Ann Thorac Surg 2006;81:541
© 2006 The Society of Thoracic Surgeons
Original article: Cardiovascular
Invited commentary
Fred H. Edwards, MD
Division of Cardiothoracic Surgery,University of Florida Health Science Center,Division of Cardiothoracic Surgery,653-2 West 8th St,Jacksonville, FL 32209-6511
(Email: fhe{at}comcast.net).
The authors [1] correctly point out that "risk stratification models have been criticized for reduced applicability when used in different populations..." The importance of their work comes from the fact that a model developed from one narrow population (New Jersey) seems to produce accurate results when applied to a population on another continent. Not only is the geographic area different, but the timeframe is significantly different as well.
Although this is an important aspect of the Bernstein-Parsonnet model, I believe the authors overstate its ease of use. In this time of ubiquitous computer access, the additive model offers little advantage in comparison with techniques that directly use computerized algorithms. In fact, it is easier to just enter patient risk factors into a computer than it is to undertake the addition and chart extrapolation required with the additive model. The fact that the additive model does not require a computer was an advantage 10 years ago, but not today.
Likewise, the authors overstate claims of superior accuracy for the Bernstein-Parsonnet model. The contention that other commonly used models are invariably less accurate cannot be substantiated. Several models have actually been shown to have more accurate results than the original Parsonnet model. In addition, the statement that multifactorial risk models are poorly integrated into clinical practice may be true in some countries, but certainly this is not true in the United States.
The Parsonnet model is an excellent, time-proven tool for risk assessment. The authors have provided valuable information regarding its global applicability, but one should not lose sight of the fact that other, more comprehensive models are being used with excellent sustained success.
 |
References
|
|---|
- Berman M, Stamler A, Sahar G, et al. Validation of the 2000 Bernstein-Parsonnet score versus the Euroscore as a prognostic tool in cardiac surgery Ann Thorac Surg 2006;81:537-541.[Abstract/Free Full Text]
Top
References