Mortality in Cardiac Surgeries in a Tertiary Care Hospital of South Brazil

Fundamentos: A prevalência de doenças cardiovasculares é muito alta e cirurgias cardíacas são comuns em centros terciários de atenção cardiovascular. Objetivo: Avaliar a mortalidade cirúrgica e comparar com a mortalidade esperada pelo EuroSCORE em pacientes de centro terciário de atenção cardiovascular. Métodos: Trata-se de coorte histórica que avaliou pacientes submetidos a cirurgias cardíacas em 2011 e 2012, em hospital terciário de referência em cardiologia no sul do Brasil. O desfecho principal foi a mortalidade por qualquer causa durante a internação hospitalar. Os desfechos secundários foram a mortalidade de acordo com o procedimento cirúrgico, o EuroSCORE total e de acordo com o procedimento, a mortalidade ajustada por EuroSCORE e o perfil de risco dos pacientes. Resultados: Foram incluídos 364 pacientes. Cirurgia de revascularização do miocárdio (CRM) isolada foi o procedimento realizado em 59,9% dos pacientes, o procedimento valvar (PV) isolado (troca ou plastia valvar) em 33,0% e a CRM + PV em 7,1% dos pacientes. A mortalidade observada foi 14,2%, sendo 5,2% nas primeiras 24 horas. A mortalidade esperada pelo EuroSCORE, por sua vez, foi 5,7±7,4%. A mortalidade ajustada foi, assim, 2,5 vezes superior à esperada, mas dentro do intervalo de dois desvios-padrão da mortalidade esperada. A mortalidade associada aos procedimentos combinados, entretanto, foi 5,2 vezes superior a do EuroSCORE esperado e superior ao intervalo de dois desvios-padrão do EuroSCORE. Conclusão: Pacientes submetidos à cirurgia cardíaca no centro estudado apresentaram mortalidade superior à esperada, em especial aqueles submetidos a procedimentos combinados.


Introduction
Cardiovascular diseases comprise a wide range of illnesses, with high prevalence and mortality rates.North American statistics show that 1:3 people have some kind of cardiovascular disease; heart illnesses were responsible for 31.3% of all deaths in 2011 1 .
The high prevalence imposes the performance of therapeutic procedures such as coronary artery bypass graft (CABG) surgeries or valve procedures (VP) (replacement or repair), which are fairly common in tertiary centers for cardiovascular care.In 2010, 7 million North Americans underwent some kind of heart surgery, fewer only to obstetric procedures 1 .
Data on mortality in heart surgeries vary widely according to the center, the number of surgeries and the type of procedure: the average mortality rate in CABG surgeries can be as low as 0.7% in North American centers, while revascularization surgeries associated to any valve procedure can reach 20.8% in some Brazilian centers [2][3][4] .
The huge variability in mortality and incidence of complications gave rise to the development of preoperative prognostic scores, such as the EuroSCORE and the Society of Thoracic Surgeons Scores (STS) [5][6][7] .Despite the many scores available in literature, the EuroSCORE seems to have the best predictive value and it is easy to be calculated 8 .
Following an anachronic path to what happens in other centers in the world; there is a gap in the evaluation of Brazilian tertiary centers that perform heart surgeries: a proper evaluation would enable an increase in the quality of data recording and a decrease in trans and postoperatory complication rates 9 .
Therefore, the purpose of this study was to analyze the surgical procedures performed in a renowned tertiary center for cardiovascular care; evaluate surgical mortality rates and compare them to preoperative EuroSCORE.

Methods
Historical cohort study that evaluated all patients undergoing CABG and/or VP, using the database of the Heart Surgery service from January 2011 to December 2012 of a hospital that is a leading center of reference in cardiology in southern Brazil.Additional epidemiological data was obtained from electronic medical records of patients, as well as manual records developed in the Coronary Care Unit, Ward, Hemodynamics department, Outpatient Clinic and the Echocardiography department.
The hospital serves patients exclusively through the Unified Health System (SUS).It is a center of reference in the state, where a team of resident heart surgeons duly qualified provides cardiac surgery services.The hospital has an average volume of 20 major surgeries (CABG and valve replacement) per month, where each surgeon performs 4 major surgeries on average.The postoperative ICU has 10 beds.The hospital also holds the medical residency program in Cardiology, Cardiac Surgery and Vascular Surgery.
The primary outcome evaluated by the study was the mortality rate, defined as all-cause death during or after the procedure and during hospitalization.Secondary outcomes were the mortality rates according to the type of surgical procedure; the procedure-related EuroSCORE and the total average EuroSCORE; the adjusted mortality rate by total EuroSCORE (defined as the quotient between the expected and observed mortality by EuroSCORE) and procedurerelated EuroSCORE; and the risk profile of patients 5,10 .Surgical records were evaluated as to the type of surgery, the nature of surgery (elective, urgent or emergency), the reason for hospitalization, length of stay and in-hospital mortality.A qualified researcher performed the calculation of EuroSCORE for each patient, regardless of the presence of the previous value on record.Additional factors taken into account: age, sex, coronary lesions on angiography, percutaneous coronary interventions, systemic arterial hypertension (SAH), diabetes mellitus (DM), kidney failure, dyslipidemia (DLP), family history of coronary artery disease (CAD), peripheral artery disease (PAD), neurological disease, heart failure, smoking, chronic obstructive pulmonary disease (COPD), ejection fraction (EF) assessed by echocardiography, pulmonary arterial hypertension (PAH) and type of valve lesion, surgery undergone in less than three months after acute myocardial infarction (AMI), surgery for unstable angina, cardiogenic shock, or infectious endocarditis.
The SPSS 13.0 software for Windows was used in the analysis.Continuous variables were analyzed using the Student t test for independent samples and expressed as mean±standard deviation, or by analysis of variance (ANOVA) with Bonferroni post hoc test, when appropriate.Categorical variables were analyzed by the Fisher exact test and expressed as frequencies.The significance level was p<0.05.

Results
The study analyzed 364 patients with mean age of 60.6±11.7 years, most of them men (64.3%).Further epidemiological characteristics of patients are shown in Table 1.
The main reasons for hospitalization of patients under study were unstable angina or AMI without ST-segment elevation (53.0%), followed by valve injuries (31.3%),AMI with ST-segment elevation (6.9%), endocarditis (5.5%), and other causes (3.0%).Patients waited an average of 38.1±25.9days to undergo surgery, with a total length of stay of 53.5±32.9days.There was no association between mortality and the number of days of hospitalization until the date of surgery (32.6±23.1 days for patients who died, and 38.8±26.3days for patients discharged, p=0.11) or the total length of stay (55.3±45.9days for patients who died, and 53.1±30.5 days for patients discharged, p=0.74).
In-hospital mortality rate was 52 (14.2%) patients, of whom 19 (5.2%) died in the first 24 hours.The mortality rate expected by logistic EuroSCORE, in turn, was 5.7±7.4% (average additive EuroSCORE of 4.5±2.9).The adjusted mortality rate was thus 2.5 times higher than the expected, however within the range of two standard deviations of the expected mortality rate (Figure 1).
The mortality and the surgical procedure-related EuroSCORE are presented in Table 2.In the population studied, the valve procedures had higher risk score, however the observed mortality was higher in the group that underwent combined procedures (CABG + VP).The combined procedure also reported an adjusted mortality rate higher than two standard deviations of the expected mortality rate (Figure 1).

Discussion
This study provides important data about heart surgeries performed in a leading center of reference in Cardiology in southern Brazil, demonstrating high surgical mortality, particularly among patients undergoing combined procedures, whose adjusted mortality rate was 5.2 times higher than expected, higher than two standard deviations of the expected mortality rate.
The mortality rate found was higher than that of other Brazilian centers: data from the Dante Pazzanese Cardiology Institute reports a total mortality rate in cardiac surgeries of 5.9%, where mortality in first-valve replacement procedures is 4.3% in mean follow-up of 10 years 11,12 .National and international data indicate, however, quite different mortality rates between centers, ranging from 5.77% to 20.8% in CABG+VP procedures [2][3][4][13][14][15] .
The total adjusted mortality rate as well as mortality in CABG or VP were higher than expected, but still within the two standard deviation range, as described for public service in other centers worldwide 10 .Mortality in combined procedures, however, shows much higher rates than the two standard deviation range from the expected mortality: these figures do not mirror any other data in existing literature, therefore prohibiting the performance of combined procedures (CABG + VP) until appropriate measures were taken.
Different reasons could explain the higher rates found, such as the technical deficiency of team members or the participation of resident physicians under training, since the learning curve could compromise the quality of surgery.In addition, patients had on average a long period of hospitalization until the date of surgery, which would give room for colonization by hospital germs in the preoperative period: the average waiting time exceeding one month is quite high compared to other data from national literature; other studies have pointed out the waiting time for the procedure as a factor for higher mortality among patients treated by the public health system 13,16,17 .It is noteworthy that despite the long-term hospitalization, there were no significant differences in the mean length of stay until the date of procedure among patients with in-hospital death.
This study has some limitations: it is a retrospective cohort study and, as such, mistakes in database records could compromise the analysis.Although not compromising the findings, it is important to emphasize that there were no specific statistical evaluations such as logistic regression to promote adjustment for potential confounders.Moreover, it was not possible to identify the immediate cause of death of each patient, and the study did not allow the evaluation of noncardiac causes of postoperative mortality, such as hospital infection.

Conclusion
Patients undergoing cardiac surgery at the center under study showed higher mortality than the expected rate, especially those undergoing combined procedures.
Given the importance of feedback for improved quality in procedures and patient safety, the head of the Cardiac Surgery Department was officially reported on the study, as well as the hospital's administration and the Committee for Ethics in Research of the institution 9 .
After the internal disclosure of data, an evaluation committee of deaths was created to point out the causes of the high mortality rate and potential failures in the service.

Figure 1
Figure 1Mortality by procedure according to the standard deviations of the expected mortality rate CABG -coronary artery bypass grafting; VP -valve procedure The study complied with the Resolution of Brazil's Health Council -CNS No. 466/2012, and the WMA Declaration of Helsinki, being approved by the Committee for Ethics in Research of the institution under No. 13546113.2.0000.0113.

Table 1 Epidemiological characteristics of the population studied according to the type of procedure
CCS -Canadian Cardiovascular Society; CAD -coronary artery disease; CABG -coronary artery bypass graft surgery; SAH -systemic arterial hypertension; DM -Diabetes mellitus; COPD -chronic obstructive pulmonary disease; SPAP -systolic pulmonary artery pressure; VP -valve procedure; RC -right coronary; AD -anterior descending artery; Cx -circumflex coronary artery; LCT -left coronary trunk; LVEF -left ventricular ejection fraction; SD -standard deviation