Key Findings and Strength of Evidence

The main findings of the treatment strategies for women with CAD are summarized in the following sections. Table 20 summarizes the strength of evidence (SOE) by effectiveness outcome for the KQs.

Table 20

Strength of evidence by effectiveness outcome.

Discussion

The findings from this systematic review on the treatment strategies for women across the spectrum of CAD presentations highlight areas for future research and for informing clinical practice. First, this review underscores the significant need for clinical researchers to provide study findings with women-specific data on the primary and secondary clinical outcomes. Overall, we were able to find only 28 relevant studies with data on either shorter term or longer term outcomes in women with CAD treated with invasive or conservative medical therapies. In addition, the representation of women enrolled in these trials was low. Melloni et al.²⁷ found similarly low rates with sex-specific results discussed in only 31 percent of the 156 primary trial publications cited by the American Heart Association’s 2007 women’s prevention guidelines. In addition, they found that enrollment of women in randomized clinical trials had increased over time (18% in 1970 to 34% in 2006) but remained low relative to their overall representation in disease populations (e.g., 25% women representation in RCTs of CAD compared with 46% women representation in the CAD population).

Second, our findings confirm current practice and evidence for care in one of the three areas evaluated. For women patients with STEMI, we found that an invasive approach with immediate PCI is superior to fibrinolysis in reducing cardiovascular events. These findings are similar to a meta-analysis¹⁰⁷ of 23 randomized trials comparing PCI with fibrinolysis for acute MI in combined populations of men and women. However, for patients with NSTEMI treated with an early invasive approach compared with a conservative or selective invasive approach, this review, while finding a trend favoring early invasive strategies, does not demonstrate a statistically significant benefit of an early invasive approach in reducing cardiovascular events in women. In contrast, the overall meta-analysis for trials of early invasive versus conservative strategies including both men and women showed a benefit of early invasive therapy.¹⁰⁸ The results from this review suggest that such a benefit may also be true in women, but the confidence intervals are too wide to support a firm conclusion.

In addition, for medical therapy alone versus revascularization plus medical therapy for patients with stable angina or high CAD burden, the findings from the current analysis suggest a benefit of revascularization in women. These findings should be viewed with caution because they are based on a limited number of studies with data on 704 (17%) women; these analyses often have both PCI and CABG together in the revascularization group, and the overall findings from these studies do not show a significant benefit beyond angina or symptom reduction for revascularization. In these studies, it is possible that women who present later in life with CAD, and with higher CAD burden, may be obtaining a greater benefit with revascularization, and the findings from this analysis should prompt further research in this area and again encourage researchers to provide data specific on women. In contrast, previous meta-analyses that combined results for men and women found similar outcomes for either treatment. The higher proportion of men enrolled in these trials (83%) may have led to the masking of the women’s results by the men’s results within a pooled analysis.

Our stakeholder group advised us to assess the effectiveness of these therapies by sex on multiple important clinical outcomes, such as nonfatal MI, death, stroke, repeat revascularization, recurrent unstable angina, heart failure, repeat hospitalization, length of hospital stay, angina relief, quality of life, or cognitive effects. A majority of sex-specific reporting was on the composite outcome of major cardiovascular adverse events (death, MI, or revascularization). Individual outcomes by sex were rarely reported, especially on heart failure, repeat hospitalization, length of hospital stay, angina relief, quality of life, or cognitive effects.

Based on the small number of studies that looked at demographic and clinical factors that influence response to treatment strategies in women, there was insufficient evidence that clinicians can use to determine if age, race/ethnicity, socioeconomic status, coronary risk factors, angiographic-specific factors, CABG-specific factors, or hospital-level characteristics should be taken into consideration when deciding a treatment strategy for women with CAD. Unfortunately, more studies are needed that evaluate the subgroups and various demographic and clinical characteristics to fully understand this evidence gap.

In addition, the safety concerns or harms of these treatment strategies are underreported for women enrolled in RCTs. It appears that the bleeding risk may be higher in women receiving fibrinolysis or PCI. Careful consideration should be given to the dose, timing, and duration of antiplatelet, antithrombotic, and anticoagulant therapies administered to women.

Limitations of This Review

With 28 studies meeting the inclusion criteria, this systematic review has several limitations. First, our search focused on comparative RCTs—the highest quality of evidence for determining the efficacy of different treatment modalities on cardiovascular outcomes. While this was adequate for evaluating the evidence to support the clinical outcomes by treatment strategy and by CAD presentation for the overall population, there were very few RCTs that reported subgroup analyses by demographic or clinical characteristics and also very few RCTs that reported the harms or risks of therapy. Most studies that reported results applicable to modifiers of effectiveness or safety did this for the overall population and did not separate the effects by sex. We are aware that there are several observational or noncomparator studies of each of the treatment modalities that address these issues in women. Because of the problems with confounding from observational studies and the difficulty of constructing reliable comparisons among single-arm studies, we did not include observational or noncomparator studies in our review.

Second, the sample size and low representation of women in most of the comparator studies may affect the study authors’ ability to analyze the results by sex, therefore reducing the number of studies reporting these findings separately (i.e., reporting bias). We excluded 355 articles due to lack of sex-specific reporting of the study results, which resulted in low numbers of studies available for analysis for each clinical presentation (STEMI, UA/NSTEMI, stable angina). Of these 355 articles, 116 were associated with the same 28 studies included in our review, but they did not report data on women separately. The remaining 239 articles were associated with 173 studies that did not report data on women. Figure 13 presents a graph of the number of articles reporting data on women per year. The percentage ranges from 0 percent (in 1992 and 1993) to 75 percent in 1995. On average, 17 percent of the articles comparing treatment strategies for CAD reported sex-specific outcomes. Of note, many articles included a multivariate analysis which included sex as a covariate in the model; the majority found no evidence of a gender effect. The result of a multivariable model is insufficient for incorporating into a meta-analysis, thus these were excluded from the review. Reporting bias in these publications therefore resulted in selection bias in this review.

Figure 13

Number of articles reporting data on women by year.

Third, the strength of our meta-analysis is limited by the different definitions of the primary composite outcome and by the timing (short-term and long-term) of those clinical endpoints. We used our best judgment in choosing which composite outcomes (e.g., death/MI/stroke and death/MI/stroke/revascularization) and time points (e.g., in-hospital and 30 days) to combine in the meta-analysis.

A final limitation is the change in PCI techniques and definition of optimal medical therapy over time. Most of the studies involved balloon angioplasty or bare-metal stents. The current era of drug-eluting stents and the use of dual antiplatelet therapy may be underrepresented. Nevertheless, the findings represent the best available evidence. While the treatment options continue to evolve over time, these older therapies (bare-metal stents, balloon angioplasty) are still being used in clinical practice, and therefore we did not downgrade the strength of evidence based on the availability of newer technologies. Medication adherence to beta blockers, angiotensin-converting enzyme inhibitors, aspirin, antiplatelet agents, and lipid-lowering agents was not reported in the studies included in this review. There was also variable reporting on the implementation of optimal medical therapy.

Many of these studies were multicenter, international RCTs with multiple countries represented. The generalizability of those studies to the United States may be of concern; however, the practice of revascularization and prescription of medical therapies are not dramatically different.

Conclusions

From a limited number of studies reporting results for women separately from the total study population, our findings confirm current practice and evidence for care in one of the three areas evaluated.

1. For women with STEMI, we found that an invasive approach with immediate PCI is superior to fibrinolysis for reducing cardiovascular events, which is similar to findings in previous meta-analyses combining results for both women and men.
2. For women with NSTEMI or unstable angina, we found evidence suggestive of a benefit of early invasive strategies for reducing cardiovascular events; however, this benefit was not statistically significant. Previous meta-analyses of studies comparing early invasive with initial conservative strategies on a combined population of men and women showed a significant benefit of early invasive therapy.
3. For women with stable angina, the few trials reporting sex-specific data on revascularization compared with optimal medical therapy showed a greater benefit with revascularization for women, while the men in the study fared equally well with either treatment. In contrast, previous meta-analyses that combined results for men and women found similar outcomes for either treatment.