International Journal of Cardiology 184 (2015) 115–121
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Trends in mortality and heart failure after acute myocardial infarction in Italy from 2001 to 2011☆ Cesare Greco a,⁎, Stefano Rosato b, Paola D'Errigo b, Gian Francesco Mureddu a, Eleonora Lacorte b, Fulvia Seccareccia b a b
Division of Cardiology, San Giovanni Hospital, Rome, Italy Istituto Superiore di Sanità, Rome, Italy
a r t i c l e
i n f o
Article history: Received 28 August 2014 Received in revised form 16 December 2014 Accepted 26 January 2015 Available online 28 January 2015 Keywords: Acute myocardial infarction Acute coronary syndromes Discharge Mortality Heart failure Trend
a b s t r a c t Background: Uncertainties on long-term outcomes after acute myocardial infarction (AMI) still exist, despite the ongoing progresses in the management of patients with AMI. Aim of the study: Our aim was to appraise both the early prognosis and prognosis at 1-year after discharge of patients hospitalized due to AMI. Methods: This is a retrospective nationwide cohort study based on data from an administrative database on patients admitted with AMI from 2001 to 2011 in all Italian hospitals sites. Mortality and readmission rates within 30 days, 60 days and 1 year were calculated, as well as re-hospitalizations for all causes and for HF. Results: A total of 1,110,822 patients were included. Index admission mortality rate (I-MR) and total in-hospital mortality rate (T-MR) at up to 1 year both decreased respectively from 11.34% to 8.99% and from 16.46% to 14.68% in the years 2001 to 2011 (both p b 0.0001), while fatal readmission rate (F-RR) at 1 year increased from 4.75% to 5.28% (p = 0.0019). Patients that developed HF during the index admission had significantly higher I-MR and F-RR. I-MR, F-RR, and T-MR, however, remained low at any time point considered (30 days, 60 days and 1 year) in a subgroup of low-risk optimally-treated patients. Conclusions: The risk of fatal readmission at 1 year increased slightly over time, in spite of the remarkable improvements currently achieved in overall prognosis after AMI. The identification of patients at high risk (mainly due to HF complicating AMI), and of patients at low risk is crucial to define and support management strategies. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Many reports published during the last decade have shown a steady reduction of in-hospital mortality due to acute myocardial infarction (AMI) [1–5]. This decrease has generally been considered as a consequence of the progresses reached in the in-hospital management of cardiovascular diseases in developed countries. These progresses are mainly due to the introduction and adoption in clinical practice of either fibrinolytic or mechanical reperfusion, and to the diffusion of pharmacological therapies including beta-blockers, angiotensinconverting enzyme inhibitors, antiplatelet agents and statins. The proportion of patients with ST segment elevation myocardial infarction (STEMI) has concurrently decreased in several countries as compared to the proportion of patients with non-ST segment elevation myocardial infarction (NSTEMI). Continuous efforts in the primary prevention of MI
☆ Primary funding source: Istituto Superiore di Sanità, Rome, Italy. ⁎ Corresponding author at: San Giovanni Hospital, Via dell'Amba Aradam, 8-00184 Rome, Italy. E-mail address: [email protected] (C. Greco).
http://dx.doi.org/10.1016/j.ijcard.2015.01.073 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
may have led to a variation in the frequency of different types of AMI, with smaller infarcts becoming more common than larger ones, in spite of a less favorable overall profile of index patients who became older and having a higher number of comorbidities [6]. However, it is still unclear if the increase in survival during index hospitalization actually continues after discharge. Apparently, while in-hospital mortality seems to decrease, some authors reported a paradoxical increase in short and long-term post-discharge deaths [2,7–10]. The increasing prevalence of heart failure (HF), mainly among the elderly with AMI, has been considered as one of the possible explanations for these unfavorable trends, along with all the potential implications on the organization of post-acute care services [7,11–14]. Most of the evidence supporting these hypotheses, however, is based on data from registries, which may be limited by selective and non-consecutive enrolment causing lower external validity [15]. We aimed at analyzing short- and long-term mortality trends among patients with AMI based on a large, comprehensive and universal administrative database including more than one million of patients over 11 years, with a special focus on the outcomes of patients surviving the index admission date and discharged with a diagnosis of heart failure.
116
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
2. Material and methods
2.4. Statistical analysis
2.1. Study design
A set of general linear models were designed to calculate annual variations in mortality and their corresponding 95% confidence intervals (CI). The goodness-of-fit of the linear regression models was assessed with the R2 test and the probability value for the F test statistic. Trends of risk ratios between F-RR and I-MR were analyzed for each selected time point. Multivariate logistic regression analyses were performed, adjusting for age, sex, comorbidities, and invasive cardiac procedures, to provide adjusted outcome data. Re-hospitalizations due to HF within the considered time spans were analyzed separately for patients with AMI who did and did not develop HF during the index admission. All assumptions of statistical methods were explicitly checked. Statistical analyses were performed using SAS 9.2 (Cary, NC, USA).
This was a retrospective cohort study based on data from an administrative database including all Italian hospital admissions and discharges. Given the retrospective nature of the study and the use of anonymized patient data, ethical approval was not required.
2.2. Study population The Italian National Registry of Hospital Discharge Records (HDR), including all records from January 1, 2001 and December 31, 2012 was used as the main source of data. All hospital discharge records of patients aged 18 to 100 years, resident in Italy, and reporting a primary diagnosis of AMI (International Classification of Disease, 9th Revision, Clinical Modification [ICD 9 CM] 410) or a secondary diagnosis of AMI with any concomitant AMI complication within the primary diagnosis (ICD-9-CM codes 411, 413, 414, 426, 427, 428, 423.0, 429.5, 429.6, 429.71, 429.79, 429.81, 518.4, 518.81, 780.01, 780.2, 785.51, 799.1, 997.02 and 998.2) were selected (Outcomes evaluation National program [PNE] Ed. 2013; available at http://95.110.213.190/PNEed13/). All patients alive at discharge (within 48 h from admission) and all patients with a previous AMI admission (within 30 days from the index admission) were excluded to minimize the inclusion of false AMIs and multiple admissions due to the same event [16,17]. For the purpose of this study AMI was classified as complicated by HF if a patient was discharged with a diagnosis of AMI and an additional diagnosis of HF (ICD-9-CM codes 428 [heart failure], 518.4 [acute pulmonary edema], and 785.51 [cardiogenic shock]). HF was chosen among all the other possible AMI complications, as the most suitable diagnosis to assess the prognosis of patients with AMI, their quality of life, and the impact on the national health system (in particular the need of re-hospitalizations) [18–20]. A group of “low-risk AMI patients” was identified as “ideal benchmark” of optimally-treated patients with which to compare the high-risk group. This “low risk” group was defined according to previously published studies as patients having b70 years, with no heart failure at the time of the index admission, and treated with PTCA within 48 h from the admission [21].
2.3. Outcomes Short- and mid-term mortality and readmission rates, and mortality and readmission within 30 days, 60 days and 1 year were considered as the main adverse outcomes. The Italian national HDR system cannot be systematically linked with comprehensive mortality registers, thus mortality rates refer only to events occurred during hospitalization. On this basis, three different in-hospital mortality estimates were computed for each time-point: index admission mortality rate (I-MR), total in-hospital mortality rate (T-MR) and fatal readmission rate (F-RR). Accordingly, I-MR refers to patients who died during the index admission for AMI (including transfers), T-MR refers to patients who died during the index admission for AMI or during any other subsequent admission within the considered time span, and F-RR refers to patients who were alive at discharged from the index admission but died during any other subsequent admission within the considered time span. Readmission rates due to HF were also considered and calculated dividing the number of patients who had at least one re-hospitalization due to HF as primary diagnosis within the considered time span by the number of patients who were alive at discharge from the index admission. Trends in the proportion of re-hospitalizations from 2001 to 2011 were then analyzed.
3. Results A total of 1,110,822 AMI events occurred from 2001 to 2011, were included. Results referring to the years 2001 to 2012 showed an increase in the proportion of NSTEMI from 23.2.% to 48.2% and a decrease in the proportion of STEMI decreased from 76.8% to 52.8%, while age remained substantially unchanged over time (mean age from 69.3 to 70.6 years) (p of trend = NS). I-MR decreased significantly from 11.34% to 8.99%, with a mean annual change of − 0.23% (CI − 0.27% to − 0.20%). T-MR showed a significant decrease within each time span, specifically, T-MR within 30 days, 60 days and 1 year from index admission decreased respectively from 12.01% to 9.67%, from 13.12% to 10.93, and from 16.46% to 14.68%. Multivariate logistic regression analyses confirmed a significant decrease in in-hospital mortality rate (I-MR) per year after index admission (odds ratio [OR], 0.977; p b 0.001), and a significant decrease in total in-hospital mortality rate (T-MR) per year within 30 days (OR, 0.977; p b 0.001) and 60 days (OR, 0.988; p b 0.0001). The Cox regression model, adjusting for the above confounders, also showed a significant decrease in total in-hospital mortality rate within 1 year from index admission (hazard ratio [HR], 0.993; p b 0.001). The percentage of patients who, between 2001 and 2011, were alive at discharge from the index hospitalization but were re-hospitalized and died during the stay, remained unchanged at 30 days (about 0.7%), and increased slightly at 60 days. However, this change of F-RR became larger and statistically significant at 1 year (mean annual change 0.04% [0.02% to 0.06%; p = 0.0019]) (Table 1). Risk ratios between 1-year F-RR and 1-year I-MR showed a steady increase over the years (Fig. 1), while both the yearly admissions due to AMI and the number of AMI patients developing HF during the index admission showed no substantial variation over the years, with around 100,000 admissions due to AMI and 20,000 patients with AMI and HF from 2001 to 2011. Therefore, our findings show a prevalence of around 20% of HF complicating AMI that remains substantially, stable over time. Patients who had a diagnosis of HF during the index admission showed an almost 4-times higher mortality rate than those who did not have a diagnosis of HF, even if favorable mortality trends were confirmed in both groups (Fig. 2). In particular, patients with HF complicating AMI, showed a decrease in I-MR from 26.53% to 23.21%, and a decrease in T-MR at 30 days, 60 days and 1 year respectively from 27.70% to 24.40%, from 29.94% to 27.07%, and from 36.08% to 33.90%. AMI patients with no diagnosis of HF at the index admission showed a decrease in I-MR from 6.97% to 4.83%, and a decrease in T-MR at 30 days, 60 days and 1 year respectively from 7.50% to 5.38%, from 8.29% to 6.22%, and from 10.82% to 9.07%. F-RR either remained stable or showed a slight increase over time, in spite of these favorable trends. F-RR, remained around 1% at 60 days in patients with no diagnosis of HF at the index admission, while raised from 2.63% to 2.95% in patients with HF complicating AMI. F-RR increased from 3.63% to 4.04% at 1 year in patients with no diagnosis of HF at the index admission and
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
117
Table 1 Trends of index in-hospital mortality rate (I-MR) total in-hospital mortality rate (T-MR) and fatal readmission rate (F-RR) after acute myocardial infarction (AMI). CI = confidence interval. Discharge year
AMI patients (N)
2001 86,107 2002 93,275 2003 100,031 2004 102,567 2005 103,847 2006 105,219 2007 105,813 2008 105,210 2009 102,094 2010 103,196 2011 103,463 Mean annual change (%) 95% CI of annual change (%) R2 p value for F test a
I-MR (%)
11.34 10.88 10.89 10.29 10.47 9.93 9.82 9.69 9.34 8.97 8.99 −0.23 (−0.27; −0.20) 0.97 b0.0001
T-MR (%) 30 days
60 days
1 year
12.01 11.55 11.67 11.06 11.21 10.64 10.54 10.43 10.05 9.70 9.67 −0.23 (−0.27; −0.20) 0.96 b0.0001
13.12 12.74 12.84 12.29 12.41 11.86 11.80 11.69 11.26 10.93 10.93 −0.22 (−0.25; −0.19) 0.97 b0.0001
16.46 16.21 16.34 15.87 16.00 15.50 15.44 15.48 14.92 14.61 14.68 −0.19 (−0.23; −0.15) 0.94 b0.0001
AMI patientsa (N) 76,345 83,126 89,134 92,009 92,975 94,775 95,423 95,017 92,557 93,935 94,166
F-RR (%) 30 days
60 days
1 year
0.67 0.67 0.78 0.77 0.74 0.72 0.72 0.74 0.71 0.73 0.69 0.00004 (−0.001; 0.001) 0.00 0.9918
1.48 1.54 1.64 1.66 1.63 1.60 1.62 1.68 1.58 1.64 1.61 0.01 (−0.004; 0.019) 0.21 0.1547
4.75 4.95 5.07 5.17 5.16 5.15 5.18 5.39 5.18 5.26 5.28 0.04 (0.02; 0.06) 0.68 0.0019
Patients survived at the index admission.
raised from 8.64% to 9.54% in patients with HF complicating AMI. However, all three in-hospital outcome measures (I-MR, F-RR, and T-MR) remained below 2% at all considered time points (30 days, 60 days and 1 year) in the subgroup of low-risk optimally-treated patients defined as “ideal benchmark” (Fig. 3). Re-hospitalizations after a previous AMI were analyzed to assess their impact on the National Health System. Results showed that almost 30,000 patients each year were re-hospitalized at least 1 time for all causes within 60 days after index admission (with about 30% surviving the stay) and more than 50,000 patients each year were re-hospitalized for all causes within 1 year from the index admission (with about 60% surviving the stay), thus confirming that re-hospitalizations are common in patients who were had been previously hospitalized due to AMI. The re-hospitalization rate showed a decrease over time only in patients with no HF complicating AMI at the index admission, with a proportion of patients re-hospitalized at least 1 time for any reason within 1 year from the index admission decreasing from 61.5% to 49.5%. Conversely, the 1-year all-cause re-hospitalization rate in patients with HF complicating AMI at the index admission, remained substantially stable in the years 2001 to 2011 (slightly decreasing from 66.7% to 64.2%) despite the decrease in in-hospital mortality rate after the index admission (Fig. 4). Finally, our data confirmed that HF is one of the most common causes of re-hospitalization, in particular in patients already diagnosed with HF during their index admission (roughly 8% within 60 days from index admission and 20% within 1 year from index admission) (Fig. 5). Re-hospitalizations were
Fig. 1. Trends of risk ratios between fatal readmission rates (F-RR) and index in-hospital mortality rates (I-MR).
generally more frequent among patients who received a diagnosis of HF at index admission (HF group), and showed a smaller decrease over time in this group (from 66.6% in 2001 to 48.7% in 2012 in the HF group versus 61.5% in 2001 to 35% in 2012 in the no-HF group). 4. Discussion There are currently two main types of risk of death in patients admitted with a diagnosis of AMI: the risk of dying during the index admission, and the risk of dying after the discharge. However, clinical research, clinical practice guidelines, health care resources allocation and process-of-care monitoring are mainly focused on the acute phase. Efforts specifically aimed at reducing post-discharge mortality are rare and uneven, if not lacking. Our analyses confirmed a decrease over the last decade in Italy in inhospital mortality of patients with AMI during index admission. Such improvement over time in the prognosis of patients with AMI could also be partly due to the increase observed in time in proportion of NSTEMI. The risk of dying after a re-hospitalization, however, remained unchanged over time. Previous studies based on administrative data showed that temporal trends of early and late mortality due to AMI are different. A nationwide Danish report [20] covering the years 1984 to 2008, reported a 52% decrease in mortality at 30 days due to AMI (from 31.4 to 14.8%), and a 28% decrease in mortality from day 31 to one year over the same period (from 15.6% to 11.1%), with short term and long term mortality converging over time. Another report on a different population of patients with AMI reported a mortality rate at 30 days and 1 year of follow-up after discharge of 1.2% and 3% respectively [22]. Other studies with different design showed similar results. Current clinical data from an Italian registry showed an unexpectedly high post-discharge mortality of patients hospitalized due to AMI [23], while a comprehensive analysis based on data from the national death registry of the mortality due to AMI in the area of Rome showed that mortality rates referring to the periods from day 1 to day 28 after discharge and from day 29 to 1 year after discharge were quite similar, being respectively 13.5% and 8% [24]. Available evidence on the incidence of HF after AMI is often inconsistent. The Framingham study showed that the incidence of HF within 30 days after AMI increased from 10% in the years 1970 to 1979 to 23.1% in the years 1990 to 1999, while incidence of HF within 5 years after AMI increased from 27.6% to 31.9% in the same years [11]. Other studies reported HF as one of the most predictive sign of cardiac death after AMI, even though its incidence showed a decreasing trend [17, 19,25]. Data from a more recent study on 3762 subjects hospitalized due to AMI in the Massachusetts medical centers in 2001, 2003, 2005,
118 C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
Fig. 2. Index admission mortality, fatal readmission and total mortality rates (I-MR, F-RR and T-MR, respectively) within 30 days, 60 days and 1 year in patients with AMI who did and did not develop heart failure during index admission.
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
I-MR
F-RR
119
T-MR 4,50
4,00
4,00
3,50
3,50
3,00
3,00
2,50
2,50
2,00
2,00
1,50
1,50
1,00
1,00
0,50
0,50
0,00
0,00
30 days
60 days
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
4,50
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
5,00
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
5,00
1 year
Fig. 3. Index admission mortality, fatal readmission and total mortality rates (I-MR, F-RR and T-MR, respectively) within 30 days, 60 days and 1 year in patients with AMI aged b70 years, who did not develop heart failure during index admission and were treated with PTCA within 48 h from the admission.
and 2007 confirmed that survival of patients with NSTEMI or STEMI and a history of heart failure is lower than that of patients with NSTEMI or STEMI and no history of HF. In particular, HF resulted as the second condition, after age, most frequently associated with post-discharge
mortality in patients with NSTEMI (OR = 1.45; 95% CI: 1.28–1.64; p b 0.001) [26]. One last population-based study on 20.812 patients with first-ever AMI in Western Australia reported a decrease in prevalence of HF after AMI from 28.1% in 1996 to 1998, to 16.5% in 2005 to
Patients without heart failure during the index admission 70,00 60,00 50,00 40,00
1 year
30,00
60 days
20,00
30 days
10,00 0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Patients with heart failure during the index admission 70,00 60,00 50,00 40,00
1 year
30,00
60 days
20,00
30 days
10,00 0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Fig. 4. Proportion of patients with acute myocardial infarction, stratified for presence or absence of heart failure during index admission, who were re-hospitalized at least one time for all causes within 30 days, 60 days and 1 year.
120
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
At least 1 readmission for heart failure within 1 year
At least 1 readmission for heart failure within 60 days 20,00
20,00
15,00
15,00
%
% 10,00
10,00
5,00
5,00
0,00
0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Patients without heart failure during index admission
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Patients with heart failure during index admission
Fig. 5. Proportion of patients with acute myocardial infarction, stratified for presence or absence of heart failure during index admission, who were re-hospitalized at least one time due to heart failure within 60 days and 1 year.
2007 (OR = 0.50; 95% CI: 0.44–0.55), and a smaller decline in crude case fatality rate at 28 days in patients with HF (from 20.5% to 15.9%; p b 0.05) than in those with no HF (from 11.0% to 4.8%; p b 0.001) [27]. Patients who were alive at 90 days, but developed HF within 90 days from index AMI had an adjusted hazard ratio of 2.7 of dying at 1 year after index AMI [27]. Our study shows a stable incidence rate of HF during index admissions due to AMI over time, while the prognosis of patients with AMI and HF can be very different. Patients who develop HF during the index admission had an almost 4 times higher mortality rate than patients who did not develop HF. Patients who developed HF during the index admission showed a smaller decrease in I-MR and T-MR at 30 days, 60 days and 1 year those who did not develop HF, while F-RR increases similarly in both groups over time. Heart failure resulted as the main cause of re-hospitalization after AMI, in particular in patients with a previous diagnosis of HF during the index admission (almost 8% within 60 days, and 20% within 1 year). However, the actual incidence of HF might be underestimated by the administrative records on which our data are based, that could also underreport or omit other conditions, such as transient or subclinical cardiac dysfunctions, which have significant prognostic impact. The T-MR of over 3% at 60 days in patients who develop HF during index admission compared to the T-MR of 2% at 1 year in optimally treated low-risk patients is a clear sign of the relevance of HF among the conditions complicating AMI. Its impact on the health system is also relevant when considering the percentage of re-hospitalizations in patients who had HF during the index hospitalization (48% at 1 year in 2012). This leads to the need of defining a comprehensive prognostic assessment and different tailored post-discharge management plans based on the risk of each patient, including the presence or absence of HF. The attention of the rehabilitation services and healthcare facilities that are in charge of patients with post-acute MI should focus on the sub-group of patients with AMI and HF. Several scores have been designed to stratify patients with AMI based on their risk at admission, and should be converted in an electronic format and applied before discharging patients after AMI [21,28,29]. On the other hand, the lowrisk subgroup, which is optimally treated while in hospital, can be an ideal reference standard, as these patients have almost no need of intensive secondary prevention or rehabilitations interventions. Patients classified in this group could be referred to general practitioners. A differential strategy implementing different paths within healthcare services for low-risk and high-risk patients could be a cost effective strategy, as it could allow to reallocate resources from the first subgroup to the second one, thus reducing the economic burden of MI on the healthcare system.
4.1. Limitations The present study was carried out based on data from hospital discharge records. The most commonly reported limitation of administrative data is their robustness as a basis for epidemiological research. One recent study, among several others, suggests that ICD-9-CM coding adequately identifies definite or possible AMIs according to the AHA criteria [30]. The lack of a systematic linkage to a comprehensive mortality registry may be considered a limitation of the study, as patients discharged after AMI could also have died out of the hospital [31]. However, published studies on Italian population registries reported a mortality rate of 8% at one year after discharge, that results in a difference of about 2.8% when compared to our data [24]. This is consistent with the reported estimate of out-of-hospital mortality in patients discharged after hospitalization due to AMI [32].
5. Conclusions The present study is a comprehensive and updated evaluation of the prognosis of Italian patients hospitalized due to AMI. Results showed that real-world patients hospitalized due to AMI are still highly at risk of long-term death, in spite of a constantly improving short-term prognosis. HF is a possible relevant complication subsequent to AMI, and deeply affects both short- and long-term prognoses. A subgroup of low-risk patients optimally treated after AMI can be identified, that has favorable short- and long-term outcomes.
Funding This study was carried out as part of the collaboration agreement between the Italian Istituto Superiore di Sanità and the GICR-IACPR (Italian Association for Cardiovascular Prevention, Rehabilitation and Epidemiology).
Conflicts of interest The authors report no relationships that could be construed as a conflict of interest.
Acknowledgments The authors thank Mrs. Gabriella Badoni for her technical support.
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
References [1] H. Kilburn Jr., L. Schoen, T. Wang, Acute myocardial infarction in New York State: 1996–2008, J. Community Health 37 (2012) 473–479. [2] H.M. Krumholz, Y. Wang, J. Chen, E.E. Drye, J.A. Spertus, J.S. Ross, J.P. Curtis, B.K. Nallamothu, J.H. Lichtman, E.P. Havranek, F.A. Masoudi, M.J. Radford, L.F. Han, M.T. Rapp, B.M. Straube, S.L. Normand, Reduction in acute myocardial infarction mortality in the United States: risk-standardized mortality rates from 1995–2006, J. Am. Med. Assoc. 302 (2009) 767–773. [3] W.D. Rosamond, L.E. Chambless, G. Heiss, T.H. Mosley, J. Coresh, E. Whitsel, L. Wagenknecht, H. Ni, A.R. Folsom, Twenty-two-year trends in incidence of myocardial infarction, coronary heart disease mortality, and case fatality in 4 US communities, 1987–2008, Circulation 125 (Apr 17 2012) 1848–1857. [4] A. Barchielli, D. Balzi, P. Naldoni, A.T. Roberts, F. Profili, F. Dima, L. Palmieri, Hospital discharge data for assessing myocardial infarction events and trends, and effects of diagnosis validation according to MONICA and AHA criteria, J. Epidemiol. Community Health 66 (2012) 462–467. [5] H. Johansen, S.E. Brien, P. Finès, J. Bernier, K. Humphries, T.A. Stukel, W.A. Ghali, Canadian Cardiovascular Outcomes Research Team, Thirty-day in-hospital revascularization and mortality rates after acute myocardial infarction in seven Canadian provinces, Can. J. Cardiol. 26 (2010) e243–e248. [6] R.W. Yeh, S. Sidney, M. Chandra, M. Sorel, J.V. Selby, A.S. Go, Population trends in the incidence and outcomes of acute myocardial infarction, N. Engl. J. Med. 362 (2010) 2155–2165. [7] W.J. Kostis, Y. Deng, S. Pantazopoulos, A.E. Moreyra, J.B. Kostis, Myocardial Infarction Data Acquisition System (MIDAS14) Study Group, Trends in mortality of acute myocardial infarction after discharge from the hospital, Circ. Cardiovasc. Qual. Outcomes 3 (2010) 581–589. [8] F.A. Masoudi, J.M. Foody, E.P. Havranek, Trends in acute myocardial infarction in 4 US states between 1992 and 2001: clinical characteristics, quality of care, and outcomes, Circulation 114 (2006) 2806–2814. [9] P. Miceli, M. Arcà, S. Pollina Addario, G. Fantaci, D. Fusco, P. Colais, C.A. Perucci, G. Dardanoni, Valutazione degli esiti dopo ricovero ospedaliero per infarto acuto del miocardio (IMA) in Sicilia: confronto preliminare fra due sistemi informativi correnti; Osservatorio Epidemiologico, , 2007, pp. 12–15 (Available at: http://www.epicentro. iss.it/regioni/sicilia/pdf/OE2007.pdf) (last accessed on January 25, 2015). [10] D.W. Baker, D. Einstadter, S.S. Husak, R.D. Cebul, Trends in postdischarge mortality and readmissions: has length of stay declined too far? Arch. Intern. Med. 164 (2004) 538–544. [11] R.S. Velagaleti, M.J. Pencina, J.M. Murabito, T.J. Wang, N.I. Parikh, R.B. D'Agostino, D. Levy, W.B. Kannel, R.S. Vasan, Long-term trends in the incidence of heart failure after myocardial infarction, Circulation 118 (2008) 2057–2062. [12] A. Torabi, A.S. Rigby, J.G.F. Cleland, Declining in-hospital mortality and increasing heart failure incidence in elderly patients with first myocardial infarction, J. Am. Coll. Cardiol. 55 (2010) 79–81. [13] L.C.J. Slobbe, O.A. Arah, J.P. Westert, Mortality in Dutch hospitals: trends in time, place and cause of death after admission for myocardial infarction and stroke. An observational study, BMC Health Serv. Res. 8 (2008) 52. [14] J. Ezekowitz, P. Kaul, J. Bakal, et al., Declining in-hospital mortality and increasing heart failure incidence in elderly patients with first myocardial infarction, J. Am. Coll. Cardiol. 53 (2009) 13. [15] F. Seccareccia, S. Rosato, P. D'Errigo, A. Maraschini, R.M. Manfredi, M. Lombardi, G. Badoni, C. Greco, In-hospital management of acute coronary syndrome patients: influence of age, gender and admission department on medical policy and outcome, G. Ital. Cardiol. 13 (2012) 124–131. [16] H. Tunstall-Pedoe, K. Kuulasmaa, P. Amouyel, D. Arveiler, A.M. Rajakangas, A. Pajak, Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents, Circulation 90 (1994) 583–612. [17] M. Shafazand, A. Rosengren, G. Lappas, K. Swedberg, M. Schaufelberger, Decreasing trends in the incidence of heart failure after acute myocardial infarction from 1993–2004: a study of 175,216 patients with a first acute myocardial infarction in Sweden, Eur. J. Heart Fail. 13 (2011) 135–141.
121
[18] A.J.C. Antoni, M.L. Antoni, G.E. Hoogslag, H. Boden, S.S. Liem, E. Boersma, K. Fox, M.J. Schalij, J.J. Bax, V. Delgado, Cardiovascular mortality and heart failure risk score for patients after ST-segment elevation acute myocardial infarction treated with primary percutaneous coronary intervention (data from the Leiden MISSION! Infarct Registry), Am. J. Cardiol. 109 (2012) 187–194. [19] Al Torabi, J.G. Cleland, N.K. Khan, P.H. Loh, A.L. Clark, F. Alamgir, J.L. Caplin, A.S. Rigby, K. Goode, The timing of development and subsequent clinical course of heart failure after a myocardial infarction, Eur. Heart J. 29 (2008) 859–870. [20] M. Schmidt, J.B. Jacobsen, T.L. Lash, H.E. Bøtker, H.T. Sørensen, 25 year trends in first time hospitalisation for acute myocardial infarction, subsequent short and long term mortality, and the prognostic impact of sex and comorbidity: a Danish nationwide cohort study, BMJ 344 (Jan 25 2012) e356. [21] C.B. Granger, R.J. Goldberg, O. Dabbous, K.S. Pieper, K.A. Eagle, C.P. Cannon, F. Van de Werf, Á. Avezum, S.G. Goodman, M.D. Flather, K.A.A. Fox, For the Global Registry of Acute Coronary Events Investigators, Predictors of hospital mortality in the global registry of acute coronary events, Arch. Intern. Med. 163 (2003) 2345–2353. [22] A.S. Adabag, T.M. Therneau, B.J. Gersh, S.A. Weston, V.L. Roger, Sudden death after myocardial infarction, JAMA 300 (2008) 2022–2029. [23] V. Rizzello, D. Lucci, A.P. Maggioni, S. Giampaoli, C. Greco, G. Di Pasquale, M.G. Pallotti, G.F. Mureddu, A. Di Chiara, A. Boccanelli, IN-ACS Outcome Investigators, Clinical epidemiology, management and outcome of acute coronary syndromes in the Italian network on acute coronary syndromes (IN-ACS Outcome study), Acute Card. Care 14 (2012 Jun) 71–80. [24] S. Picciotto, F. Forastiere, M. Stafoggia, D. D'Ippoliti, C. Ancona, C.A. Perucci, Association of area based deprivation status and individual educational attainment with incidence, treatment, and prognosis of first coronary event in Rome, Italy, J. Epidemiol. Community Health 60 (2006) 37–43. [25] K.A. Fox, P.G. Steg, K.A. Eagle, S.G. Goodman, F.A. Anderson Jr., C.B. Granger, M.D. Flather, A. Budaj, A. Quill, J.M. Gore, GRACE Investigators, Decline in rates of death and heart failure in acute coronary syndromes, 1999–2006, J. Am. Med. Assoc. 297 (2007) 1892–1900. [26] C.E. Darling, K.A. Fisher, D.D. McManus, A.H. Coles, F.A. Spencer, J.M. Gore, R.J. Goldberg, Survival after hospital discharge for ST-segment elevation and non-STsegment elevation acute myocardial infarction: a population-based study, Clin. Epidemiol. 5 (2013) 229–236. [27] J. Hung, T.H.K. Teng, J. Finn, M. Knuiman, T. Briffa, S. Stewart, F.M. Sanfilippo, S. Ridout, M. Hobbs, Trends from 1996 to 2007 in incidence and mortality outcomes of heart failure after acute myocardial infarction: a population-based study of 20812 patients with first acute myocardial infarction in Western Australia, J. Am. Heart Assoc. 2 (2013) e000172. . http://dx.doi.org/10.1161/JAHA.113.000172. [28] F. D'Ascenzo, G. Biondi-Zoccai, C. Moretti, M. Bollati, P. Omedè, F. Sciuto, D.G. Presutti, M.G. Modena, M. Gasparini, M.J. Reed, I. Sheiban, F. Gaita, TIMI, GRACE and alternative risk scores in acute coronary syndromes: a meta-analysis of 40 derivation studies on 216,552 patients and of 42 validation studies on 31,625 patients, Contemp.Clin. Trials 33 (2012) 507–514. [29] A.B. Hutchinson-Jaffe, S.G. Goodman, R.T. Yan, R. Wald, B. Elbarouni, B. Rose, K.A. Eagle, C.C. Lai, C. Baer, A. Langer, A.T. Yan, Canadian Acute Coronary Syndromes (ACS) Registry I and II Investigators and Canadian Global Registry of Acute Coronary Events (GRACE/GRACE 2) Investigators, Comparison of baseline characteristics, management and outcome of patients with non-ST-segment elevation acute coronary syndrome in versus not in clinical trials, Am. J. Cardiol. 106 (2010) 1389–1396. [30] A. Barchielli, D. Balzi, P. Naldoni, A.T. Roberts, F. Profili, F. Dima, L. Palmieri, Hospital discharge data for assessing myocardial infarction events and trends, and effects of diagnosis validation according to MONICA and AHA criteria, J. Epidemiol. Community Health 66 (2012) 462–467. [31] J. Herlitz, M. Dellborg, T. Karlsson, M.H. Evander, A. Berger, R. Luepker, Epidemiology of acute myocardial infarction with the emphasis on patients who did not reach the coronary care unit and non-AMI admissions, Int. J. Cardiol. 128 (2008) 342–349. [32] A. Adabag, T.M. Therneau, B.J. Gersh, S.A. Weston, V.L. Roger, Sudden death after myocardial infarction, J. Am. Med. Assoc. 300 (2008) 2022–2029.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Trends in mortality and heart failure after acute myocardial infarction in Italy from 2001 to 2011☆ Cesare Greco a,⁎, Stefano Rosato b, Paola D'Errigo b, Gian Francesco Mureddu a, Eleonora Lacorte b, Fulvia Seccareccia b a b
Division of Cardiology, San Giovanni Hospital, Rome, Italy Istituto Superiore di Sanità, Rome, Italy
a r t i c l e
i n f o
Article history: Received 28 August 2014 Received in revised form 16 December 2014 Accepted 26 January 2015 Available online 28 January 2015 Keywords: Acute myocardial infarction Acute coronary syndromes Discharge Mortality Heart failure Trend
a b s t r a c t Background: Uncertainties on long-term outcomes after acute myocardial infarction (AMI) still exist, despite the ongoing progresses in the management of patients with AMI. Aim of the study: Our aim was to appraise both the early prognosis and prognosis at 1-year after discharge of patients hospitalized due to AMI. Methods: This is a retrospective nationwide cohort study based on data from an administrative database on patients admitted with AMI from 2001 to 2011 in all Italian hospitals sites. Mortality and readmission rates within 30 days, 60 days and 1 year were calculated, as well as re-hospitalizations for all causes and for HF. Results: A total of 1,110,822 patients were included. Index admission mortality rate (I-MR) and total in-hospital mortality rate (T-MR) at up to 1 year both decreased respectively from 11.34% to 8.99% and from 16.46% to 14.68% in the years 2001 to 2011 (both p b 0.0001), while fatal readmission rate (F-RR) at 1 year increased from 4.75% to 5.28% (p = 0.0019). Patients that developed HF during the index admission had significantly higher I-MR and F-RR. I-MR, F-RR, and T-MR, however, remained low at any time point considered (30 days, 60 days and 1 year) in a subgroup of low-risk optimally-treated patients. Conclusions: The risk of fatal readmission at 1 year increased slightly over time, in spite of the remarkable improvements currently achieved in overall prognosis after AMI. The identification of patients at high risk (mainly due to HF complicating AMI), and of patients at low risk is crucial to define and support management strategies. © 2015 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Many reports published during the last decade have shown a steady reduction of in-hospital mortality due to acute myocardial infarction (AMI) [1–5]. This decrease has generally been considered as a consequence of the progresses reached in the in-hospital management of cardiovascular diseases in developed countries. These progresses are mainly due to the introduction and adoption in clinical practice of either fibrinolytic or mechanical reperfusion, and to the diffusion of pharmacological therapies including beta-blockers, angiotensinconverting enzyme inhibitors, antiplatelet agents and statins. The proportion of patients with ST segment elevation myocardial infarction (STEMI) has concurrently decreased in several countries as compared to the proportion of patients with non-ST segment elevation myocardial infarction (NSTEMI). Continuous efforts in the primary prevention of MI
☆ Primary funding source: Istituto Superiore di Sanità, Rome, Italy. ⁎ Corresponding author at: San Giovanni Hospital, Via dell'Amba Aradam, 8-00184 Rome, Italy. E-mail address: [email protected] (C. Greco).
http://dx.doi.org/10.1016/j.ijcard.2015.01.073 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
may have led to a variation in the frequency of different types of AMI, with smaller infarcts becoming more common than larger ones, in spite of a less favorable overall profile of index patients who became older and having a higher number of comorbidities [6]. However, it is still unclear if the increase in survival during index hospitalization actually continues after discharge. Apparently, while in-hospital mortality seems to decrease, some authors reported a paradoxical increase in short and long-term post-discharge deaths [2,7–10]. The increasing prevalence of heart failure (HF), mainly among the elderly with AMI, has been considered as one of the possible explanations for these unfavorable trends, along with all the potential implications on the organization of post-acute care services [7,11–14]. Most of the evidence supporting these hypotheses, however, is based on data from registries, which may be limited by selective and non-consecutive enrolment causing lower external validity [15]. We aimed at analyzing short- and long-term mortality trends among patients with AMI based on a large, comprehensive and universal administrative database including more than one million of patients over 11 years, with a special focus on the outcomes of patients surviving the index admission date and discharged with a diagnosis of heart failure.
116
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
2. Material and methods
2.4. Statistical analysis
2.1. Study design
A set of general linear models were designed to calculate annual variations in mortality and their corresponding 95% confidence intervals (CI). The goodness-of-fit of the linear regression models was assessed with the R2 test and the probability value for the F test statistic. Trends of risk ratios between F-RR and I-MR were analyzed for each selected time point. Multivariate logistic regression analyses were performed, adjusting for age, sex, comorbidities, and invasive cardiac procedures, to provide adjusted outcome data. Re-hospitalizations due to HF within the considered time spans were analyzed separately for patients with AMI who did and did not develop HF during the index admission. All assumptions of statistical methods were explicitly checked. Statistical analyses were performed using SAS 9.2 (Cary, NC, USA).
This was a retrospective cohort study based on data from an administrative database including all Italian hospital admissions and discharges. Given the retrospective nature of the study and the use of anonymized patient data, ethical approval was not required.
2.2. Study population The Italian National Registry of Hospital Discharge Records (HDR), including all records from January 1, 2001 and December 31, 2012 was used as the main source of data. All hospital discharge records of patients aged 18 to 100 years, resident in Italy, and reporting a primary diagnosis of AMI (International Classification of Disease, 9th Revision, Clinical Modification [ICD 9 CM] 410) or a secondary diagnosis of AMI with any concomitant AMI complication within the primary diagnosis (ICD-9-CM codes 411, 413, 414, 426, 427, 428, 423.0, 429.5, 429.6, 429.71, 429.79, 429.81, 518.4, 518.81, 780.01, 780.2, 785.51, 799.1, 997.02 and 998.2) were selected (Outcomes evaluation National program [PNE] Ed. 2013; available at http://95.110.213.190/PNEed13/). All patients alive at discharge (within 48 h from admission) and all patients with a previous AMI admission (within 30 days from the index admission) were excluded to minimize the inclusion of false AMIs and multiple admissions due to the same event [16,17]. For the purpose of this study AMI was classified as complicated by HF if a patient was discharged with a diagnosis of AMI and an additional diagnosis of HF (ICD-9-CM codes 428 [heart failure], 518.4 [acute pulmonary edema], and 785.51 [cardiogenic shock]). HF was chosen among all the other possible AMI complications, as the most suitable diagnosis to assess the prognosis of patients with AMI, their quality of life, and the impact on the national health system (in particular the need of re-hospitalizations) [18–20]. A group of “low-risk AMI patients” was identified as “ideal benchmark” of optimally-treated patients with which to compare the high-risk group. This “low risk” group was defined according to previously published studies as patients having b70 years, with no heart failure at the time of the index admission, and treated with PTCA within 48 h from the admission [21].
2.3. Outcomes Short- and mid-term mortality and readmission rates, and mortality and readmission within 30 days, 60 days and 1 year were considered as the main adverse outcomes. The Italian national HDR system cannot be systematically linked with comprehensive mortality registers, thus mortality rates refer only to events occurred during hospitalization. On this basis, three different in-hospital mortality estimates were computed for each time-point: index admission mortality rate (I-MR), total in-hospital mortality rate (T-MR) and fatal readmission rate (F-RR). Accordingly, I-MR refers to patients who died during the index admission for AMI (including transfers), T-MR refers to patients who died during the index admission for AMI or during any other subsequent admission within the considered time span, and F-RR refers to patients who were alive at discharged from the index admission but died during any other subsequent admission within the considered time span. Readmission rates due to HF were also considered and calculated dividing the number of patients who had at least one re-hospitalization due to HF as primary diagnosis within the considered time span by the number of patients who were alive at discharge from the index admission. Trends in the proportion of re-hospitalizations from 2001 to 2011 were then analyzed.
3. Results A total of 1,110,822 AMI events occurred from 2001 to 2011, were included. Results referring to the years 2001 to 2012 showed an increase in the proportion of NSTEMI from 23.2.% to 48.2% and a decrease in the proportion of STEMI decreased from 76.8% to 52.8%, while age remained substantially unchanged over time (mean age from 69.3 to 70.6 years) (p of trend = NS). I-MR decreased significantly from 11.34% to 8.99%, with a mean annual change of − 0.23% (CI − 0.27% to − 0.20%). T-MR showed a significant decrease within each time span, specifically, T-MR within 30 days, 60 days and 1 year from index admission decreased respectively from 12.01% to 9.67%, from 13.12% to 10.93, and from 16.46% to 14.68%. Multivariate logistic regression analyses confirmed a significant decrease in in-hospital mortality rate (I-MR) per year after index admission (odds ratio [OR], 0.977; p b 0.001), and a significant decrease in total in-hospital mortality rate (T-MR) per year within 30 days (OR, 0.977; p b 0.001) and 60 days (OR, 0.988; p b 0.0001). The Cox regression model, adjusting for the above confounders, also showed a significant decrease in total in-hospital mortality rate within 1 year from index admission (hazard ratio [HR], 0.993; p b 0.001). The percentage of patients who, between 2001 and 2011, were alive at discharge from the index hospitalization but were re-hospitalized and died during the stay, remained unchanged at 30 days (about 0.7%), and increased slightly at 60 days. However, this change of F-RR became larger and statistically significant at 1 year (mean annual change 0.04% [0.02% to 0.06%; p = 0.0019]) (Table 1). Risk ratios between 1-year F-RR and 1-year I-MR showed a steady increase over the years (Fig. 1), while both the yearly admissions due to AMI and the number of AMI patients developing HF during the index admission showed no substantial variation over the years, with around 100,000 admissions due to AMI and 20,000 patients with AMI and HF from 2001 to 2011. Therefore, our findings show a prevalence of around 20% of HF complicating AMI that remains substantially, stable over time. Patients who had a diagnosis of HF during the index admission showed an almost 4-times higher mortality rate than those who did not have a diagnosis of HF, even if favorable mortality trends were confirmed in both groups (Fig. 2). In particular, patients with HF complicating AMI, showed a decrease in I-MR from 26.53% to 23.21%, and a decrease in T-MR at 30 days, 60 days and 1 year respectively from 27.70% to 24.40%, from 29.94% to 27.07%, and from 36.08% to 33.90%. AMI patients with no diagnosis of HF at the index admission showed a decrease in I-MR from 6.97% to 4.83%, and a decrease in T-MR at 30 days, 60 days and 1 year respectively from 7.50% to 5.38%, from 8.29% to 6.22%, and from 10.82% to 9.07%. F-RR either remained stable or showed a slight increase over time, in spite of these favorable trends. F-RR, remained around 1% at 60 days in patients with no diagnosis of HF at the index admission, while raised from 2.63% to 2.95% in patients with HF complicating AMI. F-RR increased from 3.63% to 4.04% at 1 year in patients with no diagnosis of HF at the index admission and
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
117
Table 1 Trends of index in-hospital mortality rate (I-MR) total in-hospital mortality rate (T-MR) and fatal readmission rate (F-RR) after acute myocardial infarction (AMI). CI = confidence interval. Discharge year
AMI patients (N)
2001 86,107 2002 93,275 2003 100,031 2004 102,567 2005 103,847 2006 105,219 2007 105,813 2008 105,210 2009 102,094 2010 103,196 2011 103,463 Mean annual change (%) 95% CI of annual change (%) R2 p value for F test a
I-MR (%)
11.34 10.88 10.89 10.29 10.47 9.93 9.82 9.69 9.34 8.97 8.99 −0.23 (−0.27; −0.20) 0.97 b0.0001
T-MR (%) 30 days
60 days
1 year
12.01 11.55 11.67 11.06 11.21 10.64 10.54 10.43 10.05 9.70 9.67 −0.23 (−0.27; −0.20) 0.96 b0.0001
13.12 12.74 12.84 12.29 12.41 11.86 11.80 11.69 11.26 10.93 10.93 −0.22 (−0.25; −0.19) 0.97 b0.0001
16.46 16.21 16.34 15.87 16.00 15.50 15.44 15.48 14.92 14.61 14.68 −0.19 (−0.23; −0.15) 0.94 b0.0001
AMI patientsa (N) 76,345 83,126 89,134 92,009 92,975 94,775 95,423 95,017 92,557 93,935 94,166
F-RR (%) 30 days
60 days
1 year
0.67 0.67 0.78 0.77 0.74 0.72 0.72 0.74 0.71 0.73 0.69 0.00004 (−0.001; 0.001) 0.00 0.9918
1.48 1.54 1.64 1.66 1.63 1.60 1.62 1.68 1.58 1.64 1.61 0.01 (−0.004; 0.019) 0.21 0.1547
4.75 4.95 5.07 5.17 5.16 5.15 5.18 5.39 5.18 5.26 5.28 0.04 (0.02; 0.06) 0.68 0.0019
Patients survived at the index admission.
raised from 8.64% to 9.54% in patients with HF complicating AMI. However, all three in-hospital outcome measures (I-MR, F-RR, and T-MR) remained below 2% at all considered time points (30 days, 60 days and 1 year) in the subgroup of low-risk optimally-treated patients defined as “ideal benchmark” (Fig. 3). Re-hospitalizations after a previous AMI were analyzed to assess their impact on the National Health System. Results showed that almost 30,000 patients each year were re-hospitalized at least 1 time for all causes within 60 days after index admission (with about 30% surviving the stay) and more than 50,000 patients each year were re-hospitalized for all causes within 1 year from the index admission (with about 60% surviving the stay), thus confirming that re-hospitalizations are common in patients who were had been previously hospitalized due to AMI. The re-hospitalization rate showed a decrease over time only in patients with no HF complicating AMI at the index admission, with a proportion of patients re-hospitalized at least 1 time for any reason within 1 year from the index admission decreasing from 61.5% to 49.5%. Conversely, the 1-year all-cause re-hospitalization rate in patients with HF complicating AMI at the index admission, remained substantially stable in the years 2001 to 2011 (slightly decreasing from 66.7% to 64.2%) despite the decrease in in-hospital mortality rate after the index admission (Fig. 4). Finally, our data confirmed that HF is one of the most common causes of re-hospitalization, in particular in patients already diagnosed with HF during their index admission (roughly 8% within 60 days from index admission and 20% within 1 year from index admission) (Fig. 5). Re-hospitalizations were
Fig. 1. Trends of risk ratios between fatal readmission rates (F-RR) and index in-hospital mortality rates (I-MR).
generally more frequent among patients who received a diagnosis of HF at index admission (HF group), and showed a smaller decrease over time in this group (from 66.6% in 2001 to 48.7% in 2012 in the HF group versus 61.5% in 2001 to 35% in 2012 in the no-HF group). 4. Discussion There are currently two main types of risk of death in patients admitted with a diagnosis of AMI: the risk of dying during the index admission, and the risk of dying after the discharge. However, clinical research, clinical practice guidelines, health care resources allocation and process-of-care monitoring are mainly focused on the acute phase. Efforts specifically aimed at reducing post-discharge mortality are rare and uneven, if not lacking. Our analyses confirmed a decrease over the last decade in Italy in inhospital mortality of patients with AMI during index admission. Such improvement over time in the prognosis of patients with AMI could also be partly due to the increase observed in time in proportion of NSTEMI. The risk of dying after a re-hospitalization, however, remained unchanged over time. Previous studies based on administrative data showed that temporal trends of early and late mortality due to AMI are different. A nationwide Danish report [20] covering the years 1984 to 2008, reported a 52% decrease in mortality at 30 days due to AMI (from 31.4 to 14.8%), and a 28% decrease in mortality from day 31 to one year over the same period (from 15.6% to 11.1%), with short term and long term mortality converging over time. Another report on a different population of patients with AMI reported a mortality rate at 30 days and 1 year of follow-up after discharge of 1.2% and 3% respectively [22]. Other studies with different design showed similar results. Current clinical data from an Italian registry showed an unexpectedly high post-discharge mortality of patients hospitalized due to AMI [23], while a comprehensive analysis based on data from the national death registry of the mortality due to AMI in the area of Rome showed that mortality rates referring to the periods from day 1 to day 28 after discharge and from day 29 to 1 year after discharge were quite similar, being respectively 13.5% and 8% [24]. Available evidence on the incidence of HF after AMI is often inconsistent. The Framingham study showed that the incidence of HF within 30 days after AMI increased from 10% in the years 1970 to 1979 to 23.1% in the years 1990 to 1999, while incidence of HF within 5 years after AMI increased from 27.6% to 31.9% in the same years [11]. Other studies reported HF as one of the most predictive sign of cardiac death after AMI, even though its incidence showed a decreasing trend [17, 19,25]. Data from a more recent study on 3762 subjects hospitalized due to AMI in the Massachusetts medical centers in 2001, 2003, 2005,
118 C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
Fig. 2. Index admission mortality, fatal readmission and total mortality rates (I-MR, F-RR and T-MR, respectively) within 30 days, 60 days and 1 year in patients with AMI who did and did not develop heart failure during index admission.
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
I-MR
F-RR
119
T-MR 4,50
4,00
4,00
3,50
3,50
3,00
3,00
2,50
2,50
2,00
2,00
1,50
1,50
1,00
1,00
0,50
0,50
0,00
0,00
30 days
60 days
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
4,50
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
5,00
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
5,00
1 year
Fig. 3. Index admission mortality, fatal readmission and total mortality rates (I-MR, F-RR and T-MR, respectively) within 30 days, 60 days and 1 year in patients with AMI aged b70 years, who did not develop heart failure during index admission and were treated with PTCA within 48 h from the admission.
and 2007 confirmed that survival of patients with NSTEMI or STEMI and a history of heart failure is lower than that of patients with NSTEMI or STEMI and no history of HF. In particular, HF resulted as the second condition, after age, most frequently associated with post-discharge
mortality in patients with NSTEMI (OR = 1.45; 95% CI: 1.28–1.64; p b 0.001) [26]. One last population-based study on 20.812 patients with first-ever AMI in Western Australia reported a decrease in prevalence of HF after AMI from 28.1% in 1996 to 1998, to 16.5% in 2005 to
Patients without heart failure during the index admission 70,00 60,00 50,00 40,00
1 year
30,00
60 days
20,00
30 days
10,00 0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Patients with heart failure during the index admission 70,00 60,00 50,00 40,00
1 year
30,00
60 days
20,00
30 days
10,00 0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Fig. 4. Proportion of patients with acute myocardial infarction, stratified for presence or absence of heart failure during index admission, who were re-hospitalized at least one time for all causes within 30 days, 60 days and 1 year.
120
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
At least 1 readmission for heart failure within 1 year
At least 1 readmission for heart failure within 60 days 20,00
20,00
15,00
15,00
%
% 10,00
10,00
5,00
5,00
0,00
0,00 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Patients without heart failure during index admission
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
Patients with heart failure during index admission
Fig. 5. Proportion of patients with acute myocardial infarction, stratified for presence or absence of heart failure during index admission, who were re-hospitalized at least one time due to heart failure within 60 days and 1 year.
2007 (OR = 0.50; 95% CI: 0.44–0.55), and a smaller decline in crude case fatality rate at 28 days in patients with HF (from 20.5% to 15.9%; p b 0.05) than in those with no HF (from 11.0% to 4.8%; p b 0.001) [27]. Patients who were alive at 90 days, but developed HF within 90 days from index AMI had an adjusted hazard ratio of 2.7 of dying at 1 year after index AMI [27]. Our study shows a stable incidence rate of HF during index admissions due to AMI over time, while the prognosis of patients with AMI and HF can be very different. Patients who develop HF during the index admission had an almost 4 times higher mortality rate than patients who did not develop HF. Patients who developed HF during the index admission showed a smaller decrease in I-MR and T-MR at 30 days, 60 days and 1 year those who did not develop HF, while F-RR increases similarly in both groups over time. Heart failure resulted as the main cause of re-hospitalization after AMI, in particular in patients with a previous diagnosis of HF during the index admission (almost 8% within 60 days, and 20% within 1 year). However, the actual incidence of HF might be underestimated by the administrative records on which our data are based, that could also underreport or omit other conditions, such as transient or subclinical cardiac dysfunctions, which have significant prognostic impact. The T-MR of over 3% at 60 days in patients who develop HF during index admission compared to the T-MR of 2% at 1 year in optimally treated low-risk patients is a clear sign of the relevance of HF among the conditions complicating AMI. Its impact on the health system is also relevant when considering the percentage of re-hospitalizations in patients who had HF during the index hospitalization (48% at 1 year in 2012). This leads to the need of defining a comprehensive prognostic assessment and different tailored post-discharge management plans based on the risk of each patient, including the presence or absence of HF. The attention of the rehabilitation services and healthcare facilities that are in charge of patients with post-acute MI should focus on the sub-group of patients with AMI and HF. Several scores have been designed to stratify patients with AMI based on their risk at admission, and should be converted in an electronic format and applied before discharging patients after AMI [21,28,29]. On the other hand, the lowrisk subgroup, which is optimally treated while in hospital, can be an ideal reference standard, as these patients have almost no need of intensive secondary prevention or rehabilitations interventions. Patients classified in this group could be referred to general practitioners. A differential strategy implementing different paths within healthcare services for low-risk and high-risk patients could be a cost effective strategy, as it could allow to reallocate resources from the first subgroup to the second one, thus reducing the economic burden of MI on the healthcare system.
4.1. Limitations The present study was carried out based on data from hospital discharge records. The most commonly reported limitation of administrative data is their robustness as a basis for epidemiological research. One recent study, among several others, suggests that ICD-9-CM coding adequately identifies definite or possible AMIs according to the AHA criteria [30]. The lack of a systematic linkage to a comprehensive mortality registry may be considered a limitation of the study, as patients discharged after AMI could also have died out of the hospital [31]. However, published studies on Italian population registries reported a mortality rate of 8% at one year after discharge, that results in a difference of about 2.8% when compared to our data [24]. This is consistent with the reported estimate of out-of-hospital mortality in patients discharged after hospitalization due to AMI [32].
5. Conclusions The present study is a comprehensive and updated evaluation of the prognosis of Italian patients hospitalized due to AMI. Results showed that real-world patients hospitalized due to AMI are still highly at risk of long-term death, in spite of a constantly improving short-term prognosis. HF is a possible relevant complication subsequent to AMI, and deeply affects both short- and long-term prognoses. A subgroup of low-risk patients optimally treated after AMI can be identified, that has favorable short- and long-term outcomes.
Funding This study was carried out as part of the collaboration agreement between the Italian Istituto Superiore di Sanità and the GICR-IACPR (Italian Association for Cardiovascular Prevention, Rehabilitation and Epidemiology).
Conflicts of interest The authors report no relationships that could be construed as a conflict of interest.
Acknowledgments The authors thank Mrs. Gabriella Badoni for her technical support.
C. Greco et al. / International Journal of Cardiology 184 (2015) 115–121
References [1] H. Kilburn Jr., L. Schoen, T. Wang, Acute myocardial infarction in New York State: 1996–2008, J. Community Health 37 (2012) 473–479. [2] H.M. Krumholz, Y. Wang, J. Chen, E.E. Drye, J.A. Spertus, J.S. Ross, J.P. Curtis, B.K. Nallamothu, J.H. Lichtman, E.P. Havranek, F.A. Masoudi, M.J. Radford, L.F. Han, M.T. Rapp, B.M. Straube, S.L. Normand, Reduction in acute myocardial infarction mortality in the United States: risk-standardized mortality rates from 1995–2006, J. Am. Med. Assoc. 302 (2009) 767–773. [3] W.D. Rosamond, L.E. Chambless, G. Heiss, T.H. Mosley, J. Coresh, E. Whitsel, L. Wagenknecht, H. Ni, A.R. Folsom, Twenty-two-year trends in incidence of myocardial infarction, coronary heart disease mortality, and case fatality in 4 US communities, 1987–2008, Circulation 125 (Apr 17 2012) 1848–1857. [4] A. Barchielli, D. Balzi, P. Naldoni, A.T. Roberts, F. Profili, F. Dima, L. Palmieri, Hospital discharge data for assessing myocardial infarction events and trends, and effects of diagnosis validation according to MONICA and AHA criteria, J. Epidemiol. Community Health 66 (2012) 462–467. [5] H. Johansen, S.E. Brien, P. Finès, J. Bernier, K. Humphries, T.A. Stukel, W.A. Ghali, Canadian Cardiovascular Outcomes Research Team, Thirty-day in-hospital revascularization and mortality rates after acute myocardial infarction in seven Canadian provinces, Can. J. Cardiol. 26 (2010) e243–e248. [6] R.W. Yeh, S. Sidney, M. Chandra, M. Sorel, J.V. Selby, A.S. Go, Population trends in the incidence and outcomes of acute myocardial infarction, N. Engl. J. Med. 362 (2010) 2155–2165. [7] W.J. Kostis, Y. Deng, S. Pantazopoulos, A.E. Moreyra, J.B. Kostis, Myocardial Infarction Data Acquisition System (MIDAS14) Study Group, Trends in mortality of acute myocardial infarction after discharge from the hospital, Circ. Cardiovasc. Qual. Outcomes 3 (2010) 581–589. [8] F.A. Masoudi, J.M. Foody, E.P. Havranek, Trends in acute myocardial infarction in 4 US states between 1992 and 2001: clinical characteristics, quality of care, and outcomes, Circulation 114 (2006) 2806–2814. [9] P. Miceli, M. Arcà, S. Pollina Addario, G. Fantaci, D. Fusco, P. Colais, C.A. Perucci, G. Dardanoni, Valutazione degli esiti dopo ricovero ospedaliero per infarto acuto del miocardio (IMA) in Sicilia: confronto preliminare fra due sistemi informativi correnti; Osservatorio Epidemiologico, , 2007, pp. 12–15 (Available at: http://www.epicentro. iss.it/regioni/sicilia/pdf/OE2007.pdf) (last accessed on January 25, 2015). [10] D.W. Baker, D. Einstadter, S.S. Husak, R.D. Cebul, Trends in postdischarge mortality and readmissions: has length of stay declined too far? Arch. Intern. Med. 164 (2004) 538–544. [11] R.S. Velagaleti, M.J. Pencina, J.M. Murabito, T.J. Wang, N.I. Parikh, R.B. D'Agostino, D. Levy, W.B. Kannel, R.S. Vasan, Long-term trends in the incidence of heart failure after myocardial infarction, Circulation 118 (2008) 2057–2062. [12] A. Torabi, A.S. Rigby, J.G.F. Cleland, Declining in-hospital mortality and increasing heart failure incidence in elderly patients with first myocardial infarction, J. Am. Coll. Cardiol. 55 (2010) 79–81. [13] L.C.J. Slobbe, O.A. Arah, J.P. Westert, Mortality in Dutch hospitals: trends in time, place and cause of death after admission for myocardial infarction and stroke. An observational study, BMC Health Serv. Res. 8 (2008) 52. [14] J. Ezekowitz, P. Kaul, J. Bakal, et al., Declining in-hospital mortality and increasing heart failure incidence in elderly patients with first myocardial infarction, J. Am. Coll. Cardiol. 53 (2009) 13. [15] F. Seccareccia, S. Rosato, P. D'Errigo, A. Maraschini, R.M. Manfredi, M. Lombardi, G. Badoni, C. Greco, In-hospital management of acute coronary syndrome patients: influence of age, gender and admission department on medical policy and outcome, G. Ital. Cardiol. 13 (2012) 124–131. [16] H. Tunstall-Pedoe, K. Kuulasmaa, P. Amouyel, D. Arveiler, A.M. Rajakangas, A. Pajak, Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents, Circulation 90 (1994) 583–612. [17] M. Shafazand, A. Rosengren, G. Lappas, K. Swedberg, M. Schaufelberger, Decreasing trends in the incidence of heart failure after acute myocardial infarction from 1993–2004: a study of 175,216 patients with a first acute myocardial infarction in Sweden, Eur. J. Heart Fail. 13 (2011) 135–141.
121
[18] A.J.C. Antoni, M.L. Antoni, G.E. Hoogslag, H. Boden, S.S. Liem, E. Boersma, K. Fox, M.J. Schalij, J.J. Bax, V. Delgado, Cardiovascular mortality and heart failure risk score for patients after ST-segment elevation acute myocardial infarction treated with primary percutaneous coronary intervention (data from the Leiden MISSION! Infarct Registry), Am. J. Cardiol. 109 (2012) 187–194. [19] Al Torabi, J.G. Cleland, N.K. Khan, P.H. Loh, A.L. Clark, F. Alamgir, J.L. Caplin, A.S. Rigby, K. Goode, The timing of development and subsequent clinical course of heart failure after a myocardial infarction, Eur. Heart J. 29 (2008) 859–870. [20] M. Schmidt, J.B. Jacobsen, T.L. Lash, H.E. Bøtker, H.T. Sørensen, 25 year trends in first time hospitalisation for acute myocardial infarction, subsequent short and long term mortality, and the prognostic impact of sex and comorbidity: a Danish nationwide cohort study, BMJ 344 (Jan 25 2012) e356. [21] C.B. Granger, R.J. Goldberg, O. Dabbous, K.S. Pieper, K.A. Eagle, C.P. Cannon, F. Van de Werf, Á. Avezum, S.G. Goodman, M.D. Flather, K.A.A. Fox, For the Global Registry of Acute Coronary Events Investigators, Predictors of hospital mortality in the global registry of acute coronary events, Arch. Intern. Med. 163 (2003) 2345–2353. [22] A.S. Adabag, T.M. Therneau, B.J. Gersh, S.A. Weston, V.L. Roger, Sudden death after myocardial infarction, JAMA 300 (2008) 2022–2029. [23] V. Rizzello, D. Lucci, A.P. Maggioni, S. Giampaoli, C. Greco, G. Di Pasquale, M.G. Pallotti, G.F. Mureddu, A. Di Chiara, A. Boccanelli, IN-ACS Outcome Investigators, Clinical epidemiology, management and outcome of acute coronary syndromes in the Italian network on acute coronary syndromes (IN-ACS Outcome study), Acute Card. Care 14 (2012 Jun) 71–80. [24] S. Picciotto, F. Forastiere, M. Stafoggia, D. D'Ippoliti, C. Ancona, C.A. Perucci, Association of area based deprivation status and individual educational attainment with incidence, treatment, and prognosis of first coronary event in Rome, Italy, J. Epidemiol. Community Health 60 (2006) 37–43. [25] K.A. Fox, P.G. Steg, K.A. Eagle, S.G. Goodman, F.A. Anderson Jr., C.B. Granger, M.D. Flather, A. Budaj, A. Quill, J.M. Gore, GRACE Investigators, Decline in rates of death and heart failure in acute coronary syndromes, 1999–2006, J. Am. Med. Assoc. 297 (2007) 1892–1900. [26] C.E. Darling, K.A. Fisher, D.D. McManus, A.H. Coles, F.A. Spencer, J.M. Gore, R.J. Goldberg, Survival after hospital discharge for ST-segment elevation and non-STsegment elevation acute myocardial infarction: a population-based study, Clin. Epidemiol. 5 (2013) 229–236. [27] J. Hung, T.H.K. Teng, J. Finn, M. Knuiman, T. Briffa, S. Stewart, F.M. Sanfilippo, S. Ridout, M. Hobbs, Trends from 1996 to 2007 in incidence and mortality outcomes of heart failure after acute myocardial infarction: a population-based study of 20812 patients with first acute myocardial infarction in Western Australia, J. Am. Heart Assoc. 2 (2013) e000172. . http://dx.doi.org/10.1161/JAHA.113.000172. [28] F. D'Ascenzo, G. Biondi-Zoccai, C. Moretti, M. Bollati, P. Omedè, F. Sciuto, D.G. Presutti, M.G. Modena, M. Gasparini, M.J. Reed, I. Sheiban, F. Gaita, TIMI, GRACE and alternative risk scores in acute coronary syndromes: a meta-analysis of 40 derivation studies on 216,552 patients and of 42 validation studies on 31,625 patients, Contemp.Clin. Trials 33 (2012) 507–514. [29] A.B. Hutchinson-Jaffe, S.G. Goodman, R.T. Yan, R. Wald, B. Elbarouni, B. Rose, K.A. Eagle, C.C. Lai, C. Baer, A. Langer, A.T. Yan, Canadian Acute Coronary Syndromes (ACS) Registry I and II Investigators and Canadian Global Registry of Acute Coronary Events (GRACE/GRACE 2) Investigators, Comparison of baseline characteristics, management and outcome of patients with non-ST-segment elevation acute coronary syndrome in versus not in clinical trials, Am. J. Cardiol. 106 (2010) 1389–1396. [30] A. Barchielli, D. Balzi, P. Naldoni, A.T. Roberts, F. Profili, F. Dima, L. Palmieri, Hospital discharge data for assessing myocardial infarction events and trends, and effects of diagnosis validation according to MONICA and AHA criteria, J. Epidemiol. Community Health 66 (2012) 462–467. [31] J. Herlitz, M. Dellborg, T. Karlsson, M.H. Evander, A. Berger, R. Luepker, Epidemiology of acute myocardial infarction with the emphasis on patients who did not reach the coronary care unit and non-AMI admissions, Int. J. Cardiol. 128 (2008) 342–349. [32] A. Adabag, T.M. Therneau, B.J. Gersh, S.A. Weston, V.L. Roger, Sudden death after myocardial infarction, J. Am. Med. Assoc. 300 (2008) 2022–2029.
