Parker J Magin
This study aimed to examine the efficacy of two strategies for improving general practitioner response to a survey. A secondary aim was to assess GPs’ self reported preferred mode of survey administration.
A random sample of 1666 GPs practising in New South Wales was selected from the Australasian Medical Publishing Company database. Two randomised trials of strategies aimed at increasing response rates were embedded in a larger cross sectional survey.
Of the 1666 GPs sampled, 52 were ineligible and 500 completed the
survey. The response rates obtained in the trial of standard research group letterhead invitations alone (25.8%) versus division of general practice cover letter (32.5%) were not statistically significantly different; nor were the response rates obtained in the trial of a telephone reminder call. When asked about preferred mode of survey administration, 81.1% of respondents nominated mailed survey.
The study failed to identify strategies to improve GP participation in the survey. This survey found no basis for supporting electronic GPs surveys.
Surveys are vital for obtaining information about the knowledge, attitudes, practice patterns and needs of general practitioners. This information is often used in service and program planning and delivery. However, GP response rates to surveys are lower than those from the general population,1 and are falling.2–4 While there is no agreed standard for an acceptable minimal response rate to a survey, response rates of 70% or higher are considered good.5,6 However, published response rates with medical practitioners are often lower than 30%.1,4,6,7 Low response rates raise concerns about response bias.8 Low response rates from GPs may also mean that their voice is not heard in periods of change in delivery of primary care.
Download the PDF for the full article.