Chapter 58. Risk Stratification in Interventional Cardiology
Which risk model has the highest discrimination and calibration for 30-day mortality after transcatheter aortic valve replacement (TAVR)?
A. European System for Cardiac Operative Risk Evaluation (EuroSCORE)
B. Society of Thoracic Surgeons (STS) risk score
D. PARTNER novel risk score
EuroSCORE and STS risk models were developed for risk stratification of surgical procedures and have low discrimination and calibration for outcomes in TAVR (STS risk score c-statistic, 0.58). They are mainly used to decide best interventional approach for severe aortic valve stenosis: surgical aortic valve replacement (SAVR), TAVR, or balloon valvuloplasty. OBSERVANT risk model was shown to have high discrimination and calibration abilities (c-statistic, 0.73) when compared to EuroSCORE. However, this model needs to be validated in other cohorts. The PARTNER novel risk model was developed to predict poor outcome including 6-month mortality and Kansas City Cardiomyopathy Questionnaire–Overall Summary Scale (KCCQ-OS) score less than 45 after TAVR. This risk model has moderate discrimination (c-statistics, 0.66).
Which of the following variables is not included in all of the clinical risk models?
B. Acuity of presentation
Most of the clinical risk models are derived from similar variables, including age, acuity of presentation, left ventricular function, and periprocedural shock. Age is a universal risk marker for predication of complications after percutaneous coronary intervention (PCI). For instance, patients old than 80 years have 6.03 and 4.3 unadjusted odds ratios for mortality in Mayo Clinic and New York (NY) State risk scores. Acuity of presentation, including cardiac arrest at presentation, cardiogenic shock, and ST-segment elevation myocardial infarction (STEMI), is associated with higher complication rates. For example, the overall in-hospital mortality after PCI was 0.65% in nonurgent PCI and 4.81 in STEMI in the National Cardiovascular Data Registry (NCDR) CathPCI model. Other predictors (eg, diabetes mellitus [NY State and Beaumont models] and female sex [NY State, Cleveland Clinic, Michigan, and Beaumont models]) are not consistent across all the risk models and have a generally lower, weaker relationship with procedural complications.
Which of the following 2 models can be used to predict 30-day mortality after PCI?
A. NY State and Mayo Clinic
B. NCDR CathPCI and Mayo Clinic