Disclaimer: Health Quality Ontario uses a standardized costing method for its economic analyses of interventions. The main cost categories and the associated methods from the province’s perspective are as follows:
Hospital: Ontario Case Costing Initiative cost data are used for in-hospital stay, emergency visit and day procedure costs for the designated International Classification of Diseases (ICD) diagnosis codes and Canadian Classification of Health Interventions procedure codes. Adjustments may be required to reflect accuracy in estimated costs of the diagnoses and procedures under consideration. Due to the difficulties of estimating indirect costs in hospitals associated with a particular diagnosis or procedure, the secretariat normally defaults to considering direct treatment costs only.
Nonhospital: These include physician services costs obtained from the Ontario Schedule of Benefits, laboratory fees from the Ontario Schedule of Laboratory Fees, drug costs from the Ontario Drug Benefit Formulary, and device costs from the perspective of local health care institutions whenever possible or its manufacturer.
Discounting: For cost-effectiveness analyses, a discount rate of 5% is applied as recommended by economic guidelines.
Downstream costs: All numbers reported are based on assumptions on population trends (i.e., incidence, prevalence, and mortality rates), time horizon, resource utilization, patient compliance, health care patterns, market trends (i.e., rates of intervention uptake or trends in current programs in place in the Province), and estimates on funding and prices. These may or may not be realized by the system or individual institutions and are often based on evidence from the medical literature, standard listing references, and educated hypotheses from expert panels. In cases where a deviation from this standard is used, an explanation is offered as to the reasons, the assumptions, and the revised approach. The economic analysis represents an estimate only, based on the assumptions and costing methods that have been explicitly stated above. These estimates will change if different assumptions and costing methods are applied to the analysis.
The Toronto Health Economics and Technology Assessment (THETA) Collaborative was commissioned by Health Quality Ontario (HQO) to predict the long-term costs and effects along with the cost-effectiveness of strategies for the management and treatment of hypertension. The results of the economic analyses of the following strategies are presented: 24-hour ABPM and CBPM of hypertension. Additionally, this report reviews published economic evaluations of 24-hour ABPM and presents estimates of the budget impact of implementing the intervention for the following populations: WCH patients and any patient suspected of having hypertension.
Health Quality Ontario conducts full evidence-based analyses of health technologies being considered for use in Ontario. These analyses are then presented to the Ontario Health Technology Advisory Committee (OHTAC), whose mandate is to provide evidence-based examination of proposed health technologies in the context of existing clinical practice and provide advice and recommendations to Ontario practitioners, the broader health care system, and the Ministry of Health and Long-Term Care.
Hypertension occurs when either SBP, the pressure in the artery when the heart contracts, or DBP, the pressure in the artery when the heart relaxes between beats, are consistently high. Blood pressure that is consistently more than 140/90 mmHg (systolic/diastolic) is considered high. A lower threshold, greater than 130/80 mmHg (systolic/diastolic), is set for individuals with diabetes or chronic kidney disease.
In 2006 and 2007, the age-standardized incidence rate of diagnosed hypertension in Canada was 25.8 per 1,000 (450,000 individuals were newly diagnosed). During the same time period, 22.7% of adult Canadians were living with diagnosed hypertension.
A smaller proportion of Canadians are unaware they have hypertension; therefore, the estimated number of Canadians affected by this disease may be higher. Diagnosis and management of hypertension are important, since elevated BP levels are related to the risk of CVD, including stroke. In Canada in 2003, the costs to the health care system related to the diagnosis, treatment, and management of hypertension were over $2.3 billion (Cdn).
The objective of this economic analysis was to perform a cost-effectiveness analysis (CEA) and report costs associated with providing 24-hour ABPM to patients suspected of having hypertension, and for ongoing monitoring of BP in Ontario.
Economic Literature Review
A literature search was performed on August 4th, 2011 using OVID MEDLINE, MEDLINE In-Process and Other Non-Indexed Citations, OVID EMBASE, Wiley Cochrane, CINAHL, and Centre for Reviews and Dissemination/International Agency for Health Technology Assessment for studies published from 1948 to July week 4, 2011 for MEDLINE; and from 1980 to July week 31, 2011 for EMBASE. Included studies were those with full economic evaluations describing both costs and consequences of 24-hour ABPM for hypertension and BP monitoring; the same set of search keywords was used as for the clinical effectiveness systematic review in this report.
Several economic analyses have been performed related to home blood pressure monitoring (HBPM); however, 24-hour ABPM was not used as a comparator in these studies. (37
) The literature search found 2 cost analyses comparing 24-hour ABPM directly to CBPM (standard care) for hypertension diagnosis and treatment: Krakoff et al 2006 (41
) and Lovibond et al 2011 (41
); the latter also compared 24-hour ABPM to HBPM.
Krakoff et al calculated the cost savings for using 24-hour ABPM as a secondary diagnostic test, after an initially raised BP reading for patients. (41
) The costs of diagnosis and treatment were considered together, and cost savings ranged from approximately $85,000 (US) to $153,000 (US) per 1,000 patients when compared to CBPM. The range in savings was based on the assumption of the percentage of WCH patients confirmed to be hypertensive: $85,000 (US) per 1,000 patients would be saved if 20% of WCH patients were confirmed to be hypertensive; $153,000 (US) per 1,000 patients would be saved if 5% of WCH patients were hypertensive.
Lovibond et al presented a CEA based on a Markov economic model comparing 3 strategies for the accurate diagnosis and treatment of hypertension: CBPM, HBPM, and (24-hour) ABPM. (42
) The United Kingdom population of interest was aged 40 years or older with a screening blood pressure measurement greater than 140/90 mm Hg, with an average risk of CVD equivalent to that of the general population. Results of the study suggested ambulatory monitoring (i.e., HBPM or 24-hour ABPM) for hypertension after an initially raised CBPM reading would reduce the number of misdiagnoses and save costs. The incremental cost-effectiveness ratio (ICER) ranged from about £3,000 (GBP) to £26,000 (GBP) per quality-adjusted life year (QALY) for 24-hour ABPM compared to CBPM, based on the initial age of the patient. It was further shown that 24-hour ABPM dominated CBPM and HBPM by providing a greater increase in QALYs, but for lower costs for the average patient.
Primary Economic Evaluation
The published economic evaluations identified in the literature review addressed the intervention of interest (i.e., 24-hour ABPM). However, none of these published studies took a Canadian perspective. Due to these limitations, a primary economic evaluation of 24-hour ABPM was conducted.
As stated above, the intervention of interest is 24-hour ABPM. The primary evaluation will compare the 2 strategies of providing 24-hour ABPM or CBPM or “usual care” for hypertension. Twenty-four-hour ABPM was defined as the use of an ABPM measuring device for the diagnosis of hypertension, with a maximum of 3 diagnostic physician visits per year. Conventional blood pressure monitoring was defined as in-office measurement of BP for suspected hypertensive patients, with a maximum of 5 physician visits per year for the diagnosis of hypertension. Both strategies include subsequent, long-term consequences of developing the following CVDs: a) CHD (includes coronary death, MI, coronary insufficiency, angina); and b) cerebrovascular disease (referred to here as “stroke”, but includes ischemic stroke, hemorrhagic stroke, and transient ischemic events).
The target population of this economic analysis is patients suspected of having hypertension (i.e., management of hypertensive patients), which includes WCH patients, and aged 45 years or older with an average risk of CVD similar to the general population in Ontario.
The primary analytic perspective of the CEA is that of the Ontario Ministry of Health and Long-Term Care.
Economic Analysis Method
The current economic analysis is a cost-effectiveness analysis of 24-hour ABPM versus CBPM using health-related quality of life (HRQOL) to calculate the ICER in Canadian dollars (Cdn) per QALY. More specifically, as the current CEA analyzes QALYs gained or lost by patients as the main measure of effect, the CEA is developed as a cost utility analysis.
Discounting and Time Horizon
Costs and outcomes (QALYs) were discounted at a 5% annual rate as recommended by economic guidelines. (43
) The economic model is based on an annual cycle and aggregates patient costs and outcomes over their lifetime.
Variability and Uncertainty
Variability and uncertainty in the model were assessed using a probabilistic sensitivity analysis. Model parameter uncertainty was assessed by assigning distributions around the point estimate and results were presented in the form of probability of cost-effectiveness by ceiling ratio, i.e., willingness to pay values.
The findings of this economic analysis cannot be generalized to all patients with hypertension. They may, however be used to guide decision making about the specific patient populations addressed in the trials investigated at HQO. Note that 2 scenarios are presented in the analysis which examine the option of providing 24-hour ABPM to patients only with BP measurements that are raised or not in control, or providing 24-hour ABPM to all patients suspected of hypertension annually, but limited to 1 test per year per patient.
The model used in the CEA is Markov simulation, a decision analytic. shows a schematic representation of the model with 4 health states:
Schematic Diagram of the Decision-Analytic Markov Model Evaluating Twenty-Four-Hour ABPM
- alive with BP monitoring;
- cerebrovascular disease (stroke); and
The arrows indicate possible transitions between the 4 health states and the 3 decision-tree diamonds represent events which transition patients from health state 1 to the remaining 3 health states. The parameters informing the transition probabilities are taken from the current HQO clinical effectiveness review and are summarized in . Transitions from health state 1 to “dead” were informed by Ontario-specific life tables pooled from 2005 to 2007 as derived by Canadian Human Mortality Database and published by Statistics Canada. (44
Epidemiologic Parameters Used in the Cost-Effectiveness Analysis
Model Input Parameters
A number of different model input parameters were used to populate the model. These include variables used to model the natural history of the disease and variables used to modify the natural history model to account for the treatment effects and costs of 24-hour ABPM and CBPM.
Natural History Model Input Parameters
The main outcomes considered in this CEA were taken from the current HQO clinical effectiveness review and consist of the following as summarized in : control of BP, drug-related outcomes (i.e., patients who stopped anti-hypertensive drug therapy, change in drug intensity/dosage), and CVD-related outcomes (i.e., patients having any fatal or non-fatal CVD event including MI or stroke). In order to simplify the analysis, MI and stroke events were chosen as representative of CHD and cerebrovascular events, respectively (i.e., only MI and stroke risk and health states were modelled). The 3 main RCTs identified in the review were used to estimate the relative risk of CHD- or cerebrovascular disease–related events for 24-hour ABPM. (27
) However, the relative risks in the RCTs were estimated specifically for the MI or stroke population; it is an assumption in the current CEA that the relative risk parameters for MI or stroke are representative of those for CHD and cerebrovascular disease, in general.
The 10-year risk of CVD, CHD, or stroke listed in were taken from de Oliveira et al and derived from an Ontario-based study examining the effect of the Heart and Stroke Foundation of Ontario’s (HSFO’s) Hypertension Management Initiative (HMI) on the management of hypertension in primary care physician offices. (45
) The 10-year risk was calculated specifically from the Framingham risk score according to the framework of d’Agostino et al (46
) and used risk-factor values from the Ontario population assigned to “standard care” (CBPM) in the study by de Oliveira et al. (45
) The 10-year risk of CVD shown in is the average risk for the population of interest, whereas the actual 10-year risk values used in the economic model were age-specific (5-year age groups), but with average (population) values used for the other risk factors (i.e., BP of 134.4 mmHg, total cholesterol of 4.26 mmol/L, high-density lipoprotein cholesterol of 1.42 mmol/L, and the proportion of diabetics and smokers being 21.6% and 9.1%, respectively).
Clinical Model Input Parameters
summarizes the HRQOL utilities used for the current CEA, based on estimates derived from the CEA performed by NICE (26
) and Lovibond et al (42
), with references to Kind et al (48
), Meslop et al (49
), Tengs et al (50
), and Goodacre et al (51
). An average utility was used for health states 2 (i.e., CHD) and 3 (i.e., stroke), and was calculated as the mean of MI and other CHD utility values, and the mean of stroke and population norm utility values, respectively. The utility associated with health state 1 (i.e., alive with BP monitoring) was assigned to be the same as the general population norm. Also note that the utility values listed in represent patient population preferences in the United Kingdom and United States, which are unlikely to be significantly different from the preferences of the general population in Ontario.
Health-Related Quality of Life and Utilities Used in the Cost-Effectiveness Analysis
The cost of the 2 strategies compared for the CEA (i.e., 24-hour ABPM vs. CBPM) include physician, hospital, ambulatory monitoring device, and drug costs that are consistent with the Ministry perspective taken in the economic analysis. The costs are summarized in the tables below.
shows the annual physician costs anticipated for 24-hour ABPM and CBPM through consultations with clinical experts. Approximately 5 visits to the physician office over 6 months are often necessary for the initial, accurate diagnosis of hypertension for suspected patients using standard CBPM. The effect of using 24-Hour ABPM for the diagnosis of hypertension would reduce the number of in-office physician visits required in the first year by about 2 visits (i.e., about 3 visits would be required for 24-hour ABPM). As shown in , the difference in cost was calculated to be approximately $51.80 (Cdn) less than CBPM on average per patient for the first year of diagnosis. The physician assessment fee codes were taken from the Ontario Schedule of Benefits for Physician Services and the fee code for the interpretation of the 24-hour ABPM test was taken from the Saskatchewan Payment Schedule for Insured Services Provided by Physicians. (52
First-Year Costs of Physician Assessments for Hypertension for CBPM and Twenty-Four-Hour ABPM
Through consultations with device manufacturers, the associated costs of providing 24-hour ABPM to patients for diagnosing and monitoring hypertension were determined and are summarized in . The device costs include the actual cost of the device, analysis software, reusable cuffs and covers, and an annual maintenance contract to service the devices as necessary. shows device costs for 50 devices; the distributed cost is approximately $2,557 (Cdn) per individual device. Clinical experts suggested that two 24-hour ABPM devices could be used to diagnose or monitor patients by providing between 1.5 to 3.5 tests per week over 50 weeks per year, depending on whether the clinic would service a “low” or “high” number (volume) of patients, respectively. As a result, the average cost per test is estimated to range from about $68 to $29 (Cdn) (i.e., approximately $5,114 (Cdn) for 2 devices distributed over 75 to 175 tests per year).
Annual Device Costs per Patient for Twenty-Four-Hour ABPM
The costs associated with ongoing monitoring of BP and treatment of hypertension for CBPM and 24-hour ABPM are shown in . Physician, hospital, and drug costs were obtained from the study by de Oliveira et al (45
) and based on 30-patient-day costs calculated according to the “phase of care” method. Briefly, in this approach, the total costs per patient are divided into distinct phases over time as characterized by different patterns of resource use, which produce time-dependent cost estimates. In the case of hypertension, “stable” and “pre-death” cost phases were used corresponding to a period of relatively constant costs before death and another period of high costs just prior to death, respectively. In the current Markov model, the average costs associated with the “stable” period are accrued annually, with the costs of “pre-death” added to the annual costs in the year of death of the patient. The “initial” period for physician costs represent the costs of diagnosis of hypertension in the first year and are added only once for each patient; the “BP reassessment” cost phase is applicable for physicians reassessing patients when BP readings are raised or out of control (necessitating a possible change in drug intensity or regimen); the “stable” phase costs include other costs of patient care associated with hypertension and are accrued annually. A common stable phase cost of $223 (Cdn) was included in the analysis for other patient care costs not specific to BP monitoring, with a corresponding common pre-death cost of $3,352 (Cdn) for care not specific to BP monitoring. The costs associated with 24-hour ABPM listed in were taken from costs observed for the HSFO’s HMI program in Ontario. Whereas this is a limitation in cost estimation, the costs of hypertension management (i.e., monitoring and treatment) using 24-hour ABPM are considered to be comparable to those of the HMI program. The cost of the CBPM device was omitted, as regular CBPM equipment is already paid for by the clinic and is not part of the Ministry analytic perspective.
Annual Physician, Hospital, Device, and Drug Costs Associated With CBPM and Twenty-Four-Hour ABPM
In order to approximate the associated increase or decrease in drug cost with changing drug intensity score, a cost “multiplier” was calculated and used in the simulation model to represent the average annual effect of a decrease in intensity score for 24-hour ABPM. Specifically, the average drug costs for CBPM and 24-hour ABPM were taken from the HMI study by de Oliveira et al (45
) (i.e., $23.11(Cdn) and $20.49 (Cdn), respectively) and divided by the corresponding mean baseline intensity score as reported by Staessen et al (31
). This multiplier was used to calculate new drug costs associated with changes in intensity score for a given treatment year.
The CEA results below are presented for 2 modelling assumptions related to the frequency of follow-up or continuing use of 24-hour ABPM devices for monitoring BP. The first scenario accrues the device costs (i.e., physician interpretation and test costs) only for patients observed to have a raised BP reading, or who may require changes in drug intensity or regimen; the second scenario accrues 24-hour ABPM device costs annually for patients (i.e., maximum of 1 physician interpretation per year).
Cost-effectiveness Analysis Results
The cost-effectiveness of providing 24-hour ABPM for the management of hypertension in Ontario is show below in . It can be seen that 24-hour ABPM is very similar to CBPM when it is provided to patients only when a raised BP reading is observed or BP is not in control (i.e., drug dosage or regimen review required). Whereas the average lifetime costs of 24-hour ABPM are nearly identical to CBPM (i.e., cost difference of –$4 Cdn), the intervention provides greater effect (i.e., +0.135 QALYs) compared to CBPM and is a cost-effective strategy. In the case of providing 24-hour ABPM for the management of all (suspected) hypertensive patients annually, the results suggest it is still cost-effective, with an ICER of $4,160 (Cdn) per QALY (i.e., the ICER is well below the standard $50,000 per QALY threshold).
Annual Physician, Hospital, Device, and Drug Costs Associated With CBPM and Twenty-Four-Hour ABPM
A probabilistic sensitivity analysis was performed that allowed all clinical and cost parameters to vary according to specific probability distributions. Clinical and epidemiologic parameters found in were varied according to their 95% CIs, with proportions modelled as being normally distributed and relative risk estimates modelled with log-normal distributions. The costs of providing 24-hour ABPM (i.e., costs listed under “Hypertension (BP Monitoring)” in for the “24-Hour ABPM” scenario) were modified in the sensitivity analysis by increasing (decreasing) the listed cost by 50%; costs related to “Coronary Heart Disease (CHD)” or “Cerebrovascular Disease (Stroke)” were unmodified. In both scenarios considered above, the strategy of providing 24-hour ABPM to patients was cost-effective in approximately 97% of the randomly sampled parameter combinations at the standard $50,000 per QALY threshold willingness-to-pay. More specifically, providing a 24-hour ABPM test annually to all patients, or providing the test only for the investigation of raised BP readings, resulted in marginally increased costs (or cost savings) for patients followed over a lifetime in Ontario in 97% of cases.
Budget Impact Analysis – Ontario Perspective
A budget impact analysis of providing 24-hour ABPM to hypertension patients was calculated over the next 5 years (i.e., fiscal years FY2011–FY2016) to estimate the economic burden in Ontario. All costs are reported in current 2011 Canadian dollars. The projected Ontario population from 2011 to 2015 as developed by the Ontario Ministry of Finance was used in the current budget impact analysis for Ontario. (54
) The projected population is presented in together with the estimated hypertensive (high BP) cases expected to benefit from using 24-hour ABPM. The prevalence of high BP in Ontario was estimated as being 23.7% for the population aged 45 to 64 years, and 49.7% for the population aged 65 years or older. (55
) The incidence of diagnosed hypertension in Canada was used to estimate the number of cases in Ontario—approximately 2.21% of the population aged 20 years or older. (6
Ontario Hypertensive Population Expected to Benefit From Twenty-Four-Hour ABPM
The costs are presented below for 2 scenarios:
- provision of 24-hour ABPM for monitoring all (suspected) hypertensive patients annually, and
- provision of 24-hour ABPM for the monitoring of only patients with raised CBP readings annually.
The costs were taken from the current CEA and represent mean scenario cost differences between 24-hour ABPM and CBPM. shows the budget impact of using 24-hour ABPM for the 2 age groups mentioned above (i.e., patients aged 45–64 years, and 65+ years). The costs for the prevalent hypertensive population are distributed across the first 3 years (i.e., 2011, 2012, 2013), with the incident population being the only new cases starting in the fourth year (i.e., 2014).
Ontario Budget Impact (in Millions of Canadian Dollars) of Providing Twenty-Four-Hour ABPM Compared to CBPM
The first scenario, in which patients would receive 24-hour ABPM monitoring only for raised BP readings, was shown to save costs in the CEA. As a result, it is anticipated that approximately $19 million (Cdn) could be saved annually (i.e., average annual savings) in Ontario using ABPM. The largest cost savings would be about $34 million (Cdn) and $8 million (Cdn) in physician and drug spending, respectively. However, for the second scenario, providing 24-hour ABPM to the hypertensive population annually, the increased cost is anticipated to be approximately $37 million (Cdn) per year in Ontario. The increased cost is made up mostly of increased device (24-hour ABPM test) expenditures, which include the cost and maintenance of the device and a professional fee for the test’s interpretation.
In order to estimate the budget impact of providing 24-hour ABPM to only WCH patients, the numbers shown in Table 10 are multiplied by the prevalence of WCH among hypertensive patients. According to Staessen et al, (31
) approximately 20% to 35% (average 27.5%) of the hypertensive population can be considered to have WCH.
Twenty-four-hour ABPM was found to be cost-effective and potentially cost saving from the perspective of the Ontario Ministry of Health and Long-Term Care. However, several limitations exist in the current economic analysis:
- All cardiovascular health states were not represented in the Markov model, which may increase costs for certain diseases influenced by ongoing patient BP monitoring.
- Estimates of clinical effectiveness were mixed interchangeably, using 5-year and 6-month RCT data (i.e., used both short-term and long-term effects for model parameters affecting cost distributions).
- The cost of the 24-hour ABPM device (i.e., the cost of the device and maintenance only) was calculated as an Ontario Schedule of Benefits “technical fee” and assumes the device will be paid for within 1 year of its first use.
- The health state utilities may not be representative of Ontario and instead reflect specific preferences of hypertensive patients in the United Kingdom.
- CVD risk may not be representative of risks associated with hypertension in Ontario (i.e., applicability of the Framingham risk study and regression equations to Ontario).