Cost-utility analysis of a national project to reduce hypertension in Israel
© Yosefy et al; licensee BioMed Central Ltd. 2007
Received: 09 January 2007
Accepted: 28 November 2007
Published: 28 November 2007
Skip to main content
© Yosefy et al; licensee BioMed Central Ltd. 2007
Received: 09 January 2007
Accepted: 28 November 2007
Published: 28 November 2007
This study aims to calculate the health effects and costs of a proposed national hypertension prevention and control program.
Interventions are based on experience from our two programs: 10-year period of Ashkelon Hypertension Detection and Control Program (AHDC Program) and the Israel Blood Pressure Control (IBPC) program. The costs of a nationwide program were calculated based on economic data, training staff levels, course frequency and unit costs.
Over the next 20 years, the program should decrease the risk in one-half of the treated hypertensive cases of the following ailments: cardiovascular events such as Acute Myocardial Infarction (AMI) and Unstable Angina Pectoris (UAP) by 16.0%, stroke by 41.2%, End stage renal disease (ESRD) by 50.0% and peripheral vascular disease (PVD) by 42.6%. In total, around 2,242 lives, 35,117 years of life or 24,433 disability adjusted life years will be saved due to decreased mortality.
Program costs amount to $352.7 million. However savings ($537.6 million), from reduced medical treatment ($444.3 million) and reduced pharmaceutical use ($93.3 million) as a result of morbidity decreases, exceed costs by $185.0 million.
The project which saves both lives and resources should be extended nation-wide to reach as wide a population as possible.
Cardiovascular (CV) risk factors such as hypertension, hyperlipidemia and diabetes, are insufficiently managed, according to current guidelines . This mismanagement occurs in spite of the accepted fact that reaching the recommended target goals of the various guidelines is an effective means to reduce morbidity and mortality from atherosclerosis and vascular events [2, 3]. Reasons for this shortfall have been described in many recent studies, showing that both the American [4, 5] and the European [6, 7] guidelines, are not truly implemented.
As in other developed nations, the burden of disease attributable to Hypertension, heart disease and stroke in Israel is considerable. These diseases are also among the most expensive diseases to treat . In response to these important issues of disease burden in terms of human suffering and treatment cost, we initiated the Ashkelon Hypertension Detection and Control Program (AHDC Program): a community approach to reducing cardiovascular mortality . During a 10-year period (1980–1990), we operated a professional doctor-nurse screening team, who were instructed to find those at high risk in control of their risk factors, in order to reduce CV morbidity and mortality. They examined 12,202 subjects, (mean age 51 +/- 7 years, range 20–65 years) accounting for 23.4% of the total regional population. High risk subjects underwent an intensive CV risk factor control program. Subjects (3,506 or 28.6%) were found to have one or more CV risk factors (hypertension, obesity, smoking, hypercholesterolemia). During an average of 2 years, follow-up BP, weight reduction, and smoking cessation remained statistically significant. Total cholesterol was unchanged. Over this period, the standardized mortality ratio (SMR) in the area for acute MI fell from 100 to 76 (P < 0.01), for CV disease from 129 to 107 (P < 0.0001), and for hypertension from 121 to 87 (P < 0.1 NS). The project saved many life-years at no additional net cost to society, and cost-effectiveness analysis showed positive results. Our community approach with mainly non pharmacological treatment was proven to be feasible and cost-effective in reducing CV morbidity and mortality. Subsequent we conducted the Israel Blood Pressure Control (IBPC) program , in the year 2000 in 30 general practice clinics in Israel. After 1 year of intervention in 4848 patients according to our method , there was a significant increase in control of risk factors. In hypertension this percentage raised from 29.0 to 46.7%, in control of body weight from 36.7 to 43.8%, in Low Density Lipoprotein (LDL) levels from 31.2 to 41.7% while the percentage of patients with glucose levels above 200 mg/dl dropped from 5.2 to 3.1%. Encouraged by these results we decided to propose a nationwide program that could also be extended to act in cooperation with neighboring countries. This study was initiated to help attain this goal and convince the health authorities, by presenting the costs and the utilities if the Ashkelon project were to be extended nationwide.
Interventions will be based on experience from the following two programs:
During a 10-year period (1980–1990 screening team, (MD and RN) acted to detect those at high risk to develop CV disease in order to reduce CV morbidity and mortality. They examined 12,202 subjects, (mean age 51 +/- 7 years, range 20–65 years) accounting for 23.4% of the total regional population. High risk subjects underwent an intensive CV risk factor control program .
The IBPC program was initiated in the year 2000. in this survey, we included patients from 30 general practice clinics across Israel, directed by specialists in family medicine, each seeing 1000–5000 patients. The authors presented the program to all available general practice clinics in Israel. Those physicians who showed readiness to join the program were enrolled on the study .
The costs of a nationwide program to reduce hypertension in Israelis aged 25–64 years old were calculated based on economic data, training staff levels, course frequency and unit costs (Appendix I). All costs are presented at April 2005 price levels using a 3% discount rate for the period 2005–2024. The staff education costs included educating two members of staff from each ambulatory clinic, operational costs (including tele-medicine) as well as two half-days a year of courses to maintain the level of staff education. Additional treatment costs of patients were calculated for the estimated 320,516 hypertensive cases [11, 12] aged 25–64 living in Israel not only for the initial year, based on $230 per hypertensive cases less the costs of current existing utilization of 4 visits per year by 25% of all hypertensive cases. The costs of maintenance therapy were added, amounting to a discounted sum of $208 per treated person. The program will incur extra drug costs as 50% of hypertensive cases will be prescribed the correct preventive regimen of Ace Inhibitor, Aspirin and Satins costing around $0.92 per day. The remaining 50% would be treated by non-pharmacological methods.
Disease treatment costs, both in the presence and absence of the intervention program were based on spreadsheet modeling combining unit treatment costs (Appendix I), epidemiological (Appendix II), treatment efficacy (Appendix II) and demographic factors (Appendix III). Similarly Disability Adjusted Life Years (DALY) lost from morbidity and mortality of diseases in the presence and absence of the program were calculated using the data contained in appendices II and III. All costs and utilities used a 3% discount rate for the period 2005–2024.
Key Epidemiologic Variables. [Acute Myocardial Infarctions (AMI), Cerebrovascular Accident (CVA), End Stage Renal Disease (ESRD) and M/F = Male/Female].
AMI (20 y %risk)
Incidence data was based on national survey data , 20 year age-specific AMI and UAP risks  were conservatively assumed to increase gradually by 1% annually to reflect an increase in other risk factors (eg: obesity). A 16% decrease in relative risk of AMI and UAP was assumed to occur in those 50% who were successfully treated, based on Israeli results [9, 10] that corroborated international findings [15–17]. One year post-AMI or Post-UAP case fatality rates  were adjusted by age and gender specific national background mortality rates to estimate the one year post-AMI excess mortality rates.
Incidence data and 20 year age-specific risks were based on Tel-Aviv stroke registry data . The 1.7 relative risk of strokes in hypertensive cases  translates into a potential 41.2% reduction (1–1.0/1.7) of strokes in successfully treated hypertensive cases. As one year post-hospitalization mortality rates were not available, in-hospital mortality rates were used to estimate excess mortality .
National age and gender specific incidence rates of dialysis patients and post renal transplants were obtained . The relative risks (of 2.00) of ESRD in hypertensive cases (with blood pressure >140/>90) and persons with high normal blood pressure (with blood pressure of 130 to 139/85 to 89 mmHg.) compared to persons with blood pressure <130/85 were based on a recent study of 316,675 adults . Incidence and relative risks were used to calculate 20 year age-specific risks for hypertensive cases. Under the assumption that 25% of hypertensive became normotensive cases and 25% became controlled (9,10), as defined to be below 140/90 mmHg according to the JNC VII guidelines , then the project will reduce ESRD by around 25.0%. One year ESRD mortality rates were used to attribute mortality due to hypertension .
Twenty year PVD risks were calculated by adjusting the age and gender specific 20 year CVA risks by the relative incidence of PVD to CVA . Two groups of 25% of treated hypertensive cases were assumed to become normotensive cases and controlled hypertensive cases with a risk reduction of 56.8% and 28.4% respectively for PVD . No excess mortality was attributed to PVD.
Disease, age and gender specific decreases in incidence due to the program were multiplied by the average disease specific treatment costs obtained either from patient surveys , the Ministry of Health's official price list or from the accounts department of the Barzilai Medical Center. (Appendix I) and the standard protocols for hospital and post-hospital treatment (Appendix III).
National survey data was obtained on disease specific treatment interventions . Costs were based on the budget backed-up assumption (of the Medical Technology Assessment Division of the Ministry of Health) that Cyphers would account for 40% of all simple stent and Cypher usage in the future.
Based on experience in Barzilai Medical Center, after 8 days average length of stay, 20% of stoke patients are discharged for three months average stay in a rehabilitation facility, after which they receive 3 months physiotherapy in a home setting. The remaining 80% of patients are discharged straight to three months physiotherapy in a home setting. Post hospitalization drug usage was assumed to be as for AMI.
Currently 5.9% of ESRD patients have received transplants  and the rest are on dialysis (88% on haemodialysis, 12% on Peritoneal dialysis) for an average of 5.05 years . Time on dialysis fell with age from 9.70 years (under 44 years) and 6.11 years (45–64) to 3.59 years (65+).
Treatment costs were based on an estimated of 3.5 outpatient visits a year over 10 years for treatment of PVD, based on current Israeli protocols and practice for monitoring PVD patients for any deterioration, every 3 to 4 months.
DALY gains (subject to a 3% discount rate) were calculated by aggregating the gains from decreased mortality and morbidity as a result of the program's intervention. The average health state valuation (HSV) is less than one, especially in older age groups where people have multiple dysfunctions. In order to reflect this, DALYs saved from averted deaths were calculated by multiplying the disease specific life year gains by the age and gender specific healthy adjusted life expectancies (HALE) .
Age-specific HSVs for disease conditions were obtained from the WHO global burden of disease study . Total DALYs loss from morbidity was obtained by multiplying age-specific disease incidence by the average disability loss (1-HSV) and by the time spent in the morbid condition, obtained from the WHO  or from the survey of dialysis patients for ESRD . For PVD disease duration was assumed to be for 10 years with a HSV of 0.975 (or disability weight of 0.025).
The net cost of the program was calculated by subtracting the reduced treatment costs (including reduced use of pharmaceuticals) from the running cost of the program. The major cost-effectiveness indicator is the Cost per DALYs ratio calculated by dividing the net cost of the project by the expected number of DALYs saved. This was only able to be calculated where the project was not cost-saving.
Projects are considered to be cost saving if their net cost is negative, very cost-effective if the cost per DALYs is less than the gross domestic product (GDP) per capita, cost-effective if the cost per DALY is between one and three times the GDP per capita and not cost-effective if the cost per DALYs exceeds three times the per capita GDP .
Impact of Hypertension Control Program on incidence and mortality (2005–2024). [Acute Myocardial Infarctions (AMI), Unstable Angina Pectoris (UAP), Cerebrovascular Accident (CVA), End Stage Renal Disease (ESRD) and Peripheral Vascular Disease (PVD)].
INCIDENCE Without Program
INCIDENCE With Program
MORTALTY Without Program
MORTALITY With Program
Life years and discounted DALYs saved (2005-2024) of hypertension program. [Acute Myocardial Infarctions (AMI), Unstable Angina Pectoris (UAP), Cerebrovascular Accident (CVA), End Stage Renal Disease (ESRD) and Peripheral Vascular Disease (PVD), Disability Adjusted Life Years (DALYs)].
A smaller quantity of DALYs (13,643 discounted and 17,801 undiscounted) will be saved from reduced morbidity. While the large majority of morbidity gains (around 61.5%) are due to decreased CVA incidence, AMI/UAP and ESRD account for a further 14.3% and 14.3% respectively (Table 3).
In total, around 32,369 discounted DALYs will be saved, around 57.9% being attributable to reduced CVA mortality. Program costs amount to $352.7 million, including $9.8 million for staff education, $19.2 million for telemedicine and operational expenses, $70.0 million for patient training during the first year of treatment, $66.8 for patient maintenance visits in subsequent years and $186.2 million for additional use of pharmaceuticals.
National Program Resource Costs and Savings:
Training Hypertensive cases
Extra Drug Costs
TOTAL PROGRAM COSTS
Reduced drug Use
NET PROGRAM SAVINGS
Sensitivity analyses are reported on the parameters that were thought ex-ante to have the most potential to influence the costs and utilities.
Using no discount rate, decreasing the attainable risk reduction for each disease by 30%, or decreasing the percentage of persons successfully treated from 50% to 40%, all resulted in the project still being cost-saving. Even if the overall attained decrease in hypertension were to fall by half to 12.5%, the project would still be cost saving (saving $29 million) in addition to saving 28,052 DALYs. The cost-effectiveness indicators were particularly sensitive to the number of years people received dialysis. If the average number of years were to drop from 5 to 4 years then total DALYs would be 32,212 and cost savings would fall to $123 million. If time on dialysis would increase to six years, then DALYs saved would rise to 32,565 and cost-savings rise to $247 million.
Coronary heart disease and Cerebrovascular Accident (CVA) prevention are priority areas for improving health, especially in developed nations. World-wide targets have been set for the reducing the major Cardiovascular (CV) risk factors: Blood Pressure (BP), diabetes, dyslipidemia, smoking and diet in relation to obesity [28, 29]. Nationwide projects are called for in order to significantly shift any nation's risk factor profile towards proven goals. Health policy and economic decisions makers in each country will be more likely to accept such large preventive projects not only if it the project is proven to save lifes and reduce morbidity, but also if it cost-effective (ie: having a low or even negative cost-per DALYs).
Because mortality was only attributable for AMI and UAP for a one year post-hospitalization period. Deaths after this timeframe were conservatively not attributed to coronary syndromes. Similarly deaths due to stroke were very conservatively estimated because stroke deaths were only attributed to in-hospital mortality.
We based life years gained on the age specific life expectancies, unadjusted for a decrease of the disease in question. For example, if the program eradicates x% of AMIs, then life years saved should be based on a life expectancies adjusted upwards because of the eradication of x% of deaths from AMIs.
our model based its duration of ESRD solely on data from patients receiving dialysis. It is likely that this underestimates the true amount of life years saved as it excludes data from patients who had undergone transplants and had not reverted to dialysis . DALYs benefits will be over or underestimated to the extent that fewer or more coronary events occur in February and March, the months on which the incidence survey was based . The question whether there is a need for such a nationwide program is clear form many studies [2–5] which prove the need for risk factor control in order to achieve CV disease prevention. Despite this, there is growing evidence that the guidelines are not implemented enough and there is under treatment [5–7], poor patient compliance, physician's unawareness of the new recommendations , and non adherence to updating treatment and to targets of treatments [20, 21].
Economic-Epidemiologic analysis has become a standard method for informing health policy makers, and is also helpful for individual physicians making medical decisions . For each of the various CV risk factors, especially hypertension on account of the high prevalence, and the need for long-term, perhaps lifelong, treatment, quite minor changes in antihypertensive practices can have a significant impact on healthcare budgets .
Our first previous study, the Ashkelon Hypertension Detection and Control Program (AHDC Program), showed that a community approach to reduce cardiovascular mortality , during a 10-year period (1980–1990), can affect reduce CV morbidity and mortality. During an average of 6.4 years, follow-up BP, weight reduction, and smoking cessation, the standardized mortality ratio SMRs for MI, CVD and hypertensive diseases fell significantly. The project saved many life-years at no additional net cost to society. Also in our second project: the Israel Blood Pressure Control (IBPC) program , there was a significant increase in the percentage of patents defined as "controlled" in hypertension, Low Density Lipoprotein (LDL) levels and in the percentage of patients with diabetes.
It is important to note the impact of decreased mortality by estimating the additional Disability Adjusted Years of Life (DALYs) and not only life years added.
Each hypertension related disease encounters many measurable obstacles which patients face during their rehabilitation, which in many cases are not completely resolved. These diseases can cause pain, suffering, increasing dysfunction, temporary or permanent unemployment accompanied by attendant life stresses and major reductions in self image and in family status. In Israel, AMI patients are automatically absent from work for three months of rehabilitation as the minimum even if there are no AMI complications. This example shows that on top of the treatment and secondary prevention cost of each AMI there is also a considerable socioeconomic burden that may influence the total balance of the cost utility analysis.
Although probably underestimated the program is expected to result in 8,749 fewer AMIs, 8,134 fewer PVDs, 8,074 fewer CVAs, 6,765 fewer UAPs and 2,316 fewer cases of ESRD (Table 2). Around half of the estimated total of 2,242 fewer deaths will be due to the reduction in cerebrovascular strokes (1,114 fewer deaths), while there will be 543 fewer deaths from AMIs, 440 fewer deaths from ESRD and 145 fewer deaths from UAP. These translates into a total of 35,117 life years saved.
For CVA the whole socioeconomic benefits are even more likely to have been underestimated, as the program may also prevent work disability. The 1.7 relative risk of strokes in hypertensive patients  translates into a potential 41.2% reduction of strokes and stroke sequlae in successfully treated hypertensive cases. A further source of underestimation of DALYs saved is that our model based its duration of ESRD solely on data from patients receiving dialysis. It is likely that this underestimates the true amount of life years saved as it excludes data from patients who had undergone transplants and had not reverted to dialysis .
We feel that, consideration should be given to extending such a potentially cost-effective project (which saves life years, morbidity and resources) to a nationwide basis, in order to encompass as many people as possible.
Our study shows that a national health education program to reduce cardiovascular risk factors, over the next 20 years, should succeed in decreasing the risk in one-half of AMI and UAP by 16.0%, CVA by 41.2%, CRF by 50.0% and PVD by 42.6%. The implementation of such a program nationwide is likely to save over two thousand lives, improve the quality of life in many more people and at the same time actually save $185.0 million in health care resources alone. It is conceivable that it may be extended not only throughout our country, but also to neighboring countries.
Appendix I: Key Economic Variables
Annual Discount Rate
Exchange Rate: Shekels to USD (2005)
Per capita GDP (USD 2003)
Number of Clinics in Israel
MOH Statistical Dept.
Staff per clinic receiving training
Full Days Training in initial year
Days training in subsequent years
Cost per ticipant-day of training course
Tele-management & Operational costs per patient.
Nurse/GP guidance per patient in first year of program
Unit Care Costs
Ambulatory visit to Nurse or Nutritionist
15 mins visit & 5 mins secretarial time
Ambulatory GP visit
15 mins GP & 5 mins secretarial time
Cost of OPV
25% of general hospital cost (30)
General Hospital Cost per day
ICU cost per day
estimated as four times general per diem cost (30)
Rehabilitation Hospital Cost per day
Home Physiotherapy Visit
Devices and Procedure costs
Barzelai Hospital Accounts Dept.
Barzelai Hospital Accounts Dept.
30, includes 1 day in ICU & 3 days in cardiac ward
Average cost of Catheterisation with Stent or cypher
30 estimated as four times general per diem cost
Kidney Transplant Cost
Peritonea dialysis cost
% receiving haemodialysis
Average annual Dialysis Cost
Post-MI or stroke drugs per day per patient- Males
Post-MI or stroke drugs per day per patient- Females
Post-UAP drugs per day per patient- Males
Post-UAP drugs per day per patient- Females
Correct preventive medication per day
14, Ace inhibitor, Aspirin & Statin
Inadequate preventive medication per day
14, Assumed only 1 correct drug taken
Appendix II: Key Epidemiologic Variables
Male life expectancy at age 60 (years)
Female life expectancy at age 60 (years)
Healthy Adjusted Life expectancy at age 60 males
Healthy Adjusted Life expectancy at age 60 females
% hypertensive cases, Males 25–44
% hypertensive cases, Males 45–54
% hypertensive cases, Males 55–64
% hypertensive cases, Females 25–44
% hypertensive cases, Females 45–54
% hypertensive cases, Females 55–64
% of persons treated successfully in program
% AMI in Males
20 year MI risk at age 25
20 year MI risk at age 30
20 year MI risk at age 35
20 year MI risk at age 40
20 year MI risk at age 45
20 year MI risk at age 50
20 year MI risk at age 55
20 year MI risk at age 60
20 year MI risk at age 65
Annual increase in MI risk over 20 years
Decrease in risk of MI in persons successfully treated
Health Status Valuation of MI patient
Years in decreased health state.
Unstable Angina Pectoris
% UAP in males
Decrease in risk of UAP in persons succesfully treated
Health Status Valuation of UAP patient
Years in decreased health state.
Mean age of stroke
% strokes in males
Relative Risk of Stroke in Hypertensive cases
Stroke rate per 1000 aged 45–55
Stroke rate per 1000 aged 55–64
Stroke rate per 1000 aged 65–74
Stroke rate per 1000 aged 75–84
Decrease in Stroke risk in persons successfully treated
Health Status Valuation of first ever stroke patient
Years in decreased health state.
Chronic Renal Failure
ESRD incidence rates per 100,000 Males 45–54
ESRD incidence rates per 100,000 Males 55–64
ESRD incidence rates per 100,000 Males 65–74
ESRD incidence rates per 100,000 Males 75+
ESRD incidence rates per 100,000 Females 45–54
ESRD incidence rates per 100,000 Females 55–64
ESRD incidence rates per 100,000 Females 65–74
ESRD incidence rates per 100,000 Females 75+
Decrease in ESRD risk in persons successfully treated
Health Status Valuation of ESRD
PVD incidence per 1000 male hypertensive cases
PVD incidence per 1000 female hypertensive cases
PVD Relative Risk in Hypertensive to Normotensive cases
PVD Relative Risk in Hypertensive to controlled cases
Attainable risk reduction Hypertensive to Normotensive cases
Attainable risk reduction Hypertensive to Controlled cases
% treated who become controlled cases
% treated who become normotensive cases
Appendix III: Key Utilization variables.
% of persons transferred to NPC of hypertension
Existing ambulatory visits in first year for hypertension
Existing annual maintenance visits for hypertension
Annual maintenance visits in program
Average time of visit (minutes)
GP visits in three years prior to AMI
% AMI that are catheterised
% AMI getting stent/baloon (primary PCI)
% AMI getting stent (late PCI)
% AMI getting CABG
% AMI being discharged after catheterisation
% AMI being discharged with no catheterisation
% AMI with Thrombolysis
% AMI rehospitalised for cardiac reasons in month
% ratio of cyphers to all cypher-stent usage
MOH Medical Technology Assesment Dept.
Ashkelon Hypertension Detection and Control
Acute Myocardial Infarction
Disability Adjusted Life Years
End stage renal disease
Gross domestic product
Health Status Valuation
Israel Blood Pressure Control
Ministry of Health
Low Density Lipoprotein
Peripheral vascular disease
Unstable Angina Pectoris
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.