"We have people who are not functioning at their best in terms of taking care of patients optimally," she says. "We find people who are struggling with these issues are less likely to stay in practice. We have a workforce shortage already so we have to figure out how to allow people to function at their best and continue in practice."
The survey was sent to 25,073 surgeons, and the 7,197 surgeons who responded represent a sampling of less than 29%. A critique of the study that also appears in the February issue of Archives echoes the survey authors' concerns that the low response rate could skew the findings.
Edward Livingston, MD, a professor of surgery at the University of Texas Southwestern Medical Center in Dallas, wrote that the response rate was about half the 60% minimum required by some medical journals. Thus, he wrote that it is impossible to say if the problem is being over-reported or underreported.
"Surgeons who abuse alcohol are at risk for livelihood-threatening sanctions should this behavior be exposed. Despite reassurances of confidentiality, questionnaire responses may still be considered risky by surgeons with much to lose," Livingstone wrote.
Livingstone said that several types of bias and error may be at play in any survey, but the most obvious one is nonresponse. "Nonresponse bias is particularly salient when the topic is considered sensitive," he wrote.
Kaups concedes that the accuracy of the survey could be questioned because of the low response rate, but she's standing by the findings.
"This is a huge group of people who we studied. What we don't know and what we can theorize about is a possible sampling bias," she says. "Did we get people who respond who are having issues and feel like they can express themselves, or are these people who feel pretty good about themselves? We don't know but we think it is a reasonable survey. It was anonymous online. People are honest about what they say."