| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology Albany Medical College Albany, NY 12208
To the Editor:
The article by Dr. Laurent Glance (1) contains statements and conclusions not justified by the interpretations and analyses presented. Despite referring repeatedly to the absence of valid outcome studies comparing risk-adjusted anesthesia-related mortality rates for certified registered nurse anesthetists (CRNAs) and anesthesiologists working alone and CRNAs and anesthesiologists working together, such references are used to support his premise.
Specifically, the secondary mortality rate used to calculate probability of death is based on a retrospective, nonrandomized study by Siber et al. (2) of the differences in failure to rescue outcomes in the intensive care unit. Hospitals were assigned relative failure to rescue rates. These rates are misinterpreted to imply variability was based on provider characteristics alone, that noncertified anesthesiologists have 2.5 times higher mortality rates of certified anesthesiologists, and that CRNAs working alone have 5 times higher mortality rates of solo anesthesiologists and CRNAs working with anesthesiologists.
When valid data are available, decision analyses can integrate complex data to arrive at scientifically sound conclusions that are not otherwise self-evident. However, when baseline assumptions in the decision models are not based in fact, mathematical calculations, and the attestations that follow, are not only irrelevant but misleading and potentially detrimental to any decision making processes that result from them.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
