The Portuguese interventional cardiology database is a prospective, observational registry of PCI in interventional cardiology units in Portugal. Between January 4, 2002 and July 24, 2012, 60158 patients were included from 22 centers. We retrospectively analyzed 7544 consecutive patients undergoing PPCI in the context of STEMI of less than 12 hours evolution. STEMI was defined as the presence of symptoms consistent with myocardial ischemia lasting more than 30 minutes and persistent ST elevation (>1 mm in two contiguous leads) or new-onset or previously undocumented complete left bundle branch block. Patients who underwent thrombolysis during the index hospitalization were excluded.

No exclusion criteria were applied in terms of the complexity of STEMI presentation or type of lesion.

All patients gave their informed consent for inclusion in the Registry.

Men and women were compared in terms of demographic characteristics, risk factors, previous cardiovascular history, disease severity, location of myocardial infarction (MI) and procedural characteristics.

Data on in-hospital clinical course were obtained from the Registry.

The male and female patient groups were analyzed according to continuous and categorical variables. Categorical variables were characterized in terms of absolute and relative frequencies. The central tendency and distribution of continuous variables were estimated by sample mean and standard deviation, or median and interquartile range, depending on whether values presented a normal distribution.

Categorical variables were compared using the chi-square test or Fisher's exact test as appropriate, while continuous variables were compared using the Student's t test, or the Mann–Whitney test when criteria for use of the former were not met. The Kolmogorov–Smirnov method was used to test the normality of distribution of continuous variables and Levene's test was used for equal variances.

The influence of each variable (including gender) on mortality was assessed in univariate analysis by means of odds ratios (OR) and 95% confidence intervals (CI). In the first regression model, only gender was considered an independent variable. The impact of gender on in-hospital mortality was subsequently adjusted for other variables, producing various logistic regression models. In the second model, the impact of gender was adjusted for age >65 years; in the third, it was adjusted for age, risk factors (diabetes, hypertension, dyslipidemia and smoking) and cardiovascular history (previous MI, percutaneous or surgical revascularization, congestive heart failure, stroke, peripheral arterial disease or chronic renal failure); in the fourth, clinical variables of severity at admission (cardiogenic shock and anterior MI) were added to all the variables included in the third model. Finally, the fifth model was adjusted for all the above variables, as well as procedural characteristics (type of access, type of stent and thrombectomy).

In order to minimize selection bias in this non-randomized study, a further logistic regression analysis was performed with gender as a dependent variable to determine a propensity score, which was then used to match men and women in a ratio of 1:1. The variables used to calculate the propensity score were age, risk factors and cardiovascular history. After matching the two sexes, the estimated risk of in-hospital death in women compared to men was again adjusted by means of a logistic regression model that included the same variables as in the fourth model described above.

When constructing all the above models to determine predictors of in-hospital death, gender was forced to enter the model by being included in the first block of variables, while the other predictors were selected from the second block of variables using the forward stepwise method. The Hosmer–Lemeshow test was used to assess the goodness of fit of the logistic regression models, while their discriminatory power was assessed by the area under the receiver operating characteristic (ROC) curve.

The statistical analysis was performed using SPSS, version 19.0.0.2. The level of significance was set at 5%.

A total of 7544 consecutive patients from the Registry who underwent PPCI in the context of STEMI were included. The study population was divided into two groups according to gender: 1856 (25%) women and 5688 (75%) men. Demographic characteristics and clinical presentation at admission are shown in Table 1.

Women were older (68±14 vs. 61±13 years, p<0.001) and had a higher prevalence of diabetes (30% vs. 21%, p<0.001), hypertension (69% vs. 55%, p<0.001) and chronic renal failure (4% vs. 2%, p=0.002). On the other hand, they were less likely to smoke (22% vs. 57%, p<0.001) and less often had a history of coronary disease, including MI (11% vs. 15%, p<0.001), PCI (11% vs. 16%, p<0.001) or coronary artery bypass grafting (CABG) (1% vs. 2%, p=0.008).

The proportion of patients undergoing PPCI who went directly to a center with interventional cardiology facilities was greater among women (91% vs. 89%, p=0.014), but this did not influence door-to-balloon times, which were similar in both sexes (Table 2). Nevertheless, women were revascularized later than men (4.6 hours vs. 3.9 hours, p<0.001) as they arrived at hospital later (4.3 hours vs. 3.7 hours, p<0.001), and the percentage of men revascularized within six hours of symptom onset was higher than in women (71% vs. 63%, respectively, p<0.001).

Cardiogenic shock was more common in women (7.1% vs. 5.7%, p=0.032). There were no significant differences in the infarct-related vessel or in the extent of coronary disease, the percentage of multivessel disease being similar in both sexes.

Procedural characteristics are shown in Table 2. Femoral access was used most frequently in both groups, and radial access was used less often in women (14% vs. 17%, p=0.001). There was no difference in the type of stent employed, but thrombectomy (23% vs. 26%, p=0.003) and glycoprotein IIb/IIIa inhibitors (27% vs. 35%, p<0.001) were more often used in men. There was a tendency for final TIMI 3 flow to be more frequently achieved in men, reflecting a better angiographic result.

In-hospital mortality was higher in women (4.3% vs. 2.5%, p<0.001), the risk of death being 1.7 times higher in women than men (95% CI 1.30–2.27; p<0.001) (Table 3).

Adjustment for age >65 years revealed that this factor largely explained the increased risk in women compared to men (Table 4). On multivariate analysis (Figure 1), the independent predictors of in-hospital death were age >65 years, peripheral arterial disease, history of congestive heart failure, cardiogenic shock, and anterior MI. Dyslipidemia, use of drug-eluting stents and radial access were protective variables. The area under the ROC curve in the final model was 0.88, which reflects the excellent discriminatory power of the model.

Figure 1.

Predictors of in-hospital mortality in primary percutaneous coronary intervention (multivariate analysis). BMS: bare-metal stent; CHF: congestive heart failure; CI: confidence interval; DES: drug-eluting stent; IH: in-hospital; OR: odds ratio; PAD: peripheral arterial disease.

(0.08MB).

Based on a logistic regression model with gender as a dependent variable, a propensity score was calculated, which was then used to match 1338 women and 1338 men (a ratio of 1:1). Following this adjustment, in-hospital mortality was similar for both sexes, around 3.9% (OR 1.00, 95% CI 0.68–1.48, p=1.00).

The percentage of women in the Registry undergoing PPCI has increased linearly over the years (11% in the period 2002–2005, 15% in 2006–2008, and 19% in 2009–2012; p<0.001 for trend), similar to the trend observed in men (Figure 2). Throughout the study period, in-hospital mortality among women was higher than in men, with a more marked difference in recent years (Table 5).

Figure 2.

Prevalence of primary percutaneous coronary intervention in men and women over the years covered by the Portuguese Registry of Interventional Cardiology.

(0.05MB).

Although in-hospital mortality rates decreased in absolute terms in both sexes between 2002 and 2012, the reduction was only significant in men (Figure 3).

Figure 3.

Changes in in-hospital mortality among men and women over the years covered by the Portuguese Registry of Interventional Cardiology.

(0.06MB).

As seen in many other studies on coronary revascularization, both surgical and percutaneous, clinical presentation of coronary disease differs between the sexes.3–6,8,9 Women tend to present acute coronary syndrome at older ages, on average a decade later, and more often have atypical symptoms rather than classic angina. They also have a worse cardiovascular profile, with a higher prevalence of diabetes and hypertension, and the presence of multiple cardiovascular risk factors is common among post-menopausal women.

Despite the higher risk profile of women undergoing PPCI, disease severity is similar to that of men, with a similar prevalence of multivessel disease.9

The incidence of congestive heart failure and cardiogenic shock at admission is significantly higher in women, despite the fact that they less often have previous MI and present similar left ventricular function to men, as observed in numerous other studies.3,4,9 The most frequently suggested explanation is the greater prevalence of diastolic dysfunction due to older age and hypertensive heart disease.10 Hypertrophied myocardium may be more vulnerable in situations of transient ischemia, the decreased diastolic compliance that occurs following sudden occlusion of a vessel being less well tolerated in patients with a preexisting “stiff” ventricle.10 On the other hand, the higher incidence of hemodynamic instability at admission may be explained by the fact that women arrive in the catheterization laboratory after a longer delay from symptom onset.3 The time to mechanical reperfusion is thus longer in women due to longer pain-to-door times.3,4,11 In the 1950s and 1960s, coronary disease was seen as a disease of men, but from the 1970s onward there was an increase in coronary disease among women, which was attributed to changes in their lifestyle, with increased prevalence of smoking, use of oral contraceptives and increasing numbers of women in employment.12 This, together with the fact that cardiovascular disease is still the leading cause of death in women,1 highlights the importance of implementing measures to raise awareness among women of the symptoms of MI and of the need to seek immediate medical attention. The longer delay between symptom onset and mechanical reperfusion is one of the modifiable predictors of a worse prognosis.

There are also gender-based differences in procedural characteristics. Coronary arteries in women are of smaller diameter and this may influence the type of material used, notably reducing the use of stents.1 However, this was not the case in our study, in which stenting and types of stent were similar in both sexes. The reduction in restenosis observed with widespread use of drug-eluting stents appears to be similar in men and women.13 Smaller caliber arteries, together with a greater tendency for arterial vasospasm, may explain the less frequent use of radial access in women. PCI success rates are similar in both sexes,14 as is the use of new antithrombotic drugs.15 Nevertheless, glycoprotein IIb/IIIa inhibitors are far less often administered in women as adjuvant therapy to antithrombins in primary angioplasty. This may be because women undergoing PCI present more bleeding and vascular complications than men.16,17 A subanalysis of the ACUITY trial showed that women have more bleeding complications at 30 days compared to men, but with no differences in short- or long-term mortality. Women treated with bivalirudin have fewer bleeding complications compared to those receiving heparin together with glycoprotein IIb/IIIa inhibitors, but with no difference in ischemic events.18

Evidence shows that there are still gender-based differences in treatment of coronary disease, with women presenting STEMI and age >65 years less often receiving reperfusion therapy.1 In Portugal, there has been a progressive increase in recent years in women undergoing PPCI, with reduced in-hospital mortality for STEMI between 2002 and 2012 (in both sexes). However, despite these advances in STEMI treatment, unadjusted in-hospital mortality remains higher in women, which highlights the need for strategies to improve risk stratification, diagnosis and treatment of cardiovascular disease in women.

This analysis of the PCI Registry revealed that in Portugal female gender is not an independent predictor of in-hospital mortality in MI treated by PPCI. In our non-adjusted model, women had worse prognosis than men, but as in previous studies,5–7,19,20 after adjustment for comorbidities and other adverse factors, female gender ceased to be an independent predictor of unfavorable outcome after PCI. This finding was repeated in multivariate logistic regression analysis, as well as when adjusted to the propensity score.

In the thrombolytic era, in-hospital survival among women was consistently lower than that of men,21,22 with more associated fatal bleeding complications.19 In a multicenter study of 395 patients (27% female) with STEMI that compared PPCI with thrombolysis, primary angioplasty was an independent predictor of survival in women.19 In a substudy of the randomized GUSTO II-B trial, both sexes showed benefits from direct angioplasty compared to thrombolysis, with fewer major adverse events at 30 days, but the absolute benefit was greater in women (direct angioplasty prevented 56 events/1000 women vs. 42 events/1000 men).20 The introduction of primary angioplasty as the treatment of choice for STEMI, progressive improvements in PCI techniques and success rates, and advances in adjuvant antiplatelet and anticoagulant therapy, may all explain the differences observed between earlier and later studies. Nevertheless, the results of recent studies assessing gender differences in primary angioplasty are in disagreement with the present study, reporting female gender as an independent predictor of worse short-term prognosis, irrespective of age and baseline cardiovascular risk factors.3,4 Earlier studies suggested that pathophysiological differences between the sexes may contribute to higher mortality in women with STEMI. Estrogens, which protect women against coronary atherosclerosis,23 could make vulnerable plaques less stenotic.24 Women would therefore have less collateral circulation and less ability to develop preconditioning in STEMI, worsening their prognosis.25,26 However, this would only explain the differences observed in younger, premenopausal women. On the other hand, the present analysis of the Registry was of a larger sample (>7000 treated by PPCI) than in previous studies3,4 and included data from multiple centers, and thus reflects the situation in the country as a whole rather than of a single center; the statistical methodology used was also superior, including a propensity score; and the data presented are more up-to-date, with lower in-hospital mortality in both men and women. In a recently published risk-adjusted analysis of 13752 patients (34.6% women), Kovacic et al. confirmed that female gender was not an independent predictor of mortality after PCI in the drug-eluting stent era,27 which supports the findings of our study.

The authors declare that the procedures followed were in accordance with the regulations of the relevant clinical research ethics committee and with those of the Code of Ethics of the World Medical Association (Declaration of Helsinki).

The authors declare that they have followed the protocols of their work center on the publication of patient data.

The authors have obtained the written informed consent of the patients or subjects mentioned in the article. The corresponding author is in possession of this document.