Within each practitioner type, a ra

Within each practitioner type, a random sample was
identified to ensure adequate statistical power under a
response rate of 40 % (typical of physician response to surveys)
[21, 22]. Sampsize [23] was used to calculate ‘‘sample
size for a prevalence survey, with finite population correction’’.
The assumption was that precision = 5.00 % (95 %
CI prevalence ±5 %). A total of 2,375 licensed and practicing
OB/GYNs (700 of 1,327), CNMs (all CNMs in Ohio,
approximately 300), family practitioners (700 of 2,253), and
internists (700 of 4,726) listed in public access state licensure
files who were not part of the pilot tests were invited to
participate. From October through December 2010, surveys
were administered using a dual-mode (mail and internet)
approach to optimize response rates [24]. Providers who had
retired or moved out of state were excluded (n = 340)