Correlates of procedural complicationsand a simple integer risk scorefor percutaneous coronary intervention
Mandeep Singh, MD*,
Ryan J. Lennon, MS ,
David R. Holmes, Jr, MD, FACC*,
Malcolm R. Bell, MB, BS, FRACP, FACC* and
Charanjit S. Rihal, MD, FACC*,*
* Division of Cardiovascular Diseases and Internal MedicineRochester, Minnesota, USA
Section of Biostatistics, Mayo Clinic, Rochester, Minnesota, USA
View larger version (16K):
[in a new window]
|
Figure 1 Expected versus observed procedural complications after percutaneous coronary intervention in clinical practice. Hosmer-Lemeshow goodness-of-fit test (6 df) = 1.83, p = 0.93.
|
|
View larger version (38K):
[in a new window]
|
Figure 2 Estimated rates of procedural complications for the integer scoring system. The integers are proportional to the estimated continuous coefficient from the logistic model. Percentages at risk are shown for each of the five risk categories:  2% is very low risk for complications with coronary angioplasty; >2% to 5%, low risk; >5% to 10%, moderate risk; >10% to 25%, high risk; and >25%, very high risk. NYHA = New York Heart Association classification.
|
|
View larger version (18K):
[in a new window]
|
Figure 3 The observed and corresponding predicted procedural complications in the validation-set with the proposed risk score. The x-axis represents the risk score in the validation-set. The y-axis represents the complication rate. The solid line demonstrates the predicted procedural complications derived from the risk score. The bars represent observed procedural complications in the validation-set. Three patients had a score > 19; all three were free of complications.
|
|
|
