Portal de Periódicos da CAPES

Gender Differences in the Management and Prognosis of Myocardial Infarction Among Patients ≥ 65 Years of Age

2008; Elsevier BV; Volume: 101; Issue: 11 Linguagem: Inglês

10.1016/j.amjcard.2008.02.033

1879-1913

Soko Setoguchi, Daniel H. Solomon, Raisa Levin, Wolfgang C. Winkelmayer­,

Health Systems, Economic Evaluations, Quality of Life

Conflicting evidence exists regarding gender differences in the management and outcomes of myocardial infarctions (MIs). In this study, it was hypothesized that the management and outcomes of MIs would not differ by gender after proper adjustment for age, access to care, MI characteristics, and co-morbidities. Data from a published MI validation study, which sampled 2,200 MI hospitalizations in Medicare beneficiaries with the prescription drug benefit for detailed hospital chart review, were used. Gender differences in the use of MI-related procedures and recommended cardiovascular medications as well as short- and long-term mortality were assessed using multivariate regression. A total of 1,625 patients were identified (80% women) with confirmed MIs for whom complete clinical information was available. Compared with men, women were older and had higher body mass index. Women were more likely to have diabetes, renal dysfunction, and depression, but less likely to have had previous MIs, chronic lung disease, cancer, and to use tobacco. The characteristics of the index MIs were similar, except for non-Q-wave MIs being more common in men. The management of the MIs during admission was similar. During follow-up of up to 6.6 years, men were 40% more likely to die than women (95% confidence interval 21% to 62%), but no mortality difference was observed in patients aged 65 to 74 years (hazard ratio 0.92, 95% confidence interval 0.62 to 1.36), whereas in those aged ≥75 years, men were more likely to die than women (hazard ratio 1.54, 95% confidence interval 1.30 to 1.82). In conclusion, for older patients, the management of MIs did not significantly differ between men and women. Men, especially those aged ≥75 years, however, had worse short- and long-term prognoses compared with women of a similar age. The mortality was highest and the gender effect was more pronounced during the MI hospitalizations. Conflicting evidence exists regarding gender differences in the management and outcomes of myocardial infarctions (MIs). In this study, it was hypothesized that the management and outcomes of MIs would not differ by gender after proper adjustment for age, access to care, MI characteristics, and co-morbidities. Data from a published MI validation study, which sampled 2,200 MI hospitalizations in Medicare beneficiaries with the prescription drug benefit for detailed hospital chart review, were used. Gender differences in the use of MI-related procedures and recommended cardiovascular medications as well as short- and long-term mortality were assessed using multivariate regression. A total of 1,625 patients were identified (80% women) with confirmed MIs for whom complete clinical information was available. Compared with men, women were older and had higher body mass index. Women were more likely to have diabetes, renal dysfunction, and depression, but less likely to have had previous MIs, chronic lung disease, cancer, and to use tobacco. The characteristics of the index MIs were similar, except for non-Q-wave MIs being more common in men. The management of the MIs during admission was similar. During follow-up of up to 6.6 years, men were 40% more likely to die than women (95% confidence interval 21% to 62%), but no mortality difference was observed in patients aged 65 to 74 years (hazard ratio 0.92, 95% confidence interval 0.62 to 1.36), whereas in those aged ≥75 years, men were more likely to die than women (hazard ratio 1.54, 95% confidence interval 1.30 to 1.82). In conclusion, for older patients, the management of MIs did not significantly differ between men and women. Men, especially those aged ≥75 years, however, had worse short- and long-term prognoses compared with women of a similar age. The mortality was highest and the gender effect was more pronounced during the MI hospitalizations. We sought to examine gender differences in the management of and outcomes after myocardial infarctions (MIs) in a socioeconomically homogenous elderly population for whom precise MI-related clinical information as well as health service use were available. We hypothesized that the management and outcomes of MIs would not differ by gender in these patients after proper adjustment for age, access to care, MI characteristics, and co-morbidities. Medicare beneficiaries in Pennsylvania who were also enrolled in that state's Pharmaceutical Assistance Contract for the Elderly (a state pharmacy benefits programs for low- to middle-income elderly) in 1999 and 2000 formed the source population for a recent validation study assessing accuracy of medical claims codes for MI.1Kiyota Y. Schneeweiss S. Glynn R.J. Cannuscio C.C. Avorn J. Solomon D.H. Accuracy of Medicare claims-based diagnosis of acute myocardial infarction: estimating positive predictive value on the basis of review of hospital records.Am Heart J. 2004; 148: 99-104Abstract Full Text Full Text PDF PubMed Scopus (462) Google Scholar The study sample comprised all hospitalization episodes during 1999 or 2000 with the following criteria: International Classification of Diseases, Ninth Revision, code 410 (“acute myocardial infarction”) as a discharge diagnosis in the primary or secondary position or a diagnosis-related group code of 121, 122, or 123. Subjects with multiple events per year were included only once for their first acute MI hospitalization episode during each year. This produced a sample of 2,200 hospitalizations for MIs that were targeted for hospital record review. From the claims information, we created an initial data set that contained the subjects' dates of birth, gender, and dates of admission and discharge for the index hospitalizations and the American Hospital Association provider numbers for the treating hospitals. This information was supplied to a peer-review organization in Pennsylvania, which then sent requests for hospital records to the respective hospitals. The protocol was approved by the Centers for Medicare and Medicaid Services and the Institutional Review Board of Brigham and Women's Hospital. Hospital chart review was performed by 10 trained hospital records abstractors using a structured chart abstraction program. The chart abstraction data were entered directly into an electronic database. Data elements included gender, date of birth, detailed information relevant to the diagnosis of MI, and co-morbid conditions. The interrater agreement (κ) of the structured chart abstraction, defined as the agreement between 2 independent reviewers for all data elements in a 20-chart sample, was 0.93. The population for this study consisted of those patients with MIs confirmed by medical chart review. Patients with missing data on important predictors of mortality after MI, such as estimated glomerular filtration rate, hemoglobin concentration, and/or peak creatine kinase concentration, were excluded from the study. We assessed the administration of coronary interventions such as percutaneous coronary intervention, revascularization surgery, or the infusion of thrombolytic agents during the index MI hospitalizations. We also assessed whether the patients had ≥1 filled prescription for any statin, β blockers, angiotensin-converting enzyme (ACE) inhibitor or angiotensin receptor blocker (ARB), or nonaspirin antiplatelet agent (clopidogrel or ticlopidine) within 30 days as well as 90 days after discharge from their index MI hospitalizations. We assessed short- and long-term prognoses after the index MI. The short-term prognosis included in-hospital mortality and 30-day mortality after discharge from the index MI hospitalization, and long-term prognosis was assessed from 31 to 365 days and 1 to 4 years after discharge from the index MI hospitalization. Survival for the entire follow-up without time stratification was also estimated. All patients were censored at the end of the follow-up period (August 2005). We obtained information on each patient's age on index MI admission, gender, race (white vs nonwhite), and income from the enrollment files. We used medical chart–recorded diagnoses and extracted laboratory information to assess the presence of the important cardiovascular and noncardiovascular co-morbidities, including previous MI, previous coronary artery bypass graft, previous percutaneous coronary intervention, heart failure, cerebrovascular disease, diabetes, hypertension, atrial fibrillation, chronic pulmonary disease, dementia, renal dysfunction (defined as estimated glomerular filtration rate2Levey A.S. Bosch J.P. Lewis J.B. Greene T. Rogers N. Roth D. Modification of Diet in Renal Disease Study GroupA more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation.Ann Intern Med. 1999; 130: 461-470Crossref PubMed Scopus (13366) Google Scholar 30 kg/m2), and hemoglobin concentration. We identified the characteristics of the index MI, including length of stay, presence of ST elevation, location of MI (anteroseptal, lateral, inferior, posterior, or other), transmural versus nontransmural, and peak creatine kinase. We also extracted information from the medical records on whether patients had atrial fibrillation, ventricular tachycardia or fibrillation, second- or third-degree heart block, cerebrovascular attacks, shock, or hypotension during their index MI hospitalizations. To capture the potential confounders not recorded during the chart abstraction, we used inpatient and outpatient claims to assess the presence of other co-morbid conditions during the 12-month period preceding the discharge date. These co-morbid conditions included peripheral arterial disease,3Birman-Deych E. Waterman A.D. Yan Y. Nilasena D.S. Radford M.J. Gage B.F. Accuracy of ICD-9-CM codes for identifying cardiovascular and stroke risk factors.Med Care. 2005; 43: 480-485Crossref PubMed Scopus (585) Google Scholar arthritis,4Singh J.A. Holmgren A.R. Noorbaloochi S. Accuracy of Veterans Administration databases for a diagnosis of rheumatoid arthritis.Arthritis Rheum. 2004; 51: 952-957Crossref PubMed Scopus (129) Google Scholar dialysis, cancer, gastrointestinal bleeding, alcohol abuse, depression, and other mental diseases. Using prescription information during the 365 days before the index MI admission, we assessed the previous use of statins, β blockers, ACE inhibitors or ARBs, nonaspirin antiplatelet agents (ticlopidine or clopidogrel), nitrates, calcium channel blockers, diuretics, and warfarin. Crude short- and long-term incidence rates of death were calculated for men and women. Unadjusted frequencies of the administration of coronary interventions and the use of cardiovascular drugs after discharge from MI hospitalization were compared between men and women. Using multivariate logistic regression models, we assessed whether gender was an independent predictor of the administration of cardiac interventions and the use of drugs after discharge. The gender effect on short- and long-term prognoses was estimated using a multivariate Cox proportional-hazards model that included age, gender, race, MI characteristics, co-morbidities, and the management of MI. We sequentially introduced sets of predictors of mortality to understand how these factors confounded the gender effect. We accounted for the clustering of patients within hospitals by using the robust sandwich estimate of Wei et al.5Wei L.J. Lin D.Y. Weissfeld L. Regression analysis of multivariate incomplete failure time data by modeling marginal distributions.J Am Stat Assoc. 1989; 84: 1065-1073Crossref Scopus (1497) Google Scholar Hazard ratios (HRs) are presented together with the corresponding 95% confidence intervals (CIs). SAS for Windows version 9.2 was used for all statistical analyses (SAS Institute Inc., Cary, North Carolina). Detailed chart abstraction was conducted for 2,022 individual MI hospitalizations during 1999 and 2000, of which 93% (1,884) involved medical chart–confirmed MIs. Of those, 259 patients were excluded from the study because data on estimated glomerular filtration rate, hemoglobin concentration, and/or peak creatine kinase concentration were missing. The final study population included 1,625 patients with MIs. The median follow-up period of the study population was 2.2 years (interquartile range 4.7). The characteristics of these patients are listed in Table 1. Compared to men, women were older and more likely to have hypertension, diabetes, depression, obesity, and lower hemoglobin concentrations. Women were less likely to have previous MI, arthritis, cancer, and smoking, and fewer had undergone any previous cardiac procedures. Women dispensed more medications and, among cardiovascular medications, were more likely to be receiving calcium channel blockers before their MI admissions. The characteristics of the MIs did not significantly differ by gender, except for non-Q-wave MIs being more common in men. The characteristics of the patients who were excluded from the study cohort were not significantly different from those of the study population.Table 1Characteristics of the 1,625 study patients hospitalized for myocardial infarction during 2000 and 2001VariableTotal Population (n = 1,625)Women (n = 1,308)Men (n = 317)p Value for Gender DifferenceAge (yrs)81 ± 782 ± 780 ± 7 $10,00061%57%78%<0.001Previous MI⁎Based on medical record.40%37%52%<0.001Previous coronary bypass surgery⁎Based on medical record.14%13%18%0.03Previous percutaneous coronary intervention⁎Based on medical record.9%9%13%0.02Atrial fibrillation⁎Based on medical record.20%19%21%0.45Peripheral vascular diseases27%27%29%0.58Hypertension⁎Based on medical record.78%80%71%0.001Cerebrovascular disease⁎Based on medical record.21%20%24%0.14Heart failure⁎Based on medical record.42%43%30%0.40Diabetes mellitus⁎Based on medical record.38%39%33%0.03Chronic pulmonary diseases⁎Based on medical record.27%23%43%<0.001Renal dysfunction (estimated glomerular filtration rate <60 ml/kg/m2)68%70%61%0.003Dialysis1%1%1%0.87Arthritis47%44%60%<0.001Cancer17%15%24%<0.001Gastrointestinal bleeding15%15%15%0.91Dementia⁎Based on medical record.10%10%10%0.93Depression18%19%11% 30 kg/m2⁎Based on medical record.24%25%18%0.01Hemoglobin⁎Based on medical record.12 ± 212 ± 213 ± 2<0.001No. of physician visits10 ± 79 ± 79 ± 80.50No. of medications10 ± 612 ± 710 ± 7<0.001No. of days in hospital6 ± 126 ± 136 ± 120.82Use of other drugs (before index hospitalization) Statins22%22%22%0.98 β blockers27%28%27%0.78 ACE inhibitors/ARBs41%41%41%0.95 Calcium channel blockers43%45%37%0.01 Nitrates46%45%50%0.13 Diuretics10%11%8%0.08 Warfarin12%12%13%0.47 Nonaspirin antiplatelet agents (ticlopidine, clopidogrel)10%10%12%0.27Characteristics of index MI Length of stay (days)8 ± 58 ± 58 ± 60.51 ST elevation⁎Based on medical record.28%28%29%0.61 Anteroseptal⁎Based on medical record.10%10%9%0.61 Lateral⁎Based on medical record.35%35%31%0.13 Inferior⁎Based on medical record.3%3%2%0.35 Posterior⁎Based on medical record.8%8%10%0.21 Other location⁎Based on medical record.37%37%39%0.72 Nontransmural⁎Based on medical record.30%29%35%0.03 Atrial fibrillation during hospitalization⁎Based on medical record.23%23%26%0.19 Ventricular tachycardia/ventricular fibrillation during hospitalization⁎Based on medical record.<1%<1%<1%0.69 Cerebrovascular attack during hospitalization⁎Based on medical record.4%4%3%0.57 Mean peak creatine kinase (U/L)⁎Based on medical record.597 ± 893579 ± 848612 ± 1,0550.57Data are expressed as mean ± SD or as percentages. Nonmedical record based covariates were measured using diagnosis and procedure codes in claims data during the 12-month period before and during the index MI hospitalization. Based on medical record. Open table in a new tab Data are expressed as mean ± SD or as percentages. Nonmedical record based covariates were measured using diagnosis and procedure codes in claims data during the 12-month period before and during the index MI hospitalization. The management of MIs during and after hospitalizations in the study population and by gender is listed in Table 2. The overall use of specific cardiac procedures, including diagnostic catheterization and recommended cardiovascular drugs after discharge, appeared low. There were no significant differences in the management of MIs between men and women, except that women were more likely to receive β blockers after discharge. However, after adjusting for multiple factors listed in Table 1 in logistic regression models, gender was not independently associated with the use of any cardiac procedures or the use of any recommended cardiovascular drugs.Table 2Management of myocardial infarction during and after hospitalizationVariableTotal Population (n = 1,884)Women (n = 1,308)Men (n = 317)p Value for Crude Gender DifferenceProcedure during the index hospitalization Angiography without intervention17%18%15%0.17 Percutaneous coronary intervention11%12%10%0.40 Stent during index admission9%9%8%0.35 Cardiac surgery for MI3%3%3%0.73 Thrombolytic therapy5%4%5%0.48Cardiovascular medication use after MI (<30 days after discharge) Statins16%17%16%0.86 β blockers29%30%25%0.09 ACE inhibitors/ARBs28%29%27%0.38 Nonaspirin antiplatelet agents (ticlopidine, clopidogrel)12%12%12%0.93Cardiovascular medication use after MI (<90 days after discharge) Statins23%23%22%0.61 β blockers49%51%43%0.01 ACE inhibitors/ARBs40%41%37%0.19 Nonaspirin antiplatelet agents (ticlopidine, clopidogrel)16%16%14%0.25 Open table in a new tab Table 3 lists the number of deaths, patient-years of follow-up, and the incidence rate of death in various follow-up periods. The risk for death was highest during the index hospitalization for MI (622 per 100 patient-years) and decreased as the MI became more distant. Throughout, men had a higher risk for death during these periods.Table 3Number of deaths, patient-years of follow-up, and incidence rates after myocardial infarction in men and women during various follow-up durationsWomen (n = 1,308)Men (n = 317)DeathsTotal Patient-YearsIncidence Rate⁎Per 100 patient-years.DeathsTotal Patient-YearsIncidence Rate⁎Per 100 patient-years.Crude Rate Ratio (Men vs Women)In-hospital death142245814667991.38Death <30 days after discharge in those survived through the MI hospitalization1121159731261171.21Death 31 to 365 days after discharge in those survived for 30 days after discharge2389362568206331.30Death 1 to 4 yrs after discharge in those who survived for 1 yr after discharge3272,7591277562141.16Death during entire follow-up9633,51927252721351.28 Per 100 patient-years. Open table in a new tab Table 4 lists the HRs of short- and long-term mortality for men compared with women from multivariate Cox proportional-hazards models. After adjusting for co-morbidities and other predictors of post-MI mortality, men had 74% greater mortality rates during their MI hospitalizations compared with women. The mortality rate within 30 days after discharge in survivors of their index hospitalizations was 34% higher in men compared with women, although not statistically significant. The mortality rate was again higher in men than in women during the period of 31 days to 1 year, as well as from 1 to 4 years after discharge from MI. Adjusting for the use of cardiac procedures and recommended cardiovascular drugs did not change these estimate significantly. Overall, during the entire follow-up period (maximum 6.6 years), men were 40% more likely to die than women. The proportionality assumption was not violated (p = 0.52 for the gender-time interaction). When we examined the gender differences by age group, the overall mortality was similar in men and women in patients aged 65 to 74 years (HR 0.92, 95% confidence interval 0.62 to 1.36), but in patients aged ≥75 years, men were more likely to die (HR 1.58, 95% confidence interval 1.26 to 1.96 for those aged 75 to 84 years; HR 1.62, 95% CI 1.27 to 2.07 for those aged ≥85 years). However, a formal test for interaction between gender and age ≥75 years was not statistically significant (p = 0.57).Table 4Unadjusted and adjusted effect of gender on short- and long-term mortality after myocardial infarction (men versus women)Variables in the ModelIn-Hospital Death30-Day Mortality After Discharge31- to 365-Day Mortality After Discharge1- to 4-Year Mortality After DischargeEntire Follow-Up PeriodHR95% CIHR95% CIHR95% CIHR95% CIHR95% CIUnadjusted1.340.95–1.901.210.82–1.801.351.05–1.721.180.94–1.471.221.08–1.39Age, gender, income adjusted1.410.99–2.001.290.84–1.971.441.13–1.861.341.05–1.731.341.17–1.53Above plus MI characteristics1.491.03–2.141.310.85–2.011.471.15–1.891.371.05–1.781.351.18–1.55Above plus co-morbidities1.771.18–2.671.350.84–2.151.521.16–1.991.311.00–1.721.361.18–1.58Above plus previous cardiovascular drug use1.751.15–2.641.350.85–2.151.511.15–1.991.310.99–1.721.361.17–1.58Above plus health service use1.741.15–2.621.370.86–2.191.501.14–1.981.391.05–1.841.401.21–1.62Above plus procedures for MI1.340.83–2.151.501.14–1.981.391.04–1.84Above plus use of cardiovascular drugs <30 days after discharge1.491.11–1.981.391.04–1.85 Open table in a new tab Because assessing medication use within 30-day windows after discharge might be too stringent and cause an underestimation of the use of these drugs, all the analyses were repeated using 90-day drug ascertainment periods instead of the 30-day periods used in the primary analyses. The results of these analyses were similar to those of the main analyses (not shown). In this study with clinical, sociodemographic, and health service use information in a representative sample of low-income American seniors, we assessed gender differences in the early management of MI as well as in short- and long-term prognoses after MI. The use of procedures during the index MI hospitalization and the use of recommended cardiovascular drugs (statins, β blockers, and ACE inhibitors or ARBs) after discharge did not differ by gender. However, we found evidence that men had 34% to 74% higher mortality after MIs than women over the study period. The mortality risk was greatest and the gender effect was more pronounced during the MI hospitalization. The gender difference remained significant after accounting for possible differences in sociodemographic factors, co-morbidities, the characteristics and severity of the MIs, and the management of the MIs. We also found that the difference appeared to be driven by the group aged ≥75 years. Although a few earlier studies found higher mortality in men after MIs,6Pohjola S. Siltanen P. Romo M. Five-year survival of 728 patients after myocardial infarction A community study.Br Heart J. 1980; 43: 176-183Crossref PubMed Scopus (102) Google Scholar, 7Martin C.A. Thompson P.L. Armstrong B.K. Hobbs M.S. de Klerk N. Long-term prognosis after recovery from myocardial infarction: a nine year follow-up of the Perth Coronary Register.Circulation. 1983; 68: 961-969Crossref PubMed Scopus (114) Google Scholar more recent studies have documented worse outcomes in women after MIs or acute coronary syndromes.8Vaccarino V. Parsons L. Every N.R. Barron H.V. Krumholz H.M. Sex-based differences in early mortality after myocardial infarction National Registry of Myocardial Infarction 2 Participants.N Engl J Med. 1999; 341: 217-225Crossref PubMed Scopus (1085) Google Scholar, 9Vaccarino V. Krumholz H.M. Yarzebski J. Gore J.M. Goldberg R.J. Sex differences in 2-year mortality after hospital discharge for myocardial infarction.Ann Intern Med. 2001; 134: 173-181Crossref PubMed Scopus (298) Google Scholar, 10Vaccarino V. Krumholz H.M. Berkman L.F. Horwitz R.I. Sex differences in mortality after myocardial infarction Is there evidence for an increased risk for women?.Circulation. 1995; 91: 1861-1871Crossref PubMed Scopus (289) Google Scholar, 11Vaccarino V. Horwitz R.I. Meehan T.P. Petrillo M.K. Radford M.J. Krumholz H.M. Sex differences in mortality after myocardial infarction: evidence for a sex-age interaction.Arch Intern Med. 1998; 158: 2054-2062Crossref PubMed Scopus (170) Google Scholar, 12Reina A. Colmenero M. Aguayo de Hoyos E. Aros F. Marti H. Claramonte R. Cunat J. Gender differences in management and outcome of patients with acute myocardial infarction.Int J Cardiol. 2007; 116: 389-395Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar, 13Moriel M. Behar S. Tzivoni D. Hod H. Boyko V. Gottlieb S. Management and outcomes of elderly women and men with acute coronary syndromes in 2000 and 2002.Arch Intern Med. 2005; 165: 1521-1526Crossref PubMed Scopus (20) Google Scholar, 14Milcent C. Dormont B. Durand-Zaleski I. Steg P.G. Gender differences in hospital mortality and use of percutaneous coronary intervention in acute myocardial infarction: microsimulation analysis of the 1999 nationwide French hospitals database.Circulation. 2007; 115: 833-839Crossref PubMed Scopus (226) Google Scholar, 15Kostis J.B. Wilson A.C. O'Dowd K. Gregory P. Chelton S. Cosgrove N.M. Chirala A. Cui T. MIDAS Study GroupSex differences in the management and long-term outcome of acute myocardial infarction A statewide study.Circulation. 1994; 90: 1715-1730Crossref PubMed Scopus (233) Google Scholar In some of these studies, the findings were confined to younger women (<70 years,15Kostis J.B. Wilson A.C. O'Dowd K. Gregory P. Chelton S. Cosgrove N.M. Chirala A. Cui T. MIDAS Study GroupSex differences in the management and long-term outcome of acute myocardial infarction A statewide study.Circulation. 1994; 90: 1715-1730Crossref PubMed Scopus (233) Google Scholar <75 years,8Vaccarino V. Parsons L. Every N.R. Barron H.V. Krumholz H.M. Sex-based differences in early mortality after myocardial infarction National Registry of Myocardial Infarction 2 Participants.N Engl J Med. 1999; 341: 217-225Crossref PubMed Scopus (1085) Google Scholar, 11Vaccarino V. Horwitz R.I. Meehan T.P. Petrillo M.K. Radford M.J. Krumholz H.M. Sex differences in mortality after myocardial infarction: evidence for a sex-age interaction.Arch Intern Med. 1998; 158: 2054-2062Crossref PubMed Scopus (170) Google Scholar <60 years,9Vaccarino V. Krumholz H.M. Yarzebski J. Gore J.M. Goldberg R.J. Sex differences in 2-year mortality after hospital discharge for myocardial infarction.Ann Intern Med. 2001; 134: 173-181Crossref PubMed Scopus (298) Google Scholar and ≤85 years14Milcent C. Dormont B. Durand-Zaleski I. Steg P.G. Gender differences in hospital mortality and use of percutaneous coronary intervention in acute myocardial infarction: microsimulation analysis of the 1999 nationwide French hospitals database.Circulation. 2007; 115: 833-839Crossref PubMed Scopus (226) Google Scholar). Other studies, however, have reported no significant gender differences in outcomes after adjusting for age and other co-morbidities.16Blomkalns A.L. Chen A.Y. Hochman J.S. Peterson E.D. Trynosky K. Diercks D.B. Brogan Jr, G.X. Boden W.E. Roe M.T. Ohman E.M. et al.Gender disparities in the diagnosis and treatment of non-ST-segment elevation acute coronary syndromes: large-scale observations from the CRUSADE (Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes With Early Implementation of the American College of Cardiology/American Heart Association Guidelines) National Quality Improvement Initiative.J Am Coll Cardiol. 2005; 45: 832-837Abstract Full Text Full Text PDF PubMed Scopus (560) Google Scholar, 17Anand S.S. Xie C.C. Mehta S. Franzosi M.G. Joyner C. Chrolavicius S. Fox K.A. Yusuf S. Differences in the management and prognosis of women and men who suffer from acute coronary syndromes.J Am Coll Cardiol. 2005; 46: 1845-1851Abstract Full Text Full Text PDF PubMed Scopus (238) Google Scholar, 18Gan S.C. Beaver S.K. Houck P.M. MacLehose R.F. Lawson H.W. Chan L. Treatment of acute myocardial infarction and 30-day mortality among women and men.N Engl J Med. 2000; 343: 8-15Crossref PubMed Scopus (305) Google Scholar, 19Carrabba N. Santoro G.M. Balzi D. Barchielli A. Marchionni N. Fabiani P. Landini C. Scarti L. Santoro G. Valente S. et al.In-hospital management and outcome in women with acute myocardial infarction (data from the AMI-Florence Registry).Am J Cardiol. 2004; 94: 1118-1123Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar These studies were conducted in different populations and countries and sometimes lacked information on socioeconomic status (race or ethnicity, insurance coverage, income, and education), which may have contributed to the conflicting results. However, our results of worse prognosis in men apply only to those aged ≥65 years and do not rule out the possibility that women have worse prognoses in patients aged <65 years. Our study extends knowledge from previous studies and sheds light on gender differences in the management of and outcomes after MIs in the oldest old. Previous studies by Vaccarino et al8Vaccarino V. Parsons L. Every N.R. Barron H.V. Krumholz H.M. Sex-based differences in early mortality after myocardial infarction National Registry of Myocardial Infarction 2 Participants.N Engl J Med. 1999; 341: 217-225Crossref PubMed Scopus (1085) Google Scholar, 9Vaccarino V. Krumholz H.M. Yarzebski J. Gore J.M. Goldberg R.J. Sex differences in 2-year mortality after hospital discharge for myocardial infarction.Ann Intern Med. 2001; 134: 173-181Crossref PubMed Scopus (298) Google Scholar, 11Vaccarino V. Horwitz R.I. Meehan T.P. Petrillo M.K. Radford M.J. Krumholz H.M. Sex differences in mortality after myocardial infarction: evidence for a sex-age interaction.Arch Intern Med. 1998; 158: 2054-2062Crossref PubMed Scopus (170) Google Scholar also showed that in-hospital and 2-year mortality tended to be higher in men than women in older age groups (≥70 years,9Vaccarino V. Krumholz H.M. Yarzebski J. Gore J.M. Goldberg R.J. Sex differences in 2-year mortality after hospital discharge for myocardial infarction.Ann Intern Med. 2001; 134: 173-181Crossref PubMed Scopus (298) Google Scholar ≥75 years,11Vaccarino V. Horwitz R.I. Meehan T.P. Petrillo M.K. Radford M.J. Krumholz H.M. Sex differences in mortality after myocardial infarction: evidence for a sex-age interaction.Arch Intern Med. 1998; 158: 2054-2062Crossref PubMed Scopus (170) Google Scholar and ≥80 years8Vaccarino V. Parsons L. Every N.R. Barron H.V. Krumholz H.M. Sex-based differences in early mortality after myocardial infarction National Registry of Myocardial Infarction 2 Participants.N Engl J Med. 1999; 341: 217-225Crossref PubMed Scopus (1085) Google Scholar), suggesting the possibility that men might have higher mortality beyond 70 years of age compared with women. In our analyses, we adjusted for clinical characteristics of the index MIs, cardiovascular and noncardiovascular co-morbidities, sociodemographic factors, and past health service use. Also, our study population was quite homogenous in socioeconomic status by the nature of the drug benefit program, which had lower and upper income thresholds. Access to care was also similar, because all patients had uniform coverage for medical and drug expenses. The following limitations should be noted. First, because we did not have information on the in-hospital use of cardiovascular medications (aspirin, β blockers, ACE inhibitors or ARBs, and glycoprotein IIb/IIIa inhibitors), we could not estimate the gender differences in the use of these drugs and could not account for their influence (if any) on the differences in prognosis. Second, we also could not control for unmeasured socioeconomic factors, such as social support, marital status, and education. However, the studied population was rather homogenous regarding these domains, but we do not know whether our results also pertain to more affluent seniors. Another limitation was that our data did not have complete information on aspirin use, because aspirin is available over the counter, which leads to the underassessment of its use from claims data. However, studies have showed that the rate of aspirin use is lower in women with acute coronary syndromes16Blomkalns A.L. Chen A.Y. Hochman J.S. Peterson E.D. Trynosky K. Diercks D.B. Brogan Jr, G.X. Boden W.E. Roe M.T. Ohman E.M. et al.Gender disparities in the diagnosis and treatment of non-ST-segment elevation acute coronary syndromes: large-scale observations from the CRUSADE (Can Rapid Risk Stratification of Unstable Angina Patients Suppress Adverse Outcomes With Early Implementation of the American College of Cardiology/American Heart Association Guidelines) National Quality Improvement Initiative.J Am Coll Cardiol. 2005; 45: 832-837Abstract Full Text Full Text PDF PubMed Scopus (560) Google Scholar or coronary artery disease,20Opotowsky A.R. McWilliams J.M. Cannon C.P. Gender differences in aspirin use among adults with coronary heart disease in the United States.J Gen Intern Med. 2007; 22: 55-61Crossref PubMed Scopus (28) Google Scholar which would underestimate the true effect of male gender on mortality.