Author + information
- William Leete Hiser, MD∗ ()
- ↵∗Cardiology, Baystate Medical Center, Tufts University School of Medicine, 759 Chestnut Street, Springfield, Massachusetts 01199
George et al. (1) recently published a database review of treatment of attempted revascularization of chronic total occlusion in the United Kingdom. They have compiled a useful analysis of the outcomes and feasibility of attempted revascularization of chronic total occlusions. The paper is informative as a descriptive study. However, analyzing mortality on the basis of successful versus failed recanalization is not a fair assessment of the benefit of the procedure. The analysis is performed on the basis of an outcome variable (procedural success) rather than on an input variable (attempted revascularization).
It is not logical to suggest that a procedure is beneficial simply because those with failed procedures had worse outcomes than those with successful procedures. It is not surprising in any field of endeavor that success is associated with better outcome than failure. One should not imply that a cancer drug is beneficial simply because patients who responded to the drug did better than those that did not; one cannot imply benefit of revascularization because those with successful interventions do better than those with failed interventions.
If we are to conclude that these groups (procedural success and procedural failure) are comparable (i.e., have similar disease burdens) then we should have expected similar procedural outcomes. However, procedural failure is more likely due to the complexity of the underlying disease. It is quite likely that the difference in mortality observed is explained by differences in the severity of the underlying disease or other comorbidities rather than by any benefit of the procedure.
The authors do not make any mention of the bias introduced by inferring benefit of a procedure by comparing outcomes of failure versus success, and Dr. Mahmud, in the accompanying editorial (2), only briefly suggests there might be an interaction between the underlying disease and outcome. In this setting, there is no relevant group for comparison of revascularization benefit, as there is no group that did not have attempted revascularization. In the absence of randomization, an alternative would be to try to match patient with chronic occlusion who did not undergo attempted revascularization. In any event, to establish benefit of the procedure, there must be some comparison group that did not undergo the procedure. In the absence of a relevant control or comparison group, they should not suggest that the procedure is beneficial; they can only state that failure is worse than success.
Absent relevant comparison groups, this study stands as a very detailed description of outcomes in a large population of subjects undergoing revascularization of chronically totally occluded coronary arteries, but it does not establish a mortality benefit.
- American College of Cardiology Foundation
- George S.,
- Cockburn J.,
- Clayton T.C.,
- et al.
- Mahmud E.