Sex Differences in Cardiovascular Medication Prescription in Primary Care: A Systematic Review and Meta‐Analysis

1angiotensin‐converting enzyme inhibitors

2antihypertensive medications

3beta blocker

4calcium channel blocker

5coronary heart disease

6cardiovascular disease

7This systematic review with meta‐analysis shows that there are sex differences in cardiovascular medication prescription among patients at high risk or with established cardiovascular disease in primary care.Women were less likely to be prescribed aspirin, statin, or angiotensin‐converting enzyme inhibitor but more likely to have a prescription for diuretics. Sex differences in cardiovascular prescription in primary care need to be addressed in order to optimize the use of cardiovascular medication for both women and men.

8This systematic review with meta‐analysis shows that there are sex differences in cardiovascular medication prescription among patients at high risk or with established cardiovascular disease in primary care.Women were less likely to be prescribed aspirin, statin, or angiotensin‐converting enzyme inhibitor but more likely to have a prescription for diuretics.

9This systematic review with meta‐analysis shows that there are sex differences in cardiovascular medication prescription among patients at high risk or with established cardiovascular disease in primary care.

10Women were less likely to be prescribed aspirin, statin, or angiotensin‐converting enzyme inhibitor but more likely to have a prescription for diuretics.

11Sex differences in cardiovascular prescription in primary care need to be addressed in order to optimize the use of cardiovascular medication for both women and men.

12Sex differences in cardiovascular prescription in primary care need to be addressed in order to optimize the use of cardiovascular medication for both women and men.

13Cardiovascular disease (CVD) remains the leading cause of death worldwide, accounting for about a third of all deaths in both women and men.1 Historically, there has been a misperception that CVD predominantly affects men, which may have resulted in suboptimal management and treatment of CVD in women.2, 3 Over recent decades, substantial efforts have been made to characterize CVD in women. As a result, important differences between women and men in the presentation, diagnosis, and medical treatment of CVD have been identified.2, 4

14Most studies on sex differences in CVD management have been performed in secondary care.3, 5, 6, 7 For example, among all patients receiving statins after hospitalization for myocardial infarction in the United States, women were less likely than men to receive high‐intensity statins, despite guideline recommendations.6 Also, a study of coronary heart disease patients recruited from routine outpatient cardiology clinics in 11 countries across Europe, Asia, and the Middle East showed that women were less likely than men to reach all treatment targets set by clinical guidelines.3 Whether similar sex differences exist in primary care has not been systematically evaluated. Considering that both patients at high risk and with established CVD attended clinics in primary care to monitor their current CVD treatment, primary care visits are a key stage at which any sex inequities in treatment could and should be investigated. Comprehensive evidence on current sex differences in cardiovascular medication prescription in primary care would help to obtain a better understanding of the utilization of evidence‐based medical treatment for both sexes and encourage all health professionals to strive for sex equity in providing CVD management to their patients.

15In this study, we conducted a systematic review and meta‐analysis to determine the prevalence of common cardiovascular medication prescription in women and men in primary care and to evaluate whether prescriptions for guideline‐recommended cardiovascular medications differ between the sexes.

16The authors declare that all supporting data are available within the article and its online supplementary files.

17A systematic search of observational studies was performed in PubMed/MEDLINE and Embase for studies published between 2000 and 2019 using combined text word subject heading terms (Table S1). The reference lists of all related articles were screened for any other potentially relevant studies.

18All observational studies that reported the sex‐specific prevalence of prescriptions of cardiovascular medications (aspirin, statins, and any antihypertensive medication including beta blockers, calcium channel blockers [CCBs], angiotensin‐converting enzyme inhibitors [ACE inhibitors], and diuretics) for patients at high risk or with established CVD (coronary heart disease, stroke, heart failure, and atrial fibrillation) in primary care were included. Studies were excluded if they (1) were published in a language other than English; (2) presented an unrelated study population, outcome, or were not performed in primary care; (3) included <1000 patients; (4) reported cardiovascular medication prescription only for 1 sex; and (5) assessed cardiovascular medication not by prescription (such as self‐report or pharmacy dispensing).

19Duplicate records were removed before title and abstract screening. When there were multiple reports from the same study, the report involving the highest number of cases or most explicit participants characteristics and outcome measures was included. Four independent reviewers (M.Z., E.R.C.M., C.C., and K.H.) screened the papers by title and abstract against the inclusion and exclusion criteria. Any disagreement between reviewers was discussed and the full text was reviewed, if necessary. A similar process took place in reviewing the full text of selected papers. A tailor‐made data extraction form was used to collect information on study and participant characteristics and sex‐specific prevalence of prescriptions of cardiovascular medication (Table S2).

20Study quality was assessed using the modified Newcastle‐Ottawa scale for observational studies. This scale consists of 6 items that assess the quality of participant selection, comparability, and outcome adjudication (Tables S3 and S4).8

21The primary outcome was the women‐to‐men prescription prevalence ratio with 95% CI for each cardiovascular medication. The secondary outcomes were the sex‐specific prescription rates of each cardiovascular medication.

22In general, the included studies reported unadjusted numbers, rates, or percentages of women and men with cardiovascular medication prescriptions. If a measure of variability was not reported, these were estimated from the rate and the sample size. The women‐to‐men prevalence ratios with 95% CI were pooled across studies using random‐effects meta‐analyses with inverse‐variance weighting for each medication.9 In sensitivity analysis, we pooled the results from studies that had adjusted for age. As different studies reported on different antihypertensive medications, we also restricted the analyses on individual antihypertensive medications to studies that reported on each of the 4 antihypertensive medications. Metaregression analyses were performed to assess the impact of mean age and age difference (women minus men) on our findings. We further investigated whether there was a trend in sex differences in cardiovascular medication prescription over time. In subgroup analysis, we assessed whether the findings differed by CVD status (high risk only, prevalent CVD, and high risk and prevalent CVD combined). P<0.05 were considered statistically significant. Statistical analyses were performed by using the “metafor” package in R version 3.2.2.

23Of the 10 803 studies identified through the systematic search, 900 studies were reviewed in full text (Figure 1). Of these, 43 studies were included, including a total of 2 264 600 participants, of whom 630 111 (28%) were women. The mean age ranged from 51 to 76 years (where reported). Table shows the key characteristics of the included studies. Of the 43 studies, 18 included information on aspirin,10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 30 on statins, * 14 on any antihypertensive medications, † 21 on beta blockers, ‡ 13 on CCBs, § 21 on ACE inhibitors, || and 14 on diuretics. ¶ Eight out of 43 studies reported cardiovascular medication prescription for high‐risk patients,17, 32, 38, 47, 48, 49, 52, 53 24 for patients with established CVD, # and 11 for both high‐risk and CVD patients. **

24In women, the pooled prevalence of cardiovascular medication prescription was 41% for aspirin, 60% for statins, and 68% for overall antihypertensive medications. The corresponding rates for men were 56%, 63%, and 69%, respectively. The pooled women‐to‐men prevalence ratios were 0.81 (95% CI, 0.72–0.92) for aspirin, 0.90 (95% CI, 0.85–0.95) for statins, and 1.01 (95% CI, 0.95–1.08) for any antihypertensive medications (Figure 2).

25Figure 3 shows the women‐to‐men prevalence ratios of individual antihypertensive medication prescription. Women were less likely to be prescribed with ACE inhibitors (women‐to‐men prevalence ratio: 0.85; 95% CI, 0.81–0.89) whereas the prevalence of diuretics prescription was higher than in men (women‐to‐men prevalence ratio: 1.27; 95% CI, 1.17–1.37). There were no significant sex differences in the prescription of beta blockers and CCBs. Findings were similar in analyses restricted to studies that reported on all 4 individual antihypertensive medications (Figure S1). Findings were similar in age‐adjusted analyses, available for 31 studies (Tables S5 through S10).

26Among the 31 studies that reported a sex‐combined mean age of the study population, there was no evidence that the women‐to‐men prevalence ratio varied systematically according to the mean age (Figure S2; P values: 0.57 for aspirin; 0.24 for beta blockers; 0.27 for CCBs; 0.41 for ACE inhibitors; 0.85 for diuretics). The only exception was that in studies with older patients, women were less likely than men to be prescribed statins whereas women had a higher prevalence of statin prescription compared with men in studies including younger patients (P=0.003).

27Among the 17 studies that reported sex‐specific mean ages, there was no evidence that the prevalence ratio varied systematically according to the women to men age difference (Figure S3; P values: 0.34 for aspirin; 0.21 for statins; 0.93 for beta blockers; 0.91 for CCBs; 0.89 for ACE inhibitors). The exception was the higher prevalence of diuretics prescription in women increased as the difference between the mean age of women and the mean age of men increased (P=0.006).

28The sex differences in prevalence ratio of prescription did not significantly change over time for aspirin (P=0.92), any antihypertensive medications (P=0.99), beta blockers (P=0.43), CCBs (P=0.44), ACE inhibitors (P=0.39), and diuretics (P=0.58) (Figure S4). However, the pattern and magnitude of the sex differences in statin prescription changed over time, with an increased women‐to‐men prevalence ratio (P=0.003).

29Among patients with established CVD, women were less likely to be prescribed with aspirin (0.89, 95% CI, 0.84–0.94), statins (0.85; 95% CI, 0.80–0.90), beta blockers (0.90, 95% CI, 0.85–0.96), and ACE inhibitors (0.88, 95% CI, 0.84–0.93) (Figure S5, Table S11). In contrast, women with established CVD were more likely to be prescribed with diuretics than their male counterparts (1.25; 95% CI, 1.09–1.43). Similar pooled estimates, but with wider CIs, were found when studies included only high‐risk participants, or when studies included both participants at high risk of and with established CVD. Time trends in the women‐to‐men prevalence ratio in medication prescription did not differ materially by CVD status (Figures S6 through S11). However, the women‐to‐men ratio of statin prescription increased over time in studies among high‐risk patients but not in studies including patients with established CVD or in studies including both high‐risk and CVD patients (P for interaction=0.002).

30In this systematic review and meta‐analysis of 43 studies including over 2 million participants, we found that there were sex differences in cardiovascular medication prescription among patients at high risk or with established CVD in primary care. Compared with men, women were less likely to have a prescription for aspirin, statins, or ACE inhibitors but more likely to have a prescription for diuretics. Sex differences did not vary materially by age, but there was some evidence to suggest that the magnitude of sex differences in statin prescription increased over time.

31Previous studies in secondary care have demonstrated that women are generally less likely than men to have a prescription of guideline‐recommended cardiovascular medications after a cardiac event.2, 3, 5, 54 SUrvey of Risk Factors, a clinical audit with over 10 000 patients from 11 countries, indicated that women had a lower prevalence of cardiovascular medication use than men and were less likely to reach treatment targets.3 Similarly, a study of 36 000 patients with established coronary heart disease in the United States, showed that women were less likely than men to be prescribed with aspirin, ACE inhibitors, or statins at both acute and hospital discharge of coronary heart disease.55 A study in the United Kingdom showed that prescription rates for cardiovascular medications were about 10% lower among women than men <55 years for acute myocardial infarction.56 Furthermore, a Dutch population‐based analysis also found persistent sex differences in the use of lipid‐lowering medications for secondary prevention of CVD, particularly in younger patients.5 We did not observe that sex disparities differed between age groups, but we noticed that the sex differences in statin prescription persisted and was even larger in the more recent studies. A recent study in the United States confirmed that women were 9% less likely than men to receive high‐intensity statins, as opposed to other types of statin.57 The present study further expands these findings by showing that sex differences in medication prescription also exist among patients at high cardiovascular risk or with established CVD in a primary care setting. We also demonstrated that women were more likely to be on diuretics but less likely to be on ACE inhibitors, which is in line with other studies.56, 58, 59 Sex differences in progression and presentation of CVD and comorbidities, the efficiency of treatment, and/or adverse drug effects may lead to different requirements on antihypertensive regimens.59, 60 The reasons for the contrasting sex differences within antihypertensive medication classes require further study.

32There are several other possible explanations for the lower prescription rates of some cardiovascular medications in women than men. First, the incidence of CVD in women is, typically, about a third that of men in middle age and occurs in men about a decade earlier than women, which might have led to the misperception that CVD is less common in women and does not have to be prevented as intensively as in men.4, 34, 61 Additionally, women may have a lower awareness of the severity of their disease and of appropriate CVD treatment and receive less support from healthy providers, compared with men, resulting in lower health consciousness and less frequent use of healthcare services.5, 62, 63, 64

33Although beyond the scope of the current investigation, studies have reported a considerable delay in receiving appropriate medical treatment to reduce the risk of incident or recurrent cardiac event in women.2, 23, 62, 63 Also, women may have less belief than men in the safety and effectiveness of cardiovascular medications and have been reported to have a greater risk of suffering adverse drug reactions, which may lead to a higher discontinuation rate of cardiovascular medications.60, 65, 66, 67 Indeed, studies have shown that women have a poorer adherence to cardiovascular medication than men in primary care.68, 69 These factors would be expected to produce a wider disparity between the usage of cardiovascular medications than our study of prescriptions suggests.

34We conducted a large‐scale systematic review with meta‐analyses on sex differences in cardiovascular medication prescription among patients at high risk or with established CVD in a primary care setting. We included all major cardiovascular medications and found that our results were generally robust across patient characteristics. Limitations of this study are inherent to its design and include the differences across studies in design, population, and end point definition.9 We had no information on potential combinations of cardiovascular medications prescribed, nor were we able to adjust our findings to potentially important comorbidities or other characteristics. However, some cardiovascular medications target the same risk factor and the lower use of ACE inhibitors among women, relative to men, could be explained by women's higher use of diuretics. Also, we considered sex differences only in medication prescription and were not able to determine whether those differences, where found, resulted in different levels of risk factor control and event rates. Furthermore, patients with established CVD seen in primary care may also receive treatment from secondary care. Also, it is not clear whether general practitioners or cardiologists would be the main source of prescriptions in any individual case. Finally, as the studies included in this review were conducted in mostly high‐income countries, the generalizability of our findings to low‐ and middle‐income countries needs to be assessed.

35In conclusion, this meta‐analysis, summarizing all recent literature, shows that sex differences in cardiovascular medication prescription persist in primary care. Future research is needed to determine the underlying causes of observed sex differences and to develop tailored strategies to optimize the use of evidence‐based cardiovascular medication for both women and men.

36Zhao is supported by a grant from the Netherlands Organization for Scientific Research (NWO; grant number: 0.22.005.021). Woodward is supported by National Health and Medical Research Council (NHMRC) Australia project grant 632507 and Fellowship APP1080206. Vaartjes is supported by a grant from the Dutch Heart Foundation (grant DHF project “Facts and Figures”). Hyun is supported by National Heart Foundation Australia Postdoctoral Fellowship (102138). Peters is supported by a UK Medical Research Council Skills Development Fellowship (MR/P014550/1).

37Woodward is a consultant to Amgen and Kirin. The remaining authors have no disclosures to report.