Author + information
- Kenneth M. Kessler, MD, FACC⁎ ()
- ↵⁎University of Miami School of Medicine, Division of Cardiology, C/O 26 William Howard Drive, Glen Mills, PA 19342
The newly published guidelines for the treatment of ST-elevation myocardial infarction (STEMI) use the more detailed definitions of classes of recommendations and levels of evidence as shown in Table 1 of the report (1). Unfortunately, this system continues to conflate empiric and nonempiric information, despite the fact that these issues are incommensurable and should not be reflected within a single evidentiary hierarchy system. The result of this commingling is that nonempiric facts are erroneously assigned to the lowest level of certainty. These assignments reflect confusion between those instances where evidence is subject to, but has yet to be, scientifically tested, and those instances wherein the truth of the statement cannot be empirically proven. For example, the Class I recommendation regarding the need for an electrocardiogram (ECG) is considered a matter of expert consensus, and though this is true, the consensus does not pertain to an experimentally based question but rather to the current definition of the disease state being discussed. The ECG is the sine qua non of STEMI. All experimentation presupposes this definition. Until the definition is amended or the measurement tool changed, the certainty that an ECG is needed is absolute. This is a nonempiric statement that does not need to be, nor can it be, experimentally proven. Its certainty is parallel to, not subordinate to, the best experimental data.
Similarly, the Class III recommendation not to do coronary angiography on patients where the risks of revascularization outweigh the benefits reflects the nonempiric moral dictum to do no harm. This is a concept that is not open to experimental proof, but is an absolute that forms the basic fabric of our profession. Many other examples of undervalued nonempiric statements appear throughout cardiovascular guidelines. Future task forces/writing groups need to recognize that nonempiric concepts should be separated from empiric concepts and that nonempiric data represent levels of certainty that are parallel to, not subordinate to, empiric data.
- American College of Cardiology Foundation
- Antman E.M.,
- Anbe D.T.,
- Armstrong P.W.,
- et al.