Hypertension is a chronic condition and a leading risk factor for the development of cardiovascular (CV) events, which are a major cause of morbidity and mortality worldwide.1–4 CV events create a significant clinical management and economic burden, and are a major cause of hospitalisation and physician visits.5 The burden associated with CV events may be alleviated by treating hypertension, thereby reducing the risk of future CV events. Results from long-term randomised controlled trials (RCTs) support this strategy, showing antihypertensives to reduce the risk of stroke and coronary heart disease by around 34 and 21%, respectively.6–8 The current recommended target blood pressure (BP) in hypertensive patients without diabetes is ≤140/90mmHg, while in hypertensive patients with diabetes the target BP is ≤130/80mmHg. Management options include lifestyle intervention, but most patients inevitably require pharmacotherapy to control their BP.9 Several classes of antihypertensive medication are currently available, including diuretics, beta-blockers (BBs), angiotensin-converting enzyme inhibitors (ACEIs), calcium channel blockers (CCBs), angiotensin-II receptor blockers (ARBs) and renin inhibitors.
The renin–angiotensin system, and more specifically angiotensin-II, have long been implicated in the development of hypertension10 and CV diseases.11–13 Thus, agents that target the renin–angiotensin system, either through direct blockade of the action of angiotensin-II at the angiotensin-II receptor type 1 (AT1) with ARBs or via inhibition of endogenous angiotensin-II formation with ACEIs, have become standard antihypertensive therapies. Indeed, ARBs are utilised as part of first-, second- or third-line antihypertensive treatment strategies. Large RCTs have shown ARB-based therapies to have comparable or better effects on lowering BP and reducing CV complications than other antihypertensive classes, such as the ACEIs.14–17 Clinical trials have also shown that ARBs demonstrate a favourable safety and efficacy profile; for example, ARBs are better tolerated than ACEIs due to the former’s more specific effects on the renin–angiotensin system.11,18
RCTs generally obtain data under a tightly controlled, artificial setting. By comparison, real-life databases collect data under an uncontrolled real-world setting. Thus, real-life databases may help complement and expand on data gathered from RCTs by providing an alternative information source. There are now data available from real-life databases for the efficacy and persistence rates of ARB- and non- ARB-based antihypertensive regimens. Recent examples include analyses of the Canada-based Southwestern Ontario (SWO) database, The Health Improvement Network (THIN), the Thales–Cegedim and the IMS Disease Analyzer real-life databases.
Rationale and Design of Real-life Databases
The SWO, THIN, Thales–Cegedim and IMS Disease Analyzer databases collect data from primary care patients. An overview of the design of these databases is provided below and summarised in Table 1.
The Southwestern Ontario Database
The SWO database collects information from primary care patients ≥18 years of age who visit selected rural and urban clinical practices in London, Ontario, Canada and surrounding areas. Information accrual began in 2000 and is ongoing, with data from at least 170,000 patients currently recorded.19,20 The database is updated on a quarterly basis, with information recorded on clinical outcomes such as hospitalisation, morbidity and mortality. Information on demographics, visit diagnosis, BP, medications and regimens (monotherapy or combination therapy) and consultation notes are also recorded in the database.
The Health Improvement Network Database
THIN is a national computerised medical record database that contains longitudinal data from primary care patients in the UK.21 To date, it has recorded information for over five million patients across around 300 general practices in the UK. This database records anonymous patient information on demographics, medical history, disease diagnosis, details of prescribed drugs and biometric parameters such as BP and body mass index (BMI), with data collected in a non-interventional manner during daily record-keeping within practices.
The Thales–Cegedim Database
Thales–Cegedim is a longitudinal patient database that records information for primary care patients from general practices in Belgium, France, Germany, Italy, the UK and Spain.22–24 General practitioners voluntarily upload anonymous and coded excerpts from the medical files of patients who have visited them. Data recorded in the database includes information on drug prescriptions, hospital admissions, CV events, laboratory tests and date of death.
The IMS Disease Analyzer Database
The IMS database is a primary-care-based longitudinal patient database that is updated on a monthly basis. It provides anonymous access to a representative and valid panel of physicians and patients across Europe. In Germany, approximately 2,000 physicians record data on 10 million patients. The data recorded include patient demographics, physician characteristics, prescriptions, hospital admissions and specialist referrals.25,26
Persistence Rates Associated with Angiotensin-II Receptor Blockers – Data from Real-life Databases
Despite the demonstrated clinical efficacy of the current antihypertensive medications, around 65% of patients still have uncontrolled hypertension (i.e. ≥140/90mmHg).27 The discrepancy in the degree of BP control achieved in clinical trials versus that in real life may be partly due to poor adherence and persistence rates in actual clinical practice.28 Adherence is defined as the degree to which a patient acts according to the prescribed interval and dose of a dosing regimen, while persistence may be defined as the duration from initiating treatment to its discontinuation.29 While RCTs have shown drug discontinuation rates of 5–10% by six to 12 months of antihypertensive treatment, much higher discontinuation rates (50–60% after six months, and up to 50–60% after one year) have been reported in actual clinical practice.30,31 The higher discontinuation rates, and hence lower persistence rates, in real life may be related in particular to a lack of tolerability of some antihypertensives in actual clinical practice where patients are treated in an uncontrolled environment.
Real-life data for the persistence and adherence rates of medications are extremely important as they can aid the decision-making process for the management of chronic conditions such as hypertension. The real-life effectiveness of different antihypertensives – even those with similar clinical efficacy – may be influenced by the rate of persistence and adherence. High adherence to antihypertensives has previously been associated with improved BP control.32
A retrospective cohort study utilised data from the IMS Disease Analyzer database to analyse the persistence rates of antihypertensive medications in previously untreated hypertensive patients in Germany.26 At two-year follow-up, the persistence rates were significantly higher (p<0.05) in the ARB monotherapy group (509 days, versus 459 and 324 days for the non-ARB and diuretic groups, respectively; see Table 2). The highest persistence rate (45%) was achieved with ARB monotherapy. The adherence rate for monotherapies within two years of treatment was significantly better (all p<0.05) with ARBs versus other antihypertensives. The risk of the first hypertension-associated event was higher (all p<0.05) with diuretics, BBs, CCBs and the group of non-ARBs, and was similar with ACEIs compared with ARBs.
A population-based cohort longitudinal study utilised data from the Thales–Cegedim database to evaluate the persistence rates associated with antihypertensives in French patients with newly diagnosed hypertension (n=15,227).33 The analysis showed the one-year persistence rate to be highest for ARBs (69.8%) and lowest for diuretics (45.3%). At four-year follow-up, the persistence rates were 63.2% for ARBs, 41.2% for ACEIs, 42% for CCBs and 34.2% for diuretics. At five-year follow-up, the persistence rate was significantly greater for the ARBs versus all the other antihypertensives (p<0.0001).33
Effects of Angiotensin-II Receptor Blockers on Blood Pressure – Data from Real-life Databases
While persistence rates for antihypertensive drugs have been shown to be lower in real-life settings than in clinical trials, the potential consequence of lower persistence rates is a negative impact on the effectiveness of antihypertensives on BP control.30,31 An analysis using the SWO database on nine-month follow-up outcomes showed that a similar proportion of antihypertensive patients reached target BP with various antihypertensive monotherapy regimens, although the greatest proportion was in the ARB-treated patients (see Table 3A).34
Analysis of the THIN database has shown the percentage of target BP achievers at one year of treatment to be greatest in ARB-monotherapy- treated patients, followed by ACEI monotherapy and then the other antihypertensive monotherapy regimens, such as CCBs, BBs and diuretics.35 The reduction in BP observed in the THIN database analysis followed a similar trend: over time, the magnitude of BP reduction was statistically significant for both systolic BP (SBP: (0.75mmHg; p<0.0001) and diastolic BP (DBP: 0.5mmHg; p<0.001) for ARB monotherapy versus other antihypertensive monotherapy groups. However, these reductions in SBP and DBP were small and probably not clinically significant. As previously mentioned, data from real-life databases have shown that ARBs are associated with a higher persistence rate than other therapies.26,33 Thus, it could be argued that these data suggest a link between the degree of BP control and the persistence rate of antihypertensives.
Analysis of the SWO database has also shown that, in patients receiving ARB monotherapy (irbesartan, losartan, valsartan or candesartan), the highest proportion of patients reaching target BP were those treated with irbesartan monotherapy (p≤0.01; see Table 3B).34 The efficacy of combination therapies has also been studied. Nine-month follow-up results from the SWO database showed that there a greater proportion of patients reached target BP when treated with an ARB-based dual therapy versus non-ARB-based dual therapies (39 versus 31%; p=0.004).36 Furthermore, ARB plus hydrochlorothiazide (HCTZ) dual therapy was significantly more effective at controlling BP compared with non-ARB-based dual therapies (p≤0.03; see Table 3C).37 The differences in the effect on BP between the different dual-therapy regimens may be partly related to differences in their persistence rates. Unpublished data from a preliminary analysis of the SWO database suggest that there are differences in persistence rates between ARB-based and non-ARB-based combination therapies; full publication of these data will hopefully confirm an advantage in persistence rate for ARB-based therapies. Analysis of the SWO database also showed that patients treated with irbesartan-based dual or triple therapy were more likely to reach target BP than those treated with dual or triple therapies based on other ARBs (p=0.001; see Table 3D).36
Analyses of data from UK patients in THIN showed that in patients with hypertension (defined as BP ≥140/90mmHg) given mono- or combination therapy with ARBs, irbesartan was associated with the largest mean reductions in BP.38–40 In particular, irbesartan plus HCTZ, at two-year follow-up, achieved significantly greater mean reductions in BP (following adjustments for a number of confounding factors) than combinations of other ARBs plus HCTZ.38
Beyond Clinical Trials – the Impact of Antihypertensive Medications on Cardiovascular Events
Results from placebo-controlled trials suggest that antihypertensives can reduce the risk of major CV events in hypertensive patients.6–8,41–44 However, low persistence rates with antihypertensives have been associated with a 15 and 28% increased risk of acute myocardial infarction and stroke, respectively.45 A recent retrospective study that utilised data from 18,806 Italian patients with newly diagnosed hypertension from the Thales–Cegedim database showed a significant decrease (38%) in the risk of acute CV events in high adherers to antihypertensive medication compared with low adherers (p=0.032).22 Clinical trials have also shown that improved medication persistence and adherence rates result in improved control of BP.7 As mentioned previously, differences in BP control seen with antihypertensives may be related to variability in persistence rates. This evidence points towards a relationship between better persistence and adherence rates, improved BP control and a reduced risk of CV events – a relationship that is important to communicate to patients.
Impact of Angiotensin-II Receptor Blockers on Cardiovascular Events – Data from Real-life Databases
The impact of ARBs and other antihypertensives on the risk of CV events has recently been evaluated by analyses of real-life databases. Analysis of the SWO database has shown that a significantly lower proportion of patients (p<0.001) treated with ARB-based mono- or dual-therapy regimens experience a CV event compared with those on ACEI- or CCB-based therapies (see Table 4).46 Additionally, patients treated with irbesartan-based therapy experienced significantly fewer CV events (p<0.02) than those treated with non-ARB-based therapy (see Table 4).46
A planned analysis of the SWO database of the impact of different antihypertensive therapies on the occurrence of CV events is anticipated to provide further insights into the links between persistence rates, extent of BP control and the impact of antihypertensives on the risk of CV events. The retrospective cohort study by Mathes et al. using the German IMS Disease Analyzer database evaluated the link between adherence and risk of first hypertension-associated event.26 In this analysis, hypertension-associated events were defined as including both cardiovascular and diabetes events (new diagnosis of type 2 diabetes or diabetic retinopathy). The unadjusted risk of the time to first hypertension-associated event at nine-month follow-up was favourable for ARBs and BBs; the proportion of event-free patients was greatest with BBs and ARBs. The adjusted risk of the first hypertension-associated event was significantly lower (p<0.005) with ARBs than with BBs, diuretics, CCBs or non-ARBs, but similar to ACEIs. Similar results were found for the rate of cardiovascular complications. Importantly, high adherence was found to be associated with a reduced risk of first hypertension-associated event (p<0.05) (see Table 2). These findings from the analyses of real-life databases point towards a strong association between improved persistence and adherence rates of antihypertensives, better BP control and a reduced risk of CV events.
Cost of Cardiovascular Event Management with Antihypertensives – Data from Real-life Databases
The management of CV events carries a significant economic burden.5 Nevertheless, a reduction in the occurrence of CV events may help reduce this burden. Based on the assumption that patients were hospitalised for CV events, an analysis of the SWO database showed the average cost of treating CV events in hypertensive patients to be significantly lower in ARB-based regimens compared with non-ARB-based therapies (see Figure 1).46 Furthermore, the average cost per patient of CV event management was significantly lower with irbesartan-based therapy versus other ARB-based therapies.46 Among the ARBs, the average cost per patient of CV event management was much lower for irbesartan-based therapy than for losartan-, valsartan- and candesartan-based therapies. The cost-effectiveness of ARB-based therapy has also been evaluated from data from the IMS Disease Analyzer database.26
The estimated average cost of treating the first hypertension-associated event (CV event and/or diabetes event) was found to be lower in the ARB-treated group (€2,339.95) than in groups treated with other antihypertensives (€2,531.68 to €3,910.47). The findings of these retrospective studies indicate that the reduced risk of CV events with ARBs may help alleviate the economic burden of CV disease.
Conclusions and Future Research
Real-life primary care databases have been used to analyse the impact of ARBs and other antihypertensive therapies on the risk of CV events. ARB-based therapies have been shown to be associated with a lower risk of CV events compared with non-ARB-based therapies. Data from these real-life databases also indicate a strong association between improved persistence and adherence rates of antihypertensives (particularly ARBs), better BP control and a reduced risk of CV events. As real-life databases continue to grow, it is hoped that the data accrued may help guide the first-choice antihypertensive treatment strategy as well as the optimal dose and dosing strategy and the patient subset most suitable for ARB-based therapy. While persistence rates from real-life databases of the different ARB-based therapies exist, there is a need for real-life data to help identify the particular antihypertensive or treatment regimens that enhance persistence and adherence to therapy; better understand the impact of persistence and adherence rates on the overall efficacy of the medications; and further evaluate the association of persistence and adherence of antihypertensives, extent of BP control and risk and occurrence of CV events.