Skip main navigation

Hypertension Trends and Disparities Over 12 Years in a Large Health System: Leveraging the Electronic Health Records

Originally publishedhttps://doi.org/10.1161/JAHA.123.033253Journal of the American Heart Association. 2024;13:e033253

Abstract

Background

The digital transformation of medical data enables health systems to leverage real‐world data from electronic health records to gain actionable insights for improving hypertension care.

Methods and Results

We performed a serial cross‐sectional analysis of outpatients of a large regional health system from 2010 to 2021. Hypertension was defined by systolic blood pressure ≥140 mm Hg, diastolic blood pressure ≥90 mm Hg, or recorded treatment with antihypertension medications. We evaluated 4 methods of using blood pressure measurements in the electronic health record to define hypertension. The primary outcomes were age‐adjusted prevalence rates and age‐adjusted control rates. Hypertension prevalence varied depending on the definition used, ranging from 36.5% to 50.9% initially and increasing over time by ≈5%, regardless of the definition used. Control rates ranged from 61.2% to 71.3% initially, increased during 2018 to 2019, and decreased during 2020 to 2021. The proportion of patients with a hypertension diagnosis ranged from 45.5% to 60.2% initially and improved during the study period. Non‐Hispanic Black patients represented 25% of our regional population and consistently had higher prevalence rates, higher mean systolic and diastolic blood pressure, and lower control rates compared with other racial and ethnic groups.

Conclusions

In a large regional health system, we leveraged the electronic health record to provide real‐world insights. The findings largely reflected national trends but showed distinctive regional demographics and findings, with prevalence increasing, one‐quarter of the patients not controlled, and marked disparities. This approach could be emulated by regional health systems seeking to improve hypertension care.

Nonstandard Abbreviations and Acronyms

DBP

diastolic blood pressure

NHANES

National Health and Nutrition Examination Survey

SBP

systolic blood pressure

Clinical Perspective

What Is New?

  • Analysis of real‐world data from the electronic health record of a regional health system revealed distinctive regional findings that will inform strategies and initiatives for improving hypertension control at a regional level.

What Are the Clinical Implications?

  • Our regional analysis generated knowledge about hypertension prevalence and racial disparities that will lead to targeted quality improvement efforts.

  • Real world data can provide actionable insights about hypertension and disparities in a specific region that could inform regional system strategies and initiatives for improvement.

Hypertension is a persistent and challenging health problem in the United States, estimated to affect almost half of adult Americans.1 Of patients with hypertension, approximately half have uncontrolled blood pressure (BP),2 which disproportionately affects Black individuals.2, 3, 4 Improving the detection and treatment of hypertension is a national priority and offers an important opportunity to modify the risk for cardiovascular disease and stroke and to address a major racial disparity.5

A 2014 advisory from the American College of Cardiology, American Heart Association, and the US Centers for Disease Control and Prevention called for health system approaches for addressing hypertension.6 System‐wide strategies, along with improvements at the clinician and patient level, could help optimize BP control and effectively modify the risk associated with hypertension.6, 7 To help design health system strategies, more information is needed to characterize hypertension trends and identify opportunities for improvement at the regional health system level, ideally using readily available real‐world evidence from the electronic health record (EHR) systems.

There are opportunities to broaden the ways a regional health system might use EHR data to investigate hypertension. Converting EHR data into a common data model can provide a more agile analytical platform, which could enable a health system to take advantage of its long‐term data, explore various analytical methods, and evaluate trends and demographics that may be unique to a health system's region. Analysis of data at the regional level could enable customized regional initiatives to improve hypertension care.

Accordingly, we leveraged the EHR from a large health system caring for a diverse population to evaluate performance in the care of patients with hypertension.8 EHR data over 12 years were transformed into a harmonized and validated data set that has enabled a serial cross‐sectional analysis of hypertension in a large regional population.9 We specifically sought to identify trends using various operational hypertension definitions to investigate disparities and opportunities for improvement.

Methods

Data Source

This retrospective observational study was performed at a large nonprofit integrated health care system in Virginia and northeastern North Carolina. The system began using a centralized EHR system designed by the Epic Corporation in 2007. In 2021, the health system began extracting key clinical data from its EHR and transforming the data into a harmonized data platform using the Observational Medical Outcomes Partnership common data model, version 5.3.9

Study Population

The overall study population consisted of all adult patients (aged ≥18 years) who had at least 1 outpatient blood pressure reading recorded between January 1, 2010, and December 31, 2021. Emulating the National Health and Nutrition Examination Survey (NHANES) study,2 we conducted a serial cross‐sectional study by independently analyzing the data in 2‐year cycles from 2010 to 2011 to 2020 to 2021. In contrast to the defined clinical protocols of NHANES, our real‐world data approach required defining computational protocols to analyze recorded BP readings in the EHR to create operational definitions of hypertension.

BP readings were first analyzed at the visit level. If >1 BP reading was taken during a visit, the first reading was disregarded and the mean of the remaining BP readings from the visit defined the visit BP measurement. If multiple outpatient visits occurred for a patient on a specific date, we calculated the mean of the visit BP measurements and used the mean as the visit BP measurement for that date.

Definition of Hypertension

In each 2‐year cycle, a patient was classified as having hypertension according to the definitions listed below using either an elevated visit BP measurement or at least 1 first‐line antihypertension medication prescription recorded at any time during the 2‐year cycle, like NHANES.2 BP elevation was defined as a systolic blood pressure (SBP) ≥140 mm Hg or diastolic blood pressure (DBP) ≥90 mm Hg. First‐line antihypertension medications were defined according to the 2017 American Heart Association/American College of Cardiology hypertension guideline10 and are listed with their associated Observational Medical Outcomes Partnership concept identification numbers in Table S1.

We evaluated 4 operational hypertension definitions using 4 different methods for incorporating recorded BP measurement into the definition: (1) using at least 1 elevated visit BP measurement during a 2‐year cycle, (2) using the first visit BP measurement during a 2‐year cycle, (3) using a single randomly chosen visit BP measurement during a 2‐year cycle, or (4) using at least 2 elevated visit BP measurements during a 2‐year cycle. Thus, each of the 4 operational definitions of hypertension used either a pattern of elevated BP measurement or prescription of an antihypertension medication, or both to capture both untreated and treated patients with hypertension (Figure S1). Our rationale for using these different operational hypertension definitions was to provide definitions that are comparable with other studies and guideline recommendations. For example, our third definition would be comparable to the NHANES study,2 and our fourth definition would be comparable to the generally accepted guideline definition.10

Definitions of Outcomes

Primary outcomes were age‐adjusted hypertension prevalence and age‐adjusted rates of BP control during each 2‐year cycle. BP control was defined at the patient level as a mean SBP <140 mm Hg and mean DBP <90 mm Hg during a 2‐year cycle. To adjust for differences in age distribution among the 2‐year study cycles, direct standardization was used using 2 different standards. For age adjustment of hypertension prevalence, the standard was the average age distribution of all adults across all study cycles, whereas for age adjustment of other outcome measures, the standard was average age distribution of adults with hypertension across all study cycles. The age categories used for standardization were as follow: 18 to 44, 45 to 64, 65 to 74, and ≥75 years. The proportions used for age adjustment of all patients and of patients with hypertension for each operational hypertension definition are listed in Table S2.

Secondary outcome measures were the age‐adjusted mean SBP and DBP for patients with hypertension, and the age‐adjusted proportions of patients who were labeled with a searchable diagnosis code of hypertension in the EHR during a 2‐year cycle. The coded diagnoses of hypertension and the corresponding Observational Medical Outcomes Partnership concept identification numbers are listed in Table S3.

Statistical Analysis

Descriptive statistics were used to characterize the overall study population and the populations from each 2‐year cycle, and graphical representations were used to demonstrate temporal trends. The generalized estimation equation method was applied to assess the predictors of different outcomes while accounting for the fact that multiple observations might have been available for the same individuals at different cycles. Specifically, an exchangeable within‐person working correlation structure was specified in each generalized estimation equation model, and 95% CIs were calculated using a robust variance estimator. Models used to evaluate dichotomous outcomes included a specification for the binomial distribution of the dependent variable, and models used to evaluate continuous outcomes included a specification for the Gaussian distribution of the dependent variable. The independent variables included cycle, sex, race and ethnicity, and age group, and each variable was treated as categorical in the generalized estimation equation model. To simplify the analysis and avoid multiple comparisons, in the generalized estimation equation modeling section, we only used the operational hypertension definition that used BP measurement from a single randomly chosen visit during a 2‐year cycle. All statistical tests were 2 sided, with a level of significance of 0.05. Data collection using the Observational Medical Outcomes Partnership model was conducted with the Microsoft SQL Server, and data analysis was performed using R, version 4.2.3. The institutional review board at Eastern Virginia Medical School approved this study and issued waivers of informed consent and authorization for use of protected health information for this retrospective data analysis.

The study was reported following the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines.11 The data supporting the findings of this study were analyzed within a secure environment maintained by Sentara Health and are not publicly available because of patient privacy concerns. Limited data and further information on the analytical methods could be made available from the corresponding author (J.B.) upon reasonable request.

Results

Population Characteristics

A total of 1 376 325 unique adults met the overall study inclusion criteria. The demographics of the overall study population in each 2‐year cycle are shown in the Table. Each 2‐year cycle was analyzed independently, and individual patients could appear in one or more 2‐year cycles, depending on the care they received. Thus, the number of patients in the six 2‐year cycles varied and ranged from 395 859 to 631 892 patients per 2‐year cycle. The numbers of outpatient visits and outpatient BP measurements per patient during a 2‐year cycle were relatively constant during the study period. The median number of outpatient visits per patient was 2 during 2010 to 2011 and 3 during the remaining 2‐year cycles. The median number of outpatient BP measurements per patient was 5 during 2012 to 2013 and 2018 to 2019 and 4 during the remaining 2‐year cycles.

Table 1. Demographic Characteristics, Mean SBP, and Mean DBP of All Patients in Each 2‐Year Cycle

Variable2010–20112012–20132014–20152016–20172018–20192020–2021
Patients, n395 859429 754466 692631 892608 497619 023
Sex
Female231 203 (58)254 610 (59)275 386 (59)370 352 (59)355 759 (58)362 140 (59)
Male164 625 (42)175 130 (41)191 306 (41)261 540 (41)252 736 (42)256 874 (41)
Race and ethnicity
Non‐Hispanic Black99 779 (25)109 183 (25)122 477 (26)142 447 (23)135 880 (22)146 162 (24)
Non‐Hispanic White259 665 (66)282 167 (66)305 112 (65)438 170 (69)423 892 (70)423 186 (68)
Hispanic/Latino7342 (1.9)11 549 (2. 7)13 060 (2.8)19 449 (3.1)19 691 (3.2)21 297 (3.4)
Asian7616 (1.9)9294 (2.2)10 406 (2.2)12 698 (2.0)12 834 (2.1)13 363 (2.2)
Unknown21 457 (5.4)17 561 (4.1)15 637 (3.4)19 128 (3.0)16 200 (2.7)15 015 (2.4)
Age, mean (SD), y51.0 (18.1)51.5 (18.2)52.3 (18.2)52.9 (18.4)54.6 (18.2)55.3 (18.2)
Age group, y
18–44142 539 (36)151 749 (35)158 628 (34)210 073 (33)182 608 (30)181 485 (29)
45–64156 016 (39)165 571 (39)178 135 (38)234 314 (37)224 434 (37)221 173 (36)
65–7454 364 (14)64 801 (15)76 065 (16)109 058 (17)114 851 (19)124 125 (20)
>7542 940 (11)47 633 (11)53 864 (12)78 447 (12)86 604 (14)92 240 (15)
SBP, mean (SD)125.2 (14.9)124.8 (14.8)125.1 (14.8)126.1 (14.8)126.1 (14.2)127.5 (14.6)
DBP, mean (SD)74.7 (9.4)74.4 (9.2)74.7 (9.1)75.5 (9.0)75.3 (8.7)75.8 (8.8)

Data are given as number (percentage) unless otherwise indicated. DBP indicates diastolic blood pressure; and SBP, systolic blood pressure.

Hypertension Prevalence and Disparities

The demographics of the patients with hypertension varied slightly depending on the operational hypertension definition. The demographics of patients with hypertension (using the random BP measurement definition, which is comparable to the NHANES study) are shown in Table S4.

The age‐adjusted hypertension prevalence rates over the 12‐year study period for each operational hypertension definition are shown in Figure 1. Depending on the operational hypertension definition, the hypertension prevalence rates ranged from 36.5% to 50.9% and prevalence increased by ≈5% over the study period regardless of the operational hypertension definition.

Figure 1. Age‐adjusted hypertension prevalence rates in each 2‐year cycle by operational hypertension definition.

1BP+ indicates 1 elevated BP measurement; 2BP+, at least 2 elevated BP measurements; BP, blood pressure; and Meds, medications.

Non‐Hispanic Black patients consistently showed 12% to 14% higher age‐adjusted hypertension prevalence rates compared with the non‐Hispanic White patients (odds ratio [OR], 2.03 [95% CI, 2.02–2.04]; P<0.001; Figure 2). Hypertension prevalence rates were progressively higher by age group, peaking at 80% in patients aged >75 years during 2020 to 2021 (Figure S2). Men consistently showed ≈7% higher hypertension prevalence rates compared with women (OR, 1.41 [95% CI, 1.40–1.42]; P<0.001; Figure S3).

Figure 2. Age‐adjusted hypertension prevalence rates by race and ethnicity in each 2‐year cycle by operational hypertension definition.

1BP+ indicates 1 elevated BP measurement; 2BP+, at least 2 elevated BP measurements; BP, blood pressure; and Meds, medications.

The proportions of patients with hypertension defined by elevated BP only, by prescription of a first‐line antihypertensive medication only, or by both are shown in Figure S1. For all operational hypertension definitions, the proportions of patients with hypertension defined by medication use or combination of medication use and elevated BP increased during the study period.

BP Control Rates Among Patients With Hypertension and Disparities

Among patients with hypertension, the age‐adjusted BP control rates during the study period ranged from 61.2% to 73.3%, depending on the operational hypertension definition (Figure 3). The trend in BP control rates exhibited an initial increase from 2010 to 2011 to 2014 to 2015, followed by a decline in 2016 to 2017. Thereafter, the control rates increased again from 2018 to 2019, only to decline once more in 2020 to 2021. The age‐adjusted BP control rates were consistently ≈3% lower in men (OR, 0.89 [95% CI, 0.88–0.90]; P<0.001; Figure S4) and consistently ≈5% to 7% lower in non‐Hispanic Black patients compared with non‐Hispanic White patients (OR, 0.77 [95% CI, 0.76–0.78]; P<0.001; Figure S5).

Figure 3. Age‐adjusted BP control rates in each 2‐year cycle by operational hypertension definition.

1BP+ indicates 1 elevated BP measurement; 2BP+, at least 2 elevated BP measurements; BP, blood pressure; and Meds, medications.

Among patients with hypertension, the mean age‐adjusted SBP ranged from 132.3 to 135.4 mm Hg and the mean age‐adjusted DBP ranged from 76.8 to 78.8 mm Hg (Figure S6). Mean BP was relatively unchanged during the study time frame regardless of the operational hypertension definition. Age‐adjusted mean BP was consistently higher in non‐Hispanic Black patients compared with other non‐Hispanic White patients (SBP: 2.61 [95% CI, 2.55–2.67]; DBP: 1.31 [95% CI, 1.28–1.35]; P<0.001 for both; Figure S7).

Coded Diagnosis Rates Among Patients With Hypertension and Disparities

Among patients with hypertension, the age‐adjusted proportion of patients labeled in the EHR with a coded diagnosis of hypertension ranged from 45.5% to 68.6%, again depending on the operational hypertension definition (Figure 4). The proportion of patients diagnosed with hypertension, as indicated by coded diagnoses, exhibited an initial increase from 2010 to 2011 to 2014 to 2015, followed by a decline in 2016 to 2017. Thereafter, the proportions increased again in 2020 to 2021. The proportion of patients with a coded diagnosis of hypertension was highest in hypertension groups defined by ≥2 BP measurements than in groups defined by the other definitions. The proportion of patients with hypertension with a coded diagnosis of hypertension was similar in women and men. Rates of receiving a coded diagnosis were consistently ≈5% higher in non‐Hispanic Black patients (OR, 1.26 [95% CI, 1.25–1.27]; P<0.001) and Asian patients (OR, 1.31 [95% CI, 1.28–1.35]; P<0.001), compared with White and Hispanic patients (Figure S8).

Figure 4. Age‐adjusted hypertension coding rates in each 2‐year cycle by operational hypertension definition.

1BP+ indicates 1 elevated BP measurement; 2BP+, at least 2 elevated BP measurements; BP, blood pressure; and Meds, medications.

Discussion

In this serial cross‐sectional study, we leveraged EHRs of a large health system to analyze regional trends in hypertension using various operational hypertension definitions. We showed a marked increase in age‐adjusted hypertension prevalence rates, modest age‐adjusted BP control rates, and relatively low but upwardly trending age‐adjusted rates of coding hypertension in the EHR during the 12 years of study. In addition, we noted important disparities in age‐adjusted hypertension prevalence rates, age‐adjusted BP control rates, and age‐adjusted mean BP levels that stayed consistent over time. The age‐adjusted hypertension prevalence rates were 12% to 14% higher in non‐Hispanic Black patients, 7% higher in men, and progressively higher by age group throughout the study period.

This study demonstrates the ability to observe hypertension trends and disparities at a regional health system level using real‐world data, which could greatly facilitate locally designed system‐level interventions to address hypertension and racial disparities. We were able to create an agile analytical data platform that enabled a comparison of a variety of operational hypertension definitions and an evaluation of how age, sex, and race and ethnicity affect hypertension trends. Our analytical data platform enabled examination of multiple operational hypertension definitions, and we demonstrated substantial differences in outcomes, depending on the sensitivity of the hypertension definition.

Our study demonstrated a steady increase of ≈5% in the age‐adjusted hypertension prevalence rates throughout the study period, regardless of the operational hypertension definition. Increases in hypertension prevalence rates were seen in all age groups, both sexes, and in all race and ethnicity groups.

The average age of patients and the mean SBP and DBP increased during the study period, and because blood pressure increases with age, this may have contributed to the increase in hypertension prevalence rates. The increase in hypertension prevalence rates did not appear to be artifactually affected by the frequency of outpatient visits or BP measurements because the number of outpatient visits and BP measurements remained stable during the study period.

Our definition of hypertension included whether patients were prescribed first‐line antihypertension medications, which may have also contributed to the increase in hypertension prevalence rates. The proportion of patients who were defined as hypertensive by the medication‐use criterion increased during the 2012 to 2013, 2018 to 2019, and 2020 to 2021 2‐year cycles, possibly because of changing guidelines for hypertension. The 2017 American College of Cardiology/American Heart Association published guidelines recommended a lower BP goal, which could have caused more patients to receive first‐line antihypertensive medications and could have affected the hypertension prevalence rates during 2018 to 2019 and 2020 to 2021.10

The age‐adjusted proportion of patients with controlled BP increased from 61% to 71% in 2010 to 2011 to 64% to 73% in 2014 to 2015, decreased in 2016 to 2017, increased again to 66% to 73% in 2018 to 2019, and decreased in 62% to 68% in 2020 to 2021. The pattern was consistent among patients defined by all 4 operational hypertension definitions. Non‐Hispanic Black patients had BP control rates that were consistently ≈5% lower than White patients. The increase in control rates in all groups during 2018 to 2019 could be explained by the 2017 American College of Cardiology/American Heart Association guideline recommendations for tighter BP control.10 The decrease in control during 2020 to 2021 may have been caused by changes in health care associated with the COVID‐19 pandemic.

The proportion of patients with a coded diagnosis of hypertension among patients with hypertension increased during the study period, except during 2016 to 2017, and the trends were similar in all 4 operational hypertension definition groups. There were no differences by sex. Non‐Hispanic Black and Asian patients were 6% to 10% more likely to have a coded diagnosis of hypertension compared with other race and ethnicity groups. Interestingly, non‐Hispanic Black patients were significantly more likely to have hypertension by all the operational hypertension definitions and less likely to have BP controlled, although they were more likely to have been labeled in the EHR with a coded diagnosis of hypertension.

Our results from a regional health system provide supplemental information to the national estimates reported by the NHANES. Using a comparable operational hypertension definition (a single random visit BP measurement), our age‐adjusted hypertension prevalence rates are higher (41.1%–46.1%) and increased steadily compared with NHANES, which reported prevalence rates that were lower and remained constant (30%–32%).2 Our BP control rates (61.2%–65.9%) were also higher than those reported by NHANES (31.8%–53.8%). These differences may be because our study population was derived from patients exposed to a health care system rather than subjects randomly selected from the general population. Importantly, the proportion of non‐Hispanic Black patients in our study was approximately twice that from the NHANES study, reflecting the demographics of our region, which may have increased the hypertension prevalence rates.

Studies have suggested that hypertension prevalence based on methods using EHRs could be underestimated.12 In our study, we used the EHR, but we defined hypertension based on measurement of BP, rather than based on the coded diagnosis of hypertension, which appears to have overcome the potential of underreporting hypertension using the EHR.

Our hypertension control rates were similar to those reported by the PCORnet Blood Pressure Control Laboratory, which reported an average BP control rate of 62%.13 Their findings from 25 health systems showed different demographics than our regional analysis, with Black patients only comprising 15% of their population, compared with ≈30% of the patients with hypertension in our study. Although nationwide studies are informative, limiting the analysis to a single regional health system may provide actionable insights on disparities that are more relevant locally, based on the demographics of that regional health system.

Banerjee et al measured the hypertension coding rate in 251 590 patients with defined hypertension by ≥2 BP readings ≥140/90 mm Hg or antihypertensive medications prescribed during the study period. Among those patients, the coded diagnosis rate was 62.9%, similar to our finding among patients with a comparable operational hypertension definition.12

Our study has recognized strengths and limitations. Our approach of analyzing real‐world data from a single health system is practical and feasible, and the results are similar to the multicenter approach of the PCORnet Blood Pressure Control Laboratory.14 Our data platform allows rapid evaluation of populations at scale and could be a source of continuing evaluation with planned periodic additions to the database,15, 16 providing a powerful tool for generating actionable insights from real‐world data.17, 18 This data resource is a model of what could be a useful generator of insights for a learning health system.19, 20

Collecting high‐quality race and ethnicity data from observational databases is a known challenge.21 Following federal standards, our health system records self‐reported ethnicity and race routinely in the EHR through a variety of mechanisms, including verbal questioning, patient‐facing portals, and intake forms. For our analysis, we combined the ethnicity and race into a unified variable, consistent with prior studies.2, 4 We first determined if a patient identified as Hispanic and non‐Hispanic patients were then categorized as White, Black, Asian, or other/unknown. The number of patients reporting as Hispanic was lower than expected and may have been underreported in the EHR. The reporting of racial identity was as expected, and it is unlikely that reporting inaccuracies of ethnicity or race substantially affected our findings on racial disparities.

A possible limitation of our study, compared with the NHANES approach,2 however, is that BP was obtained in the usual care setting, rather than through a rigorous protocolized approach. Although clinicians expect and often demand accurate BP measurements in usual care settings, the BP measurements in real‐world settings may be less accurate than in the protocolized setting of a prospective research study.22 Also, the operational definition of hypertension that used any single elevated BP measurement was less specific than the other operational definitions and likely overestimated the true hypertension prevalence. In addition, the coded diagnosis of hypertension in the EHR could have been affected by recall bias and other sources of possible coding error.15 Finally, 1 criterion of our definition of hypertension was prescription of any first‐line antihypertensive medication during a 2‐year cycle. It is possible that some of these medications were used to treat other diagnoses, such as congestive heart failure, although the extent of this limitation is hard to assess because patients could have hypertension along with other diagnoses. Nevertheless, it is possible that prescription of some antihypertensive medications for other indications increased our estimates of hypertension prevalence. However, this limitation should not have affected the analysis of trends, disparities, or comparisons of the 4 operational definitions of hypertension.

In conclusion, we were able to use real‐world EHR data to establish hypertension prevalence rates, control rates, and diagnostic coding rates in a large population using the EHR from a large regional health system, showing important trends and opportunities to address racial disparities. Our study may provide a model for other health systems seeking to improve hypertension care at a regional health system level and may lead to further research on how to use real‐world data and computational approaches for addressing hypertension.

Sources of Funding

This research was funded in part by the Batten Foundation (Norfolk, VA) and the Hampton Roads Biomedical Research Consortium (Portsmouth, VA).

Disclosures

In the past 3 years, Dr Krumholz received expenses and/or personal fees from UnitedHealth, Element Science, Aetna, Reality Labs, Tesseract/4Catalyst, F‐Prime, the Siegfried and Jensen Law Firm, Arnold and Porter Law Firm, and Martin/Baughman Law Firm. He is a cofounder of Refactor Health and HugoHealth, and is associated with contracts, through Yale New Haven Hospital, from the Centers for Medicare & Medicaid Services and through Yale University from Johnson & Johnson. Dr Brush receives royalties from Dementi Milestone Publishing for the book, The Science of the Art of Medicine: A Guide to Medical Reasoning. Dr Schulz received expenses and/or personal fees from HugoHealth, Abbott, Instrumentation Laboratories, and Detect, Inc, and is a cofounder of Refactor Health. The remaining authors have no disclosures to report.

Footnotes

* Correspondence to: John E. Brush, Jr, MD, Sentara Health Research Center, 800 Independence Blvd, Virginia Beach, VA 23455. Email:

Preprint posted on MedRxiv, August 25, 2023. doi: 10.1101/2023.08.24.23294518.

* Drs Brush and Lu contributed equally as co–first authors.

This manuscript was sent to Tochukwu M. Okwuosa, DO, Associate Editor, for review by expert referees, editorial decision, and final disposition.

Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.123.033253

For Sources of Funding and Disclosures, see page 8.

References

  • 1 Tsao CW, Aday AW, Almarzooq ZI, Alonso A, Beaton AZ, Bittencourt MS, Boehme AK, Buxton AE, Carson AP, Commodore‐Mensah Y, et al. Heart disease and stroke statistics‐2022 update: a report from the American Heart Association. Circulation. 2022; 145:e153–e639. doi: 10.1161/CIR.0000000000001052LinkGoogle Scholar
  • 2 Muntner P, Hardy ST, Fine LJ, Jaeger BC, Wozniak G, Levitan EB, Colantonio LD. Trends in blood pressure control among US adults with hypertension, 1999–2000 to 2017–2018. JAMA. 2020; 324:1190–1200. doi: 10.1001/jama.2020.14545CrossrefMedlineGoogle Scholar
  • 3 Clark D, Colantonio LD, Min YI, Hall ME, Zhao H, Mentz RJ, Shimbo D, Ogedegbe G, Howard G, Levitan EB, et al. Population‐attributable risk for cardiovascular disease associated with hypertension in black adults. JAMA Cardiol. 2019; 4:1194–1202. doi: 10.1001/jamacardio.2019.3773CrossrefMedlineGoogle Scholar
  • 4 Lu Y, Wang Y, Spatz ES, Onuma O, Nasir K, Rodriguez F, Watson KE, Krumholz HM. National trends and disparities in hospitalization for acute hypertension among medicare beneficiaries (1999–2019). Circulation. 2021; 144:1683–1693. doi: 10.1161/CIRCULATIONAHA.121.057056LinkGoogle Scholar
  • 5 Bundy JD, Li C, Stuchlik P, Bu X, Kelly TN, Mills KT, He H, Chen J, Whelton PK, He J. Systolic blood pressure reduction and risk of cardiovascular disease and mortality: a systematic review and network meta‐analysis. JAMA Cardiol. 2017; 2:775–781. doi: 10.1001/jamacardio.2017.1421CrossrefMedlineGoogle Scholar
  • 6 Go AS, Bauman MA, Coleman King SM, Fonarow GC, Lawrence W, Williams KA, Sanchez E; American Heart Association , American College of Cardiology , Centers for Disease Control and Prevention . An effective approach to high blood pressure control: a science advisory from the American Heart Association, the American College of Cardiology, and the Centers for Disease Control and Prevention. Hypertension. 2014; 63:878–885. doi: 10.1161/HYP.0000000000000003LinkGoogle Scholar
  • 7 Jaffe MG, Lee GA, Young JD, Sidney S, Go AS. Improved blood pressure control associated with a large‐scale hypertension program. JAMA. 2013; 310:699–705. doi: 10.1001/jama.2013.108769CrossrefMedlineGoogle Scholar
  • 8 Lu Y, Huang C, Mahajan S, Schulz WL, Nasir K, Spatz ES, Krumholz HM. Leveraging the electronic health records for population health: a case study of patients with markedly elevated blood pressure. J Am Heart Assoc. 2020; 9:e015033. doi: 10.1161/JAHA.119.015033LinkGoogle Scholar
  • 9 Hripcsak G, Duke JD, Shah NH, Reich CG, Huser V, Schuemie MJ, Suchard MA, Park RW, Wong IC, Rijnbeek PR, et al. Observational health data sciences and informatics (OHDSI): opportunities for observational researchers. Stud Health Technol Inform. 2015; 216:574–578.MedlineGoogle Scholar
  • 10 Whelton PK, Carey RM, Aronow WS, Casey DE, Collins KJ, Dennison Himmelfarb C, DePalma SM, Gidding S, Jamerson KA, et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2018; 138:e426–e483. doi: 10.1161/CIR.0000000000000597LinkGoogle Scholar
  • 11 von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, Initiative S. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014; 12:1495–1499. doi: 10.1016/j.ijsu.2014.07.013CrossrefMedlineGoogle Scholar
  • 12 Banerjee D, Chung S, Wong EC, Wang EJ, Stafford RS, Palaniappan LP. Underdiagnosis of hypertension using electronic health records. Am J Hypertens. 2012; 25:97–102. doi: 10.1038/ajh.2011.179CrossrefMedlineGoogle Scholar
  • 13 Cooper‐DeHoff RM, Fontil V, Carton T, Chamberlain AM, Todd J, O'Brien EC, Shaw KM, Smith M, Choi S, Nilles EK, et al. Tracking blood pressure control performance and process metrics in 25 US health systems: the PCORnet blood pressure control laboratory. J Am Heart Assoc. 2021; 10:e022224. doi: 10.1161/JAHA.121.022224LinkGoogle Scholar
  • 14 Pletcher MJ, Fontil V, Carton T, Shaw KM, Smith M, Choi S, Todd J, Chamberlain AM, O'Brien EC, Faulkner M, et al. The PCORnet blood pressure control laboratory: a platform for surveillance and efficient trials. Circ Cardiovasc Qual Outcomes. 2020; 13:e006115. doi: 10.1161/CIRCOUTCOMES.119.006115LinkGoogle Scholar
  • 15 Casey JA, Schwartz BS, Stewart WF, Adler NE. Using electronic health records for population health research: a review of methods and applications. Annu Rev Public Health. 2016; 37:61–81. doi: 10.1146/annurev-publhealth-032315-021353CrossrefMedlineGoogle Scholar
  • 16 Schulz WL, Kvedar JC, Krumholz HM. Agile analytics to support rapid knowledge pipelines. NPJ Digit Med. 2020; 3:108. doi: 10.1038/s41746-020-00309-zCrossrefMedlineGoogle Scholar
  • 17 Krumholz HM. Big data and new knowledge in medicine: the thinking, training, and tools needed for a learning health system. Health Aff (Millwood). 2014; 33:1163–1170. doi: 10.1377/hlthaff.2014.0053CrossrefMedlineGoogle Scholar
  • 18 Krumholz HM. Inflection point: ideas for accelerating breakthroughs and improving cardiovascular health. Circ Cardiovasc Qual Outcomes. 2020; 13:e007615. doi: 10.1161/CIRCOUTCOMES.120.007615LinkGoogle Scholar
  • 19 Greene SM, Reid RJ, Larson EB. Implementing the learning health system: from concept to action. Ann Intern Med. 2012; 157:207–210. doi: 10.7326/0003-4819-157-3-201208070-00012CrossrefMedlineGoogle Scholar
  • 20 Foraker RE, Benziger CP, DeBarmore BM, Cene CW, Loustalot F, Khan Y, Anderson CAM, Roger VL; American Heart Association Council on Epidemiology and Prevention , Council on Arteriosclerosis, Thrombosis and Vascular Biology , et al. Achieving optimal population cardiovascular health requires an interdisciplinary team and a learning healthcare system: a scientific statement from the American Heart Association. Circulation. 2021; 143:e9–e18. doi: 10.1161/CIR.0000000000000913LinkGoogle Scholar
  • 21 Polubriaginof FCG, Ryan P, Salmasian H, Shapiro AW, Perotte A, Safford MM, Hripcsak G, Smith S, Tatonetti NP, Vawdrey DK. Challenges with quality of race and ethnicity data in observational databases. J Am Med Inform Assoc. 2019; 26:730–736. doi: 10.1093/jamia/ocz113CrossrefMedlineGoogle Scholar
  • 22 Muntner P, Shimbo D, Carey RM, Charleston JB, Gaillard T, Misra S, Myers MG, Ogedegbe G, Schwartz JE, Townsend RR, et al. Measurement of blood pressure in humans: a scientific statement from the American Heart Association. Hypertension. 2019; 73:e35–e66. doi: 10.1161/HYP.0000000000000087LinkGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.