• Cardiac risk factors and prevention
• Coronary artery disease
• Health Services

Cardiovascular disease (CVD) has historically been viewed as a man’s disease, yet more women than men die from CVD every year.1 Even with national campaigns aimed at increasing awareness of CVD in women, the number of women diagnosed with and dying from CVD continues to increase across the world.1 While biological differences between women and men likely contribute to these differences in outcomes, disparities in cardiovascular care may also play a role. Potential areas in cardiovascular prevention where disparities may occur include inadequate screening, untimely diagnosis and/or lack of appropriate treatment of CVD in women compared with men.

In their Heart paper, Hyun and colleagues evaluated potential gender disparities in cardiovascular risk screening and preventive medication use among patients with high cardiovascular risk enrolled in the Treatment of cardiovascular Risk in Primary care using Electronic Decision suppOrt (TORPEDO) study.2 3 This cluster-randomised trial conducted in 60 Australian primary care practices was designed to test whether an electronic decision support tool improves CVD risk detection and management. Using baseline enrollment data from TORPEDO, the study aimed to evaluate for gender differences in (1) recording of risk factors needed for CVD risk assessment (recorded smoking status, systolic blood pressure (BP) in previous 12 months, total and high density lipoprotein cholesterol in previous 24 months) and (2) appropriate preventive medication use (prescription for one or more BP-lowering drug and a statin for people at high risk without CVD, or a prescription for one or more BP-lowering drugs and a statin and an antiplatelet agent (unless already on an oral anticoagulant) for people with a diagnosis of CVD). High cardiovascular risk was defined by a history of CVD, presence of high risk conditions (diabetes mellitus and age >60 years, diabetes mellitus and albuminuria, stage 3B chronic kidney disease or extreme individual risk factor elevations: systolic BP ≥180 mm Hg, diastolic BP ≥110 mm Hg, total cholesterol >7.5 mmol (290 mg/dL)) or a 5-year CVD risk >15% as estimated by the Framingham risk equation.

Among the 53 085 patients enrolled, women were 12% (OR 0.88, 95% CI 0.81 to 0.96) less likely than men to have cardiovascular risk factors recorded. This difference was primarily driven by less documentation of smoking status in women compared with men (82.0% vs 85.3%, p<0.001) with smaller absolute differences in BP (89.0% vs 90%, p<0.001) or total cholesterol (75.0% vs 76.3%, p<0.001) measurement. Despite these differences, women at high risk for or with prevalent CVD were 12% more likely to have ‘appropriate’ preventive medications prescribed compared with men (OR 1.12, 95% CI 1.01 to 1.23). Notably, these gender differences were significantly dependent on age. Younger women (aged 35–54 years) were 37% less likely than younger men to have BP medications, statins or antiplatelets prescribed (OR 0.63, 95% CI 0.52 to 0.77), whereas older women (65+ years) were 34% more likely than older men to be prescribed these medications (OR 1.34, 95% CI 1.17 to 1.54) (p value for gender by age interaction <0.001). Importantly, this gender by age interaction was significant across all three classes of preventive medications studied (figure 3).3

The authors should be commended on evaluating cardiovascular prevention in a large primary care cohort. Many studies have investigated gender disparities in the acute management of cardiovascular disease with less emphasis on disparities in cardiovascular prevention.1 A particularly important finding of this study is the lower rate of smoking status screening in women compared with men (82% vs 85.3%, p<0.001). Smoking is a significant risk factor for CVD in women. Women smokers have a 25% greater risk of CVD than male smokers, independent of smoking intensity or other cardiovascular risk factors.4 Screening for smoking status and education on cessation are important for lowering cardiovascular risk for all patients, particularly women.

Certain limitations must be considered when interpreting the findings of this study. First, the study was conducted primarily in urban, teaching, primary care services throughout Australia and whether the findings apply more globally cannot be determined. Second, despite inclusion criteria requiring demonstration of regular clinic attendance (at least three visits over 24 months), up to 10% of patients had no BP recorded and over 30% had no height and weight (to estimate BMI) recorded. Since these are common or ‘protocolised’ vital signs obtained at primary care visits, the degree of missing measurement raises the possibility of (1) failure to capture the data, (2) inclusion of visits in which preventive care was not the focus (ie, acute care or problem-focused visit) or (3) substandard documentation practices. Data for TORPEDO were extracted from the electronic health record (EHR) and missing data were not imputed. If the opportunity for missing data were equal for women and men (missing at random), then the gender differences found in this study would not be affected. Alternatively, if women more often had non-preventive focused visits or attended substandard practices compared with men, then the gender differences found could be related to factors besides a failure to recognise risk in women.

Finally, another limitation of this study is whether all preventive medications studied were appropriate for all patients. The authors deemed all patients with high risk or a diagnosis of CVD as being appropriate for BP-lowering therapy, irrespective of whether they had evidence of hypertension. Clinical guidelines support BP-lowering therapy among patients with evidence of hypertension.5 Large epidemiological studies have shown a higher percentage of men than women have high BP until age 45; after age 65 a much higher percentage of women than men have high BP.6 Therefore, the gender differences in BP-lowering therapies noted in this study by age may in fact be appropriate.

Conversely, clinical guidelines support statin and antiplatelet therapy among the cohorts defined in this study. Therefore, the low rates of statin (65.9% vs 68%, p=0.01) and antiplatelet (53.4% vs 57.4%, p<0.001) therapies among women and men with indications are concerning and replicate findings in US cohorts of patients with cardiovascular disease.7 The authors acknowledge that contraindications to specific medications were not considered and may contribute to low rates overall and differences by gender. Despite these limitations, the consistent finding of low rates of preventive cardiovascular medications use among all patients with high risk and prevalent CVD highlights the need to define and overcome barriers for prescribing these important preventive therapies.

The study by Hyun et al adds to the growing evidence that many of the gender differences seen in CVD care are age dependent.8 9 Similar to what has been shown in management for hypertension and mortality after acute myocardial infarction, this study demonstrated that young women (age <54 years) with high risk of or prevalent CVD were least likely to be prescribed preventive medications.8 9 Adding to the gravity of this finding are recent US statistics showing ischaemic heart disease death rates in women between the ages of 35 and 44 are increasing, while ischaemic heart disease death rates are decreasing in men of similar age.1 10 The study supports the possibility that lower rates of preventive treatment among young women may contribute to worse long-term outcomes.

One potential explanation for age-dependent differences in preventive medication use is maintenance of a stereotype that women, and in particular young women, are not at risk for CVD. People commonly use stereotypes in their interactions with others as a mechanism to organise and simplify massive amounts of information.11 Physicians may be especially vulnerable to stereotypes in clinical settings since time pressure, brief encounters and the need to manage very complex tasks are characteristics of situations in which stereotypes are likely to be applied.12 The inappropriate application of prior knowledge or assumptions about differences between populations can result in ‘knowledge-mediated bias’.11 In other words, an assumption that all young women are at low risk for CVD resulted in lower rates of appropriate prevention therapies in women would constitute a bias.

Although gender stereotypes and bias may contribute to gender disparities in care, changing attitudes among all healthcare providers is an ambitious task. Efforts to ensure equity in care for women and men may be better spent by supporting decisions based on evidence appropriate to a patient’s objective clinical data rather than clinicians’ stereotypes. The TORPEDO trial demonstrated that a point-of-care decision support tool embedded in the EHR increased cardiovascular risk detection but did not significantly change preventive medication prescriptions.2 Comparing whether gender differences in risk estimation and management seen at enrollment into TORPEDO persisted at the end of the trial will improve our understanding of ways to potentially address gender disparities in care.