Predicting Out-of-Office Blood Pressure in the Clinic for the Diagnosis of Hypertension in Primary CareNovelty and Significance
An Economic Evaluation
Clinical guidelines in the United States and United Kingdom recommend that individuals with suspected hypertension should have ambulatory blood pressure (BP) monitoring to confirm the diagnosis. This approach reduces misdiagnosis because of white coat hypertension but will not identify people with masked hypertension who may benefit from treatment. The Predicting Out-of-Office Blood Pressure (PROOF-BP) algorithm predicts masked and white coat hypertension based on patient characteristics and clinic BP, improving the accuracy of diagnosis while limiting subsequent ambulatory BP monitoring. This study assessed the cost-effectiveness of using this tool in diagnosing hypertension in primary care. A Markov cost–utility cohort model was developed to compare diagnostic strategies: the PROOF-BP approach, including those with clinic BP ≥130/80 mm Hg who receive ambulatory BP monitoring as guided by the algorithm, compared with current standard diagnostic strategies including those with clinic BP ≥140/90 mm Hg combined with further monitoring (ambulatory BP monitoring as reference, clinic, and home monitoring also assessed). The model adopted a lifetime horizon with a 3-month time cycle, taking a UK Health Service/Personal Social Services perspective. The PROOF-BP algorithm was cost-effective in screening all patients with clinic BP ≥130/80 mm Hg compared with current strategies that only screen those with clinic BP ≥140/90 mm Hg, provided healthcare providers were willing to pay up to £20 000 ($26 000)/quality-adjusted life year gained. Deterministic and probabilistic sensitivity analyses supported the base-case findings. The PROOF-BP algorithm seems to be cost-effective compared with the conventional BP diagnostic options in primary care. Its use in clinical practice is likely to lead to reduced cardiovascular disease, death, and disability.
Hypertension is one of the most important modifiable risk factors for cardiovascular morbidity and mortality.1 Accurate measurement of blood pressure (BP) is essential to ensure that treatment is targeted appropriately. In the United Kingdom, the National Institute for Health and Care Excellence (NICE) published new guidelines on the diagnosis of hypertension in primary care in 2011.2 These recommended that all individuals with high BP readings in the clinic should be referred for ambulatory BP monitoring (ABPM) to confirm a diagnosis of hypertension, before initiating treatment. This recommendation was based on a systematic review of the clinical evidence and a Markov model-based cost–utility analysis comparing 3 different BP-monitoring methods (clinic BP monitoring [CBPM], self-monitoring at home [HBPM], and ABPM) for making a diagnosis of hypertension in individuals with a screening clinic BP measurement ≥140/90 mm Hg.3,4 ABPM was found to be the most cost-effective option across all age and sex subgroups. Despite ABPM being more expensive in terms of diagnostic costs, better targeting of treatment meant that it saved money in the long term by treating fewer individuals with white coat hypertension. Similar arguments have since been used in North America and Japan where out-of-office measurement is now also recommended.5,6
White coat hypertension is the term used to describe when an individual has raised clinic BP (≥140/90 mm Hg) but is normotensive on ABPM (≤135/85 mm Hg).7 Individuals with white coat hypertension are at lower cardiovascular disease risk compared with individuals with sustained hypertension.8 Conversely, individuals with normotensive clinic BP measurements (<140/90 mm Hg) but hypertensive ambulatory BP measurements (>135/85 mm Hg) are referred to as having masked hypertension and have an increased risk of cardiovascular events, which approaches that of overt hypertension.9,10 Individuals with potential masked hypertension were not included in the population assessed in the health economics analysis that informed the NICE guideline2 because their screening clinic BP measurement would have been <140/90 mm Hg.
The Predicting Out-of-Office Blood Pressure (PROOF-BP) algorithm calculates a predicted home clinic BP difference based on an individuals’ characteristics (age, sex, body mass index, past diagnosis of hypertension, cardiovascular disease, and antihypertensive prescription) and clinic BP, to guide utilization of ABPM (Table S1 in the online-only Data Supplement). Adding this predicted difference to the known clinic BP of an individual provides the adjusted clinic BP, which has been shown to be closer to the true out-of-office BP.11 Used as a triaging tool for ABPM (Figure 1), it has been shown to improve the accuracy of hypertension diagnosis (masked hypertension, sustained hypertension, and white coat hypertension) without appreciably increasing the use of ABPM.11 This study aimed to assess the cost-effectiveness of a strategy of targeted use of ABPM using the PROOF-BP algorithm in the diagnosis of hypertension in a primary care setting using an adaptation of the cost-effectiveness model used to inform the 2011 NICE guideline.
Full methods of the original health economic assessment that informed the NICE guideline have been described elsewhere.3,4 The data that support the findings of this study are available from the corresponding author on reasonable request.
The original model assessed the cost-effectiveness of each BP-monitoring method (CBPM, HBPM, and ABPM) for confirming a diagnosis in people with suspected hypertension (clinic BP ≥140/90 mm Hg). This model, developed in Microsoft Excel, was modified by comparing the original diagnostic strategies with use of the PROOF-BP algorithm as a comparator. The base-case model entry population was expanded to men and women aged 40 to 75 years with a screening clinic BP of ≥130/80 mm Hg although the management of those with clinic BP between 130 to 139 and 80 to 89 mm Hg was only affected in the PROOF-BP arm (see below). Model inputs were also updated where appropriate.
Data for screening patient population, defined according to their clinic BP (Table S3), were taken from the Health Survey for England,12 and adjusted clinic BPs were calculated using the PROOF-BP risk algorithm.11
The new model compared 4 methods of BP monitoring in the diagnosis of hypertension. Those approaches examined in the original model—CBPM (monthly measurements for 3 months), HBPM (measurements for a week), and ABPM (measurements for 24 hours)—were compared with the new PROOF-BP diagnostic strategy (Figure 1):
The CBPM, HBPM, or ABPM diagnostic strategies were unchanged from the original model:
i. Individuals were not considered for diagnosis or treatment if their screening clinic BP was <140/90 mm Hg.
ii. For those with a screening clinic BP of ≥140/90 mm Hg, they either underwent further clinic measurement, home monitoring, or ABPM, exactly as in the original model.
The PROOF-BP strategy operated for all with a screening clinic BP of ≥130/80 mm Hg (ie, everyone). They had an adjusted BP calculated using the PROOF-BP algorithm and then proceeded as follows:
i. If the individual had an adjusted clinic BP <130/80 mm Hg, no further action was required, and they were measured again at the next check-up period.
ii. If the individual had an adjusted clinic BP between 130/80 and 144/89 mm Hg, they received ABPM for confirmatory diagnosis.
iii. If the individual had an adjusted clinic BP ≥145/90 mm Hg, true hypertensive status was assumed, and treatment was offered without confirmatory ABPM diagnosis.
A simplified Markov model diagram of the health states and the movements between states allowed to occur in a cycle is shown in Figure 2. In keeping with the original model, a model cycle length of 3 months was chosen as that approximated the average length of time for a complete CBPM diagnosis.2 HBPM, ABPM, and the PROOF-BP algorithm were assumed to take 1 month for a complete diagnosis. In the suspected and diagnosed stages of the model, individuals could experience a fatal or nonfatal cardiovascular event (stable angina, unstable angina, stroke, myocardial infarction, and transient ischemic attack). As per the original model, after experiencing a nonfatal cardiovascular event, repeat clinical events were not modeled, and individuals remained in a postcardiovascular event state until they died.
In the model, normotensive individuals could become hypertensive over time, and those with an initial screening clinic BP of <140/90 mm Hg could move to >140/90 mm Hg. For model simplification purposes, it was assumed individuals could not become hypertensive while being assessed in the diagnostic pathway. Individuals not diagnosed with hypertension were assumed to have a BP check-up with CBPM at least every 5 years. In common with the original model, a failure rate was incorporated into ABPM. If ABPM failed (any cause of failure from a technical or a patient’s view), individuals were assumed to be put on HBPM. In the PROOF-BP algorithm strategy, if individuals had a screening clinic BP of <140/90 mm Hg and ABPM failed, it was assumed they remained undiagnosed (as in the HBPM strategy where these individuals were not considered for hypertension diagnosis), and their BP was rechecked every 5 years. This was because of a lack of data on the sensitivity and specificity of HBPM for those with a clinic BP of <140/90 mm Hg.
Clinical model parameters are detailed in Table 1. Correct diagnosis of hypertension depended on the sensitivity and specificity of the test strategy used. As in the original model, test characteristics for CBPM and HBPM were taken from a meta-analysis,13 with ABPM assumed to be the reference standard (100% sensitivity and 100% specificity). The test characteristics of the PROOF-BP algorithm with respect to their clinic BP and adjusted clinic BP categories are shown in Table S2.
Risk of coronary heart disease and stroke were calculated using the Framingham risk equations17 by combining age, sex, and BP with the general population prevalence of risk factors in the Health Survey for England.12 Individuals with masked hypertension were assumed to have the same higher risk of cardiovascular events as sustained hypertensives.25 A hypertension diagnosis resulted in prescription of antihypertensive drug therapy, and true hypertensive individuals received benefit in terms of cardiovascular risk reduction from such treatment. True normotensive individuals were assumed to receive no risk reduction from treatment (this assumption is relaxed in a sensitivity analysis). The proportion of individuals on different antihypertensive drug classes was based on treatment guidelines.2
Quality-of-life and cost data are shown in Table 2. Baseline sex- and age-specific quality-of-life (utility) weights were taken from the Health Survey for England26 and applied to the cohorts. In the base-case, individuals were assumed not to experience any quality-of-life reductions (disutility) as a result of antihypertensive treatment.
A more detailed description of costs is given in the Extended Methods on Costs section in the online-only Data Supplement. Costs were updated where necessary to 2013 to 2014 prices using the Hospital & Community Health Services index.27 Resource usage by diagnostic method and device usage assumptions were in line with the original model.4
Results were presented as the total costs and effects of each diagnostic strategy (ordered by increasing cost). Effectiveness was measured in quality-adjusted life years (QALYs). Incremental cost-effectiveness ratios were calculated from the difference in costs and effects between 2 options. Cost-effectiveness was assessed in relation to the NICE lower threshold of £20 000 per QALY.36 More costly and less effective (dominated) options were excluded from consideration. The analysis adopted a lifetime horizon, and all costs and outcomes were discounted at the standard 3.5% rate.37 Costs and outcomes were considered from a UK National Health Service/Personal Social Services perspective.
Uncertainty was explored via sensitivity analyses. Additional model runs were undertaken to determine the impact of changing key parameters on the model results. The following univariate sensitivity analysis was undertaken on all cohorts: model entry criteria were varied up and down, expanded to a screening clinic BP ≥120/70 mm Hg population (Table S4) and then restricted to a screening clinic BP ≥140/90 mm Hg population (Table S5). In line with the original model, sensitivity analyses were performed using the males aged 60 years subgroup. The following scenarios were explored:
A treatment disutility of 1% was assumed. This was equivalent to a quarter of the individuals experiencing a quality-of-life reduction of 4%, and everyone else experiencing no ill effects of treatment.
A treatment disutility of 2% was assumed. This was equivalent to a quarter of the individuals experiencing a quality-of-life reduction of 8% and everyone else experiencing no ill effects of treatment;
Antihypertensive treatment risk reduction was based on half doses of medication;
Antihypertensive drug costs were increased by 30%;
ABPM (reference) strategy to confirm diagnosis was undertaken in all patients with a screening clinic BP of ≥130/80 mm Hg;
The BP check-up frequency for those not diagnosed with hypertension was reduced from every 5 years to every 3 years;
The prevalence of masked hypertension was increased and decreased by 25%, respectively;
Antihypertensive treatment risk reduction for masked hypertension was based on half doses;
Antihypertensive treatment risk reduction for all treated people (ie, those who are not truly hypertensive also gain benefit from BP reduction);
Antihypertensive treatment risk reduction assumed to be same as intensive treatment from the SPRINT trial (Systolic Blood Pressure Intervention Trial).38
The failure rate of ABPM was increased from 5% to 17%.39
Where available, data were inputted into the model as distributions to fully incorporate the uncertainty around parameter values for a probabilistic sensitivity analysis. The probabilistic sensitivity analysis ran for 1000 iterations across all cohorts for the 3 different model entries, respectively. The number of times a strategy was the most cost-effective diagnostic option for each simulation (ie, produced the highest net benefit) was expressed as a percentage for all cohorts. Positive count data from the PROOF-BP risk algorithm test characteristics formed the parameters for a Dirichlet distribution.
In the base-case analysis (Table 3), the use of the PROOF-BP algorithm to triage for ABPM was cost-effective in all age and sex cohorts compared with the current NICE standard ABPM strategy and dominated the other comparators (saved costs and increased QALYs). This was because of the influence of treating otherwise unrecognized cases of masked hypertension (see Table S6 for number of initial misdiagnosis by strategy). For example, in a cohort of 1000 males aged 60 years, with a screening BP of ≥130/80 mm Hg, using the PROOF-BP algorithm would result in 62 more true hypertension cases detected, 5 more cardiovascular disease events prevented (excluding fatal chronic heart disease and cardiovascular disease deaths), 19.6 QALYs gained, and increased total costs by £32 929 compared with standard ABPM (£1680 per QALY gained). A detailed breakdown of costs and events for all age cohorts can be seen in Tables S7 and S8. Using the PROOF-BP algorithm, with a screening BP of ≥130/80 mm Hg, reduced the number of ABPM investigations in all age cohorts except for 40-year-old cohort (males and females) and 50-year-old females compared with the NICE standard ABPM strategy (Table S9).
The PROOF-BP algorithm was also cost-effective when the model entry was widened to individuals with a screening BP ≥120/70 mm Hg (Table S10). The probabilistic sensitivity analysis results indicated that for the base-case and ≥120/70 mm Hg model populations, PROOF-BP was the most cost-effective option in all iterations. When entry to the model was restricted to individuals with a screening BP ≥140/90 mm Hg (Table S11), PROOF-BP was the most cost-effective option except in the 40-year-old female cohort. Univariate sensitivity analysis (Table 4) demonstrated that the model was sensitive to the assumption of quality-of-life reduction from treatment. For example, if a quarter of the individuals experienced a quality-of-life reduction of 8% and everyone else experienced no ill effects of treatment, PROOF-BP was dominated (more costly, less health gain) by the standard ABPM strategy. Use of the PROOF-BP algorithm was also cost-effective compared with a strategy of utilizing ABPM in all individuals with a screening BP of ≥130/80 mm Hg, which was cheaper but resulted in fewer QALYs gained.
This represents the first economic evaluation to compare the cost-effectiveness of using the PROOF-BP algorithm with strategies to diagnose hypertension, which includes the consideration of individuals with potential masked hypertension. Targeted use of ABPM using the PROOF-BP algorithm was the most cost-effective diagnostic option for individuals presenting with a screening clinic BP of ≥130/80 mm Hg. The increased quality of life arising from the use of the PROOF-BP algorithm was mainly because of identification and treatment of masked hypertension (and the subsequent cardiovascular disease events avoided), which was ignored by the other strategies. The results were robust to several sensitivity analyses examining treatment disutility caused by side effects to medication, adjusting the masked hypertension prevalence, higher treatment costs, and increased use of ABPM in individuals with apparently normal screening BPs (<140/90 mm Hg). The findings suggest that a strategy of targeted use of ABPM in individuals with high or high–normal screening BP is likely to be cost-effective at a willingness to pay £20 000 per QALY gained and results in increased quality of life and lower mortality rate for individuals with hypertension.
Strengths and Weaknesses
The major strength of this work is that it represents a direct update of the cost-effectiveness model that informed the NICE hypertension guidance and which currently underpins the use of ABPM in routine clinical practice in the United Kingdom.2 This means that this new strategy of targeted use of ABPM using the PROOF-BP algorithm can be directly compared with the current UK reference standard approach for diagnosis of hypertension. A large number of sensitivity analyses were considered to test the robustness of assumptions in the model and consistently supported the base-case findings.
The original analysis included a detailed discussion including potential limitations with the original model, and these are not repeated here.4 Additional points are discussed below.
One limitation of the model as used here is that it assumed that individuals derived the same benefit from treatment of masked hypertension as applies to those with sustained hypertension. Although this has been suggested in many observational studies,25,40 no randomized trial of treatment versus no treatment in individuals with masked hypertension has yet been reported. One previous study did examine the efficacy of treatment based on ABPM rather than clinic readings and reported similar levels of BP control at follow-up but less treatment in the intervention arm.41 However, this study did not include individuals with masked hypertension. A trial of treatment of masked hypertension is currently underway in the United States42; however, this plans to enroll individuals with existing hypertension who are apparently controlled according to clinic BP, so the findings will not be directly relevant in the diagnostic scenario examined here. Until a randomized clinical trial of treatment in drug-naive individuals with masked hypertension is conducted, the true benefits of treatment will remain unknown. However, because the relationship between BP and vascular outcomes seems log-linear and predictable in epidemiological studies,19 it is reasonable to assume that treatment of masked (but true) hypertension carries similar benefit.
The present study used a prevalence of masked hypertension from the International Database on Ambulatory BP in relation to Cardiovascular Outcomes (IDACO).14 In fact, because of the difficulty recognizing masked hypertension in routine clinical practice, the true prevalence has been shown to vary, with estimates ranging from 8.5% to 16.6%.40,43 We examined the impact of varying prevalence in a sensitivity analysis, and the PROOF-BP algorithm remained cost-effective across the range assessed.
As with the previous model that informed the latest NICE hypertension guidance,4 and in keeping with the results of the recent HOPE-3 trial,44 the present analysis assumed that there was no benefit from treatment in individuals who were truly normotensive. This assumption has been challenged by the meta-analysis by Law et al19 and more recently the SPRINT trial,38 which support the prescription of treatment to those with BP levels of ≥130/80 mm Hg. However, SPRINT was a trial of individuals at a high risk, and <10% were treatment naive at baseline, limiting the applicability of those results to a modeled population of undiagnosed individuals undergoing screening for hypertension. Sensitivity analyses undertaken in the present study also revealed assuming equal risk reduction in normotensive patients or those undergoing intensive BP-lowering regimes would actually reduce the incremental cost-effectiveness ratio in favor of the PROOF-BP algorithm.
Although the PROOF-BP algorithm was cost-effective when the model entry was widened to individuals with a screening BP ≥120/70 mm Hg, the increased primary care workload burden from the additional ABPM investigations (Table S9) would likely make implementation infeasible (between 1.57 and 6.85 times more ABPM investigations by cohort than current practice).
Findings in the Context of Existing Literature
There are many economic analyses examining the cost-effectiveness and cost benefit of different BP-monitoring strategies in the diagnosis of hypertension. Previous studies from Australia, United States, and Europe have compared ABPM with CBPM,45–48 and further studies from Japan and the United States have compared HBPM with CBPM.49,50 The original cost-effectiveness model developed for NICE,2 which formed the basis for the present analyses, was the first to compare all 3 strategies. All previous analyses found diagnosis with out-of-office monitoring to be cost-effective but only examined individuals with a high screening BP and examined strategies that targeted the use of ABPM or HBPM monitoring at those most likely to benefit. A recent analysis compared the cost-effectiveness of central BP monitoring with CBPM and found the former to be cost-effective, although they did not compare it with ABPM or HBPM.51
The present analysis examined the cost-effectiveness of a new strategy designed to target the use of ABPM at those displaying a potential white coat or masked effect, something which has not been attempted before. Utilization of the PROOF-BP algorithm was found to be cost-effective at all ages and in males and females, primarily because of treatment of masked hypertension. Some variation by sex was observed, which may be attributable to the varying Framingham risk profile17 between sexes: females had a lower cardiovascular risk that limited the benefits of antihypertensive treatment.
Scenarios where the PROOF-BP risk algorithm was not the most cost-effective option centered on adjustments in treatment disutility. All strategies that increased the proportion of individuals receiving treatment (in the case of PROOF-BP, treating masked hypertension) were disadvantaged when quality-of-life decrement penalties attributable to treatment side effects were assumed. The level of treatment disutility associated with antihypertension medication is a matter of debate and may vary with age. The noninclusion of disutility in the present analysis base-case was consistent with previous modeling, which argued that where side effects exist, individuals can switch to alternative drugs.4
Implications for Clinical Practice
The present analyses suggest that using the PROOF-BP algorithm was likely to result in slightly higher healthcare costs (because of increased utilization of treatment in masked hypertensives) but improved quality of life in individuals screened for hypertension. The PROOF-BP algorithm is not currently used in routine clinical practice, but implementation would be possible with relative ease: automated BP monitors that take ≤3 consecutive readings (required for the decision tool) are now cheap and routinely available. The prediction algorithm is already available as an online calculator (https://sentry.phc.ox.ac.uk/proof-bp) and could easily be incorporated into general practice computer systems or built into smartphones linked to BP monitors. This strategy has the potential for individuals with apparently normal clinic BP to end up on treatment (if they have masked hypertension), which represents a notable shift from the current practice model and therefore would require some buy in from both patients and practitioners. Presenting the evidence and treatment options clearly, perhaps through formal patient and practitioner education, may be required, in much the same way that it accompanied the adoption of ABPM into routine primary care.
Current guidelines recommend use of out-of-office measurements to confirm hypertension diagnosis for individuals with raised clinic BP readings. The PROOF-BP algorithm considers both normal and raised clinic BP individuals with less reliance on out-of-office measurements to confirm hypertension diagnosis. Targeted use of ABPM (PROOF-BP algorithm) in the diagnosis of hypertension seems to be cost-effective compared with the conventional BP diagnostic options in primary care and would lead to reduced death and disability. Limitations of the model include the lack of data on the assumed efficacy of antihypertensive treatment for masked hypertension, which requires further investigation.
We would like to thank the researchers, patients, and practices who took part in the original BP-Eth (Blood Pressure in Ethnic groups), CAMBO (Conventional Versus Automated Measurement of Blood Pressure in the Office), HITS (Telemonitoring-Based Service Redesign for the Management of Uncontrolled Hypertension), Oxford self-monitoring study, TASMINH2 (Telemonitoring and Self-Management in the Control of Hypertension), and TASMINHSR (Targets and Self-Management for the Control of Blood Pressure in Stroke and at Risk Groups) studies without whom this work would have been impossible. We would also like to thank Roger Holder and David Yeomans for their assistance and support during the original conception, design, and dissemination of this study. The Predicting Out-of-Office Blood Pressure (PROOF-BP) investigators are J.P. Sheppard, R. Stevens, P. Gill, U. Martin, M. Godwin, J. Hanley, C. Heneghan, F.D.R. Hobbs, J. Mant, B. McKinstry, M. Myers, D. Nunan, B. Williams, S. Fleming, S. Stevens, R. Perera-Salazar, A. Ward, and R.J. McManus.
Sources of Funding
This work was funded by the National Institute for Health Research (NIHR) via a Programme Grant (RP-PG-1209–10051). J.P. Sheppard held a Medical Research Council (MRC) Strategic Skills Postdoctoral Fellowship (MR/K022032/1) and is now funded by the NIHR Oxford Collaborations for Leadership in Applied Health Research and Care (CLAHRC). R.J. McManus holds an NIHR Professorship (RP-02-12-015) and leads the self-management theme of the NIHR Oxford CLAHRC. F.D.R. Hobbs acknowledges support from the NIHR as Director of the NIHR School for Primary Care Research, Director of the NIHR CLAHRC Oxford, Theme Leader of the NIHR Oxford Biomedical Research Centre, NIHR Oxford Diagnostic Evidence Co-operative, and also from Harris Manchester College. B. Williams is a NIHR Senior Investigator and his research is supported by the NIHR University College London Hospitals Biomedical Research Centre. J. Mant is an NIHR Senior Investigator, and the Cambridge Primary Care Unit receives some infrastructure funding from the NIHR National School for Primary Care Research. K. Lovibond is employed by the National Guideline Centre, hosted by the Royal College of Physicians, which is funded by the National Institute for Health and Care Excellence. S. Greenfield is part funded by the NIHR Collaboration for Leadership in Applied Health Research and Care West Midlands.
R.J. McManus has received equipment for research purposes from Omron. C. Heneghan has received expenses and payments for his media work from Channel 4, British Broadcasting Corporation, FreshOne Television productions, and the Guardian and also expenses from the World Health Organization and the US Food and Drug Administration. He is also an expert witness in an ongoing medical device legal case, has received payment from British United Provident Association for analyzing and appraising guidelines and income from the publication of a series of tool kit books published by Blackwells. F.D.R. Hobbs has received limited research support in terms of blood pressure devices from Microlife and BpTRU. B. Williams works in academic collaboration with Healthstats, Singapore, in developing novel blood pressure–monitoring approaches. The other authors report no conflicts.
The views and opinions expressed are those of the authors and do not necessarily reflect those of the Medical Research Council, National Health Service, National Institute of Health and Care Excellence, Royal College of Physicians, National Institute for Health Research, or the Department of Health.
The online-only Data Supplement is available with this article at http://hyper.ahajournals.org/lookup/suppl/doi:10.1161/HYPERTENSIONAHA.117.10244/-/DC1.
- Received August 29, 2017.
- Revision received September 14, 2017.
- Accepted November 12, 2017.
- © 2017 American Heart Association, Inc.
- 2.↵National Institute for Clinical Excellence. Clinical guidance 127: hypertension (partial update). 2011. http://www.nice.org.uk/guidance/cg127. Accessed July 6, 2015.
- 3.↵National Clinical Guideline Centre. Appendix J: Cost-effectiveness analysis – blood pressure monitoring for confirming a diagnosis of hypertension. In: Hypertension: Clinical Management of Primary Hypertension in Adults. London, UK: National Clinical Guideline Centre; 2011:446–502.
- Lovibond K,
- Jowett S,
- Barton P,
- Caulfield M,
- Heneghan C,
- Hobbs FD,
- Hodgkinson J,
- Mant J,
- Martin U,
- Williams B,
- Wonderling D,
- McManus RJ
- Dasgupta K,
- Quinn RR,
- Zarnke KB,
- et al
- Pickering TG,
- Davidson K,
- Gerin W,
- Schwartz JE
- Sheppard JP,
- Stevens R,
- Gill P,
- Martin U,
- Godwin M,
- Hanley J,
- Heneghan C,
- Hobbs FD,
- Mant J,
- McKinstry B,
- Myers M,
- Nunan D,
- Ward A,
- Williams B,
- McManus RJ
- 12.↵NatCen Social Research and University College London. Department of Epidemiology and Public Health, Health Survey for England, 2013 [computer file]. Colchester, Essex, UK: UK Data Archive; 2015.
- Hodgkinson J,
- Mant J,
- Martin U,
- Guo B,
- Hobbs FD,
- Deeks JJ,
- Heneghan C,
- Roberts N,
- McManus RJ
- Conen D,
- Aeschbacher S,
- Thijs L,
- et al
- 15.Office for National Statistics. National Life Tables, England & Wales, 1980–82 to 2011–13. 2014. http://www.ons.gov.uk/ons/rel/lifetables/national-life-tables/2011–2013/rft-ew.xls. Accessed July 3, 2015.
- 16.Office for National Statistics. Deaths Registered in England and Wales, 2013. 2014. http://www.ons.gov.uk/ons/rel/vsob1/death-reg-sum-tables/2013/rft-deaths-summary-tables-2013.xls. Accessed July 3, 2015.
- Law MR,
- Morris JK,
- Wald NJ
- 21.National Institute for Health and Clinical Excellence. Unstable Angina and NSTEMI: Early Management (CG94). London, UK: NICE; 2010.
- Dennis M,
- Bamford J,
- Sandercock P,
- Warlow C
- 26.↵NatCen Social Research and University College London. Department of Epidemiology and Public Health, Health Survey for England, 2012 [computer file]. Colchester, Essex, UK: UK Data Archive; 2014.
- Curtis L
- Luengo-Fernandez R,
- Gray AM,
- Rothwell PM
- 29.National Institute for Health and Care Excellence. Lipid Modification: Cardiovascular Risk Assessment and the Modification of Blood Lipids for the Primary and Secondary Prevention of Cardiovascular Disease. London, UK: NICE; 2014.
- 30.National Institute for Health and Clinical Excellence. Clopidogrel and Modified-Release Dipyridamole for the Prevention of Occlusive Vascular Events: Review of NICE Technology Appraisal Guidance 90. London, UK: NICE; 2010.
- 31.Joint Formulary Committee. British National Formulary. 69th Ed. London, UK: BMJ Group and Pharmaceutical Press; 2015.
- Palmer S,
- Sculpher M,
- Philips Z,
- Robinson M,
- Ginnelly L,
- Bakhai A,
- Packham C,
- Abrams K,
- Cooper N,
- Alfikah K,
- Flather M
- 34.Department of Health. NHS Reference Costs 2013/2014. 2014. https://www.gov.uk/government/publications/nhs-reference-costs-2013-to-2014, 2015. Accessed April 14, 2015.
- 35.National Institute for Health and Care Excellence. Management of Stable Angina. London, UK: NICE; 2011.
- Appleby J,
- Devlin N,
- Parkin D
- 37.↵National Institute for Health and Care Excellence. Guide to the Methods of Technology Appraisal. London, UK: NICE; 2013: https://www.nice.org.uk/process/pmg9/. Accessed June 19, 2015.
- Wood S,
- Greenfield SM,
- Sayeed Haque M,
- Martin U,
- Gill PS,
- Mant J,
- Mohammed MA,
- Heer G,
- Johal A,
- Kaur R,
- Schwartz C,
- McManus RJ
- Staessen JA,
- Byttebier G,
- Buntinx F,
- Celis H,
- O’Brien ET,
- Fagard R
- Drawz PE
- Mancia G,
- Facchetti R,
- Bombelli M,
- Grassi G,
- Sega R
- Krakoff LR
- Pierdomenico SD,
- Mezzetti A,
- Lapenna D,
- Guglielmi MD,
- Mancini M,
- Salvatore L,
- Antidormi T,
- Costantini F,
- Cuccurullo F
- Fukunaga H,
- Ohkubo T,
- Kobayashi M,
- Tamaki Y,
- Kikuya M,
- Obara T,
- Nakagawa M,
- Hara A,
- Asayama K,
- Metoki H,
- Inoue R,
- Hashimoto J,
- Totsune K,
- Imai Y
- Arrieta A,
- Woods JR,
- Qiao N,
- Jay SJ
Novelty and Significance
What Is New?
This study considered the merits (cost-effectiveness) of using the Predicting Out-of-Office Blood Pressure algorithm to triage for ambulatory blood pressure monitoring in the diagnosis of hypertension.
This is the first economic evaluation of different methods of diagnosing hypertension, which includes the consideration of individuals with potential masked hypertension.
What Is Relevant?
Predicting Out-of-Office Blood Pressure algorithm considered a broader screening population in terms of clinic blood pressure for diagnosis but necessitated less ambulatory blood pressure monitoring investigations in most age cohorts compared with current guidelines.
Economic modeling suggest that such an approach would be cost-effective compared with conventional blood pressure diagnostic options in primary care.