In the January-February 2015 online edition of the Journal of Healthcare Quality, a nurse, Donald E. Beissel, reported complication rates for 1,788 fluoroscopic guided interlaminar lumbar epidural steroid injections performed by 13 nurse anesthetists. The article neither discusses the appropriate patient indications for the procedure nor addresses the efficacy of the procedures performed. Although the safety of epidural steroid injections in the lumbar spine (LESIs) is very favorable, the medical decision making on whom to perform the procedure is one of the most challenging aspects of pain medicine and cannot be learned in a weekend training course.
The article includes methodological flaws that threaten the statistical and internal validity of the study and that could render inferences or conclusions drawn from the study invalid.
* The survey results presented in the article are not generalizable to the overall population of nurse anesthetists. The analytic sample includes only 13 nurse anesthetists who volunteered to participate in a survey and self-report. The article provides no evidence that these 13 individuals are representative of the population of nurse anesthetists performing fluoroscopic guided interlaminar LESIs, and the validity of inferences are threatened by potential selection and reporting bias in this limited sample.
* Inappropriate statistics are used to describe the subject demographics. The subject demographics in Table 1 are nonparametric; they do not represent normally distributed data. The correct descriptive statistics for nonparametric data are medians and interquartile ranges, not means and standard deviations.
* The sample size is too small to provide meaningful results. The nurse anesthetists self-reported performing a total of 1,788 fluoroscopic guided interlaminar LESIs with reported complication rates ranging from 0 to 89 per 10,000. This sample size is too small to comment on complication rates. For example, to detect a 10% difference in complication rates between two equal-sized samples when the population complication rate is around 50 per 10,000 would require samples of more than 330,000 cases each.
* The study does not account for the clustering of data among the 13 participants. The calculation of standard deviation for purposes of calculating confidence intervals must account for this clustering as samples with clustered data will have broader confidence intervals.
There are substantial complexities in diagnosing and developing an effective plan to treat chronic pain. When interventional measures are considered as part of the treatment plan, the decision to incorporate them should be based on a careful medical evaluation performed by an appropriately trained physician who considers the relative risks and potential benefits.
Nurse anesthetists are performing LESIs although they do not have the comprehensive training that pain medicine physicians have to
* complete a medical assessment,
* adequately assess whether the patient is appropriate for the procedure,
* safely perform the procedure,
* develop an effective multidisciplinary treatment plan for the patient's care, and
* provide long-term ongoing care.
Given the complexity of spinal pain diagnoses and the large economic burden suffered by patients with spinal pain, it is imperative that highly trained physicians provide interventional pain medicine services in the context of an overall treatment plan. Similarly, the importance and complexity of interventional pain medicine procedures necessitate that research on outcomes be held to the highest standards. We are disappointed that the Journal for Healthcare Quality would publish this nonscientific piece. No conclusions can be or should be drawn from this poorly conceived study.