ABC | Volume 113, Nº4, October 2019

Original Article Cruz et al. Surgical mitral valvuloplasty in paediatric patients Arq Bras Cardiol. 2019; 113(4):748-756 Methods A retrospective cohort study was performed. Data were collected by reviewing information on medical records (physical and electronic). The collection was performed by four researchers after standardized training. The Escola Bahiana de Medicina e Saúde Pública Research Ethics Committee approved this study together with CAAE from 64019316.0.0000.5544. Population The study included 54 patients with mitral insufficiency of rheumatic etiology who underwent surgical correction by MVP technique, from March 2011 to January 2017. Preoperative evaluation Patients were clinically identified using the New York Heart Association (NYHA) Functional Classification. 9 All medications that patients used continuously for at least one month were recorded. Valvular lesions were assessed by preoperative transthoracic echocardiography, classifying the lesions as “absent/discrete” or “moderate/significant”. Patients who presented another cause of valve damage at the time of surgical correction by MVP (infective endocarditis; congenital, post-traumatic, degenerative lesions or dystrophic lesions; cardiomyopathies or inflammatory or ischemic disease) or who underwent aortic valve surgery or other procedures in the same surgical time of MVP or an undocumented previous MVP, or patients who did not reach 60 postoperative days until January 2017 were excluded from the study. Surgical technique The reconstructive valve surgery technique was MVP, described by Carpentier, 10 which includes annuloplasty and commissurotomy. The patients studied were preferably operated by the same medical team. The intraoperative data collected were: surgical technique used, duration of cardiopulmonary bypass (CPB), duration of anoxia and presence of atrial fibrillation. The intraoperative outcomes studied were: arrhythmia, cardiorespiratory arrest (CRA) and bleeding. Follow-up Follow-up was carried out within 60 days after surgery, in an outpatient setting, in a single centre. The predictors of risk for unfavourable outcomes studied were: age, ejection fraction, type of valve lesion, degree of mitral insufficiency (MI), left chamber dilatation, NYHA Preoperative Functional Classification, surgical technique used, duration of CPB, duration of anoxia, presence of atrial fibrillation, presence of pulmonary hypertension (PH) (sPAP > 35 mmHg) and presence of tricuspid insufficiency (TI). Early (up to 7 postoperative days) and late (>7 days post‑operative) outcomes related to heart valve disease were studied. The following were investigated: death, heart failure, cardiogenic shock, endocarditis, mitral valve damage, sepsis, stroke, bleeding, reoperation and valve replacement. The presence of any of these outcomes alone or in combination would characterize an unfavourable outcome as a single dependent variable. Statistical analysis The Statistical Package for Social Sciences (SPSS Inc., Chicago, IL, USA), version 14.0 for Windows, was used for the elaboration of the database and for descriptive analysis. The results were presented in tables. Categorical variables were expressed in frequencies and percentages. Continuous variables with normal distribution were expressed as mean and standard deviation; those with non-normal distribution were expressed in median and interquartile range. The normality of the numerical variables was verified through descriptive statistics, graphical analysis and the Kolmogorov-Smirnov test. The independent Student’s t test was used to compare groups of numerical variables with normal distribution (age, weight, body mass index – BMI, duration of anoxia, duration of CPB, ejection fraction). The Mann-Whitney test was used to compare numerical variables with asymmetric distribution, such as duration of use of vasoactive drugs (VAD). The χ 2 test was used to compare the use of medications in the preoperative and upon hospital discharge, and the intergroup comparison of the following categorical variables: gender, origin, outpatient follow-up, reoperations, functional classification, surgical team, events during surgery, duration of extubation and echocardiographic variables. When the distribution showed n < 5 individuals in each category, Fisher’s exact test was used. The paired Student’s t-test was used for the numerical variable “ejection fraction” in the comparison of the paired groups (pre- and postoperative), and the McNemar test was used to compare the categorical variables of the echocardiogram. For all univariate analyses, a value of p < 0.05 was established. The logistic regression model was used to evaluate the predictive variables for unfavourable outcomes in children and adolescents who underwent surgical mitral valvuloplasty secondary to rheumatic heart disease. After the univariate analysis, the independent variables were included in the logistic model if they presented p < 0.05, remaining in the model if they remained significant (p < 0.05). The manual procedure for insertion and withdrawal of the variables was adopted. Results were presented using Odds Ratio (OR) and their respective 95% confidence intervals (95%CI). Results FromMarch 2011 to January 2017, 90 patients underwent surgery by the MVP technique in the tertiary hospital where the present study was carried out. Of these, 36 were excluded, of which 7 had congenital mitral lesions (mitral dysplasia), 8 had underwent aortic valve replacement associated with MVP and 21 due to loss to follow-up or incomplete data. Characteristics of the sample studied Table 1 shows clinical and demographic aspects of the 54 patients included, of which 29 (53.7%) were female, with a mean age of 10.5 ± 3.2 years. Of these patients, 34 (64.2%) lived on the countryside of the State of Bahia, 5 (9.4%) were from the Metropolitan Region and 14 (26.4%) were from the capital, Salvador. The mean BMI was 15.7 ± 3.5 kg/m 2 . 749