Table 4.
EuroSCORE II Population Compared With External Validation Populations, Stratified by Relatedness
| Statistica | EuroSCORE II | Validation Populationsb, c | |
|---|---|---|---|
| Related | Distantly Related | ||
| Total patients (n) | 16 828 | 14 382 | 98 744 |
| Total validations (n) | NA | 5 | 73 |
| Age, y | Mean (SD): 64.6 (12.5) | 63.4 (62.7–67.0) | 67.1 (61.1–80.5) |
| Number of events (n) | 656 | 123 (53–215) | 27 (12–57) |
| Event rate, % | 3.9 | 5.7 (5.7–6.1) | 6.3 (3.0–10.5) |
| Sex reported, n (%) | NA | 5 (100%) | 50 (68%) |
| Men, % | 69.1 | 65.2 (62.5–66.5) | 52.5 (46.8–64.1) |
| Type of intervention, n (%) | |||
| Surgery | 1 (100%) | 5 (100%) | 52 (71.2%) |
| Percutaneous | 0 (0%) | 0 (0%) | 20 (27.4%) |
| Both | 0 (0%) | 0 (0%) | 1 (1.4%) |
| Valve‐related, % | 53.3 | 56 (54.6–56.1) | 100 (100–100) |
| Enrollment, y (range) | 2010 | 2005 to 2013 | 1999 to 2015 |
| C‐statistic | 0.8095 | 0.82 (0.76–0.85) | 0.72 (0.67–0.78) |
| C‐statistic (range) | NA | 0.737 to 0.861 | 0.50 to 0.95 |
| Any calibration reported, n (%) | 0 (0%) | 4 (80%) | 65 (89%) |
| Change in discrimination,d % | NA | 2.6 (−16.0–13.1) | −28.9 (−45.3–−9.5) |
EuroSCORE indicates European System for Cardiac Operative Risk Evaluation.
All values are reported as median (interquartile range) unless otherwise specified.
Validation data is reported at the population level only; patient‐level data was not available.
Validation population are “related” if it meets all of the following criteria: (1) same type of intervention (eg, both surgical populations), (2) ±10% absolute difference in the proportion of isolated valve procedure (eg, derivation population was 100% isolated valve and validation population was 95% isolated valve), and (3) overlapping years of enrollment. A validation population that does not meet all 3 criteria is “distantly related.”
Change in discrimination is calculated as [(Validation AUC−0.5)−(Derivation AUC−0.5)]/(Derivation AUC−0.5)×100.