Letter by Ahmed Regarding Article, “Second Internal Thoracic Artery Versus Radial Artery in Coronary Artery Bypass Grafting: A Long-Term, Propensity Score-Matched Follow-Up Study”
To the Editor:
I would like to congratulate the authors on providing a useful insight on the controversial issue of the benefit of radial artery (RA) or right internal thoracic artery as the second conduit in multiple arterial revascularization.1
I appreciate that it is difficult to analyze retrospective data from multiple surgeons over a 9-year period, and it appears that the authors did their level best using propensity matching to reduce confounding variables.
However, it is of some concern that during the study period there appeared to be a heavy bias toward RA use in comparison with right internal thoracic artery, with more than two thirds of multiple arterial revascularization cases having RA as the second arterial conduit. This inherently suggests an institutional bias that is difficult to correct for. In addition, the use of data from 16 surgeons over this 9-year period equates to an average of only 5 cases per year for each surgeon involved. One can interpret that, although the surgeons had >6 years surgical experience, they may not be regarded as expert in multiple arterial revascularization with these numbers.
More importantly, one must appreciate that propensity scoring provides a means for adjusting for selection bias in observational studies of causal effects. Randomized trials ensure that subjects receiving different treatments are comparable, but, in observational studies, propensity scoring is used to try to compare equal groups and ensure covariates are balanced. However, if the propensity model misses an important reason why subjects are exposed to a particular treatment, then this is an important confounding factor that could affect the results.
One such reason is the level of native disease in the vessel being grafted. There are data to demonstrate that degree of native stenosis significantly impacts RA patency. Desai et al2 demonstrated that target coronary vessel lesions with stenosis of ≥90% were associated with a lower rate of occlusion of the RA graft. These data were consolidated by the Radial Artery Versus Saphenous Vein Patency (RSVP) trial investigators.3 In addition, no information is given regarding the preparation of the RA. Moderate supraphysiological distension causes a 70% loss of contractility and therefore decreased propensity to spasm in this muscular artery, which can affect early outcomes.3
It would be useful if the authors could provide this information without which one must assume a level of confounding that has not been accounted for (hence, reducing the strength of the propensity matching), which could affect the outcome and strength of this important analysis.
Ahmed Ishtiaq Mehmood, BSC(Hons), MBChB, MRCS, PhD, FRCS (C-Th)
James Cook University Hospital
Department of Cardiac Surgery
Middlesbrough, United Kingdom
- © 2012 American Heart Association, Inc.
- Ruttmann E,
- Fischler N,
- Sakic A,
- Chevtchik O,
- Alber H,
- Schistek R,
- Ulmer H,
- Grimm M
- Collins P,
- Webb CM,
- Chong CF,
- Moat N