Maintaining Goal Blood Pressures After Switching From Olmesartan to Other Angiotensin Receptor Blockers

Abstract

This study assessed blood pressure (BP) goal maintenance in patients controlled with olmesartan monotherapy after switching to another angiotensin type II receptor blocker (ARB). Hypertensive patients prescribed olmesartan monotherapy were identified from GE Healthcare's Centricity electronic medical record between 2007 and 2011. After documentation of BP goal (<140/90 mm Hg) attainment, patients were placed into the continuation cohort if olmesartan monotherapy was maintained or into the switch cohort if they were changed to irbesartan, losartan, or valsartan. Follow‐up assessments were the first BP measurement 28 to 390 days after attaining BP goal (continuation cohort) or after prescribing an alternative ARB (switch cohort). Of 3412 patients included (3027 continuation cohort, 385 switch cohort), 52% were women and mean age was 58.0 years. In the switch cohort, 310 (80.5%) were switched to losartan (n=236), irbesartan (n=58), or valsartan (n=16) monotherapy and 75 (19.5%) were switched to combination antihypertensive therapy. Mean baseline and follow‐up BP were 122.5/75.8 mm Hg and 126.6/77.6 mm Hg, respectively, in the continuation cohort (P<.001) and 123.5/75.4 mm Hg and 129.6/78.5 mm Hg, respectively, in the switch cohort (P<.001). BP goal maintenance was 78.7% and 72.2% in the continuation and switch cohort, respectively (odds ratio, 0.707; 95% confidence interval, 0.555–0.899). Patients who continued on olmesartan monotherapy after attaining BP goal had a higher percentage of BP goal maintenance than patients who switched therapy.

Hypertension is a leading risk factor for cardiovascular events and a major cause of morbidity and mortality.1, 2 The costs of hypertension have been estimated to be $131 billion annually in healthcare expenditures in the United States.3 Data from the 2009–2010 National Health and Nutrition Examination Survey indicate that nearly one third of US adults had hypertension but less than half (47.2%) had it controlled to a blood pressure (BP) goal of <140/90 mm Hg.4 However, 60.3% of these hypertensive patients had controlled BP if they were on antihypertensive treatment. Uncontrolled hypertension among adults with hypertension is directly associated with increased mortality. Adequate hypertension treatment and control of BP can help reduce the incidence of first and recurrent heart attacks and strokes.1, 5 However, many patients receiving treatment fail to attain recommended BP goals of <140/90 mm Hg, and these estimates are even lower for patients with diabetes mellitus or chronic kidney disease who may have lower BP goals.4

Angiotensin type II receptor blockers (ARBs) are highly effective in the management of hypertension and they are a well‐tolerated drug class overall.6 Eight ARBs are commercially available in the United States. Olmesartan medoxomil is a widely prescribed ARB and is the second newest agent within this drug class. Clinical trials have demonstrated that olmesartan medoxomil (referred to as olmesartan) is considered effective as monotherapy and is considered cost‐effective within a managed care health system.7 It has been evaluated in head‐to‐head studies against other ARBs and has demonstrated efficacy in lowering BP.8, 9

A number of the ARBs have become generic. Losartan was the first generic ARB but recently, in 2012, both valsartan (certain versions) and irbesartan lost patent exclusivity. The availability of generic ARBs has prompted many payers to force conversions of hypertensive patients taking other brand name ARBs or even other therapeutic drug classes to switch to these generic ARBs as cost‐saving measures. This retrospective database analysis study was conducted to compare maintenance of BP control in patients who continue olmesartan monotherapy with patients who switch to therapy with another ARB. The primary objective of this study was to evaluate BP control after switching of ARB therapy in patients who had achieved BP control on olmesartan monotherapy.

Methods

Study Design

A retrospective analysis was performed using the GE Centricity electronic health record (EHR) database, with data from 2007 to 2011. The data collected included medical coding (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD‐9‐CM]), clinical measurements, and prescribing data. These data were used to identify diagnoses, determine BP values, and ascertain antihypertensive medication prescribing, respectively.

Inclusion Criteria

The study population included patients aged 18 to 89 years with a diagnosis of hypertension (ICD‐9‐CM 401.x). Patients were then included if they were prescribed olmesartan as their only antihypertensive therapy (monotherapy) for a period of at least 14 days before the BP measurement and if they had controlled BP based on a BP measurement that was <140/90 mm Hg. The encounter at which patients met this inclusion criterion for controlled BP was identified as their BP control index date. Patients were only included if they were active in the GE Centricity EHR before the follow‐up BP measurement. A follow‐up BP measurement was between 28 and 390 days after the first BP measurement that met the inclusion criterion of <140/90 mm Hg (BP control index date).

Exclusion Criteria

Patients were excluded if they had a history of diabetes (ICD‐9‐CM 250.x) or heart failure (ICD‐9‐CM 428.x). Patients with diabetes were excluded due to evolving data evaluating different BP goals in this population, which has questioned standards of care. Patients with heart failure were excluded because antihypertensive drugs are often prescribed to treat disease beyond control of BP.

Patient Cohorts

Patients were placed into one of two cohorts based on prescribing of ARB therapy after their BP control index date.

The continuation cohort included patients who were prescribed ongoing olmesartan monotherapy after their BP control index date. The switch cohort included patients who were prescribed either losartan, irbesartan, or valsartan therapy after their BP control index date.

Outcome Measurements

BP values documented in the EHR were used as the primary clinical outcome measurements. The following outcomes were determined for each cohort: (1) mean systolic and diastolic BP values after the BP control index date, (2) changes in BP values after the BP control index date, and (3) percentage of patients maintaining BP values <140/90 mm Hg after the BP control index date. The first BP measurement documented in the EHR between 28 and 390 days after BP control index date was used to determine these outcomes.

Statistical Analysis

All analyses were performed using sas 9.2 (SAS Institute Inc, Cary, NC). Descriptive statistics were used to compare outcomes between the two cohorts. We also compared demographic characteristics such as age, sex, race, smoking status, body mass index, and insurance type as relevant characteristics between cohorts.

Adjusted analyses were conducted using a logistic regression comparing the two cohorts. For the adjusted analysis, dichotomous outcome variables of maintaining BP goal were regressed against demographic characteristics of sex, race, age, and Charlson Comorbidity Index (divided into three groups of 0, 1–2, and ≥3). Odds ratios were calculated to assess for differences between cohorts.

Results

A total of 3412 patients met the inclusion criteria. Within this population, 3027 patients were in the continuation cohort and 385 were in the switch cohort. The baseline demographics and insurance types were similar between the two cohorts (Table 1).

Table 1.

Baseline Characteristics

Characteristic	Continuation Cohort, No. (%)	Switch Cohort, No. (%)	P Value
Total cohort size	3027 (100.0)	385 (100.0)
Age, y	57.7	60.0	.001
Sex
Female	1547 (51.1)	216 (56.1)	.06
Male	1480 (48.9)	169 (43.9)	.06
Body mass index, kg/m2	30.1	29.9	.55
Race
Asian	24 (0.8)	5 (1.3)	.31
Black	124 (4.1)	13 (3.4)	.50
Hispanic	38 (1.3)	7 (1.8)	.37
Not entered	199 (6.6)	19 (4.9)	.21
Other	24 (0.8)	2 (0.5)	.56
Unknown	1224 (40.4)	126 (32.7)	.004
White	1394 (46.1)	213 (55.3)	.0006
Smoking
Unknown	351 (11.6)	28 (7.3)	.01
Current	299 (9.9)	42 (10.9)	.52
Former	829 (27.4)	116 (30.1)	.26
Never	1548 (51.1)	199 (51.7)	.84
Insurance type
Commercial	1027 (33.9)	128 (33.2)	.78
Medicaid	23 (0.8)	0 (0)	.96
Medicare	561 (18.5)	98 (25.5)	.001
Other/unknown/missing	1416 (46.8)	159 (41.3)	.04

Comorbidities are identified by International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis codes.

By definition, all of the patients in the continuation cohort were on olmesartan monotherapy after the BP control index date. However, of the 385 patients in the switch cohort, 310 (80.5%) were switched to losartan, irbesartan, or valsartan monotherapy. Despite being at goal BP prior to the index date, 75 (19.5%) patients in this switch cohort were taking another ARB (losartan, irbesartan, or valsartan) in combination with another antihypertensive agent from a different drug class, as combination therapy. Most patients were switched to losartan monotherapy. The distribution of prescribing patterns in the switch cohort is depicted in Table 2.

Table 2.

Description of Antihypertensive Use in the Switch Cohort

Switch Cohort (n=385)	Patients, No. (%)
Switched to another ARB as monotherapy	310 (80.5)
Losartan	236 (61.3)
Irbesartan	58 (15.1)
Valsartan	16 (4.2)
Switched to another ARB in combination with another antihypertensive agent	75 (19.5)

Abbreviation: ARB, angiotensin 2 receptor blocker.

Most patients were prescribed the 20 mg olmesartan dose at the BP control index date in both cohorts. In the continuation cohort, 1783 and 642 patients were prescribed the 20 mg and 40 mg dose of olmesartan (602 had either no dose listed or were prescribed the 5 mg dose), respectively. In the switch cohort, 223 and 90 patients were initially prescribed the 20 mg and 40 mg dose of olmesartan (72 had either no dose listed or were prescribed the 5 mg dose), respectively. Among patients switched to losartan, the 25 mg, 50 mg, and 100 mg doses were prescribed in 14, 58, and 48 patients, respectively (116 had no dose listed). Among those switched to valsartan, the 40, 80, 160, and 320 mg doses were prescribed in 2, 30, 3, and 8 patients, respectively (15 had no dose listed). Among patients switched to irbesartan, the 75, 150, and 300 mg doses were prescribed in 2, 8, and 3 patients, respectively (3 had no dose listed).

Maintenance of BP response was assessed in several ways. Patients in both the continuation cohort and the switch cohort had significant increases in mean systolic and diastolic BP values after the index date. Mean BP values before and after the index date for patients prescribed ARB monotherapy throughout the study duration are listed in Table 3. All of the increases in BP were greater in the switch cohort. The mean increase in systolic BP was 4.1 mm Hg and 6.1 mm Hg in the continuation and switch cohorts, respectively. The mean increase in diastolic BP was 1.8 mm Hg and 3.1 mm Hg in the continuation and switch cohorts, respectively.

Table 3.

Mean Blood Pressure (BP) Values Before and After the BP Control Index Date

	Before	After	Change	P Value
Continuation cohort (n=3027)
Systolic BP, mm Hg	122.5	126.6	4.1	<.001
Diastolic BP, mm Hg	75.8	77.6	1.8	<.001
Switch cohort: monotherapy (n=385)
Systolic BP, mm Hg	123.5	129.6	6.2	<.001
Diastolic BP, mm Hg	75.4	78.5	3.1	<.001

To assess maintenance of BP control after the index date, the percentage of patients maintaining BP values <140/90 mm Hg after the BP control index date was calculated. Higher goal BP attainment rates were observed in the continuation cohort (Table 4). This was observed for systolic BP goal maintenance, diastolic BP goal maintenance, and both systolic/diastolic BP goal maintenance. The odds ratio of maintaining BP control in the switch cohort compared with the continuation cohort is displayed in Table 5. Compared with the continuation group, patients in the switch group had 23.4%, 38.5% and 29.3% lower odds of maintaining systolic BP goal, diastolic BP goal, and both, respectively.

Table 4.

Percentage of Patients Maintaining BP Goal After the Index Date

	Continuation Cohort, No. (%)	Switch Cohort, No. (%)
Total cohort size	3027 (100.0)	385 (100.0)
Systolic BP <140 mm Hg	2510 (82.9)	301 (78.2)
Diastolic BP <90 mm Hg	2731 (90.2)	332 (86.2)
Both systolic BP <140 mm Hg and diastolic BP <90 mm Hg	2382 (78.7)	278 (72.2)

Blood pressure (BP) goal defined as <140/90 mm Hg based on the first measurement occurring between 28 and 390 days after the BP control index date.

Table 5.

Odds Ratio for Maintaining BP Goal After the Index Date in the Switch Group Compared With the Continuation Group

	Odds Ratio (95% Confidence Interval)
Systolic BP <140 mm Hg	0.766 (0.589–0.996)
Diastolic BP <90 mm Hg	0.615 (0.445–0.847)
Both systolic BP <140 mm Hg and diastolic BP <90 mm Hg	0.707 (0.555–0.899)

Blood pressure (BP) goal defined as <140/90 mm Hg based on the first measurement occurring between 28 and 390 days after the BP control index date.

Discussion

The overall benefits of appropriately treating hypertension have been definitively demonstrated in numerous clinical trials.1, 10 However, implementing an antihypertensive regimen and titrating that regimen can be challenging. Many patients with hypertension are on antihypertensive drug therapy yet have not achieved their goal BP value.11 Switching antihypertensive therapy after achieving goal BP values is sometimes done for financial reasons in an effort to convert patients to less expensive generic products. Our findings indicate that patients who had achieved BP goal values with olmesartan monotherapy were at risk for decreased control whether they continued olmesartan therapy or when switched to other antihypertensive drugs. However, they were more likely to have decreased control if switched to a different ARB. The relationship between BP and cardiovascular disease is well‐known.12 It is possible that loss of BP control may result in additional medical harm.

The differences between the two cohorts in BP values after goal attainment were small but may be clinically meaningful. The change in systolic BP was approximately 4 mm Hg in the continuation cohort and 6 mm Hg in the switch cohort. The clinical impact of small differences in BP has been widely debated. In a meta‐analysis that included 12.7 million person‐years of data from 61 prospective observational studies, elevated BP was strongly and directly related to an increased risk of stroke mortality, ischemic heart disease mortality, and other vascular mortality.13 Similar differences in diastolic BP have also been demonstrated.14, 15 A systolic BP difference of 2 mm Hg can result in a patient not being considered at their goal BP value and may necessitate a change in drug therapy (increased dosage or addition of a new drug). However, it is not known whether this small difference in BP is clinically meaningful.

The impact of switching drug therapy for the management of chronic diseases has been evaluated in the literature. In the treatment of dyslipidemia, patients who were switched from a higher potency statin to a lower potency statin experienced worse disease control.16 Switch programs with statin therapy that involve allowing physicians and patients to have complete control of the switch process have been shown to decrease drug costs.17 However, the direct costs of implementing this type of a program may be a barrier, as it was estimated that the cost was $131,000 to facilitate switching among 1710 patients. Additionally, another barrier is that even when patients are given a choice, only 50% in this type of model chose to voluntarily switch their therapy. It has also been demonstrated that statin‐switching programs that utilize a therapeutic conversion approach with prospective clinical evaluation result in better control of dyslipidemia than usual care switching programs that are based on equipotent conversions.18

There are several similarities between statin therapy for dyslipidemia and ARB therapy for hypertension. Both diseases are generally asymptomatic and have long‐term risk for cardiovascular events that are associated with more severe disease.1, 19 For both hypertension and dyslipidemia, achieving goal values can be challenging, and many patients are treated yet not controlled.4, 11 Moreover, once goals are attained, the standard of care is to continue therapy unless there is a reason to switch or alter treatment. When comparing statins with ARBs, both represent a drug class that is proven to reduce risk of cardiovascular events, both classes have generic and brand name agents, and both classes have demonstrated differences in efficacy within class.6, 20 Moreover, both classes are targets for switch programs to minimize drug costs. Therefore, much of the knowledge gained from statin switch programs can be applied to ARB switch programs.

The effects of restrictive drug benefit designs for antihypertensive agents have been published in the literature.21, 22, 23 One study compared employers who implemented an antihypertensive step therapy program (11,815 patients) with employers who did not (30,882 patients).21 The step therapy program demonstrated small initial reductions in costs, but also a decrease in adherence (based on medication fills). Moreover, patients in the step therapy programs also experienced an increase in inpatient admissions and emergency department visits. Similar findings demonstrating decreased adherence and increased medical costs have also been demonstrated in a switch program that converted patients from valsartan to another ARB.22 Other step programs that have encouraged switching an ARB to another antihypertensive drug class have demonstrated risks such as patients being treated with no antihypertensive therapy.23 Switching is a complicated process. There are no exact conversion algorithms for converting patients from one antihypertensive agent to another. Because of this lack of an exact conversion, switching programs may result in loss of BP control.

Study Limitation

There are several limitations to our data. This was a retrospective analysis of one EHR database. Our findings may not be reproducible to all populations. In our database, medication use was defined based on prescribing from this EHR, and BP values were based on what was entered into the EHR. Accuracy of BP measurements that are documented in the EHR cannot be assured. The time that BP was reassessed after achievement of goal BP was not systematically evaluated in both of the cohorts. There is a possibility that variability in the time of these assessments could have influenced the BP measurement. However, the intention of this study was not to control for the follow‐up period, but rather assess real‐world patterns once controlled patients were switched to a different ARB. We did not assess dosing of ARB therapy, which could have influenced BP changes and BP goal maintenance. Due to the nature of EHR data, many patients did not have the exact ARB dose documented. In addition, patients with lower copays with generic ARBs might be more adherent with therapy due to lower financial burden, but the relationship between copayments and adherence could not be assessed within this study. These limitations are typical for these types of database analyses but nonetheless represent real‐life data.

Our findings highlight the potential shortcomings of switching antihypertensive therapy in patients who have achieved control of their hypertension. It is important to note that even when therapy is continued after achieving BP control, there is a risk of losing control. This risk was amplified when patients were switched from olmesartan to another ARB‐based therapy. An explanation for the loss of control observed in the switch cohort could have been converting olmesartan to a dose of another ARB that is not similar in antihypertensive efficacy. Additionally, patients in the switch cohort were older, with more receiving Medicare benefits. It is also possible that the patients in the switch group were influenced to change therapy because of intervention programs that promote generic utilization, but this cannot be determined. While most of the patients in our switch cohort were switched to another ARB as monotherapy, many were switched to combination antihypertensive therapy. This would suggest that evaluating the effect of switching using real‐world data may be complicated.

Conclusions

Our data indicate that more hypertensive patients who had attained their BP goal while treated with olmesartan monotherapy maintained control by continuing olmesartan monotherapy than by switching to losartan, irbesartan, or valsartan. Respectively, our study showed that the odds of maintaining goal BP were 29.3% lower in patients who switched to losartan, irbesartan, or valsartan as either monotherapy or in combination antihypertensive therapy.

J Clin Hypertens (Greenwich). 2013;15:888–892. DOI: 10.1111/jch.12197. ©2013 Wiley Periodicals, Inc.