ABC | Volume 114, Nº3, March 2020

Original Article Pivatto Júnior et al. Risk scores for surgery in endocarditis Arq Bras Cardiol. 2020; 114(3):518-524 issues, but also by systemic infective and microbiological factors. 25 More recently, new IE-specific risk scores have been developed. They incorporate some IE-specific factors (such as microbiological cultures, abscess formation and sepsis) that are known to be independent predictors of mortality. IE-specific scores have demonstrated greater accuracy for mortality prediction than classical risk scores. 26 Among the IE-specific scores analyzed, only the PALSUSE 10 and RISK-E 12 scores had derivation cohorts limited to patients with active IE. The PALSUSE score, 10 which incorporates the EuroSCORE in its composition, was derived from a prospective cohort study including 437 patients who underwent surgery in the acute phase of IE. Data were collected in 26 Spanish Table 3 – Observed/expected mortality ratio and ROC curve analysis for the studied scores SCORE O/E MORTALITY* 95%CI p AUC 95%CI p NON-SPECIFIC SCORES Logistic EuroSCORE 1.33 0.92-1.87 0.123 0.77 0.66-0.87 <0.001 EuroSCORE II 2.46 1.70-3.45 <0.001 0.69 0.58-0.80 0.002 IE-SPECIFIC SCORES STS-IE 2.50 1.73-3.50 <0.001 0.67 0.56-0.79 0.005 PALSUSE 1.01 0.70-1.42 0.919 0.68 0.57-0.79 0.003 RISK-E 1.53 1.05-2.18 0.029 0.71 0.60-0.81 0.001 AEPEI 1.71 1.18-2.40 0.006 0.65 0.53-0.77 0.017 EndoSCORE 2.90 2.00-4.06 <0.001 0.76 0.66-0.86 <0.001 *Observed mortality was 29.0%, except for the RISK-E score, which was 28.4% (5 right-sided infective endocarditis cases were excluded, since they are not included in this score analysis). O/E: observed/expected; AUC: area under the curve; CI: confidence interval; IE: infective endocarditis. Figure 1 – Observed and expected hospital mortality according scores. *Observed mortality was 29.0%, except for the RISK-E score, which was 28.4% (5 right-sided infective endocarditis cases were excluded, since they are not included in this score analysis). Error bars represent 95% confidence intervals. hospitals. In-hospital mortality was 24.3%, ranging from 0% in patients with a score of 0 to 45.4% in those with a score ≥ 4. AUC was 0.84 (95%CI: 0.79-0.88), indicating satisfactory discriminatory ability. The RISK-E score 12 was developed from research performed in three tertiary care centers in Spain, which sought to predict in-hospital mortality in 424 patients with active left-sided IE undergoing cardiac surgery. AUC was 0.82 (95%CI: 0.75-0.88). The predicted probability of postoperative mortality ranged from 3% for a patient with a score of 0 to 97% for a patient with the highest possible score of 68. A comparison of AUCs showed a statistically significant superior predictive performance of the RISK-E score (p = 0.01) when compared with EuroSCORE, EuroSCORE II, or PALSUSE. 522