Skip Navigation LinksHome > March 2014 - Volume 32 - Issue 3 > Cost-effectiveness of Barostim therapy for the treatment of...
Journal of Hypertension:
doi: 10.1097/HJH.0000000000000071
ORIGINAL PAPERS: Hypertension management

Cost-effectiveness of Barostim therapy for the treatment of resistant hypertension in European settings

Borisenko, Olega; Beige, Joachimb; Lovett, Eric G.c; Hoppe, Uta C.d; Bjessmo, Staffane

Open Access
Supplemental Author Material
Article Outline
Collapse Box

Author Information

aSynergus AB, Stockholm, Sweden

bDepartment of Nephrology, Kuratorium for Dialysis and Transplantation (KfH), Hospital St Georg, Leipzig, Leipzig, Germany

cCVRx Inc., Minneapolis, Minnesota, USA

dDepartment of Internal Medicine II, Paracelsus Medical University Salzburg, Salzburg, Austria

eSynergus AB and Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm, Sweden

Correspondence to Oleg Borisenko, MD, PhD, Synergus AB, Svardvagen 19, 182 33, Danderyd, Sweden. Tel: +46 8 544 767 50; fax: +46 8 544 767 59; e-mail:

Abbreviations: ACE, angiotensin-converting enzyme; ESRD, end-stage renal disease; ICER, incremental cost-effectiveness ratio; LoS, length of stay; LYG, life-year gained; MI, myocardial infarction; OMT, optimal medical treatment; post-MI, post-myocardial infarction; PSA, probabilistic sensitivity analysis; QALY, quality-adjusted life years; RR, relative risk; RRT, renal replacement therapy; TIA; transient ischaemic attack

This study has not been presented at any scientific conferences or published in biomedical journals.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivitives 3.0 License, where it is permissible to download and share the work provided it is properly cited. The work cannot be changed in any way or used commercially.

Received June 18, 2013

Accepted November 7, 2013

Collapse Box


Objective: The purpose of this study is to simulate the cost-effectiveness and the long-term clinical performance of the Barostim neo System for the treatment of resistant hypertension when compared to optimal medical treatment.

Methods: A decision analytic model with a combination of a decision tree and Markov process was used to evaluate the cost-effectiveness of Barostim. The clinical effectiveness of Barostim was based on the results of the randomized, placebo-controlled Rheos trial and the follow-up substudy of the DEBuT-HT trial. The cost-effectiveness was modelled from a German societal perspective over a lifetime horizon. Patients with high SBP levels have an increased risk of myocardial infarction, stroke, heart failure and end-stage renal disease.

Results: In a simulated cohort of 50-year-old patients at high risk of end-organ damage, Barostim therapy generated 1.66 additional life-years and 2.17 additional quality-adjusted life years with an incremental cost of €16 891 when compared with continuation of medical management. Barostim was estimated to be cost-effective compared with optimal medical treatment with an incremental cost-effectiveness ratio of €7 797/QALY. In the model, Barostim reduced over a lifetime the rates of myocardial infarction by 19%, stroke by 35%, heart failure by 12% and end-stage renal disease by 23%. The cost-effectiveness of Barostim can be greater in younger patients with resistant hypertension and in patients with significant risk factors for end-organ damage.

Conclusion: Barostim may be a cost-effective treatment when compared with optimal medical management in patients with resistant hypertension.

Back to Top | Article Outline


Hypertension is a significant worldwide health problem that leads to end-organ damage including myocardial infarction, stroke, heart failure and end-stage renal disease. The risk of cardiovascular disease doubles with each increment of 20 mmHg of SBP or 10 mmHg of DBP, starting as low as 115/75 mmHg [1]. Throughout the development of end-organ damage, hypertension causes approximately 13% of the global deaths annually, and its related morbidity leads to a significant cost to society. Hypertension widely affects the European population with a prevalence rate of 55% in Germany, 47% in Spain and 42% in the UK [2]. Hypertension is often inadequately treated due to patients’ lack of compliance, inadequate dosages of the drugs or simply resistance to medications.

Resistant hypertension is defined as the failure to control blood pressure despite the use of three or more antihypertensive medications from different classes including a diuretic at the maximal or the highest tolerated dose [3]. The prevalence of drug-resistant hypertension ranges between 6 and 27% in different studies. Current medical options are inadequate to treat resistant hypertension, which remains a serious societal and healthcare problem.

Barostim (CVRx Inc., Minneapolis, Minnesota, USA) is a minimally invasive device that reduces high blood pressure and improves cardiovascular function. Barostim technology triggers the natural blood pressure regulation system by electrically activating the carotid baroreceptors. In the latest European Society of Hypertension/European Society of Cardiology (ESH/ESC) Guidelines for the management of arterial hypertension, carotid baroreceptor stimulation is mentioned as one of the options to treat resistant hypertension [4].

The objective of this economic evaluation is to determine, from a societal perspective, the cost-effectiveness of Barostim as a second-line treatment in adult patients with resistant hypertension (defined as SPB ≥140 mmHg while receiving at least three appropriate antihypertensive medications including a diuretic) compared to an appropriate pharmacological multidrug treatment in a German population.

Back to Top | Article Outline


Model description

The design of the study was a decision analytic model. A combination of a decision tree and a Markov model was used to evaluate the cost-effectiveness of Barostim compared with optimal medical treatment [5]. The cycle length was 1 month. The decision tree structure included four branches for the Barostim arm (no complications, device pocket haematoma, wound complication and wound pain) and a single branch for the optimal medical treatment arm (alive with hypertension; Fig. 1).

Figure 1
Figure 1
Image Tools

All patients entered the Markov process which covers the most important end-stage organ damage including myocardial infarction, stroke and transient ischaemic attack (TIA), heart failure and end-stage renal disease (Fig. 2). During each cycle, patients in an initial hypertensive state could experience a nonfatal myocardial infarction, stroke or TIA, heart failure, end-stage renal disease or die from cardiovascular or other conditions. Patients could also experience a hypertensive crisis requiring hospitalization while being in a hypertensive state. Patients who experienced a nonfatal myocardial infarction could develop heart failure, stroke, survive or die from a noncardiovascular condition. Survivors of a myocardial infarction could experience a recurrent fatal or nonfatal myocardial infarction, heart failure and stroke or remain in a post-myocardial infarction health state. Patients who experienced a nonfatal stroke could only proceed to a post-stroke health state or die from a noncardiovascular condition. Survivors of stroke could experience a fatal or nonfatal recurrent stroke, die or remain in a post-stroke health state. Patients in a heart failure state could develop end-stage renal disease, stroke, die or remain in this state. End-stage renal disease was defined as kidney failure requiring renal replacement therapy. Patients in an end-stage renal disease state could receive renal transplant, die or remain in the same state. Patients who underwent a renal transplantation could die or survive. Survivors of a renal transplantation could die or remain in a post-transplant health state.

Figure 2
Figure 2
Image Tools

The economic analysis was conducted from a societal perspective. Cost-effectiveness was estimated over a lifetime horizon.

Back to Top | Article Outline
Input data
Clinical effectiveness and safety data

The clinical effectiveness of Barostim in the context of resistant hypertension was determined by a decrease in SBP and DBP. Several cardiovascular risk prediction models were used to transform the ability of Barostim to lower blood pressure to the hard clinical outcomes.

Fatal cardiovascular events were determined by using the Systematic Coronary Risk Evaluation (SCORE) project data [6] as recommended by the clinical ESH/ESC Guidelines [4]. Data were modelled for low-risk populations, which are relevant to the German setting. The risks of a nonfatal myocardial infarction, stroke and heart failure were based on the Framingham equations [7–10]. Equations using SBP were employed. The risk of end-stage renal disease according to SBP was determined using data from a large cohort study [11]. Risks were updated annually for the initial 4 years after start of the model (the longest follow-up reported for Barostim) and after that every 10 years. Age-specific and sex-specific noncardiovascular mortality was estimated using German life tables for 2010 with subtracted mortality for cardiovascular conditions [12,13].

The ability of Barostim to lower blood pressure in the simulation was estimated from the randomized, double-blind, sham-controlled (RCT) Rheos Pivotal trial [14]. In the trial, 265 patients were randomized to either the activation of the implanted Barostim at 1 month or 6 months postimplant. Outpatient office blood pressure was measured using a standardized automated device technique in which six measurements were performed and the average of the last five measurements was recorded. Initially, all patients included in the study received an implant. The use of a sham-control, instead of a pharmacological placebo group, leads to an underestimation of the effect of Barostim at 6 months when compared to the sham-control group. Patients with an inactive system in the control group experienced a significant blood pressure-lowering effect, which may be explained by the mechanical impact of the implant on the carotid sinus resulting in a possible baroreflex mechanical activation. To adjust the results of the trial for this mechanical effect, the effectiveness of Barostim in the model was taken from the active implant arm using SBP changes at 6 and 12 months, and the clinical effectiveness was then compared to the preimplant level. By definition, patients with resistant hypertension on intensive medication therapy do not experience any improvements. Data of impact of medication therapy on SBP were derived from the control group of the Symplicity HTN-II trial and used to determine the changes in the level of SBP in the optimal medical treatment arm in the simulation [15]. During the 1st model cycle, there was no decrease in SBP levels in any of the arms. Beginning with the 2nd until the 12th cycle, patients in the Barostim arm experienced a decrease of SBP when using the 6-month RCT data. The level of SBP at 12 months and until 24 months was determined by the decrease demonstrated at 12 months in the trial. The level of SBP between 24 and 48 months was determined by the results of the prospective case series study with the longest follow-up reported for Barostim – substudy of DEBuT-HT trial [16]. Data for SBP reduction at 48 months were used to extrapolate the clinical effectiveness until the end of the model as long-term follow-up studies demonstrated no rebound of the clinical effect for Barostim [17,18]. Key clinical data and transition probabilities are provided in Table 1. Other data are provided in the Supplemental Digital Content,

Table 1
Table 1
Image Tools

Short-term adverse events were included for the second generation of Barostim, and were based on a recent prospective study [19]. The invasiveness of the procedure as well as Barostim properties changed dramatically with the second generation of the device, so the Rheos data were not relevant. The probability of a hypertensive crisis was based on the results of the Rheos RCT [14].

For end-stage organ damage health states, increased mortality was determined using the relative risks of all-cause mortality when compared to the general population. Relative risks were derived from large epidemiological studies [20,25–27,29]. Time-dependent and age-dependent mortality risks were used for end-stage renal disease (mortality decreases over years with renal replacement therapy and is higher for younger patients) [30]. Time-dependent mortality risk was used for post-transplant health states (mortality decreases for survivors in the first year after a renal transplant) [32]. Data from the normal population in German life tables were used to calculate the mortality in the model for the end-stage organ damage health states [13]. Probabilities of acute mortality for recurrent myocardial infarction and stroke were also obtained from the literature [22,23]. Details of transition probabilities are provided in Table 1. Additional clinical data are provided in the Supplemental Digital Content,

Transition probabilities from one end-stage organ damage state to another were based on the literature [25–27,29,31].

In summary, the effectiveness of Barostim in lowering SBP in patients with resistant hypertension was evaluated using well studied and validated risk prediction models for the occurrence of the first adverse health event (myocardial infarction, stroke, heart failure or end-stage renal disease). When patients experienced an adverse event, disease-specific risks were used to model the increased mortality or transition to other negative health states. Disease-specific risks did not account for the level of SBP.

Transformation of transition probabilities into monthly probabilities for different time horizons was performed using a standard approach [33].

Back to Top | Article Outline
Resource utilization and cost data

Evaluation of resource utilization data with cost was performed using German sources. Both direct and indirect costs were used. Indirect costs were calculated using the human capital method.

The cost of Barostim implantation procedure included the cost of the full Barostim system, the cost of the procedure, the cost of treatment for complications and the cost of outpatient follow-up visits to a surgeon. The costs of implantation and complications were obtained from the G-DRG 901D. The required hospitalization for the implantation procedure was considered to be 2 days. The treatment of wound complications required 3 additional hospital days, the treatment of pocket haematoma required 2 additional days and device repositioning due to wound pain required 1 additional day. In the Barostim arm, two visits to the cardiologist and two visits to the technician (for the device half and full activation) were required during the 2nd month. Currently, the manufacturer of Barostim is providing activation services and no additional costs were assumed for the device activation. Battery life was 6 years. The cost of the battery replacement consisted of the Barostim battery cost and the cost of the procedure. No lost productivity was assumed for the Barostim implantation procedure. Resource utilization for Barostim implantation and follow-up was validated by a German physician familiar with the Barostim system. Details of resource and cost data are provided in Table 2.

Table 2
Table 2
Image Tools

The basic management of hypertension included pharmacological therapy and biannual visits to a general practitioner. A distribution of the different pharmacological agents used was derived from the Rheos RCT [14] with a slightly lower monthly cost for pharmacological therapy in the Barostim arm when compared with the optimal medical treatment arm.

The costs of end-stage organ damage health states were obtained from the German literature [34–39]. The cost of Barostim was provided by CVRx Inc., the manufacturer of the Barostim system. The cost of medications was based on PharmNet.Bund-Drug Information System of the German Institute of Medical Documentation and Information (DIMDI) [40]. Only reimbursable prices were taken into account. The cost of outpatient medical services was based on EBM (Einheitlicher Bewertungsmaßstab) data [41].

The human capital method was used to calculate indirect costs. Lost productivity from paid and unpaid work due to acute illness and from early retirement was considered. Productivity was the sum of paid and unpaid work. Productivity from paid work was determined using average labour costs from German employees. Standard German retirement age (60 for women and 65 for men) was used as a threshold for the accounting of paid work. After standard retirement age, only unpaid work was considered. Productivity from unpaid work was determined using the net income of a social worker. The following activities were considered relevant to household work (according to Harmonized European Time Survey terminology): food preparation, dish washing, dwelling cleaning, other household upkeep, laundry, ironing, construction and repairs, shopping and services along with other domestic work. Harmonized European Time Survey data (2007) were used to calculate the time spent on unpaid household work (Supplemental Digital Content,, and it was assumed to be independent of age. Data on the proportion of patients who did not return to work after major events and the timing for the return to work for the remaining patients were derived from the literature (Supplemental Digital Content, Data concerning average gross labour cost and net labour cost for German workers were derived from the Federal Statistical Office.

Cost data are presented in 2011 Euros. Inflation adjustment was performed using the German consumer price index [42].

Back to Top | Article Outline
Utility data

Health-related quality of life was expressed on the basis of the generic HRQoL instrument, EQ-5D [43–46] and Health Utility Index [47] scores. Preferences, measured by a visual analogue scale in hypertensive patients, were used for the utility of the hypertensive states [48]. Patients’ preferences were used for the utility for all health states. For the renal transplantation state, a utility decrement of 0.3 was used as no published sources of the utility of the renal transplant state were found. Face validity of utility inputs was evaluated. Details of utility data are provided in Table 3.

Table 3
Table 3
Image Tools
Back to Top | Article Outline
Cohort description

A single cohort of patients at high risk of end-organ damage was simulated. A cohort representative was a 50-year old smoking man with hyperlipidaemia, and no history of coronary heart disease (CHD) and atrial fibrillation. He has a SBP of 170 mmHg, a heart rate of 79 beats/min, a BMI of 32.6 kg/m2, lung vital capacity of 2.5 l, cholesterol level of 9.06 mmol/l, high-density lipoprotein level of 1.32 mmol/l, no sign of cardiomegaly on radiograph and left ventricular hypertrophy (LVH) on ECG. Cohort characteristics were based on the Barostim RCT [14] and, for the parameters that were not reported in the RCT, on the large German epidemiological Heinz Nixdorf Recall Study [50].

Back to Top | Article Outline

The incremental cost-effectiveness ratio (ICER) was calculated by comparing the difference in average total costs with the difference in average quality-adjusted life years (QALY) among the study cohorts. The intervention was considered cost-effective if the ICER was below €35 000 per QALY [51,52]. All costs and outcomes beyond the first year were discounted 3.0% annually based on the recommendations of the German National Institute for Quality and Efficiency in Healthcare (IQWiG) [53].

In addition to the cost-effectiveness, clinical effectiveness was evaluated by analysing the cumulative rates of adverse events and the relative risk of adverse events at 10 years and over a lifetime.

The model was constructed using Microsoft Excel 2010 (Microsoft Corp., Redmond, Washington, USA).

Back to Top | Article Outline
Model validation

The validation of the model was performed through one of the largest prospective studies in hypertension in the European population – the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT-BPLA) [54,55]. Details of the model validation process are presented in the Supplemental Digital Content,

Back to Top | Article Outline
Sensitivity analysis

One-way sensitivity analysis was performed to assess the impact of varying the model parameters while holding other variables fixed at base-case values. Cost drivers (variables with a major input to the costs) were identified and results are presented by means of a Tornado diagram.

In addition to a one-way sensitivity analysis, a number of additional scenarios were tested: limiting the effectiveness of Barostim to the results of Rheos RCT only with the extrapolation of 1-year follow-up results; assuming the fade-out effect of Barostim effectiveness between 1 and 5 mmHg annually from the latest available observation (4 years); assuming no change of SBP in the optimal medical treatment arm; limiting the effectiveness of Barostim to only the period of observation (4 years) with the assumption that the SBP returned to the baseline level; and applying different annual discount rates (costs – 5%, benefits – 5%; costs – 0%, benefits – 0%; costs – 3%, benefits – 0%).

A probabilistic sensitivity analysis (PSA) was also performed using Monte Carlo simulations. Ten thousand simulations were performed. Beta distribution was used for the probabilities. Gamma distributions were used for the longitudinal data (SBP level), for the cost data with descriptive statistics available and for the utility data. Lognormal distributions were used for the relative risks. Uniform distributions were used for parameters based on expert or analyst assumptions and for parameters for which descriptive statistics were not available in the original publications.

Back to Top | Article Outline


In the base-case analysis, Barostim was shown to be cost-effective against continued optimal medical treatment in patients with resistant hypertension.

In simulation, Barostim was calculated to prevent a significant number of the first and recurrent events over a lifetime. Thus, Barostim was calculated to decrease the number of first episodes of myocardial infarction by 12% and recurrent episodes by 46% (overall rate reduction of 19%), the number of first episodes of stroke by 33% and recurrent episodes by 41% (overall rate reduction of 35%), the number of heart failure cases by 12% and the number of end-stage renal disease cases by 23%. The number of events over a lifetime is provided for both comparative arms in Table 4.

Table 4
Table 4
Image Tools

Over a lifetime, Barostim was shown to increase survival from 15.93 to 17.59 years (an increment of 1.66 life-years) and quality-adjusted survival from 13.94 to 16.10 years (an increment of 2.17 QALYs) with an incremental cost of €16 891 compared with continued optimal medical treatment. The ICER for the base-case analysis was 7 797 €/QALY, which was significantly lower than the standard ICER threshold of 35 000 €/QALY.

Back to Top | Article Outline
Model validation

Results of the model validation against ASCOT-BPLA are presented in the Fig. 3. Results of validation for the primary parameters (overall mortality, cardiovascular mortality and combined outcome of cardiovascular death, myocardial infarction and stroke) showed that the model predicted outcomes with a high degree of precision, although the combined outcome of cardiovascular death, myocardial infarction and stroke was overestimated because of the higher rate of myocardial infarction and stroke in the model. In the model, the rate of heart failure was very similar to the ASCOT-BPLA results.

Figure 3
Figure 3
Image Tools
Back to Top | Article Outline
Sensitivity analysis

To verify the robustness of the analysis, one-way deterministic and probabilistic sensitivity analyses were carried out.

The model was robust in the one-way sensitivity analysis with no single parameter affecting the cost-effectiveness of Barostim. The most sensitive parameters were the baseline age and SBP, labour cost, Barostim effectiveness and the cost of Barostim battery, although the ICER did not exceed 14 000 €/QALY for any of the tested variables. The cost-effectiveness of Barostim increased with the addition of risk factors to the cohort characteristics. Results of the one-way sensitivity analysis are provided in the Tornado diagram (Fig. 4).

Figure 4
Figure 4
Image Tools

When the effectiveness of Barostim was limited to the 1-year results of the Rheos RCT and extrapolation of the follow-up results, Barostim demonstrated high effectiveness and cost-effectiveness. Over a lifetime, Barostim was shown to decrease the total number of myocardial infarctions by 12%, the total number of strokes by 22%, the number of heart failure cases by 8% and the number of end-stage renal disease cases by 16%. Barostim was calculated to produce additional 1.19 life years and 1.55 QALYs with an incremental cost of €23 791 compared with continued optimal medical treatment. The ICER for this analysis was 15 345 €/QALY, which is significantly lower than the standard ICER threshold of 35 000 €/QALY.

Sensitivity analysis on fade-out effect of Barostim effectiveness between 1 and 5 mmHg annually from the latest available observation (4 years) showed that Barostim remains cost-effective even with the rapid fade-out of effectiveness. Even with a maximal 5 mmHg annual fade-out effect in the Barostim arm, Barostim remained a cost-effective option with an ICER of 14 286 €/QALY. Detailed results of the sensitivity analysis on fade-out effect are presented in the Supplemental Digital Content,

Sensitivity analysis with no change from baseline SBP in the optimal medical treatment arm showed that the cost-effectiveness of Barostim changed slightly with an ICER of 8200 €/QALY.

Results of the very conservative scenario that assumed the immediate return of SBP to the baseline level from the time of the latest available observation (4 years) showed that Barostim remained cost-effective with an ICER of 15 328 €/QALY. In this scenario, Barostim was shown to decrease, over a lifetime, the total number of myocardial infarction by 12%, the total number of stroke by 15%, the number of heart failure cases by 8% and the number of end-stage renal disease cases by 14%.

In the sensitivity analysis with different discount rates, the model was robust with a better cost-effectiveness when lower discount rates were used. Detailed results of the sensitivity analysis on different discount rates are presented in the Supplemental Digital Content,

The PSA demonstrates that Barostim has an 82% probability of being cost-effective with an ICER threshold of 10 000 €/QALY, although the probability of it being cost-effective is dramatically increased after this point, and, already at an ICER threshold of 14 000 €/QALY, Barostim has a 99% probability of being cost-effective. A cost-effectiveness acceptability curve is presented in Fig. 5.

Figure 5
Figure 5
Image Tools

A cost-effectiveness acceptability plane (Fig. 6) demonstrates that Barostim provides a stable beneficial effect in 10 000 simulations with incremental QALY between 1.68 and 2.77 at an incremental cost in the range of € −1092 (cost-savings) and € 31 366.

Figure 6
Figure 6
Image Tools
Back to Top | Article Outline


The present economic evaluation supports Barostim as a cost-effective option for the treatment of resistant hypertension. The one-way sensitivity analysis demonstrates that the model is robust and no single parameter alters the cost-effectiveness of the technology. The PSA shows that the probability of Barostim to be cost-effective increases dramatically from 82% at an ICER threshold of 10 000 €/QALY to 99% at an ICER threshold of 14 000 €/QALY.

There are few economic evaluations published in the setting of resistant hypertension. One such evaluation published in 2009 is a US-based cost-effectiveness analysis performed from a payer perspective that evaluated Barostim versus optimal treatment with the renin inhibitor, aliskiren [56]. At the time of the model development, only limited feasibility data were available for Barostim, and the authors reported a gain of 16.868 QALYs in the Barostim arm compared with 16.584 QALYs in the optimal medical treatment arm. In a recently published economic evaluation of renal denervation in patients with resistant hypertension, authors reported a gain of 13.17 QALYs in the renal denervation arm and a gain of 12.07 QALYs in the standard of care arm [57]. Our results are similar to the latter study as we found a gain of 16.10 QALYs in the Barostim arm compared with a gain of 13.94 QALYs in the optimal medical treatment arm. Compared with renal denervation, simulation of the risk of adverse events showed similar or slightly beneficial results with Barostim, with lifetime relative risks of myocardial infarction, stroke, heart failure and end-stage renal disease of 0.85, 0.83, 0.92 and 0.81 compared with 0.81, 0.65, 0.88 and 0.77, respectively. The clinical benefits of Barostim are even greater with a 10-year time horizon with relative risks of myocardial infarction, stroke, heart failure and end-stage renal disease of 0.62, 0.53, 0.67 and 0.68, respectively. The relative clinical benefits of Barostim in the model are lower after 5–15 years due to significantly lower morbidity and mortality in the Barostim arm as more patients remain in an ‘hypertensive’ state or alive and, therefore, eligible for transition into adverse health states. This is a well known feature of closed cohort Markov simulations.

Compared with previously published models, our model is different in several ways. First, our model allows for the possibility of recurrent stroke and myocardial infarction with the risk of events independent of the level of blood pressure. Second, our model does not incorporate nonacute conditions like chronic kidney disease and angina pectoris. Third, the well known restriction of Markov models is the ‘memory-less’ or so-called the Markov assumption which means that once patients move to the next health state, the model does not ’remember’ the timing of transition or from which health state the patient moved from [58]. To attenuate the impact of the assumption, we built in a number of states to allow the incorporation of available clinical and cost data. The distinctive feature of the present model is its large number of health states (166 states), which significantly increases the preciseness of the disease history and better reflects the costs associated with adverse events. Fourth, another specific feature of the present model is the use of the SCORE project data to predict cardiovascular mortality instead of the Framingham equation used in other models. There are a number of cardiovascular risk prediction models available [the Framingham risk score, SCORE, the Prospective Cardiovascular Munster (PROCAM) score, the QRESEARCH cardiovascular risk algorithms and the Reynolds risk score], although the Guidelines for Management of Hypertension of the ESH/ESC mentioned only two of them: the SCORE project and the Framingham risk score [4]. The SCORE system is based on 12 datasets of more than 200 000 persons representing more than 2.1 million person-years of observation [6]. The SCORE project system was chosen for the prediction of cardiovascular mortality due to its highest potential of accurate risk prediction for different European populations. Similar to other models, our model used multivariate risk equations obtained from observational studies which, on one hand, may decrease the demonstrated clinical benefits compared with the results from experimental studies but, on the other hand, may predict risk more precisely and to a broader extent than experimental studies that usually have a limited timeframe.

Our study has several limitations. First, the impact of Barostim on the level of SBP was incorporated using changes between 6 and 48 months compared with the preimplant baseline values. Although, traditionally, comparative effectiveness is assessed through the comparison of the effect between groups at follow-up, it may lead to a significant underestimation of the effectiveness of Barostim. Rheos RCT demonstrated that the blood pressure-lowering effect of Barostim originated from both the mechanical and electrical stimulation of the baroreceptors. In the sham-group, a SBP decrease of 17 ± 29 mmHg was demonstrated at 6 months, which cannot be solely explained by the placebo effect, especially in patients with drug-resistant hypertension. Therefore, data from the control group of the recent Symplicity HTN-II trial were used to determine the blood pressure change in the optimal medical treatment arm of our model [15], although unadjusted comparison using results from another study is a well known methodological limitation. Second, the present model evaluates the cost-effectiveness of the second generation of the Barostim System, whereas clinical effectiveness data are based on the first generation of the device. This situation is quite common in the area of medical devices with frequent implementation of new features and device development. The second generation of Barostim has several features that improve the risk–benefit ratio of the device, but the most important are: the unilateral implantation, the smaller device size and the increased battery life, which significantly reduce the time and extent of the procedure and dramatically improve the safety and longevity of the device. A recently published case series study of the second generation of Barostim provided similar clinical results confirming the applicability of the effectiveness data from the RCT to the second generation of the device [19]. Third, the maximal length of observation in the Barostim studies is 4 years, whereas clinical effectiveness was extrapolated over a lifetime. Extrapolation was supported by the 2–4 years of follow-up data, which showed no rebound of clinical effectiveness. This was also supported by clinical experts, although long-term effectiveness should be evaluated in future studies. Fourth, due to practical reasons, our model incorporates a limited number of pathways (e.g. patient in the end-stage renal disease state can experience stroke or progress to renal transplant, but cannot experience myocardial infarction or heart failure). We only incorporated the main pathways with the highest probability and clinical relevance to the research question. Fifth, we assumed a constant consumption of medications in both arms over a lifetime, although clinical studies showed a decrease in the number of medications used [e.g. from 5.3 ± 1.9 to 4.7 ± 2.1 (P ≤0.001) at 12 months for the responders enrolled in the RCT [59]], which can lead to a lower cost of treatment and better quality of life due to fewer side-effects of medication therapy. The current recommendation of CVRx Inc. is to continue medications with the same optimal scheme after device implantation and activation.

In conclusion, Barostim may be cost-effective compared with optimal medical management for treatment of resistant hypertension in European settings.

Back to Top | Article Outline


The authors would like to thank Laura Geissler, MSc and Farhang Modaresi, MD, MBE for support in clinical and cost data collection. They thank Daniel Adam, MSc and Zeynep Colpan, MSc for support in data analysis. They thank Danielle Libersan, PhD for her help in the proof-reading of the article.

Back to Top | Article Outline
Conflicts of interest

O.B. and S.B. are employees of Synergus AB, a Med Tech consulting company. Synergus AB was paid by CVRx Inc. to develop a cost-effectiveness model. U.C.H. received scientific support from CVRx Inc. E.L. is an employee of the CVRx Inc. and has stock options of CVRx Inc. J.B. received honoraria from CVRx Inc. for the conductance of clinical studies and for holding scientific lectures. This study was supported by CVRx Inc.

Back to Top | Article Outline
Reviewer's Summary Evaluation Reviewer 2

Strenghts: the paper contains interesting new data demonstrating that Barostim may be cost-effective treatment when compared with optimal medical management in patients with resistant hypertension.

Weaknesses: the length of observation of Barostim studies was 4 years and clinical effectiveness was extrapolated over lifetime. This was supported by 2–4 years of follow-up. As the authors admit, the long-term effectiveness should be confirmed in future studies.

Back to Top | Article Outline


1. Kannel WB. Blood pressure as a cardiovascular risk factor: prevention and treatment. JAMA 1996; 275:1571–1576.

2. Wolf-Maier K, Cooper R, Kramer H, Banegas J, Giampaoli S, Joffres M, et al. Hypertension treatment and control in Five European countries, Canada, and the United States. Hypertension 2004; 43:10–17.

3. Sander GE, Giles TD. Resistant hypertension: concepts and approach to management. Curr Hypertens Rep 2011; 13:347–355.

4. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Bohm M, et al. The Task Force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology (ESC). ESH/ESC Guidelines for the management of arterial hypertension. J Hypertens 2013; 31:1281–1357.

5. Drummond MF, Sculpher MJ, Torrance GW, O’Brian BJ, Stoddart GL. Methods for the economic evaluation of healthcare programmes. New York:Oxford University Press; 2005.

6. Conroy RM, Pyorala K, Fitzgerald AP, Sans S, Menotti A, De Backer G, et al. Estimation of ten-year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 2003; 24:987–1003.

7. Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular disease risk profiles. Am Heart J 1991; 121:293–298.

8. D’Agostino RB, Wolf PA, Belanger AJ, Kannel WB. Stroke risk profile: adjustment for antihypertensive medication. The Framingham Study. Stroke 1994; 25:40–43.

9. Kannel WB, D’Agostino RB, Silbershatz H, Belanger AJ, Wilson PW, Levy D. Profile for estimating risk of heart failure. Arch Intern Med 1999; 159:1197–1204.

10. Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham study. Stroke 1991; 22:312–318.

11. Hsu CY, McCulloch CE, Darbinian J, Go AS, Iribarren C. Elevated blood pressure and risk of end-stage renal disease in subjects without baseline kidney disease. Arch Intern Med 2005; 165:923–928.

13. Life tables. German Federal Statistics Office 2008-2010. [Accessed 18 June 2013].

14. Bisognano JD, Bakris G, Nadim MK, Sanchez L, Kroon AA, Schafer J, et al. Baroreflex activation therapy lowers blood pressure in patients with resistant hypertension: results from the double-blind, randomized, placebo-controlled Rheos Pivotal trial. J Am Coll Cardiol 2011; 58:765–773.

15. Esler MD, Krum H, Sobotka PA, Schlaich MP, Schmieder RE, Bohm M. Renal sympathetic denervation in patients with treatment-resistant hypertension (the Symplicity HTN-2 trial): a randomised controlled trial. Lancet 2010; 376:1903–1909.

16. Kroon ASJ, Scheffers I, Tordoir J, Mohaupt M, Allemann Y, Jordan J, et al. Sustained blood pressure reduction by Baroreflex activation therapy with a chronically implanted system: 4-year data of Rheos Debut-HT study in patients with resistant hypertension. J Hypertens 2010; 28:e441.

17. Scheffers I, Kroon SJ, Tordoir AA, Mohaupt J, Allemann M, Jordan Y, et al. Sustained blood pressure reduction by baroreflex hypertension therapy with a chronically implanted system: 3-year data from the Rheos DEBuT study in patients with resistant hypertension. J Hypertens 2009; 27 (suppl 4):S421.

18. Scheffers IJ, Kroon AA, Schmidli J, Jordan J, Tordoir JJ, Mohaupt MG, et al. Novel baroreflex activation therapy in resistant hypertension: results of a European multicenter feasibility study. J Am Coll Cardiol 2010; 56:1254–1258.

19. Hoppe UC, Brandt MC, Wachter R, Beige J, Rump LC, Kroon AA, et al. Minimally invasive system for baroreflex activation therapy chronically lowers blood pressure with pacemaker-like safety profile: results from the Barostim neo trial. J Am Soc Hypertens 2012; 6:270–276.

20. Rosengren A, Wilhelmsen L, Hagman M, Wedel H. Natural history of myocardial infarction and angina pectoris in a general population sample of middle-aged men: a 16-year follow-up of the Primary Prevention Study, Goteborg, Sweden. J Intern Med 1998; 244:495–505.

21. Zanchetti A, Hansson L, Dahlof B, Elmfeldt D, Kjeldsen S, Kolloch R, et al. Effects of individual risk factors on the incidence of cardiovascular events in the treated hypertensive patients of the Hypertension Optimal Treatment Study. HOT Study Group. J Hypertens 2001; 19:1149–1159.

22. Krumholz HM, Wang Y, Chen J, Drye EE, Spertus JA, Ross JS, et al. Reduction in acute myocardial infarction mortality in the United States: risk-standardized mortality rates from 1995-2006. JAMA 2009; 302:767–773.

23. Dennis MS, Burn JP, Sandercock PA, Bamford JM, Wade DT, Warlow CP. Long-term survival after first-ever stroke: the Oxfordshire Community Stroke Project. Stroke 1993; 24:796–800.

24. Ovbiagele B, Diener HC, Yusuf S, Martin RH, Cotton D, Vinisko R, et al. Level of systolic blood pressure within the normal range and risk of recurrent stroke. JAMA 2011; 306:2137–2144.

25. Seliger SL, Gillen DL, Longstreth WT, Kestenbaum B, Stehman-Breen CO. Elevated risk of stroke among patients with end-stage renal disease. Kidney Int 2003; 64:603–609.

26. Witt BJ, Ballman KV, Brown RD, Meverden RA, Jacobsen SJ, Roger VL. The incidence of stroke after myocardial infarction: a meta-analysis. Am J Med 2006; 119:354 e1–354 e19.

27. Witt BJ, Gami AS, Ballman KV, Brown RD, Meverden RA, Jacobsen SJ, et al. The incidence of ischemic stroke in chronic heart failure: a meta-analysis. J Cardiac Fail 2007; 13:489–496.

28. Arnold JM, Yusuf S, Young J, Mathew J, Johnstone D, Avezum A, et al. Prevention of Heart Failure in Patients in the Heart Outcomes Prevention Evaluation (HOPE) Study. Circulation 2003; 107:1284–1290.

29. Velagaleti RS, Pencina MJ, Murabito JM, Wang TJ, Parikh NI, D’Agostino RB, et al. Long-term trends in the incidence of heart failure after myocardial infarction. Circulation 2008; 118:2057–2062.

30. Villar E, Remontet L, Labeeuw M, Ecochard R. Effect of age, gender, and diabetes on excess death in end-stage renal failure. J Am Soc Nephrol 2007; 18:2125–2134.

31. de Silva R, Nikitin NP, Witte KK, Rigby AS, Goode K, Bhandari S, et al. Incidence of renal dysfunction over 6 months in patients with chronic heart failure due to left ventricular systolic dysfunction: contributing factors and relationship to prognosis. Eur Heart J 2006; 27:569–581.

32. Rabbat CG, Thorpe KE, Russell JD, Churchill DN. Comparison of mortality risk for dialysis patients and cadaveric first renal transplant recipients in Ontario, Canada. J Am Soc Nephrol 2000; 11:917–922.

33. Briggs A, Claxton K, Sculpher M. Decision modelling for health economic evaluation. New York:Oxford University Press; 2006.

34. Brüggenjürgen B, Rupprecht HJ, Willich SN, Spannagl M, Ehlken B, Smala A, et al. Cost of atherothrombotic diseases—myocardial infarction, ischaemic stroke and peripheral arterial occlusive disease—in Germany. J Public Health 2005; 13:216–224.

35. Dodel RC, Haacke C, Zamzow K, Paweilik S, Spottke A, Rethfeldt M, et al. Resource utilization and costs of stroke unit care in Germany. Value Health 2004; 7:144–152.

36. Kolominsky-Rabas PL, Heuschmann PU, Marschall D, Emmert M, Baltzer N, Neundorfer B, et al. Lifetime cost of ischemic stroke in Germany: results and national projections from a population-based stroke registry – the Erlangen Stroke Project. Stroke 2006; 37:1179–1183.

37. McMurray JJ, Andersson FL, Stewart S, Svensson K, Solal AC, Dietz R, et al. Dialysis: a regional population-based analysis. Nephrology Assessment of Reduction in Mortality and Morbidity (CHARM) programme. Eur Heart J 2006; 27:1447–1458.

38. Icks A, Haastert B, Gandjour A, Chernyak N, Rathmann W, Giani G, et al. Costs of dialysis: a regional population-based analysis. Nephrol Dial Transplant 2010; 25:1647–1652.

39. Hagenmeyer EG, Haussler B, Hempel E, Grannas G, Kalo Z, Kilburg A, et al. Resource use and treatment costs after kidney transplantation: impact of demographic factors, comorbidities, and complications. Transplantation 2004; 77:1545–1550.

40. PharmNet. Bund-Drug Information System. German Institute of Medical Documentation and Information 2012. [Accessed 18 June 2013].

41. Einheitlicher Bewertungsmaßstab. 2012. [Accessed 18 June 2013].

43. Calvert MJ, Freemantle N, Cleland JF. The impact of chronic heart failure on health-related quality of life data acquired in the baseline phase of the CARE-HF study. Eur J Heart Fail 2005; 7:243–251.

44. Dorman P, Dennis M, Sandercock P. Are the modified ‘simple questions’ a valid and reliable measure of health related quality of life after stroke? J Neurol Neurosurg Psychiatry 2000; 69:487–493.

45. Haacke C, Althaus A, Spottke A, Siebert U, Back T, Dodel R. Long-term outcome after stroke. Stroke 2006; 37:193–198.

46. Lee AJ, Morgan CL, Conway P, Currie CJ. Characterisation and comparison of health-related quality of life for patients with renal failure. Curr Med Res Opin 2005; 21:1777–1783.

47. Mahoney EM, Jurkovitz CT, Chu H, Becker ER, Culler S, Kosinski A, et al. Cost and cost-effectiveness of an early invasive vs conservative strategy for the treatment of unstable angina and non-ST-segment elevation myocardial infarction. JAMA 2002; 288:1851–1858.

48. Stein JD, Brown GC, Brown MM, Sharma S, Hollands H, Stein HD. The quality of life of patients with hypertension. J Clin Hypertens 2002; 4:181–188.

49. Sullivan PW, Slejko JF, Sculpher MJ, Ghushchyan V. Catalogue of EQ-5D scores for the United Kingdom. Med Decis Making 2011; 31:800–804.

50. Erbel R, Eisele L, Moebus S, Dragano N, Mohlenkamp S, Bauer M, et al. The Heinz Nixdorf Recall study. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2012; 55:809–815.

51. Eichler HG, Kong SX, Gerth WC, Mavros P, Jönsson B. Use of cost-effectiveness analysis in health-care resource allocation decision-making: how are cost-effectiveness thresholds expected to emerge? Value Health 2004; 7:518–528.

52. Belgian Healthcare Knowledge Centre. Threshold values for cost-effectiveness in healthcare; 2008.

53. General Methods for the Assessment of the Relation of Benefits to Costs Version 1.0. German National Institute for Quality and Efficiency in Healthcare; 2009.

54. Dahlöf B, Sever PS, Poulter NR, Wedel H, Beevers DG, Caulfield M, et al. Prevention of cardiovascular events with an antihypertensive regimen of amlodipine adding perindopril as required versus atenolol adding bendroflumethiazide as required, in the Anglo-Scandinavian Cardiac Outcomes Trial-Blood Pressure Lowering Arm (ASCOT-BPLA): a multicentre randomised controlled trial. Lancet 2005; 366:895–906.

55. Sever PS, Dahlöf B, Poulter NR, Wedel H, Beevers G, Caulfield M, et al. Rationale, design, methods and baseline demography of participants of the Anglo-Scandinavian Cardiac Outcomes Trial ASCOT investigators. J Hypertens 2001; 19:1139–1147.

56. Young KC, Teeters JC, Benesch CG, Bisognano JD, Illig KA. Cost-effectiveness of treating resistant hypertension with an implantable carotid body stimulator. J Clin Hypertens 2009; 11:555–563.

57. Geisler BP, Egan BM, Cohen JT, Garner AM, Akehurst RL, Esler MD, et al. Cost-effectiveness and clinical effectiveness of catheter-based renal denervation for resistant hypertension. J Am Coll Cardiol 2012; 60:1271–1277.

58. Siebert U, Alagoz O, Bayoumi AM, Jahn B, Owens DK, Cohen DJ, et al. State-Transition Modeling: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force-3. Value Health 2012; 15:812–820.

59. Bakris GL, Nadim MK, Haller H, Lovett EG, Schafer JE, Bisognano JD. Baroreflex activation therapy provides durable benefit in patients with resistant hypertension: results of long-term follow-up in the Rheos Pivotal trial. J Am Soc Hypertens 2012; 6:152–158.


baroreflex activation therapy; Barostim; cost-effectiveness analysis; resistant hypertension

Supplemental Digital Content

Back to Top | Article Outline

© 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins


Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.