Effectiveness and cost-effectiveness of team-based care for hypertension: a meta-analysis and simulation study

Abstract

Background:

Team-based care (TBC), a team of two or more healthcare professionals working collaboratively toward a shared clinical goal, is a recommended strategy to manage blood pressure (BP). However, the most effective and cost-effective TBC strategy is unknown.

Methods:

A meta-analysis of clinical trials in US adults (aged ≥20 years) with uncontrolled hypertension (≥140/90 mmHg) was performed to estimate the systolic BP (SBP) reduction for TBC strategies vs. usual care at 12 months. TBC strategies were stratified by the inclusion of a non-physician team member who could titrate antihypertensive medications. The validated BP Control Model-Cardiovascular Disease (CVD) Policy Model was used to project the expected BP reductions out to 10 years, and simulate CVD events, direct healthcare costs, quality-adjusted life years (QALYs), and cost-effectiveness of TBC with physician and non-physician titration.

Results:

Among 19 studies comprising 5,993 participants, the 12-month SBP change vs. usual care was −5.0 (95% confidence interval: −7.9 to −2.2) mmHg for TBC with physician titration and −10.5 (−16.2 to −4.8) mmHg for TBC with non-physician titration. Relative to usual care at 10 years, TBC with non-physician titration was estimated to cost $95 (95% uncertainty interval: -$563 to $664) more per patient and gain 0.022 (0.003 to 0.042) QALYs, costing $4,400/QALY gained. TBC with physician titration was estimated to cost more and gain fewer QALYs than TBC with non-physician titration.

Conclusions:

TBC with non-physician titration yields superior hypertension outcomes compared with other strategies and is a cost-effective way to reduce hypertension-related morbidity and mortality in the US.

INTRODUCTION

Blood pressure (BP) control rates worsened by 10% from 2013-2014 to 2017-2018 despite the availability of effective antihypertensive medications.¹ In order to improve BP control and reduce associated cardiovascular disease (CVD) morbidity and mortality, it is important to understand the most efficient and effective care delivery strategies to guide program design and health policy development.¹

Team-based care (TBC), i.e., healthcare services delivered by two or more providers working collaboratively, is a hypertension control strategy recommended by the 2017 American College of Cardiology (ACC)/American Heart Association (AHA) BP guideline and the 2020 US Surgeon General’s Call to Action to Control Hypertension.^2-4 TBC has been associated with improved rates of BP control compared with usual primary care.^{5, 6} Additionally, TBC with non-physician team members who can manage medications (e.g., pharmacists, nurse practitioners), is associated with greater BP reductions compared with programs where only physicians can manage medications.⁶ However, training and deploying non-physician team members to manage antihypertensive medications may be costly to clinics and healthcare systems.

Reversing recent BP control trends and meeting quality benchmarks may require healthcare organizations to substantially change usual hypertension care processes.⁷ However, the most clinically effective and cost-effective TBC design, including TBC impact on hypertension care processes (e.g., provider visit frequency, clinical inertia, and optimal length of the program) is unknown. We therefore performed a meta-analysis and simulation to estimate the effectiveness and cost-effectiveness of TBC program strategies for hypertension management.

METHODS

The data that support the findings of the meta-analysis are available from Dr. Bellows upon reasonable request. The simulation model and key inputs used for the cost-effectiveness analysis are available on reasonable request and approval by the research team. Interested researchers will be asked to submit a 1- to 2-page research proposal and collaboration plan to Dr. Bellows and will be required to sign a Creative Commons agreement.

Meta-Analysis

Data Searches and Study Inclusion

Randomized controlled trials were identified through a literature search performed using PubMed and from prior systematic reviews and meta-analyses (Supplemental Methods).^{5, 6, 8} Studies were included if they were (1) performed in the US, (2) published after 1/1/2003, (3) had a duration of ≥6 months, (4) included a TBC arm, and (5) included participants with uncontrolled BP at baseline, defined as systolic BP (SBP) ≥140 mmHg or ≥130 mmHg in those with diabetes or chronic kidney disease. Studies were excluded if they did not report sufficient data on BP outcomes, did not report outcomes at six months, or did not report information such as standard deviations or confidence intervals for BP outcomes.

Data Extraction

Data extraction was performed using a standardized tool in the Covidence systematic review software (Veritas Health Innovation, Melbourne, Australia). Data were extracted by authors KBB, ASR, LPC, and BKB, with all data extracted verified by a second team member. All data extracted are described in the Supplemental Methods.

Outcomes

The primary outcomes were SBP change at six and twelve months for TBC vs. usual care (Supplemental Methods). Secondary outcomes included diastolic BP (DBP) change and BP control at six and twelve months. All outcomes assessed compared with usual care, which was generally defined across studies as hypertension management by a primary care physician without non-physician team member support.

Quality Assessment

We used the Cochrane Risk of Bias 2 tool for randomized controlled trials to assess the quality of the data collected.⁹ Each paper was assessed by two independent reviewers (KBB, ASR, LPC, and BKB), with differences resolved by consensus with a third reviewer.

Statistical Analyses

The meta-analysis adhered to practices recommended in the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (Table S1). The meta-analysis was performed using Stata version 16.1 (StataCorp LLC, College Station, TX) using random effects models to generate weighted mean differences of SBP and DBP outcomes and percent difference in BP control. Analyses were stratified by whether a physician or non-physician team (e.g., pharmacist, nurse practitioner) member performed antihypertensive medication titration. When there were two or more intervention arms meeting inclusion criteria within a given study, they were each included in the meta-analysis, and labeled sequentially. Funnel plots were used to examine the potential for publication bias. A “leave-one-out” sensitivity analysis was performed to examine the impact of excluding each study from the analysis one at a time to identify studies that might have skewed the results. Meta-regression was used to examine heterogeneity by baseline SBP.

Cost-Effectiveness Analysis

Model Overview

The BP Control-CVD Policy Model (BP-CVDPM) is a hybrid, discrete event simulation model combining the independently validated BP Control Model and CVD Policy Model (Figure S1).¹⁰ The BP-CVDPM predicts long-term BP outcomes by simulating modifiable hypertension care management processes and subsequent effect on CVD events, direct healthcare costs, survival, and quality-adjusted survival (Supplemental Methods).

Population

The model simulated a nationally representative cohort of 100,000 US adults (aged ≥20 years) with uncontrolled BP (SBP/DBP ≥140/90 mmHg) identified from the National Health and Nutrition Examination Survey (NHANES) 1999-2018 cycles. Included individuals from NHANES had lifetime CVD risk factor trajectories multiply imputed from participants in the National Heart, Lung, and Blood Institute Pooled Cohorts Study (Supplemental Methods).^10-14

Comparators

In the primary analysis, three strategies were compared: usual care, TBC with physician antihypertensive medication titration, and TBC with non-physician antihypertensive medication titration (Table S2). The usual care strategy was the same as previously published.^{7, 10, 15} All TBC visits were assumed to be in addition to usual care visits. For TBC with physician antihypertensive medication titration, it was assumed individuals had TBC visits with a registered nurse where BP was measured and hypertension counseling, including medication adherence, was provided. If BP was uncontrolled at these visits, it was assumed that this was communicated to the physician who then determined if treatment should be intensified. TBC visits with non-physician medication titration were assumed to be the same except that care was provided by a pharmacist who could also intensify treatment if BP was uncontrolled.

In the primary analysis, TBC visits were assumed to occur for twelve months, matching the meta-analysis; this was explored in sensitivity analysis. After TBC visits had ended, individuals were assumed to return to usual care. Both TBC arms were assumed to have beneficial effects on medication adherence (Supplemental Methods), which decreased linearly over one year after the end of TBC visits until no adherence benefits over usual care remained.^{10, 15}

Model Inputs

In the primary analysis, the frequency of physician office visits (once every 11.5 weeks) and TBC visits (once every six weeks) matched the frequency from the meta-analysis intervention during the first twelve months (Table 1). After the first twelve months, TBC visits ended, and usual care visit frequency was dependent upon BP control and patient and visit characteristics.^{7, 10}

Table 1.

Selected BP-CVDPM team-based care inputs.

Team-Based Care or Self-Monitoring Intervention Components	Base-case	Lower Range	Upper Range	Source
Processes of Hypertension Care
Weeks between visits during intervention				Current meta-analysis
TBC visit	6.0	2.0	12.0
Physician visit	11.5	-	-
Probability of treatment intensification for TBC with non-physician titration when SBP				Bryant et al.,10 calibrated
140-149 mmHg	0.713	0.330	1.000
≥150 mmHg	0.950	0.330	1.000
Probability of treatment intensification by physician when BP uncontrolled				Bellows et al.7
First antihypertensive medication
≥160 mmHg without diabetes or CKD, ≥140/90 mmHg with diabetes or CKD	0.333	0.313	0.440
BP otherwise uncontrolled	0.208	0.207	0.310
All subsequent medication intensifications	0.130	0.065	0.195
Relative reduction in non-adherence with intervention	0.400	-	-	Calibrated, Assumption
Costs (2021 USD)
Minutes per first TBC visit	42.5	25.0	60.0	Bryant et al.10
Minutes per TBC visit when measured BP
<140/90 mmHg	15.0	12.5	45.0
≥140/90 mmHg	30.0	15.0	52.5
Pharmacist hourly wage	$60	$41	$79	Bureau of Labor Statistics
Nurse hourly wage	$38	$26	$56
Physician review following TBC with physician titration	$23	$20	$30	CMS Physician Fee Schedule, CPT 99211
Employee benefit rate	30.0%	-	-	Assumption

BP – blood pressure, CKD – chronic kidney disease, SBP – systolic blood pressure, TBC – team-based care, USD – United States dollars.

Notes: The table shows the mean base-case values, and the upper and lower values used in sensitivity analysis, for selected model inputs related to the TBC interventions.

Costs for TBC visits were calculated using the national median hourly wage for registered nurses and pharmacists, derived from the US Bureau of Labor Statistics, and duration of the visit derived from interviews with providers in prior analyses, and an assumed 30% rate for the cost of employee benefits (Table 1).^{10, 16, 17} For TBC with physician titration, an additional cost was added for physicians to review notes and determine if treatment intensification was needed using (CPT code: 99211). All other model inputs, including other costs, were derived from published literature and analysis of publicly available databases (Table S3).

Outcomes

The primary outcomes were total direct healthcare costs (2021 USD), quality-adjusted life years (QALYs), and incremental cost-effectiveness ratios (ICERs) at 10 years from a healthcare sector perspective.¹⁸ All future costs and QALYs were discounted 3% annually.¹⁸ TBC was classified as highly cost-effective when the ICER was <$50,000 per QALY gained and at least intermediately cost-effective when <$150,000 per QALY gained.¹⁹ Secondary outcomes included CVD events, survival, SBP changes, proportion with controlled BP (<140/90 mmHg), disaggregated costs, and disaggregated QALYs.

Statistical Analyses

As in prior analyses, the BP-CVDPM was calibrated to match contemporary US CVD event and mortality rates and reproduce the difference in SBP with TBC interventions vs. usual care at six and twelve months from the meta-analysis (Supplemental Methods, Figures S2-S3, Table S4).^{10-12, 15} The analyses adhered to practices recommended in the Consolidated Health Economic Evaluation Reporting Standards (Table S5).²⁰ The BP-CVDPM was programmed in TreeAge Pro 2021 (TreeAge Software Inc, Williamstown, MA) with additional analyses performed using R version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria). To ensure stable cost-effectiveness estimates, 100,000 individuals were simulated in all analyses. The mean and uncertainty interval (UI) for all outcomes were derived from 100 probabilistic iterations, where input parameters were randomly sampled from pre-specified statistical distributions. Uncertainty in the ICER was reported as the proportion of iterations with an ICER <$50,000 and <$150,000 per QALY gained.

Sensitivity Analyses

One-way sensitivity analyses were used to examine the effect of independently varying the model inputs shown in Table 1 across plausible ranges on the cost-effectiveness of TBC, while holding all other inputs at their base-case values (Supplemental Methods). Additional analyses examined the impact of extending the duration over which TBC visits occurred to 5 and 10 years, with the same visit frequency as occurred in the first year. Separately, a “light touch” scenario was simulated in which individuals only had one TBC visit per year after the initial twelve months with 5- and 10-year durations of TBC visits. As costs beyond those required to directly provide patient care may be incurred with TBC, a scenario analysis included administrative costs of TBC (Supplemental Methods).¹⁰

RESULTS

Meta-Analysis

A total of 19 studies were identified from the prior meta-analyses and updated literature search, comprising 5,993 individual participants (Figure S4).^{17, 21-39} The mean (SD) baseline age was 58.9 (6.8), 35.3% were female, 37.7% were Black, and mean baseline SBP was 148.1 (10.3) mmHg (Tables S6-S7). The components included in TBC (e.g., web-based communications, self-monitoring of BP) and workflow for hypertension management varied across trials (Table S7). The overall risk of bias was rated as low in 63% of the studies, some concerns of bias in 32%, and high in 5% (Figure S5).

A total 12 studies (3,322 participants) and 7 studies (3,128 participants) were included in the meta-analyses of SBP change at six and twelve months, respectively. At six and twelve months, TBC overall had 7.2 mmHg (95% confidence interval [CI] −10.6, −3.8) and 8.1 mmHg (95% CI −12.1, −4.1) greater reductions in SBP than usual care (Figure 1), respectively. SBP reductions vs. usual care varied by the team member titrating antihypertensive mediations; physician titration resulted in 4.7 mmHg (95% CI −8.1, −1.4) and 5.0 mmHg (95% CI −7.9, −2.2) greater reductions, and non-physician titration in 11.5 mmHg (95% CI −15.7, −7.4) and 10.5 mmHg (95% CI −16.2, −4.8) greater reductions, at six and twelve months, respectively.

Figure 1. Systolic blood pressure changes with team-based care vs. usual care.

A) Six months

B) Twelve months

WMD – weighted mean difference.

Notes: The figure shows the results of the meta-analysis for the change in systolic blood pressure for team-based care vs. usual care at 6 months in Panel A and 12 months in Panel B.

The results for DBP and BP control followed the same patterns as SBP (Figures S6-S7). For the outcomes of SBP change and BP control at 6 months, there did not appear to be evidence of publication bias. However, for other outcomes, asymmetry in the funnel plots was observed (Figure S8). The “leave-one-out” sensitivity analysis indicated the SBP reduction at twelve months for TBC with non-physician titration was sensitive to the inclusion of the Victor et al. 2019 study, decreasing to −8.3 mmHg (95% CI −10.3, −6.3) when removed (Figure S9).¹⁷ The meta-regression showed that greater reductions in SBP at six and twelve months with TBC vs. usual care were associated with higher baseline SBP (Figure S10).

Cost-effectiveness

Primary Analysis

The mean (SD) baseline age of the simulated population of US adults with uncontrolled hypertension was 59.7 (16.4) years, 50.1% were female, and mean baseline SBP was 150.6 (14.3) mmHg (Table S8). At 10 years, the mean SBP with usual care was projected to be 137.9 (95% UI 136.6, 139.9) mmHg; twelve months of TBC was estimated to reduce SBP relative to usual care at 10 years by 0.8 (95% UI −1.0, −0.7) mmHg with physician titration and 3.3 (95% UI −3.9, −2.6) mmHg with non-physician titration (Figure 2, Table 2). TBC was estimated to reduce the number of individuals with ≥1 CVD event by 5 (95% UI 3, 8) per 1000 treated with physician titration and 11 (95% UI 8, 14) with non-physician titration.

Figure 2. Projected systolic blood pressure outcomes and cost-effectiveness of team-based care.

A) Projected systolic blood pressure over time

B) Incremental costs and quality-adjusted life years vs. usual care

C) Cost-effectiveness acceptability curve

QALY – quality-adjusted life years, SBP – systolic blood pressure, TBC – team-based care, USD – United States dollars.

Notes: The figure shows the projected changes in SBP with each strategy over the simulated time horizon in Panel A, the incremental costs and incremental QALYs for TBC vs. usual care in Panel B, and the probability of TBC and usual care being the preferred strategy as the cost-effectiveness threshold is varied in Panel C. The shaded region in Panel A shows the 95% uncertainty interval from 100 probabilistic iterations of the model.

Table 2.

Clinical and cost-effectiveness outcomes at 10 years.

	Usual Care	TBC with Physician Titration	TBC with Non- physician Titration
Clinical Outcomes
SBP (mmHg)	137.9 (136.3, 139.5)	137.0 (135.5, 138.7)	134.6 (133.5, 135.8)
Difference	Ref	−0.8 (−1.0, −0.7)	−3.3 (−3.9, −2.6)
BP control	66.6% (57.8%, 75.4%)	70.0% (61.2%, 78.8%)	80.6% (73.5%, 86.7%)
Difference	Ref	3.4% (2.9%, 3.9%)	13.9% (10.8%, 17.0%)
Incident CVD events	16.9% (14.2%, 20.1%)	16.4% (13.6%, 19.5%)	15.3% (12.7%, 18.1%)
Difference	Ref	−0.5% (−0.8%, −0.3%)	−1.1% (−1.4%, −0.8%)
Cost-effectiveness Outcomes
Mean total direct healthcare costs (2021 USD)	$105,085 ($99,127, $110,330)	$105,548 ($99,798, $110,760)	$105,180 ($99,731, $110,359)
Difference	Ref	$463 ($126, $770)	$95 (-$563, $664)
TBC visit costs*	Ref	$373 ($369, $378)	$231 ($226, $235)
Medication costs*	Ref	$105 ($94, $117)	$388 ($340, $436)
CVD costs*	Ref	−$123 (-$212, -$24)	−$430 (-$586, -$304)
Mean QALYs	6.116 (5.926, 6.292)	6.128 (5.937, 6.3)	6.138 (5.945, 6.314)
Difference	Ref	0.012 (0.001, 0.021)	0.022 (0.003, 0.042)
ICER ($/QALY gained)	Ref	Dominated**	$4,390
Probability preferred strategy at cost-effectiveness threshold of $50,000/QALY	0%	0%	100%

BP – blood pressure, CVD – cardiovascular disease, ICER – incremental cost-effectiveness ratio, QALY – quality-adjusted life year, Ref – reference group, SBP – systolic blood pressure, TBC – team-based care, USD – United States dollars.

^*Only selected cost types are shown and thus do not sum to the overall difference. Additional costs include, background (non-hypertension related) healthcare, non-TBC physician office visits, and adverse events.

^**TBC with physician titration is dominated by TBC with non-physician titration, meaning it costs more and is less effective.

Notes: The table shows the projected clinical and economic outcomes of usual care, TBC with physician titration of medications, and TBC with non-physician titration of medications at 10 years. TBC was assumed to occur during the first year of the simulation. After the first year, hypertension care processes (e.g., visit frequency, probability of medication titration, medication adherence) were assumed to return to usual care. Values are mean and 95% uncertainty interval (2.5th to 97.5th percentile) derived from 100 probabilistic iterations of 100,000 individuals.

The estimated per-patient cost for TBC visits over twelve months was $373 (95% UI $369, $378) with physician titration and $231 (95% UI $226, $235) with non-physician titration. Relative to usual care at 10 years, TBC with non-physician titration was estimated to cost $4,400/QALY gained and was the preferred strategy in 100% of model iterations at a cost-effectiveness threshold of <$50,000/QALY gained (Figure 2). TBC with non-physician titration was estimated to dominate (i.e., cost less and be more effective) TBC with physician titration in 30% of probabilistic analyses.

Sensitivity Analysis

The cost-effectiveness of TBC with non-physician titration was most sensitive to the probability of non-physician medication titration when BP was uncontrolled and frequency of TBC visits (Figure S11). Relative to usual care, TBC with non-physician titration remained highly cost-effective even with only one TBC visit every 12 weeks. In the scenario analyses with increased duration of TBC to 5 and 10 years, TBC with physician titration continued to be dominated by TBC with non-physician titration, which was highly or intermediately cost-effective vs. usual care (Table S9). In the “light touch” scenarios with 5 and 10 years of TBC but with only one visit per year in subsequent years, TBC with non-physician titration continued to be highly cost-effective relative to usual care. In the scenario analysis including administrative costs, TBC with non-physician titration remained highly cost-effective with an ICER $17,400/QALY gained and continued to dominate TBC with physician titration. If TBC with non-physician titration could be implemented for <$1,100 or <$3,300 per patient enrolled, it would remain highly or intermediately cost-effective, respectively.

DISCUSSION

Hypertension TBC with a non-physician team member (e.g., pharmacist) titrating antihypertensive medications yields better BP outcomes than usual care or TBC with physician medication titration. Additionally, TBC with a non-physician team member titrating medications is highly cost-effective vs. usual care (ICER $4,400/QALY gained) and is estimated to cost less and be more effective than TBC with physician titration. Health care organizations that have capacity to deploy non-physicians to titrate antihypertensive medications as part of a TBC program may face up-front implementation costs, but our analysis suggests that this approach will yield health and economic value over time.

The greater BP reduction with TBC with non-physician medication titration that we estimated is consistent with prior meta-analyses upon which our analysis expanded.^{5, 6} Few prior economic analyses have estimated the impact of TBC for hypertension management. One analysis of US adults estimated that 1-year of hypertension TBC delivered by a pharmacist or nurse would decrease SBP by 8.1 mmHg relative to usual care at a per-patient cost of $1,046 (inflated to 2021 USD) for office and telephone visits.⁴⁰ Another analysis of a pharmacist-led home BP telemonitoring intervention found a decrease in SBP by 9.7 mmHg relative to usual care at 1 year with a total per-patient cost of $1,457 (inflated to 2021 USD). This estimate did not substantially differ from usual care costs and included pharmacist visits, telemonitoring costs, and savings in medical care but did not include longer-term projections or impact on quality of life.^{35, 41} The current study builds upon these analyses by estimating the impact of varying the processes of care driving changes in BP, which may be useful for health systems in designing and implementing TBC that fits into existing frameworks.

Healthcare systems are often reimbursed relative to their success in controlling BP, among other quality metrics, introducing financial incentives to optimize care delivery. Our analysis indicates that use of non-physician team members with the ability to prescribe and titrate antihypertensive medications, e.g., clinical pharmacists under a collaborative practice agreement, may improve BP and CVD outcomes in a cost-effective manner. However, health systems may encounter barriers to implementing these programs in a sustainable fashion due to challenges with team member availability, variability in state laws allowing collaborative practice agreements between physicians and pharmacist, insufficient reimbursement for pharmacist services, and high upfront training and implementation costs. However, successful TBC programs demonstrate that these obstacles can be overcome.^{35, 42, 43} With persistent decline in BP control rates on a national scale and disruptions to care because of the COVID-19 pandemic, establishing a framework for TBC for hypertension control is imperative to improve CVD outcomes.

This work is strengthened by use of the BP-CVDPM and the ability to simulate processes of hypertension care that drive changes in BP. However, these results should be considered in the context of the following limitations. There is a lack of representation of racial and ethnic minority populations in the original trials, as most studies in our meta-analysis had a majority white population, with few exceptions.^{17, 27} Additionally, the impacts of social determinants of health were not captured in our analysis. We were therefore unable to explore TBC designs that account for different outcomes that may result from differences in access to care, affordable medications, adequate nutrition, and safe, stable environments. Future research should examine the impact of these factors and how TBC may be implemented in ways that promote health equity and decrease disparities.⁴⁴ Cost data from individual trials were not available, trials may not have intended to perform cost-effectiveness analyses, and cost-effectiveness components were not necessarily systematically reported; therefore, national estimates were used in our analysis. Though our analysis indicates TBC with a non-physician who can titrate medications improves BP control, detailed workflows and intervention components (e.g., inclusion of self-monitoring of BP) varied across trials and their contribution to BP lowering with TBC were not assessed in our analysis. Meta-analyses show the intensity of a self-monitoring of BP intervention is positively associated with BP lowering.⁸ There may be a synergistic effect on BP lowering achieved by combining TBC with self-monitoring of BP.^{8, 15} As uptake of self-monitoring increases both because of improved guideline adherence and transition to and persistence of telemedicine in the wake of the COVID-19 pandemic, more data are likely to be available to better understand the impact of self-monitoring of BP on TBC.

This work was conducted from a healthcare sector perspective, which includes direct healthcare costs regardless of the payer, and used national averages for costs of providers and medications, which may not be applicable to local health systems. Training and licensing costs for non-physicians were not included and other costs of implementing TBC (e.g., time for scheduling TBC visits, BP measurement equipment) may not be fully accounted for in our analysis. However, incorporating administrative costs for TBC in our analysis did not alter cost-effectiveness conclusions.

NOVELTY AND RELEVANCE.

What Is New?

• Hypertension care provided by a team of providers that includes a non-physician who can titrate antihypertensive medications cost-effectively reduces systolic blood pressure and cardiovascular disease risk

What Is Relevant?

• Hypertension control has worsened in the US over the last decade and team-based care is recommended by hypertension treatment guidelines to improve control

Clinical Implications?

• Health systems providing team-based care for hypertension should prioritize inclusion of non-physician providers who can titrate antihypertensive medications

PERSPECITVES.

TBC with non-physician titration is a clinically effective and cost-effective strategy to improve hypertension control and help reverse the current national trends. These results should be used to inform policy with the potential to restructure reimbursement to facilitate implementation. Clinics and health systems actively implementing TBC can use these results to guide program design.

NON-STANDARD ABBREVIATIONS

ACC

American College of Cardiology

AHA

American Heart Association

BP-CVDPM

Blood Pressure Control-Cardiovascular Disease Policy Model

ICER

incremental cost-effectiveness ratio

NHANES

National Health and Nutrition Examination Survey

QALY

quality-adjusted life year

TBC

team-based care

UI

uncertainty interval