Thank you for the opportunity to respond to the letter from the physician anesthetists J. P. Abenstein and Richard Rosenquist. It is always interesting to hear the voices of those in the same practice as nurse anesthetists.
The stated purpose of my study was to explore complication rates of certified registered nurse anesthetists (CRNAs) providing these procedures. This is one aspect of the pain management puzzle. This aspect was chosen because of undocumented claims of "paralysis and death" to legislators and stakeholders that anecdotal evidence was not supporting. I do not understand why decision making is brought up as a concern regarding this study, because it was not the primary topic; however, the abilities of nurse anesthetists to provide these services are clearly addressed in the Discussion and Implications for Practice sections.
The physician anesthetists' mantra of "weekend course" training is addressed in the reporting of the number of postgraduate hours in pain management training reported in the article. The mean for these 13 providers is 204 hours, which is clearly more than "a weekend." Five of the 13 participants (38%) hold earned doctoral degrees, which represent considerably more training than a "weekend course." This shows that the nurse anesthetists commonly performing these procedures are obtaining much more training than what is claimed by physician anesthetists. It is interesting that it is very common for physicians to obtain training in "weekend courses," but it seems to be unacceptable to physician anesthetist for any other professional.
The sample in the study by Manchikanti and colleagues (2012) only includes four physician anesthetists performing these procedures. If anything, my study is more generalizable to CRNAs than the comparison study is to physician anesthetists. Both studies investigate providers actively engaged in chronic pain management. The volunteer participants are representative of nurse anesthetists across the United States who are actively engaged in providing interventional pain care. To my knowledge, there is no study that shows that these procedures are safe in the hands of all physicians or physician anesthetists. It is usual and customary to study complication rates in the providers who most often provide the procedures.
Self-reporting bias is a legitimate concern; however, it is also a concern for the study by Manchikanti and colleagues (2012) of physician complication rates, which served as the basis of comparison to this study. I am willing to trust both groups of professionals to be honest about their complication rates in the interest of patient safety. In addition, I was the only one who knew the rates before publication, so the participants were not cued to reduce their numbers to present favorable outcomes.
Parametric statistics were not used in this study because doing so would result in trying to prove the null hypothesis that there is no difference between the two groups. It was expected, and shown, that both physician and nurse anesthetists performing this procedure have similar and low complication rates. I considered performing and reporting tests of noninferiority. However, the confidence interval (CI) comparison more concisely told the tale of safe care by both providers without the statistical methods being the center of the discussion.
Parametric statistics were not used on purpose, so the comment on sample size is not germane to the analysis. It is of interest that the nurse anesthetists provided 338 more procedures than those in the study by Manchikanti and colleagues.
I consciously chose to not apply cluster methods because of how broad the CIs would become. If the CIs were widened (in both directions) using this method, then nurse anesthetist's rates would likely overlap all physician CIs and this would be cited as a statistical weakness. It was also considered that the physician data would be clustered, but I did not have access to the raw data of Manchikanti and colleagues (2012), and multiple requests for information from the authors of the study by Manchikanti and colleagues. went unanswered. Therefore, the more conservative method of comparison was chosen. This potential clustering was also examined by attempts to correlate individual nurse anesthetist complication rates with volunteer demographics. These correlations were not reported because none of significance were found, giving credence to the idea that the data are not clustered.
The rest of the claims made by my physician anesthetist colleagues are beyond the scope of the study at hand. The concerns that they raise are addressed in the Discussion and Implications for Practice Sections and I will allow these remarks to stand on their own high merit. Unsupported physician-centric assertions regarding care will not change the fact that nurse anesthetists can and do provide safe interventional pain care.
I look forward to the completion of the follow-up to this study where a much broader range of procedures, in greater numbers, will be investigated.