminders over time, and differences in response rates based on survey mode and primary care specialty. As part of a study to assess vaccine policy issues, 13 separate surveys were conducted by internet and mail over the period of 2008 to 2013. Surveys were conducted among pre-recruited networks of pediatricians, family physicians and general internists. Each network was active for 2 years and responded to 3-6 surveys. Physicians who indicated preference to respond through an online survey received up to 9 e-mailed requests to complete the questionnaire and up to 2 mailed questionnaires. Physicians who chose to respond by mail received up to 3 mailed questionnaires and a reminder postcard. For 6 of the 13 surveys conducted over the 6 year period, an additional mailing using a hand-addressed envelope was mailed to non-responders at the end of the usual protocol. Effectiveness of survey methods was measured by response rates. The overall response rates varied from 66‒83%. Response rates declined 17 percentage-points on average between the first and last surveys administered within each physician network. The internet group consistently had higher response rates than the mail group (74% vs. 62% on average). An additional mailing in a hand-written envelope boosted the final response rate by 11 percentage-points. Self-selection of survey mode, multiple reminders, and hand-written envelopes are effective methods for maximizing response rates in physician surveys.