ABC | Volume 114, Nº1, January 2019

Original Article Otto et al. Aortic prosthesis mismatch in public health system Arq Bras Cardiol. 2020; 114(1):12-22 Table 1 – Baseline clinical characteristics of the PPM groups (176 patients Definition #2) No PPM 30.4%(54) Moderate PPM 36.2% (64) Severe PPM 33.4% (58) p Age (years) 55 ± 17 ┼ 60 ± 15 ┼* 52 ± 16 * 0.0335 Female/male gender % 11/19 15/21.2 14/20.4 0.78 BSA (m 2 ) 1.70 ± 0.24 1.71 ± 0.17 * 1.8 ± 0.21 * 0.016 BMI (kg/m 2 ) 25 ± 3.37 ┼ 26 ± 4.42 ┼* 27 ± 5.17 * 0.03 SBP (mmHg) 120 ± 15 ┼ 117 ± 18 ┼* 111 ± 15 * 0.03 DBP (mmHg) 72 ± 14 ┼ 66 ± 11 ┼ 68 ± 13 0.028 HR bpm 83 ± 14 83 ± 14 87 ± 13 0.26 Hypertension% 17.6 22.1 17.6 0.54 Diabetes% 2.8 5.1 4.6 0.77 CABG% 6.3 8 4 0.25 Renal Disease 1 2 1 0.4 Aortic Root enlargement % 1.7 1.1 3.4 0.27 Mitral valve surgery % 0 1.14 0.57 0.63 Valve disease Etiology (%) 0.003 Rheumatic 4(9.5) 8(14.55) ┼ 22(43.1) ┼ Degenerative 21(50) 29(52.8) 19(37.3) Congenital (Bicuspid) 11(26.2) 10(8.2) 7(13.7) Aortic Root Dilation 6(14.3) 8(14.6) 3(5.9) Type of Prosthesis Biop./Mech. % 24/6.4 31/5.2 31.4/2 0.27 ┼ p < 0.05 between no PPM and moderate PPM. * p < 0.05 between moderate and severe PPM. BSA: body surface area; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic blood pressure; HR: heart rate; CABG: coronary bypass graft; Biop.: biological; Mech: mechanical. Preoperative Determinants of Severe PPM Determinants of severe PPM according to Definition #1 (indexed EOA< 0.65 cm²/m²) In univariate analysis (Table 4), there was an association between severe PPM and the following variables: age < 60 years, BSA > 1.74 m 2 , rheumatic heart disease as the etiology of aortic valve disease and not performing aortic root enlargement. Multivariable analysis (Table 4) revealed that preoperative variables independently were the same as in univariate analysis, except for not performing aortic root enlargement. The tolerance indicator for multicollinearity was 0.78, indicating that there is no strong multicollinearity among the independent variables. Determinants of severe PPM according to Definition #2 (indexed EOA<0,65 cm 2 /m 2 for patients with BMI <30 kg/m 2 and EOA < 0.55 cm 2 /m 2 for BMI > 30 kg/m 2 In addition to the independent variables described in the analysis above using the cut-off value of < 0.65 cm 2 /m 2 for severe PPM, we found that male gender is an independent determinant of PPM when BMI is considered as a parameter for reclassification of severe PPM to iEOA ≤ 0.55 cm 2 /m 2 . However, BSA was not an independent variable within this new model. Univariate and multivariate analysis are shown in Table 5. Determinants of severe PPM according to Definition #3 (mean prosthesis gradient ≥ 20 mmHg and iEOA ≤ 0,65 cm 2 /m 2 ) With this definition, only age < 60 years (PR: 3.33; IC 95%: 1.56-7.12) and LVOT diameter < 2.1 cm (PR = 1.68; IC 95%: 0.87-3.21) were independently associated with severe PPM. Complete analysis is described in Table 6. Accuracy and Mathematical model for Prediction of Severe PPM with preoperative variables We tested the accuracy of the predictive model for severe PPM using Definition # 2, for its precision in identifying more independent variables compared with the other definitions. The area under the ROC curve was 0.82 (Figure 1). In addition, to calculate the individual risk of a patient to develop severe PPM, we built a mathematical model summarized by a formula based on multivariate logistic regression analysis (Table 7). With this formula, it is possible to calculate the individual risk of PPM for each patient before surgery (Table 7). Discussion One of the main findings of this study is that frequency of severe PPM is high after AVR in patients treated in the Brazilian Public Health System in a representative tertiary center. 14