Trends and Hospital Mortality in Myocardial Revascularization Procedures Covered by the Brazilian Unified Health System in Rio de Janeiro State from 1999 to 2010

cerebrovascular diseases were responsible for 26.1% of deaths.1 Many techniques including coronary artery bypass grafting (CABG) and coronary angioplasty (CA) have been developed in attempt to minimize individual and collective health problems caused by IHD.2 However, although these procedures were introduced in Brazil over 30 years ago, and despite their increasing use in clinical practice and costs for the Brazilian Unified Health System (UHS), studies evaluating the performance of these procedures in terms of efficacy and effectiveness are still scarce.3 Introduction


Introduction
Circulatory system diseases (CSD) are the main cause of death in the state of Rio de Janeiro (RJ) and in Brazil. 1 According to DATASUS, in 2013, 38,172 and 339,672 individuals died of CSD in RJ and Brazil, respectively, accounting for nearly 30% of all deaths in both places. 1 Results of these procedures in the Brazilian population are crucial for clinical and administrative decision making.Therefore, the aim of this study was to describe temporal evolution and hospital mortality of CABG and CA performed in RJ, covered by the UHS, in the period from 1999 to 2010.

Methods
In this study, we analyzed data on CABG and CA obtained from medical records database.Only data of procedures covered by the UHS during the period from 1999 to 2010, available at the "Authorization for Hospital Admission (AHA)" database, and provided by the UHS Department of Informatics were selected, and those performed in patients residents in RJ aged 20 years or older, even in institutions located out of RJ were included.
CA were divided into three groups according to AHA database codes: a) CA without stent, including CA without mentioning stent (32023014, 48030066, 0406030014) and CA in coronary artery bypass graft without stent (48030090, 0406030065); b) CA with stent, including CA with an intraluminal prosthesis or a stent (48030074, 0406030030), CA with double-lumen prosthesis or two stents (32035012, 48030082, 0406030022), and CA in coronary artery bypass graft with a stent 48030104, 0406030073); and c) primary CA (48030112, 0406030049).
From the authorized, UHS-covered procedures, the following data were obtain: diagnosis of admission according to the 10th international classification of diseases (ICD-10), 4 date of birth, date of admission, sex, CABG or CA between 1999 and 2010, time of hospitalization, hospital type (public or private), hospital discharge (death or other).
Descriptive statistics were performed using Stata 12, 5 and expressed as frequencies (percentage), mean ± SD or median (interquartile range), as appropriate.Since the study was not conducted on a sample of patients, but rather on the total of patients who underwent myocardial revascularization covered by the UHS from 1999 to 2010 in RJ, [6][7][8] p-values were not calculated for patients' characteristics.Trends analysis was performed by estimation of the mean annual increment by linear regression analysis.
In the study period, 1999-2010, sufficient information was identified to estimate not only the mortality per procedure, but also the mortality per individual.For this reason, the last year of the series was 2010.

Results
In the study period, 34,413 patients (65.4% men) underwent 38,509 procedures, 66.3% of them CA and the rest of them CABG.Two or more procedures were performed in 10.2% of patients.
With respect to the annual number of procedures, there was a considerable increase of CA performed in 2010 (3,633 CA) as compared with 1999 (940 CA), with fluctuations throughout the period (Figure 1).Mean annual growth was 15.8% greater than that of the population of RJ, including individuals of all age ranges and both sexes (Table 1).Considering only the census years, CA increased 156.7% from 2000 (8.9 CA per 100,000 inhabitants) to 2010 (22.7 CA per 100,000 inhabitants), and such increase was even larger in the three last years of the study, corresponding to a mean annual growth of 30.9% (Figure 1).This tendency, however, was different between the CA groups.CA without stent progressively increased from 1999 to 2003 (from 940 to 2,278) and decreased from 2004 to 2010 (from 748 to 145).CA with stent were rare in the period from 1999 to 2003, with 255 procedures in this five-year period, and increased from 2004 (1,009 procedures) to 2010 (3,210 procedures).CA without stents were clearly substituted with CA with stents in this period.It was only in 2004 that the number of primary CA had a modest increase, from 55 in 2004 to 278 in 2010.
With respect to CABG, the increase in the number of procedures was not as great as that observed for CA in the comparison between the initial and the final year of the study (Figure 1).In 1999 and 2010, 943 and 1,269 CABG  The number and percentage of procedures by hospital admission (in biennium) and age ranges for men and women are presented in Tables 2 and 3, respectively.The increase in the number of CA with time was markedly higher than CABG at any age range, for both sexes.The percentages of women aged 70 years or older were higher than men at the same age range for both procedures in all two-year periods.The relative increase in the number of both CA and CABG performed between the initial and final biennium in patients aged 20-49 years was greater among women (2.3 times for CA and 1.5 times for CABG) than men (1.7 time for CA and 1.0 for CABG), whereas in patients aged 50-69 years, the increase was greater in men (3.4 times for CA and 1.5 time for CABG) than women (3.1 times for CA and 1.3 time for CABG).For patients aged 70 years or older, the increase was similar in both sexes for CA (3.5 times for men and 3.6 times for women), and higher among women (1.6 time) than men (1.4 time) for CABG.Despite these differences, mean age (years) was similar between men and women for both CA (59.9 ± 10.5 in men and 61.9 ± 10.8 in women) and CABG (61.0 ± 9.7 in men and 61.9 ± 9.9 in women).
Admission diagnoses related to CA were: acute IHD in 60.6%, chronic IHD in 38.6% and others in 0.8% of admissions.The frequency (%) of CA tended to increase in case of admissions for acute IHD over the years; this percentage was 51.9% in 1999-2000 and 61.1% in 2009-2010.In CABG, the diagnosis of acute IHD was also more prevalent than chronic IHD (57.9% and 40.2%, respectively, and 1.9% of others).Similar to CA, the percentage of hospitalizations for acute IHD was greater in 2009-2010 (68.1%) than in 1999-2000 (57.2%).These percentages were similar between men and women.The number of CA and CABG performed, by admission diagnosis, is shown in Figure 2.
Of 38,509 procedures carried out in patients living in RJ, 99.6% were performed in RJ, and only a few of them were performed in Sao Paulo (151 procedures), Minas Gerais (13 procedures) and Espirito Santo (9 procedures).In addition, 64.3% (24,765 procedures) of them were conducted in private hospitals (68.1% of 25,531 CA and 56.8% of 12,978 CABG), 96.7% in the countryside.On the other hand, 98.6% of 13,744 procedures performed in public hospitals were carried out in the capitals.
Median hospital stay was 2 days, and the 25 and 75 percentiles were 1 and 3 days, respectively for CA, and was 10 days for CABG, with 25 and 75 percentiles of 6 and 18 days, respectively.In CA, hospital stay was similar between sexes and age ranges, whereas in CABG, hospital stay was longer in women (median 10 days) than in men (9 days).In CABG, there was a direct relationship between age and hospital stay duration; in patients aged less than 50 years, median duration was 8 days; in those aged 50-69 years, the median was 10 days, and in those aged 70 years or older, the median was 11 days.Median hospital stay in primary CA was 4 days, and in CA with and without stent, the median was 2 days.The median hospital stay was 9 days in CABG without ECC, and 10 days in CABG with ECC.In private hospitals, median hospital stay was 2 days for CA, and only 1 day in public hospitals.However, in CABG, hospital stay was longer in public hospitals (17 days) than in private hospitals (7 days).
Hospital mortality in CA and CABG by sex, year of hospital admission (in biennium) and age ranges is presented in Table 4. Hospital mortality in CA was 1.7% among men and 2.0% among women.Apart from primary CA, in which hospital mortality was 8.1% and 9.5% among men and women, respectively, hospital mortality was 1.4% in men and 1.7% in women.In CABG, global hospital mortality was 5.8% among men and 9.1% among women.Hospital mortality in patients aged 70 years or older was 2.9 times higher among men and 3.0 times higher among women than in patients younger than 50 years in CA, regardless of admission biennium.In CABG, this increase was five times greater in men and 2.7 times greater in women.By comparing the initial and final biennium of the study, an increase in hospital mortality was observed in women at all age ranges in CA, in contrast   3 and 4. In CA, the age range of 70 years or older, primary CA and acute IHD had greater hospital mortality in both sexes, with few exceptions.In CABG, hospital mortality was higher among women, patients aged 70 years or older, and in those with diagnosis of acute IHD at hospital admission, also with few exceptions.

Discussion
During the period from 1999 to 2010, the number of  In the USA, a study conducted on Medicare & Medicaid Services' beneficiaries from 2001 to 2009 showed an increase in the number of CA until 2004, followed by a mean annual decrease of 2.5%. 10 An Australian study on data recorded in the Melbourne Interventional Group registry from 2004 to 2008 showed a decrease in the number of CA from 2006. 11From this same year on, there was a stagnation in the number of CA performed in Sweden, according to data registered in the Swedish Coronary Angiography and Angioplasty Registry from 1990 to 2010. 12Studies showing no benefit of CA on medium and long term survival of patients with stable angina or silent ischemia under clinical treatment, even in the era of stents, [13][14][15] have been pointed out as one of the main causes for the decrease of CA performed in chronic IHD patients.Nevertheless, in RJ, an increase, rather than a decrease in the number of CA was observed until 2010, probably because the results of these previous studies [13][14][15] had no influence on the clinical practices in the state.Other possible explanations include a lower financial ceiling restriction and higher accessibility to CA covered by the UHS.Also, the elderly population, and consequently the prevalence of IHD increased in RJ. 16 It is also plausible   Hospital mortality (%) de Souza e Silva et al.

Original Article
Regarding the CABG, there was a modest increase in the number of these procedures performed in RJ compared with population growth.A North American study on data obtained from Agency for Healthcare Cost and Utilization Project-Nationwide Inpatient Sample evaluated revascularization procedures performed from 2001 to 2008 and detected a decrease by nearly one third in CABG procedures. 19In Canada, according to Canadian Institute of Health Information, the number of CABG remained stable, varying from 75.6 procedures per 100,000 inhabitants per year in 1994 to 70.8 procedures per 100,000 inhabitants per year in 2005. 20The difference between the annual growth of CA and CABG performed in RJ and in other regions may be related to the substitution of surgical treatment with percutaneous treatment in certain cases of IHD, 19 since hospital stay and short-term complication rates are lower in CA.Besides, in RJ, the higher cost of CABG as compared with CA, in addition to financial issues such as the limited financial ceiling for CABG (covered by the UHS), may have contributed to a discrete growth of CABG in comparison with CA.Nevertheless, one should emphasize that, compared with CA, CABG is still a better option of myocardial revascularization in certain conditions, including three-vessel coronary disease and left main coronary artery disease, even in the era of stents. 21,22th respect to admission diagnoses, acute IHD were the most frequent diagnosis in CA from 2002, and in CABG, the condition remained stable until 2007, and became predominant in 2008 onwards.The possibility that these diagnoses referred to events occurring prior to myocardial revascularization procedures cannot be ruled out, and hence the use of diagnoses recorded in the AHA database as indication for CA and CABG may not be reliable.
Furthermore, these high-complexity procedures were mostly performed in private hospitals in the countryside of RJ or in public hospitals of the capitals, which may be explained by the inefficient connection of primary health care with medium-and high-complexity services.Historically, high-complexity, public hospitals were created almost exclusively in the city of Rio de Janeiro (as the Federal District and Capital of Republic); the access of the population to high-complexity services, and frequently to the UHS resulted from a spontaneous, but excessive demand for these services, since most of these patients would have their needs met by primary health care.In attempt to ensure the universal coverage of the population, agreements with private hospitals had to be firmed in RJ countryside.Unfortunately, this results increased cost of health care, often impracticable for the UHS. 23-related hospital mortality in RJ in the study period was similar to that reported in previous observational studies on short-term outcomes of CA.Choi et al. 24 evaluated annual data of myocardial revascularization of the Health Insurance Review and Assessment Service in South Korea from 2006 to 2010 and found hospital mortality rates of 1.5-1.8%related to CA. Aggarwal et al. 25 assessed data of patients who had underwent CA between 2009 and 2011 in a North American tertiary hospital, and observed a 30-day mortality of 2%.However, these values are higher than short-term mortality rates reported in some controlled, clinical trials.A meta-analysis by Pursnani et al. 26 showed that, in one year, i.e., a follow-up period longer than in our study, the mortality of CA in chronic IHD reported in clinical trials such as MASS-1 27 and COURAGE 28 varied from 0.0 to 2.0%.26 Regarding primary CA in acute IHD only, Hannan et al. 29 found a hospital mortality of 5.8%, which was lower than the 8.8% reported in our study for the same hospital admission diagnosis.
CABG-related hospital mortality in RJ was higher than that reported in both observational and controlled studies.In the study by Choi et al., 24 mortality rate ranged from 2.8% to 3.9% between 2006 and 2010, whereas ElBardissi et al. 30 found a 30-day mortality of CABG performed in the hospitals of the Society of Thoracic Surgeons of 2.4% in the 2000s, and 1.9% in 2009.In the controlled study by Lamy et al., 31 a 2.5% 30-day mortality rate of CABG with and without ECC.
In addition, hospital mortality, especially CABG-related mortality, was predominant among women.This cannot be explained by differences in age profiles, since they were similar between procedures.3][34][35] However, Peterson et al., 36 Argulian et al. 37 and Abramov et al. 38 showed that female sex-related factors, such as larger body surface area, age and higher number of comorbidities would influence short-term mortality in CA and CABG.These factors, however, were not assessed in our study, since these clinical variables are not available in the AHA database.
Another important variable for hospital mortality was age.Hospital mortality in patients aged 70 years or older was nearly 3 times higher than in patients aged less than 50 years for both sexes and procedures.In comparison with patients aged 50-69 years, hospital mortality in older patients was approximately 2.5 times higher in males in both procedures, but nearly 3.5 times higher in CA and 2 times higher in CABG in females.Tadei et al. 39 observed that hospital mortality was the CA outcome mostly affected by age, having increased from 0.28% to 3.45% in patients aged less than 50 years to patients aged older than 80 years.
With respect to CABG, a study conducted in hospitals in Canada and USA revealed increased hospital mortality in patients aged 75 years or older as compared with younger patients. 40Regardless of age, hospital mortality reduced from 1999 to 2010 for both CABG and CA, except for the CA performed in female patients.de Noronha et al. 41 showed that the increased volume of CABG covered by the Ministry of Health in Brazilian hospitals from 1996 to 1998 had a negative association with hospital mortality.Such association, however, was not observed between the volume of CA and CABG and mortality in hospitals in RJ from 1999 to 2003. 42Therefore, other factors that may explain the reduced mortality in CABG are the learning curve, a better control of myocardial revascularization techniques, 43 and the better health care provided, including the monitoring of patients after the procedures. 44,45is study has some limitations.In this study, only data of CA and CABG covered by the UHS performed in public and private hospitals were collected, and hence one cannot infer that the results of this study reflect those of the same procedures covered exclusively by private resources.However, data from the Supplemental Health Care Agency show that 25% and 35% of the population had a health insurance plan in 2000 and 2010, respectively. 46Nevertheless, not all these beneficiaries have the coverage for high-complexity procedures such as CA and CABG, and we speculate that at least 70% of CA and CABG performed in RJ residents were covered by the UHS.
Besides, the AHA database has been created for administrative purposes and do not include detailed clinical information such as the use of medications, number of vessels affected, comorbidities and socioeconomic conditions of patients, and complications of the procedures (except for the occurrence of death during hospitalization).Data obtained from secondary databases do not follow strict protocols as in randomized clinical trials, and may be considered as of poorer quality and less complete.However, these data allow the assessment of a larger number of patients who undergo therapies normally adopted in the clinical practice, yielding results that are not always the same as those obtained from controlled studies.Today, the AHA database is the best tool available in the public health system for this type of study, due to its comprehensiveness and accessibility.
Further studies on a longer follow-up of patients undergoing CA and CABG are needed to establish the survival after hospital discharge.

Conclusion
The tendency of an increase in the number of myocardial revascularization in RJ, especially in CA with stent, differ from the trends observed in the OECD countries in the first decade of the 21 st century, marked by stagnation or a modest growth.Although many factors may have contributed for such increase, arbitrary indications for this procedure, without scientific basis, have probably contributed to it and need to be investigated.In addition, according to controlled clinical trials, mortality rates are still higher than expected.The low efficiency of coronary revascularization procedures covered by public health care highlights the need for a regular performance evaluation for adjustments in the healthcare planning and decision-making in clinical practice.
de Souza e Silva et al.

Figure 1 -
Figure 1 -Trends in the number of coronary angioplasty and coronary artery bypass grafting procedures in the State of Rio de Janeiro from 1999 to 2010.CA: coronary angioplasty; CABG: Coronary artery bypass grafting; ECC: extracorporeal circulation.
CA performed in RJ considerably increased compared with the population growth, especially from 2008 to 2010, differently from what was observed in other parts of the world.Mean annual growth of CA in RJ between

Figure 2 -
Figure 2 -Number of coronary angioplasty and coronary artery bypass graftingw procedures performed according to the year (biennium) and admission diagnosis.IHD: Ischemic heart disease.

Figure 3 -
Figure 3 -Hospital mortality in the three groups of coronary angioplasty according to the admission diagnosis, sex and age range.CA: coronary angioplasty; IHD: ischemic heart diseases.

deFigure 4 -
Figure 4 -Hospital mortality in coronary artery bypass grafting with and without extracorporeal circulation according to the admission diagnosis, sex and age range.IHD: Ischemic heart diseases; CABG: coronary artery bypass grafting; ECC: extracorporeal circulation.

Population mean annual growth (%) from 2000 to 2010 Sex 20-49 years 50-69 years ≥ 70 years Total
to what was observed in CA performed in men and CABG in both sexes.Hospital mortality by CA and CABG by admission diagnosis, sex, and age range of patients is de Souza e Silva et al.