Current sedation practices: lessons learned from international surveys.

Limitations are inherent to surveys. Most surveys have low response rates, which raises the issue of responder bias. Another limitation of self-report surveys stems from the possible differences between stated and actual practice. That is, what physicians report that they do in surveys often contrasts significantly with what they do in observational studies, as highlighted by the Canadian surveys conducted in 2002 and 2008. Some surveys report estimates provided by ICU nurse managers or physician directors, potentially resulting in inaccurate estimates or data reflecting the individuals practice rather than the entire ICU. Surveys may not reflect how different specialists practice; for example, the German surveys collected data only in ICUs run by anesthesiologists.Notwithstanding these limitations, surveys provide a wealth of information on current practice and determinants of practice, and serve as a useful tool to guide future research and educational interventions. The authors identified substantial international variation in the use of sedative and analgesic drugs, and marked changes over the last 10 years. Overall, there is a trend toward lighter sedation, along with a shift from benzodiazepines toward propofol, and from morphine toward fentanyl and remifentanil. Despite the publication of numerous studies and guidelines for sedation and analgesia, actual practice differs from recommended practice, suggesting that the impact of clinical trials and guidelines on physician practice is quite low. It is clear that there remain substantial barriers to the incorporation of sedation scales, protocols,and daily interruption into routine ICU care.