Comparative Cost-Effectiveness of Conservative or ...hyper.ahajournals.org/content/hypertensionaha/68/1/88.full.pdf · Pressure Intervention Trial (SPRINT) found that targeting an

88

For ≈4 decades, the Joint National Committee (JNC) on the Detection, Evaluation and Treatment of High Blood

Pressure (BP) supported formulation of US hypertension treat-ment guidelines. From 1977 to 2003 (JNC1 to JNC7), the guidelines progressively lowered diagnostic thresholds and treatment targets, effectively expanding the treatment-eligible population. The 2014 hypertension guidelines (referred to here as JNC8) recommended higher BP goals compared with JNC7, so that ≈5.8 million fewer adults were eligible for antihyper-tensive medication treatment.1,2 JNC8’s less intensive BP goal recommendations for patients aged ≥60 years and those with diabetes mellitus or chronic kidney disease (CKD) provoked

controversy and uncertainty.3 More recently, the Systolic Blood Pressure Intervention Trial (SPRINT) found that targeting an intensive systolic BP (SBP) goal of 120 mm Hg in patients with high cardiovascular disease (CVD) risk and baseline SBP ≥130 mm Hg reduced CVD events by 25% and all-cause mor-tality by 27%, compared with a 140-mm Hg goal.4

The objective of this study was to project the potential value of adding intensive SBP goals in high-risk patients to the JNC7 or JNC8 guidelines in a contemporary population of untreated hypertensive individuals aged 35 to 74 years. We also assessed if the incremental cost-effectiveness of intensive BP goals remained sensitive to the costs of more frequent monitoring or

Abstract—The population health effect and cost-effectiveness of implementing intensive blood pressure goals in high-cardiovascular disease (CVD) risk adults have not been described. Using the CVD Policy Model, CVD events, treatment costs, quality-adjusted life years, and drug and monitoring costs were simulated over 2016 to 2026 for hypertensive patients aged 35 to 74 years. We projected the effectiveness and costs of hypertension treatment according to the 2003 Joint National Committee (JNC)-7 or 2014 JNC8 guidelines, and then for adults aged ≥50 years, we assessed the cost-effectiveness of adding an intensive goal of systolic blood pressure <120 mm Hg for patients with CVD, chronic kidney disease, or 10-year CVD risk ≥15%. Incremental cost-effectiveness ratios <$50 000 per quality-adjusted life years gained were considered cost-effective. JNC7 strategies treat more patients and are more costly to implement compared with JNC8 strategies. Adding intensive systolic blood pressure goals for high-risk patients prevents an estimated 43 000 and 35 000 annual CVD events incremental to JNC8 and JNC7, respectively. Intensive strategies save costs in men and are cost-effective in women compared with JNC8 alone. At a willingness-to-pay threshold of $50 000 per quality-adjusted life years gained, JNC8+intensive had the highest probability of cost-effectiveness in women (82%) and JNC7+intensive the highest probability of cost-effectiveness in men (100%). Assuming higher drug and monitoring costs, adding intensive goals for high-risk patients remained consistently cost-effective in men, but not always in women. Among patients aged 35 to 74 years, adding intensive blood pressure goals for high-risk groups to current national hypertension treatment guidelines prevents additional CVD deaths while saving costs provided that medication costs are controlled. (Hypertension. 2016;68:88-96. DOI: 10.1161/HYPERTENSIONAHA.115.06814.) • Online Data Supplement

Key Words: cardiovascular diseases ■ cost-benefit analysis ■ hypertension ■ guideline ■ policy

Received November 20, 2015; first decision December 22, 2015; revision accepted April 15, 2016.From the Division of General Medicine, Department of Medicine, Columbia University Medical Center, New York, NY (N.M., C.N.K.-L., K.Y.T.,

A.E.M.); Department of Evidence Based Medicine, Cardiovascular Institute and Fu Wai Hospital of the Chinese Academy of Medical Sciences, Beijing, China (C.H.); Department of Epidemiology, Fuwai Hospital, Peking Union Medical College and Chinese Academy of Medicine Science, Beijing, China (C.H.); George Institute for Global Health, Sydney, New South Wales, Australia (A.R.); Health Economics & Health Technology Assessment, Institute of Health and Wellbeing, College of Medical, Veterinary and Life Sciences and College of Social Sciences, University of Glasgow, Glasgow, United Kingdom (C.N.K.-L.); Division of General Medicine, Department of Medicine, University of California at San Francisco (P.G.C., K.B.-D); and College of Physicians and Surgeons, Columbia University, New York, NY (L.G., A.E.M.).

The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA. 115.06814/-/DC1.

Correspondence to Andrew E. Moran, Division of General Medicine, Department of Medicine, Columbia University Medical Center, 622 W 168th St, PH 9 E 105, New York, NY 10032. E-mail [email protected]

Comparative Cost-Effectiveness of Conservative or Intensive Blood Pressure Treatment Guidelines

in Adults Aged 35–74 YearsThe Cardiovascular Disease Policy Model

Nathalie Moise, Chen Huang, Anthony Rodgers, Ciaran N. Kohli-Lynch, Keane Y. Tzong, Pamela G. Coxson, Kirsten Bibbins-Domingo, Lee Goldman, Andrew E. Moran

© 2016 American Heart Association, Inc.

Hypertension is available at http://hyper.ahajournals.org DOI: 10.1161/HYPERTENSIONAHA.115.06814

Health Economics

by guest on May 25, 2018

http://hyper.ahajournals.org/D

ownloaded from

by guest on M

ay 25, 2018http://hyper.ahajournals.org/

Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

by guest on M


Dow

nloaded from



ownloaded from

mailto:[email protected]

http://hyper.ahajournals.org/

























Moise et al Comparative Cost-Effectiveness of Hypertension Guidelines 89

high medication prices. Patients aged ≥75 years were excluded from this analysis because of uncertainty about the tradeoff of risks and benefits of antihypertensive therapy in that population.

Methods

CVD Policy ModelThe CVD policy model is a computer-simulation, state-transition (Markov cohort) model of incidence, prevalence, mortality, and costs of CVD in US adults (Methods section in the online-only Data Supplement).5,6 Means or proportions and joint distributions of risk fac-tors, including BP, cholesterol, hypertension medication use, smoking, diabetes mellitus, and CKD status were estimated from pooled National Health and Nutrition Examination Surveys (NHANES) 2007 to 2010. Default multivariate stroke and coronary heart disease incidence func-tions were estimated in original Framingham Heart Study analyses.

The CVD policy model predicts life years, CVD events (myocardial infarction and stroke), coronary revascularization procedures, CVD mortality (stroke [International Classification of Diseases-10 codes I60–I69], coronary heart disease [I20–I25 and two thirds of I49, I50, and I51], hypertensive heart disease deaths [I11.0 and I11.9]), and non-CVD deaths (remainder of International Classification of Diseases codes). Reductions in heart failure deaths because of hypertension treat-ment were calculated by adding prevented ischemic heart failure deaths (I50 with coronary heart disease) and hypertensive heart disease deaths (I11.0 and I11.9; Methods section in the online-only Data Supplement).

Model Calibration and ValidationDefault model input parameters were calibrated, so that 2010 coronary heart disease and stroke incidence predictions matched hospitalized myocardial infarction and stroke rates observed in the 2010 National Hospital Discharge Survey, and mortality predictions were within 1% of age-specific 2010 CVD vital statistics mortality rates. Age- and sex-specific SBP and diastolic BP β-coefficients from the Prospective Studies Collaboration7 were calibrated, so that CVD Policy Model age-weighted relative risks with BP reduction fell within the 95% con-fidence interval of the overall relative risk estimates for the same BP reduction observed in a large meta-analysis of randomized controlled hypertension treatment trials (Methods section in the online-only Data Supplement; Tables S1–S3 in the online-only Data Supplement).8 To test predictive validity, we populated the model with the Systolic Hypertension in the Elderly Program (SHEP) trial cohort and simu-lated the BP reduction achieved in the active treatment arm of the trial for 5-years of follow-up. Our estimates accurately reproduced the risk reduction observed in the original trial (Table 1; Methods section in the online-only Data Supplement; Table S4).9

Model InputsJNC7 recommended a goal BP <130/80 mm Hg for diabetes mellitus or CKD and BP <140/90 mm Hg for all others. JNC8 recommended a goal <140/90 mm Hg for diabetes mellitus or CKD, diastolic BP <90 mm Hg if age is <60 years, and BP <150/90 mm Hg if age is ≥60 years and without diabetes mellitus or CKD. On the basis of SPRINT, intensive interventions were applied to adults aged ≥50 years with pretreatment SBP ≥130 mm Hg and either existing CVD, CKD, or 2013 American Heart Association/American College of Cardiology Pooled Cohorts 10-year CVD risk ≥15%. Using these categories and BP and treatment status information from NHANES, we estimated the number of currently untreated US adults eligible for treatment under JNC7 and JNC8 with and without the intensive intervention in selected high-CVD risk individuals (Table S5).

BP change caused by antihypertensive medications was determined by pretreatment BP and the number of standard doses of medications needed to reach the guideline BP goal according to a trial-based for-mula.10 BP changes were calculated based on pretreatment BP, age, and sex. We assumed the same BP reduction per standard dose of the main drug classes and did not include non-BP–lowering benefits of specific agents (Table 1; Methods section in the online-only Data Supplement; Table S5).8,10

We expected that CVD risk is reduced log linearly in relation to BP reduction (mm Hg) down to SBP 120 mm Hg in high-CVD risk patients in intensive strategies, 130/80 mm Hg in select JNC7 groups (diabetes mellitus or CKD), and SBP 140 mm Hg in those aged 60 to 74 years but without diabetes mellitus or CKD.8,10,11 Hypertension treatment costs included monitoring, side effect, and averaged wholesale drug costs. Quality of life penalties were applied for side effects.10 A medication adherence rate of 75% estimated in a meta-analysis of clinical trials was assumed because it corresponded to risk reduction associated with treatment estimated in the same meta-analysis (Table 1).8

A status quo simulation projected CVD events, CVD deaths, heart failure deaths, costs, and quality-adjusted life years (QALYs) for adults aged 35 to 74 years with untreated hypertension from 2016 to 2026. Adults aged ≥75 years were excluded from this analysis be-cause of variable medication-related adverse event risk in this group.12 Guideline simulations modeled treatment according to JNC7 or JNC8. Incremental to JNC7 or JNC8, intensive strategies targeted an SBP of 120 mm Hg goal in high-CVD risk patients, limiting to 5 antihyperten-sive drugs maximum.4 Incremental cost-effectiveness ratios (ICERs) were calculated as change in costs divided by incremental change in QALYs. ICERs <$50 000 per QALY gained were considered cost-ef-fective, ≥$50 000 and <$150 000 of intermediate value, and ≥$150 000 of low value.13 All analyses were approached from a payer’s perspec-tive. Future costs and QALYs were discounted at 3% per year.

Subgroup and Sensitivity AnalysesJNC7 and JNC8 with and without intensive treatment in selected high-CVD risk individuals were compared within age groups. One-way sensitivity analyses assessed cost-effectiveness assuming lower and upper uncertainty boundaries of the main inputs, including increased monitoring costs for the intensified treatment strategies (Table 1). We also modeled medication adherence as low as 40%.14 Main analyses did not include patients with treated but uncontrolled hypertension because it was not clear what proportion of poor control was because of underuse of combination therapy, poor adherence, or resistant hypertension.6 Nonetheless, we repeated the analyses in the entire population with uncontrolled hypertension, including previ-ously treated and uncontrolled hypertension.

Probabilistic AnalysesProbabilistic (Monte Carlo) simulation sampled across uncertainty distributions of antihypertensive drug BP-lowering effectiveness, CVD relative risk reduction with treatment, quality of life penal-ties, costs related to side effects, and drug and monitoring costs. Uncertainty distributions were randomly sampled 1000×, and 95% uncertainty intervals were calculated for all model outputs. Cost-effectiveness acceptability curves were constructed to illustrate the probability that each hypertension treatment strategy would be cost-effective at different willingness-to-pay thresholds.

Results

Main and Probabilistic ResultsCompared with no treatment, JNC8 would increase the annual number of newly treated adults aged 35 to 74 years by ≈12 mil-lion and would avert ≈65 000 CVD events and 17 000 CVD deaths annually. Compared with JNC8, JNC7 would recom-mend treatment for nearly twice the number of untreated patients (21 million) and add substantial treatment costs but would avert 24 000 additional CVD events and 5000 additional CVD deaths annually (Table 2). Incremental to JNC8, JNC8 plus intensive treatment in selected high-risk groups (JNC8+intensive) would prevent 43 000 additional annual CVD events and 15 000 CVD deaths. Incremental to JNC7, JNC7+intensive would lead to 35 000 fewer annual CVD events and 14 000 fewer CVD deaths. Total annual heart failure deaths avoided ranged from



ownloaded from


90 Hypertension July 2016

Table 1. Main Assumptions for the Comparative Effectiveness Analysis of Adding Intensive Blood Pressure Goals for High-Risk Patients to Current US Hypertension Treatment Guidelines

Variable

Estimate (Range in Main Estimate If a Variation Assumed According to Age and Sex)

SourcesMain Lower Upper

Effectiveness Law et al8,10; Systolic Hypertension in the Elderly Program Trial9

Average RR per 5-mm Hg reduction in DBP or 10-mm Hg reduction in SBP (age, 35–59 y)*

CHD 0.74 (0.71–0.77) 0.71 (0.66–0.75) 0.78 (0.76–0.79)

Stroke 0.66 (0.62–0.70) 0.61 (0.56–0.67) 0.71 (0.69–0.73)

All-cause mortality 0.89 (0.83–0.89) 0.78 0.95

Average RR per 10-mm Hg reduction in SBP or 5-mm Hg reduction in DBP (age, 60–74 y)*

CHD 0.78 (0.76–0.81) 0.76 (0.74–0.78) 0.80 (0.79–0.83)

Stroke 0.71 (0.69–0.77) 0.68 (0.66–0.70) 0.77 (0.73–0.84)

All-cause mortality 0.92 0.81 1.02

Average SBP-lowering effect, mm Hg† Law meta-analysis8,10

Stage 2 hypertension

Pretreatment ≥160 mm Hg, JNC7 diabetes mellitus or CKD target 130 mm Hg or intensive intervention in age ≥50 y with existing CVD, CKD or CVD risk ≥15%, target 120 mm Hg (4–5 standard dose medications)

38.4–42.3 32.8–42.3 44.0–48.3

JNC8 age <60 y or JNC7 no risk factors, pretreatment ≥160 mm Hg, target 140 mm Hg (3–4 standard dose medications)

31.0–34.7 26.0–29.4 36.0–39.9

JNC8 age ≥60 y, pretreatment ≥160 mm Hg, target 150 mm Hg (2–3 standard dose medications)

22.1–24.2 18.1–18.9 27.2–29.2


JNC7 diabetes mellitus or CKD, pretreatment 140–159 mm Hg, target 130 mm Hg (2–3 standard dose medications)

16.1–19.2 11.7–15.2 19.3–25.1

JNC8 age <60 y or JNC7 no risk factors, pretreatment 140–159 mm Hg, target 140 mm Hg (0.5–2.0 standard dose medications)

7.9–10.9 5.9–8.3 9.9–13.4

Age ≥60 y, pretreatment 150–159 mm Hg, target 150 mm Hg (0.5 standard dose medications)

7.1 3.2 11.0

Intensive intervention in age ≥50 y with existing CVD, CKD or Framingham risk ≥15%, pretreatment ≥130 mm Hg, target 120 mm Hg

25.5–29.0 22.6–25.8 28.3–32.1

Prehypertension 6.7 4.7 8.7

JNC7 diabetes and/or CKD, pretreatment 130–139 mm Hg target 130 mm Hg (0.5 standard dose medications)

DBP-lowering effect, mm Hg† Law meta-analysis8


JNC7 diabetes mellitus or CKD, pretreatment ≥100 mm Hg, target 80 mm Hg (3 standard dose medications)

19.2–21.2 16.2–17.9 22.2–24.4

JNC8 all ages, JNC7 no risk factors, stage 2 hypertension (≥100 mm Hg), target 90 mm Hg (1–2 standard dose medications)

15.5–17.4 12.9–14.6 18.1–20.2


(Continued )



ownloaded from



JNC7 diabetes mellitus or CKD, pretreatment ≥ 90 mm Hg, target 80 mm Hg (1–2 standard dose medications)

8.0–9.6 6.4–7.7 9.7–11.5

JNC8 all ages, JNC7 no risk factors, stage 1 hypertension (90–99 mm Hg), target 90 mm Hg (1 standard dose medications)

4.5–5.9 3.3–4.5 5.7–7.4

Prehypertension

JNC7 diabetes mellitus or CKD, prehypertension (80–89 mm Hg), target 80 mm Hg (0.5 standard dose medications)

3.4 2.3 4.4

JNC7 diabetes mellitus or CKD target 130 mm Hg or intensive intervention in age ≥50 y with existing CVD, CKD or CVD risk ≥15%, target 120 mm Hg

13.7 9.6 17.8

Annual costs per person treated (2010 costs; inflated to 2014 costs in all results)

MD office visit ALLHAT trial29; JNC7 recommendation

Treatment monitoring visits (number)

Stage 2 hypertension 4 3 5

Stage 1 hypertension 3 2 4

Cost per routine monitoring visit $71 Not modeled Not modeled Outpatient visit, Medicare Physician fee schedule (code 99213, nonfacility limiting charge)30

Hospitalization National Inpatient Sample survey

Average cost (used for infrequent hospitalized drug-related adverse events)

$11 994 Not modeled Not modeled

High cost (used for rare hospitalized drug-related adverse events)

$20 680

Laboratory test (electrolytes monitoring on treatment)

No. of tests 1 1 2 JNC7, Centers for Medicare and Medicaid

Cost per test $15 Not modeled Not modeled Laboratory fee schedule31

Antihypertensive drug costs (total daily doses)‡ Average wholesale prices reported by manufacturers (Red Book; 2010)32

0.5 standard doses $120 Not modeled $287

1.0 standard dose $161 $351

1.5 standard doses $208 $4395





Pharmacy dispensing fees $27 $33

Acute and chronic CVD treatment costs California Office of Statewide Health Planning and Development

(OSHPD) hospital data, 200833

Myocardial infarction hospitalization

Nonfatal $33 000

Table 1. Continued

Variable



(Continued )



ownloaded from



Fatal $46 000

Coronary revascularization procedures

Percutaneous coronary intervention $21 000 to $23 000

Coronary artery bypass graft surgery $57 000 to $59 000

Stroke

Fatal $21 000 to $26 000

Nonfatal $15 000 to $21 000

Chronic CHD costs US Medical Expenditure Panel 1998–200834

First year $11 000

Subsequent years $2000

Chronic poststroke costs US Medical Expenditure Panel 1998–200834

First year $16 000

Subsequent years $5000

Inflation from 2010 to 2014 costs 9% Main=change in general US consumer price index

Serious adverse effects of medications (incidence per 100 000 person-years)

Common, outpatient management Law 20038,10

3 standard doses 10 039.20 6950.21 12 742.06

2 standard doses 7572.41 5242.43 9611.13

1 standard dose 5200.00 3600.06 6600.00

1-half dose 2600.00 1800.00 3300.00

Infrequent, hospitalized Trials, medication labels, postmarketing reports

3 standard doses 193.06 19.31 965.31

2 standard doses 145.62 14.56 728.12

1 standard dose 100.00 10.00 500.00

1-half standard dose 50.00 5.00 250.00

Rare, hospitalized/severe

3 standard doses 1.93 0.0193 19.31

2 standard doses 1.46 0.0146 14.56

1 standard dose 1.00 0.0100 10.00


Death

3 standard doses 0.0193 0.0002 0.1931

2 standard doses 0.0146 0.0001 0.1456

1 standard dose 0.0100 0.0001 0.1000


Utility (QALY weight penalty [where 1.00=perfect health], duration)

Drug side effects managed as outpatient, 1 d 0.23 Montgomery35

0.50 Clinical judgment

Drug side effect requiring hospitalization, 1d GBD 2010 Study36

Table 1. Continued

Variable



(Continued )



ownloaded from



≈2000 under JNC8 alone to ≈4000 under JNC7+intensive (Table S6 in the online-only Data Supplement).

In men, implementing JNC7 in addition to JNC8 would be cost-effective (ICER, ≈$7000 per QALY gained; Table 2). Incremental to JNC8, JNC7+intensive and JNC8+intensive strategies would be cost saving in men aged 35 to 74 years. At a willingness to pay threshold of $50 000 per QALY gained, the probability JNC7+intensive was more cost-effective than any other strategy in men was 100% (Figure 1, cost-effectiveness

acceptability curve). At a lower willingness to pay thresh-old of <$25 000, JNC8+intensive was more likely to be cost-effective than the JNC7+intensive strategy (>50%, probability more cost-effective). In women, JNC7 was borderline cost-effective compared with JNC8 (≈$52 000 per QALY gained). Adding intensive treatment of high-risk patients was cost-effective in women incremental to JNC8. At a willingness-to-pay threshold of $50 000 per QALY gained, the probability that JNC8+intensive was the most cost-effective strategy for women

Acute stroke, 1 m 0.86

Chronic stroke survivors, 12 m 0.85–0.88

Acute myocardial infarction, 1 m 0.91

Acute unstable angina, 1 m 0.95

Chronic CHD, 12 m 0.91–0.98

Death 1.00

Adherence to medications (percent of patients continuing prescribed treatment)

75% 25% or 50% lower than

observed in trials

Not modeled Law meta-analysis for main estimate8,10

ALLHAT indicates Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial; CHD, coronary heart disease; CKD, chronic kidney disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; GBD, Global Burden of Disease; JNC Joint National Committee; MD, medical doctor; OSHPD, Office of Statewide Health Planning and Development; QALY, quality-adjusted life years; RR, relative risk; and SBP, systolic blood pressure.

*RR reductions vary by age and sex category (see Methods and online-only Data Supplement for details).†BP change dependent on age- and sex-specific distribution of baseline blood pressures within stage 1 or stage 2 category and number of standard dose

antihypertensive medications.‡Standard dose medications used to estimate costs: captopril 25 mg twice daily, nifedipine 30 mg daily, amlodipine 5 mg daily, hydrochlorothiazide 25 mg daily,

and atenolol 50 mg daily.

Table 2. Annual Population and Cost-Effectiveness of Implementing Alternative National Hypertension Guidelines in Untreated Hypertensive Patients With and Without Intensive Treatment of High-CVD Risk Individuals Incremental to the 2014 JNC8 Guidelines

Strategy

Annual Number Eligible for

Treatment, Millions

Annual CVD Events Averted

(vs JNC8; 95% UI)

Annual CVD Deaths Averted

(vs JNC8; 95% UI)*Annual QALYs Gained

(vs JNC8; 95% UI)

Annual Costs, Millions of $US

(vs JNC8; 95% UI)ICER

(vs JNC8)

Men (age, 35–74 y)

JNC 7 13.3 −15 000 (−4000 to −27 000)

−3000 (−400 to −6000)

28 400 (11 000 to 47 000)

+$190 (−$1000 to +$1300)

$7000/QALY gained

JNC 8+intensive treatment in high CVD risk†

11.2 −29 000 (−17 000 to −47 000)

−11 000 (−2000 to −18 000)

40 100 (23 000 to 47 000)

−$1240 (−$3000 to −$200)

Cost saving


16.2 −39 000 (−23 000 to −55 000)

−13 000 (−7000 to −16 000)

59 500 (36 000 to 85 000)

−$1030 (−$3000 to +$300)

Cost saving

Women (age, 35–74 y)

JNC 7 9.5 −9000 (−2000 to −17 000)

−2000 (−4000 to +76)

17 500 (7000 to 30 000)

+$920 (+$200 to +$1500)

$52 000/QALY gained


6.3 −14 000 (−5000 to −23 000)

−4000 (−2000 to −6000)

14 500 (4000 to 28 000)

−$290 (−$1400 to +$200)

Cost saving


13.5 −20 000 (−11 000 to −31 000)

−6000 (−3000 to −8000)

29 200 (16 000 to 44 000)

+$600 (−$500 to +$1300)

$21 000/QALY gained

ICER <$50 000 per QALY gained is considered cost-effective. CVD indicates cardiovascular disease; ICER, incremental cost-effectiveness ratio; JNC, Joint National Committee; QALY, quality-adjusted life years; SBP, systolic blood pressure; and UI, uncertainty interval.

*Overall CVD death includes heart failure deaths. As heart failure deaths were estimated not simulated, 95% UI do not account for heart failure deaths.†Intensive treatment is defined as an SBP goal of <120 mm Hg in adults aged 50 y and older with SBP ≥130 mm Hg and existing CVD, chronic kidney disease, or

2013 American Heart Association/American College of Cardiology Pooled Cohorts 10-year CVD risk ≥15%.

Table 1. Continued

Variable





ownloaded from



was 81.7%, whereas the probability that the JNC7+intensive strategy most cost-effective was 18.3% (Figure 2).

Subgroup AnalysesIncremental to JNC8, JNC7 would be cost saving in men aged 60 to 74 years, cost-effective in men aged 45 to 59 years (ICER, ≈15 000 per QALY gained) and in women 60 to 74 years (ICER, ≈30 000 per QALY gained), but of intermediate and low value in men and women aged 35 to 44 years, respectively. Incremental to JNC8 alone, JNC8+intensive would be cost saving in all age groups, whereas JNC7+intensive would be cost saving in all men and in women 60 to 74 years old, but cost-effective in women aged 45 to 59 years (ICER, 44 000 per QALY gained; Table S7).

Sensitivity AnalysesAssuming 20% less CVD risk reduction per BP change (in mm Hg), more frequent monitoring plus double the drug costs, or 40% medication adherence, adding JNC8+intensive or JNC7+intensive remained cost-saving or cost-effective in most instances (ICERs <$50 000; Table S8). High drug costs plus higher monitoring frequency or 40% adherence made JNC7+intensive of intermediate or low value in women. JNC7 alone was sensitive to high drug costs incremental to JNC8.

Adding treatment of treated but uncontrolled hypertension would double the population eligible for treatment to BP con-trol under all strategies and lead to 60 000 to 91 000 fewer CVD events with intensive strategies compared with JNC8 alone. ICERs for the comparison of JNC7 versus JNC8 with and without intensive strategies remained similar when the previ-ously treated and uncontrolled group was added (Table S9).

DiscussionWe projected that adding intensive strategies to JNC hyper-tension treatment guidelines would be cost saving in men and cost-effective in women aged 35 to 74 years, which held true even in the event of higher monitoring costs. From a payer’s perspective, JNC8+intensive would most likely be the high-est value strategy in women, whereas JNC7+intensive would most likely be the highest value strategy for men.

The committee appointed by the JNC8 recommended an SBP target of 150 mm Hg among individuals aged 60 years and older and a target of 140 mm Hg for patients with diabetes mellitus or CKD, based on selected hypertension medication treatment tri-als. SPRINT results were released after the JNC8 published its recommendations and suggested greater CVD benefit from an SBP goal of 120 mm Hg, as opposed to 140 mm Hg in patients at

Figure 1. Cost-effectiveness acceptability curves for the probability of selecting Joint National Committee (JNC)-7+intensive over JNC8+intensive treatment in high-risk men aged 35 to 74 years (high risk: ≥50 years old with one of the following: existing cardiovascular disease [CVD], 2013 American College of Cardiology/American Heart Association Pooled Cohorts 10-year CVD risk ≥15%, or chronic kidney disease). JNC7 alone and JNC8 alone were dominated and do not appear on the plot.

Figure 2. Cost-effectiveness acceptability curves for the probability of selecting Joint National Committee (JNC)-7+intensive over JNC8+intensive treatment in high-risk women aged 35 to 74 years (high risk: ≥50 years old with one of the following: existing cardiovascular disease [CVD], 2013 American College of Cardiology/American Heart Pooled Cohorts 10-year CVD risk ≥15%, or chronic kidney disease). JNC7 alone and JNC8 alone were dominated and do not appear on the plot.



ownloaded from



high CVD risk.4 SPRINT reinforced evidence favoring a lower BP goal in selected high-risk patients.15–18 Concerns about the risks of intensive treatment persist.19,20 The bulk of randomized trial evidence demonstrates reduction in major CVD events, renal out-comes, and retinopathy from BP lowering well below the 140/90 mm Hg threshold without clear effects on CVD or noncardiovas-cular death, and the size of these benefits is consistent with epi-demiological associations.8,21,22 The more recent Heart Outcomes Prevention Evaluation (HOPE)-3 trial found that BP lowering con-ferred no appreciable benefit in intermediate risk patients (mean 10-year CVD risk, ≈10%), except for those with pretreatment systolic BP >144 mm Hg.23 Therefore, treatment of patients with pretreatment systolic BP 130 to 139 mm Hg and 10-year CVD risk <15% according to JNC7 remains controversial.

Our study had several limitations. Hypertension treat-ment guideline effectiveness and cost-effectiveness may vary among specific population groups with higher hyper-tension prevalence, such as blacks16 and subgroups at high risk for CVD, in whom greater benefits may derive from hypertension treatment. Although we estimated the effect of hypertension treatment on ischemic heart failure hospitaliza-tions and deaths, coronary heart disease hospitalizations and deaths involving heart failure are difficult to accurately mea-sure based on International Classification of Diseases–coded data. We projected heart failure deaths prevented because of hypertension treatment, but we did not simulate heart fail-ure incidence or capture heart failure states directly, and we may have underestimated reduced heart failure burden attributable with hypertension treatment. We did not account for non–blood pressure–lowering benefits of certain anti-hypertensive drug classes, such as angiotensin-converting enzyme inhibitors, in patients with heart failure or past myo-cardial infarction. We did assume that most CVD patients would require >1 medication to reach the BP goal, one of those being an angiotensin-converting enzyme inhibitor. We may also have underestimated monitoring costs, including personnel, technology, or additional office visits needed to achieve intensive goals.

We followed the decision of the SPRINT trial and did not target an SBP 120 mm Hg goal in patients with diabetes mel-litus. Uncertainty persists about benefits and risks of intensive BP lowering in these patients.24,25 Intensive BP lowering consis-tently lowered stroke risk in trials enrolling older patients with diabetes, but results for coronary heart disease were variable.23,24 Patients with stroke were excluded from SPRINT; our decision to target intensive BP goals in patients with stroke is supported by suggestion of a benefit from intensive treatment in patients with stroke enrolled in the Secondary Prevention of Small Subcortical Strokes (SPS3) trial.26 SPRINT included partici-pants aged ≥75 years, but we excluded elderly patients from our analysis because of uncertainty about risks and benefits of intensive BP lowering in the frail elderly.27

JNC recommendations have increased hypertension awareness, treatment, and control in the US population and likely contributed to the decline in CVD mortality during the past 4 decades.28 Our results suggest that targeting an inten-sive goal of 120 mm Hg in selected high-CVD risk patients in addition to the standard JNC guidelines would be cost saving if high drug costs can be controlled.

PerspectivesHypertension treatment is inexpensive, safe, and effective. Guidelines should not be applied blindly, without considering the balance between benefits and harms in individual patients. However, in robust otherwise healthy patients aged <75 years, targeting more intensive blood pressure treatment goals in high CVD risk patients would be cost saving if monitoring and drug costs could be contained.

AcknowledgmentsN. Moise and A.E. Moran had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. L. Goldman, N. Moise, and A.E. Moran contributed to the study concept and design. A.E. Moran, C. Huang, P.G. Coxson, and N. Moise contributed to acquisition of data. N. Moise, A.E. Moran, C. Huang, C.N. Kohli-Lynch, and A. Rodgers contributed to analysis and interpretation of data. N. Moise, A.E. Moran, and A. Rodgers con-tributed to drafting of the article. N. Moise, A.E. Moran, P.G. Coxson, K. Bibbins-Domingo, A. Rodgers, and L. Goldman contributed to criti-cal revision of article for important intellectual content. N. Moise, A.E. Moran, P.G. Coxson, and C. Huang contributed to statistical analysis. A.E. Moran obtained funding and supervised the study.

Sources of FundingThis work was supported by funds from Health Resources and Services Administration (T32HP10260), National Heart, Lung, and Blood Institute (R01 HL107475-01), and the American Heart Association Founder’s Affiliate (10CRP4140089).

DisclosuresThis article was prepared using Framingham Cohort and Framingham Offspring Research Materials obtained from the US National Heart, Lung and Blood Institute (NHLBI) Biological Specimen and Data Repository Information Coordinating Center and does not necessarily reflect the opinions or views of the Framingham Cohort, Framingham Offspring, or the NHLBI. K. Bibbins-Domingo is a member of the US Preventive Services Task Force (USPSTF) and current co-Vice Chair. This work does not necessarily represent the views and policies of the USPSTF.

References 1. James PA, Oparil S, Carter BL, et al. 2014 evidence-based guideline for

the management of high blood pressure in adults: report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA. 2014;311:507–520. doi: 10.1001/jama.2013.284427.

2. Navar-Boggan AM, Pencina MJ, Williams K, Sniderman AD, Peterson ED. Proportion of US adults potentially affected by the 2014 hyper-tension guideline. JAMA. 2014;311:1424–1429. doi: 10.1001/jama. 2014.2531.

3. Wright JT Jr, Fine LJ, Lackland DT, Ogedegbe G, Dennison Himmelfarb CR. Evidence supporting a systolic blood pressure goal of less than 150 mm Hg in patients aged 60 years or older: the minority view. Ann Intern Med. 2014;160:499–503. doi: 10.7326/M13-2981.

4. Wright JT Jr, Williamson JD, Whelton PK, et al; The SPRINT Research Group. A randomized trial of intensive versus standard blood-pres-sure control. N Engl J Med. 2015;373(22):2103–2016. doi:10.1056/NEJMoa1511939.

5. Weinstein MC, Coxson PG, Williams LW, Pass TM, Stason WB, Goldman L. Forecasting coronary heart disease incidence, mortality, and cost: the Coronary Heart Disease Policy Model. Am J Public Health. 1987;77:1417–1426.

6. Moran AE, Odden MC, Thanataveerat A, Tzong KY, Rasmussen PW, Guzman D, Williams L, Bibbins-Domingo K, Coxson PG, Goldman L. Cost-effectiveness of hypertension therapy according to 2014 guidelines. N Engl J Med. 2015;372:447–455. doi: 10.1056/NEJMsa1406751.

7. Prospective Studies Collaboration. Collaborative overview (‘meta-anal-ysis’) of prospective observational studies of the associations of usual blood pressure and usual cholesterol levels with common causes of death:



ownloaded from



protocol for the second cycle of the Prospective Studies Collaboration. J Cardiovasc Risk. 1999;6:315–320.

8. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 ran-domised trials in the context of expectations from prospective epidemio-logical studies. BMJ. 2009;338:b1665. doi: 10.1136/bmj.b1665.

9. Prevention of stroke by antihypertensive drug treatment in older per-sons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). SHEP Cooperative Research Group. JAMA. 1991;265:3255–3264.

10. Law MR, Wald NJ, Morris JK, Jordan RE. Value of low dose combination treatment with blood pressure lowering drugs: analysis of 354 randomised trials. BMJ. 2003;326:1427. doi: 10.1136/bmj.326.7404.1427.

11. Sundstrom J, Arima H, Woodward M, Jackson R, Karmali K, Lloyd-Jones D, Baigent C, Emberson J, Rahimi K, MacMahon S, Patel A, Perkovic V, Turnbull F, Neal B; The Blood Pressure Lowering Treatment Trialists Collboration. Blood pressure-lowering treatment based on cardiovascular risk: a meta-analysis of individual patient data. Lancet. 2014;384:591–598. doi: 10.1016/S0140-6736(14)61212-5.

12. Odden MC, Peralta CA, Haan MN, Covinsky KE. Rethinking the association of high blood pressure with mortality in elderly adults: the impact of frailty. Arch Intern Med. 2012;172:1162–1168. doi: 10.1001/archinternmed.2012.2555.

13. Anderson JL, Heidenreich PA, Barnett PG, Creager MA, Fonarow GC, Gibbons RJ, Halperin JL, Hlatky MA, Jacobs AK, Mark DB, Masoudi FA, Peterson ED, Shaw LJ. ACC/AHA statement on cost/value method-ology in clinical practice guidelines and performance measures: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures and Task Force on Practice Guidelines. J Am Coll Cardiol. 2014;63:2304–2322. doi: 10.1016/j.jacc.2014.03.016.

14. Vrijens B, Vincze G, Kristanto P, Urquhart J, Burnier M. Adherence to prescribed antihypertensive drug treatments: longitudinal study of elec-tronically compiled dosing histories. BMJ. 2008;336:1114–1117. doi: 10.1136/bmj.39553.670231.25.

15. Brugts JJ, Ninomiya T, Boersma E, Remme WJ, Bertrand M, Ferrari R, Fox K, MacMahon S, Chalmers J, Simoons ML. The consistency of the treatment effect of an ACE-inhibitor based treatment regimen in patients with vascular disease or high risk of vascular disease: a combined analysis of individual data of ADVANCE, EUROPA, and PROGRESS trials. Eur Heart J. 2009;30:1385–1394. doi: 10.1093/eurheartj/ehp103.

16. Sleight P, Yusuf S, Pogue J, Tsuyuki R, Diaz R, Probstfield J; Heart Outcomes Prevention Evaluation (HOPE) Study. Blood-pressure reduc-tion and cardiovascular risk in HOPE study. Lancet. 2001;358:2130–2131. doi: 10.1016/S0140-6736(01)07186-0.

17. Svensson P, de Faire U, Sleight P, Yusuf S, Ostergren J. Comparative effects of ramipril on ambulatory and office blood pressures: a HOPE substudy. Hypertension. 2001;38:E28–E32.

18. Yusuf S, Sleight P, Pogue J, Bosch J, Davies R, Dagenais G. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med. 2000;342:145–153. doi: 10.1056/NEJM200001203420301.

19. Kaplan NM. The diastolic J curve: alive and threatening. Hypertension. 2011;58:751–753. doi: 10.1161/HYPERTENSIONAHA.111.177741.

20. O’Connor PJ, Narayan KM, Anderson R, Feeney P, Fine L, Ali MK, Simmons DL, Hire DG, Sperl-Hillen JM, Katz LA, Margolis KL, Sullivan MD. Effect of intensive versus standard blood pressure control on depression

and health-related quality of life in type 2 diabetes: the ACCORD trial. Diabetes Care. 2012;35:1479–1481. doi: 10.2337/dc11-1868.

21. Lewington S, Clarke R, Qizilbash N, Peto R, Collins R; Prospective Studies Collaboration. Age-specific relevance of usual blood pressure to vascular mortality: a meta-analysis of individual data for one million adults in 61 prospective studies. Lancet. 2002;360:1903–1913.

22. Lv J, Neal B, Ehteshami P, Ninomiya T, Woodward M, Rodgers A, Wang H, MacMahon S, Turnbull F, Hillis G, Chalmers J, Perkovic V. Effects of intensive blood pressure lowering on cardiovascular and renal outcomes: a systematic review and meta-analysis. PLoS Med. 2012;9:e1001293. doi: 10.1371/journal.pmed.1001293.

23. Lonn E, Bosch J, Lopez-Jaramillo P, et al. Blood-pressure lowering in inter-mediate risk persons without cardiovascular disease [published online ahead of print April 2, 2016]. N Engl J Med. doi: 10.1056/NEJMoa1600175. http://www.nejm.org/doi/full/10.1056/NEJMoa1600175. Accessed April 30, 2016.

24. Emdin CA, Rahimi K, Neal B, Callender T, Perkovic V, Patel A. Blood pressure lowering in type 2 diabetes: a systematic review and meta-analy-sis. JAMA. 2015;313:603–615. doi: 10.1001/jama.2014.18574.

25. Cushman WC, Evans GW, Byington RP, et al; The ACCORD Study Group. Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med. 2010;362:1575–1585. doi: 10.1056/NEJMoa1001286.

26. Benavente OR, Coffey CS, Conwit R, Hart RG, McClure LA, Pearce LA, Pergola PE, Szychowski JM; The SPS3 Study Group. Blood-pressure tar-gets in patients with recent lacunar stroke: The SPS3 randomised trial. Lancet. 2013;382:507–515. doi: 10.1016/S0140-6736(13)60852-1.

27. Odden MC, Pletcher MJ, Coxson PG, Thekkethala D, Guzman D, Heller D, Goldman L, Bibbins-Domingo K. Cost-effectiveness and population impact of statins for primary prevention in adults aged 75 years or older in the United States. Ann Intern Med. 2015;162:533–541. doi: 10.7326/M14-1430.

28. Cutler DM, Long G, Berndt ER, Royer J, Fournier AA, Sasser A, Cremieux P. The value of antihypertensive drugs: a perspective on medical innovation. Health Aff (Millwood). 2007;26:97–110. doi: 10.1377/hlthaff.26.1.97.

29. Heidenreich PA, Davis BR, Cutler JA, Furberg CD, Lairson DR, Shlipak MG, Pressel SL, Nwachuku C, Goldman L. Cost-effectiveness of chlortha-lidone, amlodipine, and lisinopril as first-step treatment for patients with hypertension: an analysis of the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). J Gen Intern Med. 2008;23:509–516. doi: 10.1007/s11606-008-0515-2.

30. Lazar LD, Pletcher MJ, Coxson PG, Bibbins-Domingo K, Goldman L. Cost-effectiveness of statin therapy for primary prevention in a low-cost statin era. Circulation. 2011;124:146–153. doi: 10.1161/CIRCULATIONAHA.110.986349.

31. Greving JP, Visseren FL, de Wit GA, Algra A. Statin treatment for primary prevention of vascular disease: whom to treat? Cost-effectiveness analy-sis. BMJ. 2011;342:d1672. doi: 10.1136/bmj.d1672.

32. Red Book Drug References. 2014. http://redbook.com/redbook/awp/. Accessed December 2, 2014.

33. California Office of Statewide Health Planning and Development Hospital Discharge Survey, 2008.

34. U.S. Medical Expenditure Panel Survey. http://meps.ahrq.gov/mepsweb/. Accessed December 1, 2014.

35. Montgomery AA, Harding J, Fahey T. Shared decision making in hyper-tension: the impact of patient preferences on treatment choice. Fam Pract. 2001;18:309–313.

36. Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and 33 injuries in 21 regions, 1990–2010: a system-atic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2197–2223. doi: 10.1016/S0140-6736(12)61689-4.

What Is New?•This is the first study to compare the cost-effectiveness of implementing

an intensive systolic blood pressure goal of 120 mm Hg in high-cardio-vascular disease risk hypertensive patients in addition to implementing 2014 Joint National Committee (JNC)-8 or 2003 JNC7 guidelines alone.

What Is Relevant?•Changes in national hypertension treatment guidelines led to uncertainty

about the safest, most effective approach to achieving hypertension control.

•More recent evidence suggests that intensive blood pressure lowering leads to net health gains compared with more conservative goals in high cardiovascular disease risk patients.

Summary

Adding intensive blood pressure goals for high-risk patients to cur-rent national hypertension treatment guidelines prevents additional cardiovascular disease deaths while saving costs provided that medication costs are controlled.

Novelty and Significance



ownloaded from


Pamela G. Coxson, Kirsten Bibbins-Domingo, Lee Goldman and Andrew E. MoranNathalie Moise, Chen Huang, Anthony Rodgers, Ciaran N. Kohli-Lynch, Keane Y. Tzong,

74 Years: The Cardiovascular Disease Policy Model−Guidelines in Adults Aged 35 Comparative Cost-Effectiveness of Conservative or Intensive Blood Pressure Treatment

Print ISSN: 0194-911X. Online ISSN: 1524-4563 Copyright © 2016 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Hypertension doi: 10.1161/HYPERTENSIONAHA.115.06814

2016;68:88-96; originally published online May 15, 2016;Hypertension.

http://hyper.ahajournals.org/content/68/1/88World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://hyper.ahajournals.org/content/suppl/2016/05/15/HYPERTENSIONAHA.115.06814.DC1Data Supplement (unedited) at:

http://hyper.ahajournals.org//subscriptions/

is online at: Hypertension Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialHypertensionin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:



ownloaded from

http://hyper.ahajournals.org/content/68/1/88

http://hyper.ahajournals.org/content/suppl/2016/05/15/HYPERTENSIONAHA.115.06814.DC1

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://hyper.ahajournals.org//subscriptions/


SUPPLEMENTARY MATERIAL

Comparative cost-effectiveness of conservative or intensive blood pressure treatment

guidelines in adults aged 35-74 years: The CVD Policy Model

LIST OF INVESTIGATORS

Authors Nathalie Moise, MD1; Huang, Chen, MD2; Anthony Rodgers, MD3 ; Ciaran N. Kohli-

Lynch, MS4; Keane Y. Tzong, MS1; Pamela G. Coxson, PhD5; Kirsten Bibbins-Domingo, MD,

PhD5; Lee Goldman, MD, MPH6; Andrew E. Moran, MD, MPH1,6

Affiliations:

1. Division of General Medicine, Columbia University Medical Center, New York, NY, USA

2. Department of Evidence Based Medicine, Cardiovascular Institute and Fu Wai Hospital of the

Chinese Academy of Medical Sciences, Beijing, China and National Center for Cardiovascular

Diseases, Beijing, China

3. George Institute for Global Health, Sydney, Australia

4. Health Economics & Health Technology Assessment, University of Glasgow, UK.

5. Division of General Medicine, University of California at San Francisco, San Francisco, CA,

USA

6. College of Physicians and Surgeons, Columbia University, New York, NY, USA

1

SUPPLEMENTARY METHODS Structure of the Model The Cardiovascular Disease (CVD) Policy Model is a computer-simulation, state-transition (Markov cohort) model of coronary heart disease and stroke incidence, prevalence, mortality, and costs in the U.S. population over age 35 years.1-3 The Demographic-Epidemiologic Submodel predicts coronary heart disease and stroke incidence and non-CVD mortality among subjects without CVD, stratified by age, sex, and up to 8 additional categorized risk factors estimated from weighted United States National Health and Nutrition Examination Survey (NHANES) data from 2007-2010. Risk factors include: systolic blood pressure (<140, 140-159.9, ≥160), isolated diastolic blood pressure (normal systolic blood pressure and diastolic blood pressure (90-99 or ≥100 mmHg), antihypertensive medication treatment status (self-reported), smoking status (active smoker, non-smoker with exposure to environmental tobacco smoke, non-smoker without environmental exposure), high density lipoprotein (HDL) cholesterol (<1.0, 1.0-1.5, ≥1.6 mmol/L; <40, 40-59.9, ≥60 mg/dL), low-density lipoprotein (LDL) cholesterol (<2.6, 2.6-3.3, ≥3.4 mmol/L; <100, 100-129.9, ≥130 mg/dL), body mass index (<25, 25-29.9, ≥30 kg/M2), diabetes mellitus (yes or no), and chronic kidney disease [estimated glomerular filtration rate of < 60 ml/min (using the Modifcation of Diet in Renal Disease formula), and/or proteinuria (spot albumin/creatinine ration > 200 μg/g)]. After CVD develops, the Bridge Submodel characterizes the initial stroke or coronary heart disease event (cardiac arrest, myocardial infarction, or angina) and its sequelae for 30 days. Then, the Disease History Submodel predicts subsequent CVD events, coronary revascularization procedures, CVD mortality, and non-CVD mortality among patients with CVD, stratified by age, sex, and history of events. The general chronic CVD categories are coronary heart disease only, stroke only, and combined prior coronary heart disease and prior stroke. Each state and event has an annual cost and quality-of-life adjustment as well as an annual probability of a repeat event and/or transition to a different CVD state. All population distributions, risk factor levels, coefficients, event rates, case fatality rates, costs, and quality-of-life adjustments can be modified for forecasting simulations. Data Sources Version 4 of the CVD Policy Model includes data from prior versions as well as many updates and upgrades.1-3 The 2010 U.S. Census provides the baseline population4 and number of 35 year-olds projected to enter the model population from 2010-2060.5, 6 CHD and stroke deaths in 2010 were extracted from U.S. Vital Statistics.7 Deaths were categorized according to the International Classification of Diseases (ICD) 10 codes8: I20-I25 and two-thirds of I49, I50, and I51 were used to estimate coronary heart disease deaths,9 I60-I69 were used to estimate stroke deaths, and all other deaths were considered non-CVD deaths. Ischemic heart failure deaths were calculated based on the proportion of 2010 coronary heart disease deaths with primary ICD-10 codes I50.x. The incidence of coronary heart disease and stroke were based on competing risk Cox proportional hazards analysis of the Framingham Heart Study10 and the Framingham Offspring Study11 cohorts from 1988-2007, with further adjustment for risk factor differences between the Framingham cohorts and the contemporary U.S. population represented by the NHANES. Incident coronary heart disease events were allocated to angina pectoris, hospitalized myocardial

2

infarction, or cardiac arrest. Prevalence, joint distributions and means of U.S. risk factor values were estimated from pooled, survey design-weighted NHANES, 2007-10.12 Annual transition rates between risk factor levels were calculated to preserve age-range trends over time. Risk function betas for non-blood pressure risk factors were estimated separately for the risk of incident coronary heart disease events, incident strokes , and non-CVD deaths, using examinations 1-8 of the Framingham Offspring cohort.11 The Framingham coefficients have been useful across many populations.13-16 Risk factors were assumed to affect the incidence of myocardial infarction, arrest, and angina in proportion to the overall incidence of coronary heart disease, except tobacco smokers were assumed to have a higher relative risk for infarction and arrest (17; personal communication, Sean Coady, National Heart, Lung, and Blood Institute, February, 2006 ) and a proportionately lower coefficient for angina. Environmental tobacco exposure was assumed to carry a relative risk of 1.26 for myocardial infarction and cardiac arrest compared with non-exposed non-smokers18 but not to influence angina. Baseline CVD Policy Model inputs for the year 2010 were within 1% of all targets obtained from U.S. national data sources (Table S1). Preliminary analyses suggested that the strength of the observational association between blood pressure and CVD risks in the Framingham Heart Study fell short of the strength of association observed in antihypertensive treatment trials (basis of comparison was the Law, Morris, and Wald trials meta-analysis summary estimates) .19 Starting with CVD Policy Model default blood pressure beta coefficients estimated from Framingham Heart Study data, we calibrated the coefficients in order to 1) reproduce the association of change in systolic BP estimated from both the Prospective Cohort Studies Collaborative, and 2) a large pooled analysis of BP treatment trials by Law, Morris, and Wald.19 A U.S. NHANES-based cohort representing the age and sex structure and mean systolic BP of the U.S. stage one and stage two hypertensive population was prepared for calibration simulations. In order to simulate the average trial pooled in the Law, Morris and Wald study, five year treatment duration was simulated. First, a base case was simulated for 2010-2014 with no change in 2010 blood pressure levels. A 2010-2014 intervention simulation followed in which systolic BP was lowered by approximately 10 mm Hg or diastolic blood pressure was lowered 5 mm Hg in each age and sex category. Beta coefficients were calibrated until the Prospective Cohort Studies Collaboration age- and sex- specific relative risks were matched within 0.02 or less (Table S2). Summary (age and sex weighted) relative risk with a 10 mm Hg systolic or 5 mm Hg diastolic blood pressure change for ages 35-74 years of 0.75 was within the 95% confidence limits of the treatment trials relative risks estimated in the Law, Morris, and Wald meta-analysis for coronary heart disease (target interval 0.73-0.83), as was the estimate for stroke (estimate 0.64; target interval 0.52-0.67; Table S3). In order to ensure that the systolic BP relative risk inputs were not over-fitted to the calibration targets, and not representative of the results of a real clinical trial, we set out to use the model to simulate the landmark Systolic Hypertension in the Elderly Program (SHEP), a double-blind placebo-controlled trial of antihypertensive adults aged 60 and older. By analyzing individual participant-level data, we were able to enter the characteristics of the SHEP intervention arm and control arms (age, baseline systolic BP, mean HDL and LDL cholesterol, and smoking status of participants). We then simulated the SHEP trial over a five-year period (mean follow up in

3

SHEP lasted 4.5 years). Cox proportional hazards analysis of SHEP was performed in order to ensure that we could reproduce the estimates reported in the 1999 SHEP trial report, and we performed original analyses of SHEP data in order estimate the total stroke (fatal and nonfatal), and total and fatal coronary heart disease relative risks associated with BP treatment consistent with CVD Policy Model definitions. For each of stroke and CHD, we simulated RRs of stroke and CHD, assuming baseline characteristics of the SHEP cohort and systolic blood pressure beta coefficients used in our main analysis (these based on the Prospective Cohort Studies Collaboration and Law, Morris, Wald meta-analysis of trials as described above). Stroke and CHD RRs resulting from the simulated trial were then compared with the main estimates and 95% confidence intervals of the observed effects reported by the SHEP trial (Table S4). Compared with the results observed in the actual trial, our five-year simulation of the SHEP trial resulted in nearly perfectly matched results for reductions in the rate of coronary heart disease events with treatment. For stroke, our simulations yielded a 30% reduction in strokes (relative risk 0.70) compared with the 36% reduction observed in the trial (relative risk 0.64). SHEP excluded potential participants if they had a history of atrial fibrillation, but our Model cannot selectively do so. Since five-year effects of antihypertensive treatment are unlikely to lower risk for stroke caused by atrial fibrillation, the inclusion of people with atrial fibrillation likely explains why our simulated reduction of nonfatal stroke was less than that observed in SHEP. If we adjust our total stroke relative risk for an assumed 15% of all strokes (fatal and nonfatal) due to atrial fibrillation,20 and unaffected by the BP lowering intervention, our simulated trial would yield a relative risk of total stroke of 0.66 (calculated as: exp[1.15*ln(0.70)]), very close to the SHEP total stroke relative risk of 0.64. The number of hospitalized myocardial infarctions was obtained from discharges coded as ICD-9 code 410 in the 2010 National Hospital Discharge Survey (NHDS)21 adjusted for likely miscoding,22 such as patients who were discharged alive after two days or fewer without a percutaneous coronary intervention, and transfer patients. Case-fatality rates and rates of myocardial infarction in subgroups were estimated from national data21 and a variety of complementary sources.23-25 Prehospital arrest deaths were estimated from the U.S. Vital Statistics,26 and out-of-hospital cardiac arrests surviving to hospital discharge were estimated from national data21. Survival after a coronary heart disease event was estimated using California data on the ratio of in-hospital survival to 30 day survival27 and data from Medicare and Seattle, Washington.28, 29 Rates of coronary revascularizations were estimated from the National Hospital Discharge Survey,21 with mortalities estimated from aggregrated historical data. Stroke incidence was assumed to independent of the risk of new onset coronary heart disease in the same year. The number of hospitalized strokes was also obtained from the 2010 NHDS. Positive predictive values of specific ICD-9 stroke hospital diagnosis codes (inclusive of ICD 9 codes 430-438) were derived by pooling several studies of stroke incidence that compared hospital diagnoses with a gold standard (e.g., stroke ascertained by Atherosclerosis in Communities Study, the Rochester Epidemiology Study or similar criteria).30 The positive predictive values were applied to age- and sex-specific NHDS cases in order to estimate total

4

stroke event rates (inclusive of first-ever and recurrent stroke events). Applying 30-day case fatality rates based on the Atherosclerosis in Communities Study31, 32 yielded annual mortality rate estimates within the range of stroke rates reported by the U.S. Centers for Disease Control (CDC Wonder) for 2010. Incidence calibration assumed that 77% of all strokes are incident (first ever),33 but it was assumed that the proportion first ever/total diminished with age (i.e., >90% of all strokes are first strokes in 35-44 year olds and 50% are first strokes in 85-94 year olds). The resulting incidence of hospitalized stroke approximated age and sex specific stroke incidence rates observed in U.S. stroke cohort and surveillance studies. The annual probabilities of stroke after myocardial infarction34 and the probability of coronary heart disease in stroke patients was based on natural history studies.35-40 The background prevalence of CVD by age, sex, and CVD disease state (stroke, coronary heart disease, or both stroke and coronary heart disease) in 2010 was estimated from the National Health Interview Survey data from 2009-2011,41 assuming that the imperfect positive predictive value of survey data is offset by its imperfect sensitivity.42-44 Age-specific prevalences for individual CVD disease states were fitted with polynomial or spline functions of age to obtain smooth, monotonically increasing prevalences. The background prevalence of prior coronary revascularization was estimated from revascularizations before 2010 and estimated survival after revascularization, while model projections were used to infer the distribution of revascularization by CVD state. Heart failure deaths were estimated by applying age- and sex-specific proportions of hypertensive heart disease (ICD-10 I11.0, I11.9) to extrapolate total ischemic heart disease and congestive heart failure deaths. The proportion of hypertensive heart disease and ischemic heart disease were then used to inflated coronary heart disease deaths similated by the CVD Policy Model and estimate number of heart failure deaths in total. Age and sex specific health care costs were estimated using national data.45 Hospitalized stroke and coronary heart disease costs and acute stroke rehabilitation costs were estimated using 2008 California hospital data,46 deflated using cost to charge ratios47 and the ratio of the U.S. national average costs to the California average.48 Chronic outpatient CVD costs additional to average background health care costs for the first year after the event and for subsequent years were estimated for patients with a stroke or coronary heart disease diagnosis surveyed in the U.S. Medical Expenditure Panel Surveys (MEPS) pooled from 1998-2008. Average annual non-cardiovascular (background) costs were also estimated from the MEPS.49 All model costs were indexed to the year 2010 using the medical component of the consumer price index. Health-related quality-of-life weights and severity distributions for disease states were based on the Global Burden of Disease disability weights study.50-51 Blood pressure treatment assumptions Each recommendation made by the 2014 U.S. hypertension guidelines committee was backed by review of multiple studies and assessment of the variable quality of the evidence was made explicit. However, for a quick overview of the committee’s treatment recommendations, the reader may access the 2014 Hypertension Guideline Management Algorithm (http://files.jamanetwork.com/jn/Algo_image.png ). In brief, after a trial period of lifestyle interventions geared at reducing blood pressure, patients with hypertension selected for

5

pharmacologic treatment were grouped into those with or without diabetes or chronic kidney disease (CKD). The blood pressure targets for those treatment groups are shown in manuscript Table 1. Those without those conditions were categorized into Nonblack and Black. For initial choice of antihypertensive agent, Nonblack patients were recommended to start with any of thiazide diuretic, angiotensin converting enzyme (ACE) inhibitor, angiotensin receptor blocker (ARB) or calcium channel blocker (CCB). Black patients were recommended the same initial choice of therapies, except were not recommended to start with an ACE. Patients without diabetes but without CKD were recommended the same options as above for non-diabetics (options based on if the patient is Nonblack or Black). Patients with chronic kidney disease were recommended to start with an ACE or an ARB. Should the patient require more than one medication to reach target blood pressure, the guideline recommended any of the following: 1) maximize the first medication to full dose, then add a second agent; 2) add the second agent before maximizing the dosage of the first, or 3) start with two medication classes in combination. Our analysis used recommendations from the prior national guideline and the observed number of visits needed to achieve hypertension control in the ALLHAT trial (manuscript Table 1). We did not choose to simulate the effects of any particular medication; instead we simulate “standard dose” effects and assumed average drug prices across classes. Examples of standard doses for different antihypertensive medications can be found at http://www.wolfson.qmul.ac.uk/bpchol/a3.pdf, and include nifedipine 30 mg daily, amlodipine 5 mg daily, lisinopril 10 mg daily, hydrochlorothiazide 25 mg daily, and losartan 50 mg daily, Especially in patients with higher BPs (e.g. stage two hypertension), medications are added sequentially and BP lowered gradually over time. The amount of blood pressure change was assumed to be a function of the pre-treatment or baseline BP and the effect of a standard-dose antihypertensive agent at that pre-treatment level (Table S5). It is important to note that for patients with very high BPs (mean systolic BP of 185 mm Hg or more) it was assumed that even with taking four standard dose medications, these patients would on average achieve a BP of about 143 mm Hg, near, but not at, the target of 140 mmHg.

6

Supplementary References

1. Bibbins-Domingo K, Coxson P, Pletcher MJ, Lightwood J, Goldman L. Adolescent overweight and future adult coronary heart disease. The New England Journal of Medicine. 2007;357:2371-2379.

2. Hunink MGM, Goldman L, Tosteson ANA, Mittleman MA, Goldman PA, Williams LW, Tsevat J, Weinstein MC. The recent decline in mortality from coronary heart disease, 1980-1990: The effect of secular trends in risk factors and treatment. JAMA. 1997;277:535-542.

3. Weinstein MC, Coxson PG, Williams LW, Pass TM, Stason WB, Goldman L. Forecasting coronary heart disease incidence, mortality, and cost: the Coronary Heart Disease Policy Model. American Journal of Public Health. 1987;77:1417-1426.

4. U.S. Census Bureau Population Division. Monthly Postcensal Resident Population by Single Year of Age, Sex, Race and Hispanic Origin for the United States: July 1, 2010 to December 1 2010 (NC-EST2011-ALLDATA-R-File02). 2011; http://www.census.gov/popest/data/national/asrh/2011/files/NC-EST2011-ALLDATA-R-File02.csv. Accessed January 15, 2013.

5. U.S. Census Bureau Population Division. Projected Population by Single Year of Age, Sex, Race, and Hispanic Origin for the United States: 2012 to 2060 (NP2012_D1). 2012; http://www.census.gov/population/projections/files/downloadables/NP2012_D1.csv Accessed January 15, 2013.

6. U.S. Census Bureau Population Division. Methodology and Assumptions for the 2012 National Projections. 2012; December 2012:http://www.census.gov/population/projections/files/methodology/methodstatement12.pdf. Accessed January 15, 2013.

7. Centers for Disease Control and Prevention; National Center for Health Statistics. Underlying cause of death 1999-2010 on CDC WONDER online database, released 2012. 2012; http://wonder.cdc.gov/ucd-icd10.html. Accessed January 15, 2013.

8. National Center for Health Statistics. ICD10 Codes. 2004; http://ftp.cdc.gov/pub/Health_Statistics/NCHS/Publications/ICD10/each10.txt. Accessed July 12, 2006.

9. Consensus recommendations for the management of chronic heart failure. On behalf of the membership of the advisory council to improve outcomes nationwide in heart failure. The American Journal of Cardiology. 1999;83:1A-38A.

10. Dawber TR. The Framingham Study: the epidemiology of atherosclerotic disease. Cambridge, MA: Harvard University Press; 1980.

11. Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham Offspring Study. Design and preliminary data. Prev Med. 1975;4:518-525.

12. National Center for Health Statistics. National Health and Nutrition Examination Survey, 2009-2012. http://wwwn.cdc.gov/nchs/nhanes/search/nhanes09_10.aspx & http://wwwn.cdc.gov/nchs/nhanes/search/nhanes11_12.aspx, 2012.

13. Brindle P, Emberson J, Lampe F, Walker M, Whincup P, Fahey T, Ebrahim S. Predictive accuracy of the Framingham coronary risk score in British men: prospective cohort study. BMJ (Clinical research ed.). 2003;327:1267-1270.

14. D'Agostino RB, Grundy S, Sullivan LM, Wilson PW. Validation of the Framingham Coronary Heart Disease Prediction Scores. JAMA.. 2001;286:180-187.

7

15. Liu J, Hong Y, D'Agostino RBS, Wu Z, Sun J, Wilson PW, Kannel WB, Zhao D. Predictive value for the Chinese population of the Framingham CHD risk assessment tool compared with the Chinese Multi-Provincial Cohort Study. JAMA. 2004;291:2591-2599.

16. Wilson PW, D'Agostino RB, Levy D, Belanger AM, Silbershaltz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation. 1998;97:1837-1847.

17. Parish S, Collins R, Peto R, Youngman L, Barton J, Jayne K, Clarke R, Appleby P, Lyon V, Cederholm-Williams S. Cigarette smoking, tar yields, and non-fatal myocardial infarction: 14,000 cases and 32,000 controls in the United Kingdom. The International Studies of Infarct Survival (ISIS) Collaborators. BMJ (Clinical research ed.). 1995;311:471-477.

18. Law MR, Morris JK, Wald NJ. Environmental tobacco smoke exposure and ischaemic heart disease: an evaluation of the evidence. BMJ (Clinical research ed.). 1997;315:973-980.

19. Law MR, Morris JK, Wald NJ. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. BMJ. 2009;338:b1665.

20. When the beat is off--atrial fibrillation. http://www.strokeassociation.org/STROKEORG/LifeAfterStroke/HealthyLivingAfterStroke/UnderstandingRiskyConditions/When-the-Beat-is-Off---Atrial-Fibrillation_UCM_310782_Article.jsp. Accessed August 29, 2014.

21. National Center for Health Statistics. National Hospital Discharge Survey. ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHDS/ Accessed March 29, 2012.

22. Petersen LA, Wright S, Normand S-LT, Daley J. Positive predictive value of the diagnosis of acute myocardial infarction in an administrative database. Journal of General Internal Medicine. 1999;14:555-558.

23. Canto JG, Rogers WJ, Goldberg RJ, Peterson ED, Wenger NK, Vaccarino V, Kiefe CI, Frederick PD, Sopko G, Zheng ZJ. Association of age and sex with myocardial infarction symptom presentation and in-hospital mortality. JAMA. 2012;307:813-822.

24. Rieves D, Wright G, Gupta G, Shacter E. Clinical trial (GUSTO-1 and INJECT) evidence of earlier death for men than women after acute myocardial infarction. The American Journal of Cardiology. 2000;85:147-153.

25. Vaccarino V, Parsons L, Peterson ED, Rogers WJ, Kiefe CI, Canto J. Sex differences in mortality after acute myocardial infarction: changes from 1994 to 2006. Archives of Internal Medicine. 2009;169:1767-1774.

26. Centers for Disease Control and Prevention; National Center for Health Statistics. Underlying cause of death 1999-2010 on CDC WONDER online database, released 2012. 2012; http://wonder.cdc.gov/mcd-icd10.html. Accessed February 13, 2013.

27. Medical Expenditure Panel Survey. Medical Expenditure Panel Survey Public Use Files 1996-2001. http://www.meps.ahrq.gov/Puf/PufSearch.asp?SearchOption=Keyword. Accessed February 13, 2013.

28. Groeneveld PW, Heidenreich PA, Garber AM. Racial disparity in cardiac procedures and mortality among long-term survivors of cardiac arrest. Circulation. 2003;108:286-291.

29. Lenfant C, Chobanian AV, Jones DW, Roccella EJ, Joint National Committee on the Prevention DE, Treatment of High Blood P. Seventh report of the Joint National Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7): resetting the hypertension sails. Hypertension. 2003;41:1178-1179.

8

30. Williams GR, Jiang JG, Matchar DB, Samsa GP. Incidence and occurrence of total (first-ever and recurrent) stroke. Stroke. 1999;30:2523-2528.

31. Rosamond WD, Folsom AR, Chambless LE, Wang CH, McGovern PG, Howard G, Copper LS, Shahar E. Stroke incidence and survival among middle-aged adults: 9-year follow-up of the Atherosclerosis Risk in Communities (ARIC) cohort. Stroke. 1999;30:736-743.

32. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. American Journal of Epidemiology. 1989;129:687-702.

33. Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Blaha MJ, Dai S, Ford ES, Fox CS, Franco S, Fullerton HJ, Gillespie C, et al. Heart disease and stroke statistics--2014 update: a report from the American Heart Association. Circulation. 2014;129:e28-e292.

34. Witt BJ, Brown RD, Jr., Jacobsen SJ, Weston SA, Yawn BP, Roger VL. A community-based study of stroke incidence after myocardial infarction. Annals of Internal Medicine. 2005;143:785-792.

35. Amarenco P, Bogousslavsky J, Callahan A, 3rd, Goldstein LB, Hennerici M, Rudolph AE, Sillesen H, Simunovic L, Szarek M, Welch KM, Zivin JA. High-dose atorvastatin after stroke or transient ischemic attack. The New England Journal of Medicine. 2006;355:549-559.

36. Appelros P, Gunnarsson KE, Terent A. Ten-year risk for myocardial infarction in patients with first-ever stroke: a community-based study. Acta Neurologica Scandinavica. 2011;124:383-389.

37. Behar S, Tanne D, Abinader E, Agmon J, Barzilai J, Friedman Y, Kaplinsky E, Kauli N, Kishon Y, Palant A. Cerebrovascular accident complicating acute myocardial infarction: incidence, clinical significance and short- and long-term mortality rates. The SPRINT Study Group. The American Journal of Medicine. 1991;91:45-50.

38. Lakshminarayan K, Schissel C, Anderson DC, Vazquez G, Jacobs DR, Jr., Ezzeddine M, Luepker RV, Virnig BA. Five-year rehospitalization outcomes in a cohort of patients with acute ischemic stroke: Medicare linkage study. Stroke. 2011;42:1556-1562.

39. Prosser J, MacGregor L, Lees KR, Diener HC, Hacke W, Davis S. Predictors of early cardiac morbidity and mortality after ischemic stroke. Stroke. 2007;38:2295-2302.

40. Touze E, Varenne O, Chatellier G, Peyrard S, Rothwell PM, Mas JL. Risk of myocardial infarction and vascular death after transient ischemic attack and ischemic stroke: a systematic review and meta-analysis. Stroke. 2005;36:2748-2755.

41. National Center for Health Statistics. National Health Interview Survey, 2009-2011. ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NHIS/. Accessed June 12, 2012.

42. Bergmann MM, Byers T, Freedman DS, Mokdad AH. Validity of self-reported diagnoses leading to hospitalization: a comparison of self-reports with hospital records in a prospective study of American adults. American Journal of Epidemiology. 1998;147:969-977.

43. Ford ES, Giles WH. Changes in prevalence of nonfatal coronary heart disease in the United States from 1971-1994. Ethnicity and Disease. 2003;13:85-93.

44. Gross R, Bentur N, Elhayany A, Sherf M, Epstein L. The validity of self-reports on chronic disease: characteristics of underreporters and implications for the planning of services. Public Health Reviews. 1996;24:167-182.

45. Dhamoon MS, Moon YP, Paik MC, Boden-Albala B, Rundek T, Sacco RL, Elkind MS. Long-term functional recovery after first ischemic stroke: the Northern Manhattan Study. Stroke. 2009;40:2805-2811.

9

46. Bath PM, Cotton D, Martin RH, Palesch Y, Yusuf S, Sacco R, Diener HC, Estol C, Roberts R. Effect of combined aspirin and extended-release dipyridamole versus clopidogrel on functional outcome and recurrence in acute, mild ischemic stroke: PRoFESS subgroup analysis. Stroke. 2010;41:732-738.

47. California Office of Statewide Health Planning & Development. Hospital financial data for cost to charge ratio, CA inpatinet discharage data hospital annual financial data, pivot profiles. 1999-2000; http://www.oshpd.ca.gov/HQAD/Hospital/financial/hospAF.htm Accessed June 15, 2003.

48. Average cost to community hospitals per patient, by state (Table 204). Statistical Abstract of the United States. Bureau of the Census. Washington, D.C.: Government Printing Office; 1998.

49. Medical Expenditure Panel Survey. Medical Expenditure Panel Survey Public Use Files 1998-2008. http://meps.ahrq.gov/mepsweb/. Accessed June 15, 2003.

50. Moran AE, Forouzanfar MH, Roth G, Mensah G, Ezzati M, Murray CJ, Naghavi M. Temporal Trends in Ischemic Heart Disease Mortality in 21 World Regions, 1980-2010: The Global Burden of Disease 2010 Study. Circulation. 2014;129:1493-501.

51. Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2197-2223.

10

Table S1. Comparisons of selected CVD Policy Model simulation outputs for 2010 (model base year) with national targets

Age and

sex

category

Total myocardial

infarctions

Target sources:

NHDS

Total strokes

Target source:

NHDS

CHD deaths

Target source:

national vital

statistics

Stroke deaths

Target source:

national vital

statistics

All-cause deaths

Target source:

national vital

statistics

Target Model Target Model Target Model Target Model Target Model

Males

35-44 13,979 13,839 16,535 16,553 4,783

4,862 1,027 1,031 43,345

43,335

45-54 56,129 55,811 43,493 43,710 19,489

19,594 3,298 3,301 111,981

111,933

55-64 77,992 77,395 67,863 68,497 38,032

38,065 6,159 6,133 190,845

190,629

65-74 75,804 75,689 79,450 79,239 45,700

46,096 9,350 9,265 231,327

231,231

75-84 62,982 63,063 76,205 76,436 64,610

65,097 16,215 16,240 312,778

312,873

85-94 37,568 37,483 38,943 39,247 64,071

63,958 15,318 14,742 264,705

263,235

Females

35-44 6,259 6,144 6,390 6,387 1,710 1,822 873 875 26,538 26,619

45-54 17,071 17,035 36,952 37,083 6,858

6,969 2,609 2,764 71,145

71,352

11

55-64 40,246 40,403 42,966 43,222 15,122

15,265 4,622 4,605 122,502

122,546

65-74 43,843 43,898 69,473 69,659 24,964

25,137 8,504 8,308 178,530

178,342

75-84 60,097 60,043 93,040 93,434 53,247

53,600 21,492 21,541 313,803

313,894

85-94 57,661 57,403 77,481 77,883 99,680

98,988 35,416 36,233 448,864

447,244

Deviation

from

target -0.26% 0.39% 0.27% 0.12% -0.14%

12

Table S2. Results of the systolic blood pressure calibration exercise. FHS = Framingham Heart Study. PCS= Prospective Cohort

Studies Collaboration. CVDPM = CVD Policy Model inputs and simulation outputs.

Age/sex Mean

category change

in SBP

(mm Hg)

Beta coefficients (per 1.0 mm Hg systolic BP) Relative risks (after 10 mm Hg SBP lowering or 5 mm

Hg diastolic BP lowering)

CHD Stroke CHD Stroke

FHS

default

PCS

target

CVDPM

(input)

FHS

default

PCS

target

CVDPM

(input)

FHS

default

PCS

target

CVDPM

(output)

FHS

default

PCS

target

CVDPM

(output)

Males

35-44 9.57 .0217 .0361 .0325 .0215 .0513 .0500 .78 .71 .73 .73 .61 .60

45-54 10.38 .0176 .0353 .0310 .0196 .0496 .0460 .81 .70` .72 .75 .61 .59

55-64 10.38 .0135 .0330 .0300 .0178 .0453 .0420 .78 .71 .73 .76 .63 .62

65-74 10.88 .0094 .0285 .0265 .0159 .0385 .0370 .75 .74 .76 .76 .66 .66

Females

35-44 9.87 .0217 .0360 .0320 .0215 .0513 .0470 .71 .70 .72 .79 .60 .60

45-54 10.35 .0176 .0353 .0320 .0196 .0496 .0450 .72 .70 .72 .77 .61 .60

55-64 10.94 .0135 .0330 .0295 .0178 .0452 .0414 .73 .70 .72 .73 .62 .61

13

65-74 11.68 .0094 .0285 .0245 .0159 .0385 .0345 .73 .73 .75 .77 .65 .65

14

Table S3. Comparison of CVD Policy Model trial simulation to Law Morris Wald meta-analysis of anti-hypertensive

medication treatment trials, both assuming a systolic blood pressure change of 10 mm Hg or a diastolic blood pressure change

of 5 mm Hg.

Outcome

Law, Morris, and Wald

meta-analysis estimate

CVD Policy Model

main estimate

Males

CVD Policy Model

main estimate

Females

Coronary heart

disease

0.78 (0.73—0.83) Ages 35-74 years: 0.75

Ages 35-64 years: 0.74



Stroke 0.59 (0.52—0.67) Ages 35-74 years: 0.64




15

Table S4. Results of a simulated trial of blood pressure treatment in participants in the Systolic

Hypertension in the Elderly Program (SHEP) trial, using relative risk inputs and transition probabilities in

the CVD Policy Model version used for the analysis.

SHEP Population RR (SBP lowering of 12 mmHg for 5 years)

Outcome Actual SHEP paper RR and 95% CI

CVD Policy Model five year

simulation results with

PSC calibrated SBP betas

[overall rate ratio (95%

uncertainty interval)*]

Incident stroke 0.64 (0.50-0.82) 0.70 (0.66-0.75)

Incident coronary heart disease† 0.75 (0.60-0.94) 0.76 (0.73-0.79)

*95% uncertainty intervals derived from 1,000 probabilistic simulations that sampled from within the 95% confidence

intervals of age- and sex-specific Prospective Cohorts Studies Collaboration beta coefficients.

†SHEP incident coronary heart disease defined as first-in-trial nonfatal acute myocardial infarction, coronary

revascularization, or coronary heart disease death. CVD Policy Model incident coronary heart disease defined as first-ever

nonfatal acute myocardial infarction, angina pectoris with or without coronary revascularization, or coronary heart disease

death.

16

Table S5. Sequential changes in blood pressure with successive standard dose medications, based on the trials-based blood pressure

change prediction formulas of Law, Morris, and Wald.19

Standard dose

Mean

start

BP

BP

change

1

drug

new

BP

BP

change

2nd

drug

new

BP

BP

change

3rd

drug

new

BP

BP

change

4th

drug

new

BP

Final BP change

Systolic blood

pressure (mm Hg

changes)

< 60 years

old

≥ 60 years

old

effect of 4 std dose 185 12.2 172.8 11.0 161.8 9.9 151.9 8.9 143.0 42.0 33.1

effect of 3.5 std doses 175 11.2 163.8 10.1 153.7 9.1 144.6 6.3 138.4 36.6 21.3

179 11.6 167.4 10.4 157.0 9.4 147.6 6.5 141.1

effect of 3 std doses 165 10.2 154.8 9.2 145.6 8.3 137.4 27.6 19.4

169 10.6 158.4 9.5 148.9 8.6 140.3

effect of 2 std doses 155 9.2 145.8 8.3 137.5 17.5

effect of 1 std dose 155 9.2 145.8 9.2

effect of 1 std dose 147 8.4 138.6 8.4

effect of 0.5 std dose 142 6.1 135.9 6.1


17

Diastolic blood

pressure (mm Hg

changes)

All ages

effect of 3 std doses 107 6.6 100.5 5.9 94.6 5.2 89.4 17.7






effect one std dose 94 5.2 89.0 5.2





18

Table S6. Average annual ischemic and hypertensive heart disease (HHD) deaths averted due to hypertension treatment compared with no treatment by age and sex from years, over 2016-2026.*

Deaths Averted compared with no treatment

Males 35-59 years

Females 35-59 years

Males 60-74 years

Females 60-74 years

JNC 8 compared to no treatment 430 160 730 440

JNC 7 compared to no treatment 550 200 980 560

JNC 8 plus intensified treatment†

JNC 7 plus intensified treatment |†

710

790

290

300

1690

1800

820

900

*Hypertensive Heart Disease (CDC Wonder) includes those with and without congestive heart failure (ICD10 I11.0, I11.9). Ischemic heart failure deaths were calculated by applying the proportion of coronary heart disease deaths that are heart failure deaths (based on proportions observed in U.S. vital statistics; see Data Sources section, page 3 of this Materials) to the total number of coronary heart disease deaths.

† Intensive treatment is defined as an SBP goal of <120 mmHg in adults aged 50 years and older with SBP ≥ 130 mmHg and existing CVD, chronic kidney disease and/or ACC/AHA 10 year CVDrisk ≥15%.

CVD Cardiovascular disease; JNC Joint National Committee

19

Table S7. Cost-effectiveness of JNC7 rand JNC8 recommendations with and without intensive interventions in adults at

high risk for cardiovascular disease incremental to JNC8 recommendations within subgroups; ICERS (95% uncertainty intervals)

ICER compared to JNC8

Males Females

ages 35-44 ages 45-59 ages 60-74 ages 35-44 ages 45-59 ages 60-74

Complete JNC 7 $94,000

$15,000

cost saving

$269,000

$75,000

$30,000

JNC 8+ Intensive Intervention*

N/A cost saving cost saving N/A cost saving cost saving

JNC 7+ Intensive Intervention*

N/A cost saving cost saving N/A $44,000 cost saving

*Intensive intervention to <120 mmH in selected high risk individuals with systolic blood pressure 130 mmHg or greater.

JNC Joint National Committee; ICER Incremental Cost Effectiveness Ratio

20

Table S8. One-way sensitivity analyses of the cost-effectiveness of implementing JNC7 and JNC8 with and without intensive treatment in selected high CVD risk individuals compared with JNC8 in all previously untreated adults with hypertension aged 35-74 years, the CVD Policy Model.

One Way sensitivty analyses assumptions (vs. JNC8) Total Males Females

JNC7 $24,000 $7,000 $52,000

Effectiveness

Upper bound of relative risk for CHD or stroke with change in BP $30,000 $13,000 $62,000

Costs

High drug costs $80,000 $53,000 $121,000

Adherence

Assume 40% adherence $73,000 $43,000 $125,000

JNC8 + Intensive Treatment in Selected High Risk* cost saving cost saving cost saving

Effectiveness Upper bound of relative risk for

CHD or stroke with change in BP cost saving cost saving cost saving

Costs

21

High drug costs plus high monitoring frequency $9,000 cost saving $36,000

High frequency of monitoring† cost saving cost saving $600 Adherence

Assume 40% adherence $6,000 $800 $22,000

JNC7+ Intensive Treatment in Selected High Risk* cost saving cost saving $21,000

Effectiveness Upper bound of relative risk for CHD

or stroke with change in BP $37,000 $16,000 $100,109

Costs

High drug costs plus high monitoring frequency $24,000 $6,000 $69,000

High frequency of monitoring† $6,000 cost saving $33,000 Adherence

Assume 40% adherence $32,000 $12,000 $77,000

*Intensive intervention to <120 mmH in selected high risk individuals with systolic blood pressure 130 mmHg or greater

†Monitoring costs were applied only to the intensive strategies.

22

Table S9. Annual population and cost-effectiveness of implementing each hypertension treatment strategy for primary prevention in treated but uncontrolled and previously untreated adults with hypertension aged 35-74 years, the CVD Policy Model.

Guideline Strategy vs. JNC8

Annual Number

eligible for treatment (millions)

Annual CVD

events averted

(vs. JNC8)

Annual CVD

deaths averted

(vs. JNC8)

Annual QALYs Gained

(vs. JNC8)

Annual costs

(millions of $US) (vs.

JNC8)

ICER (vs. JNC8)

Males Age 35-74 years

JNC 7 20.2 -28,000 -7,000 51,500 +$90

$1,800/QALY

gained

JNC 8+ intensive treatment in high

CVD risk* 17.9 -40,000 -13,000 58,200 -$1100

cost-saving

JNC 7 + intensive treatment in high

CVD risk*

24.7 -57,000 -17,000 93,000 -$14,810 cost-saving

Females Age 35 -74 years

JNC 7 15.6 -17,000 -4,000 31,700 +$1500 $48,000/QALY

gained

JNC 8+ intensive treatment in high

CVD risk* 12.1 -20,000 -6,000 24,600 +$60

$2,000/QALY gained

JNC 7 + intensive treatment in high

CVD risk*

17.2 -34,000 -9,000 53,800 +$1470 $27,000/QALY

gained

*Intensive treatment is defined as an SBP goal of <120 mmHg in adults aged 50 years and older with SBP ≥ 130 mmHg and existing CVD, chronic kidney disease and/or 10 year CVD risk ≥15%. CS cost saving; CVD Cardiovascular Disease; ICER Incremental Cost-Effectiveness Ratio; JNC Joint National Committee; QALY Quality adjusted life years; ICER <$50,000 per QALY gained is considered cost-effective; SBP Systolic Blood Pressure; UI Uncertainty Interval

23

Documents

Comparative Cost-Effectiveness of Conservative or ...hyper.ahajournals.org/content/hypertensionaha/68/1/88.full.pdf · Pressure Intervention Trial (SPRINT) found that targeting an