Produção Nacional
Revisado por pares
Mortality related to cardiac surgery in Brazil, 2000-2003
2006; Elsevier BV; Volume: 131; Issue: 4 Linguagem: Inglês
10.1016/j.jtcvs.2005.11.022
ISSN1097-685X
AutoresAntônio Luiz Pinho Ribeiro, Saverio Paulo Laurito Gagliardi, Jose Luiz Santos Nogueira, Lídia Marques Silveira, Enrico A. Colosimo, Carlos Armando Lopes do Nascimento,
Tópico(s)Maternal and Neonatal Healthcare
ResumoCardiovascular disease (CVD) is now a leading cause of death in the developing world, where, because of demographic shifts of the population age profile, a staggering additional increase in CVD morbidity and mortality can be predicted.1Reddy K.S. Yusuf S. Emerging epidemic of cardiovascular disease in developing countries.Circulation. 1998; 97: 596-601Crossref PubMed Scopus (972) Google Scholar Cardiac surgery (CS) remains a standard treatment for select patients with advanced CVD, although data about surgical results in developing countries are scanty. The aim of this study was to determine mortality related to CS in adults according to sex and type of surgery by studying its relation to age and hospital volume in a large nationwide dataset.See related editorial on page 773. See related editorial on page 773. Data of open CS performed from January 2000 to December 2003 in adults (>17 years) were extracted from the Hospital Information System of Brazilian Ministry of Health (SIH/DATASUS) dataset, an established and audited nationwide database with administrative and epidemiologic data on every hospital admission performed by the Brazilian Universal Health System (SUS; Sistema Único de Saúde).2Leao B.F. Bernardes M.M. Levin J. Moura L. Bandarra E. Modesto L.M. et al.The Brazilian National Health Informatics Strategy.Medinfo. 2001; 10: 38-42Google Scholar In Brazil, 75.4% of the population is covered exclusively by SUS, and highly complex procedures, such as CS, are performed mainly under the public system. SIH/DATASUS dataset files contain summarized hospital-discharge records for the hospitalization of all SUS patients, including procedure code, hospital identification, and patient information, such as the city of residence, age, sex, and vital status at discharge. Hospital volume was expressed as the average number of procedures per year, and, to analyze hospital volume, only hospitals with 5 or more procedures per year were included. Surgical procedures were grouped into coronary artery surgery, heart valve surgery, surgery for congenital disease in the adult, and complex CS, including combined procedures and aortic surgery. Pacemaker implantation and other minor CS procedures were excluded. Mortality was defined as death before hospital discharge. Univariate analysis included correlation analysis, logistic regression, and χ2 test. Hospital volume and age were ranked in order and divided into 4 groups (quartiles) for further analysis. Multivariate logistic regression was used to examine relationships between CS-related mortality and sex, type of surgery, age, and hospital volume. From 115,021 CS performed between 2000 and 2003, death occurred in 9211 cases (8.0%). Both univariate and multivariate analyses showed that CS mortality increases with age (Table 1 and Figure 1), is higher in women (8.8% vs 7.4%; P < .001), and increases significantly for valvular and complex CS and, after adjustment, decreases for coronary artery surgery. There is a significant relationship between hospital volume and CS mortality (rs = −0.326; P < .001; Figure 1) that remains significant after adjustment for age, sex, and type of operation (Table 1).TABLE 1Mortality in cardiac surgery related by age, sex, type of operation, and hospital volume in Brazil, 2000 to 2003VariableNo. cases (%)Mortality (%)Unadjusted ORAdjusted ORAge (y) 18-4527,857 (24.2)5.8 46-5730,721 (26.7)6.41.12 (1.04-1.19)1.50 (1.39-1.61) 58-6628,871 (25.1)8.01.41 (1.31-1.50)2.01 (1.87-2.16) 67-9927,572 (24.0)12.02.22 (2.09-2.36)3.24 (3.03-3.48)Sex Male66,710 (58.0)7.4 Female48,311 (42.0)8.81.21 (1.16-1.26)1.20 (1.15-1.26)Type of operation Congenital5806 (5.0)6.1 Coronary65,716 (57.1)7.01.17 (1.05-1.31)0.72 (0.64-0.81) Valvular38,789 (33.7)8.91.51 (1.35-1.69)1.22 (1.09-1.37) Complex4710 (4.1)16.53.04 (2.66-3.47)2.30 (2.01-2.63)Hospital volume (CS/y) 1-13028,562 (24.8)9.71.73 (1.63-1.84)1.81 (1.70-1.93) 131-27029,244 (25.4)8.41.47 (1.38-1.57)1.52 (1.42-1.62) 271-34027,867 (24.7)8.11.43 (1.34-1.53)1.49 (1.39-1.56) >34129,348 (25.5)5.8Values are expressed as absolute numbers (%) and odds ratios (OR) with 95% confidence intervals. The adjusted OR refers to the OR obtained in a model that included all 4 variables. CS, Cardiac surgery. Open table in a new tab Values are expressed as absolute numbers (%) and odds ratios (OR) with 95% confidence intervals. The adjusted OR refers to the OR obtained in a model that included all 4 variables. CS, Cardiac surgery. Classic determinants of an increased risk of death related to CS, such as age, sex, and hospital volume,3Society of Cardiothoracic Surgeons of Great Britain and Ireland. National Adult Cardiac Surgical Database Report 2000-2001. Available at: http://www.scts.org. Accessed: May, 29, 2005.Google Scholar, 4The Society of Thoracic Surgeons. STS Adult CV Surgery National Database. Fall 2004 Executive Summary contents. Available at: http://www.ctsnet.org/file/STS-ExecutiveSummary-Fall2004.pdf. Accessed: May, 29, 2005.Google Scholar are also related to an increased risk of death after CS in a developing country's nationwide database. Moreover, valve, combined, and aortic surgery were more risky than coronary artery revascularization. The overall mortality rate for each group of conditions observed in this dataset is grossly higher than in the most recent surgical registries, such as The Society for Thoracic Surgeons database in the United States3Society of Cardiothoracic Surgeons of Great Britain and Ireland. National Adult Cardiac Surgical Database Report 2000-2001. Available at: http://www.scts.org. Accessed: May, 29, 2005.Google Scholar and the UK Cardiac Surgical Register.4The Society of Thoracic Surgeons. STS Adult CV Surgery National Database. Fall 2004 Executive Summary contents. Available at: http://www.ctsnet.org/file/STS-ExecutiveSummary-Fall2004.pdf. Accessed: May, 29, 2005.Google Scholar Nonetheless, these results should be cautiously analyzed. Considering that both The Society for Thoracic Surgeons and UK Cardiac Surgical Register are voluntary registries, whereas the DATASUS is an administrative dataset, it is inappropriate to formally compare these surgical results obtained in Brazil with those observed elsewhere. Hospital volume calculated in our study considered surgical procedures financed only by the public health system, but this is probably a minor cause of error, because other sources provide less than one fifth of the CS performed in Brazil. Surgical results may be influenced by several factors, including differences in the socioeconomic status of the local population, the severity of cardiac illness and of comorbidities, standards of medical care, and adequacy of facilities and staffing.5Reis A.A.C. Nascimento C.A.L. Ribeiro A.L.P. Daher C.V. Costa H.S. Nogueira J.L.S. Política Nacional de Atenção Cardiovascular de Alta Complexidade. Ministério da Saúde, Brasília2004: 125Google Scholar Important practical considerations emerged from this study. Because the choice of surgical treatment for a specific heart disease relies on the balance between surgical risk and benefit in comparison with more conservative strategies, prudence is desirable in the appliance of guidelines developed in Europe and North America in other countries' settings, in which CS-related mortality might be relatively increased. On the administrative and regulatory fronts, the Brazilian Health Ministry defined, in conjunction with the scientific societies, a national policy of high-complexity cardiovascular care, which comprises a series of minimal prerequisites, both in facilities and staffing levels, as well as a minimum surgical volume to allow a cardiovascular service to operate in the public health system. A national registry, the Brazilian Cardiac Surgery Registry, which is being implemented as part of the national policy, will permit, in the future, the comparison of Brazilian CS mortality with data from other international databases. Should cardiac surgery be performed in low-volume hospitals?The Journal of Thoracic and Cardiovascular SurgeryVol. 131Issue 4PreviewIn the "Brief Communications" section of this issue of The Journal there is a report on hospital mortality after cardiac surgery in Brazil.1 An analysis of the database of the Brazilian Universal Health System, which apparently covers 75.4% of the population and more than 80% of all cardiac operations, revealed that from 2000 to 2003, 115,021 patients underwent cardiac surgery with an overall hospital mortality of 8%. The hospital mortality was 6.1% for congenital heart surgery, 7% for coronary artery bypass, 8.9% for heart valve surgery, and 16.5% for "complex" operations (eg, thoracic aorta and combined procedures). Full-Text PDF
Referência(s)