Targeting of UnControlled Hypertension in the Emergency Department (TOUCHED): Design of a randomized controlled trial

Abstract

Background:

Uncontrolled or undiagnosed hypertension (HTN) is estimated to be as high as 46% in emergency departments (EDs). Uncontrolled HTN contributes significantly to cardiovascular morbidity and disproportionately affects communities of color. EDs serve high risk populations with uncontrolled conditions that are often missed by other clinical settings and effective interventions for uncontrolled HTN in the ED are critically needed. The ED is well situated to decrease the disparities in HTN control by providing a streamlined intervention to high risk populations that may use the ED as their primary care.

Methods:

Targeting of UnControlled Hypertension in the Emergency Department (TOUCHED), is a two-arm single site randomized controlled trial of 770 adults aged 18–75 presenting to the ED with uncontrolled HTN comparing (1) usual care, versus (2) an Educational and Empowerment (E2) intervention that integrates a Post-Acute Care Hypertension Consultation (PACHT-c) with a mobile health BP self-monitoring kit. The primary outcome is differences in mean systolic blood pressure (SBP) at 6-months post enrollment. Secondary outcomes include differences in mean SBP and mean diastolic BP (DBP) at 3-months and mean DBP at 6-months. Additionally, improvement in cardiovascular risk score, medication adherence, primary care engagement, and HTN knowledge will also be assessed as part of this study.

Conclusions:

The TOUCHED trial will be instrumental in determining the effectiveness of a brief ED-based intervention that is portable to other urban EDs with high-risk populations.

Introduction

Approximately, 77.9 million adults in the United States have high BP with a projected increase of 8.4% by 2030.[1] High BP has an estimated national cost of $46.4 billion annually and can cause a myriad of health complications.[1] Uncontrolled hypertension (HTN) is the leading risk factor for the development of cardiovascular complications.[2] In the US, the prevalence of HTN among adults 20 years of age or older is an estimated 46%.[3] Within the hypertensive population, racial disparities have been well documented.[4–6] HTN prevalence is higher among non-Hispanic black (40.3%) than non-Hispanic white (27.8%), Hispanic (27.8%) or non-Hispanic Asian (25.0%) adults.[5] Additionally, the prevalence of uncontrolled or undertreated HTN is highest among Hispanics (52.6%) and non-Hispanic blacks (51.5%) as compared to non-Hispanic whites (44.3%).[6]

There were 145.6 million emergency department (ED) visits in the US in 2016.[7] Elevated BP is common in ED visits and historically referred to as asymptomatic hypertension, often attributed to pain or other acute issue and not diagnosed as essential HTN.[8] Chronic conditions that have traditionally been managed in primary care clinics, such as HTN, are being evaluated and treated in the ED with greater frequency.[9] Trends in ED utilization have identified disparities in overutilization and routine, non-emergent use of EDs among black, Hispanic, uninsured, and low-income patients.[10] EDs are well situated to serve a high risk population not readily captured in other clinical settings and initiate streamlined hypertension screening methods.[11]

One of the most effective strategies to improve BP control has been team-based care, particularly the utilization of pharmacists and nurses.[12, 13] Clinical pharmacists have been integrated within EDs since the late 1970’s and studies have found improved cardiovascular outcomes when pharmacists assist with medication management.[14–17] The primary mechanism for improved BP control with team-based care involves medication intensification and optimization, along with strategies to improve adherence.[18, 19] These findings involving team-based care were the basis for including clinical pharmacists and advanced practice nurses to deliver the intervention in the present trial.

Targeting of UnControlled Hypertension in the Emergency Department (TOUCHED) trial methods were designed to build on two NIH-funded pilot studies from the same principle investigator.[20][21] In 2014, a 1-year study entitled “Longitudinal Study of Asymptomatic Hypertensive Patients with Subclinical Heart Disease in an Urban Emergency Department Population: A Pilot Study” [20] was funded by the National Center for Advancing Translational Sciences. A cohort of patients (N=10) presenting to the ED with HTN was selected to return for a brief HTN intervention. The intervention included a bedside echocardiogram and HTN education session. At 9 months post-enrollment, 85.7% of participants had readings of <140/90mmHg. The results suggested EDs could serve as a point of screening and intervention for HTN patients.

In 2015, the NHLBI funded “A Hypertension Emergency Department Intervention Aimed at Decreasing Disparities (AHEAD2)” [21]. This study was a three-arm, single-site randomized pilot trial for feasibility conducted in an urban academic ED. We enrolled 150 predominantly ethnic minorities (70% Black and 17.3% Latinx) with no primary care provider and severely elevated BP (≥160/100 mmHg). Subjects were randomized into one of three study arms: (1) Enhanced Usual Care (EUC); (2) ED-initiated Screening, Brief Intervention, and Referral for Treatment (ED-SBIRT); or (3) ED-SBIRT plus a 48–72 hours Post-Acute Care Hypertension Transition Clinic (ED-SBIRT+PACHT-c). The mean age of study participants was 48 +9.1 years and 112 (75%) completed the trial. The multicomponent intervention resulted in a SBP decrease of 7.9 mmHg (95% CI: −0.6, 16.4) when compared to EUC at 9 months. The results suggested ED visits present an opportunity to implement risk reduction strategies for diverse, high-risk patients with elevated BP.

The current study builds on experience from the past pilot studies. The primary objective of the TOUCHED trial is to test the effectiveness of an ED-based patient education and empowerment education on BP over 6 months compared with usual care. This is the first randomized, controlled trial funded by the NHLBI addressing HTN disparities in an urban ED setting.

Methods

Trial Design

The TOUCHED trial is a single site, two-arm randomized controlled trial with a recruitment goal of 770 participants. The study aims to determine the effectiveness of an ED-initiated educational and empowerment intervention and primary care referral program for asymptomatic uncontrolled HTN. The Education and Empowerment (E2) intervention involves (1) a brief Post-Acute Care HyperTension consultation (PACHT-c) with a clinical pharmacist or advanced practice nurse to reinforce the educational component and to initiate or modify antihypertensive medications if appropriate; and (2) a smartphone-enabled Blood Pressure Monitoring Kit consisting of a Bluetooth-enabled BP monitor, the Withings Health Mate app for daily self-monitoring of BP, and behavior change text messages. It compares: (1) usual care (preprinted discharge instructions and standard outpatient referral), and (2) ED-initiated E2 intervention coupled with a PACHT-c intervention and referral for primary care follow-up.

Figure 1 illustrates the study components. The left panel shows the current standard of care for a patient presenting to the ED. The right panel shows the research related tasks. Research Assistants (RAs) scan patients in the ED for eligibility criteria using the information obtained by the standard of care process, if criteria is met the RA approaches the patient about participation, interested participants are consented and then randomized into one of the study arms. The usual care group is referred to a primary care provider (PCP) prior to discharge, if they indicate they do not have a PCP. The intervention group receives the education and empowerment intervention along with the pharmacy consultation and are referred to a PCP, if they indicate they do not have one, prior to discharge. All patients are asked to return to the ED 3-months and 6-months after being discharged for follow up research BP measurements.

Figure 1:

The TOUCHED Program Procedures

Patient Population

Participant recruitment is conducted at an inner-city, academic ED. The ED is a Level II trauma center, comprehensive stroke center and a regional tertiary care center. Geographically, it is located within a predominantly Black and Hispanic neighborhood in Chicago. The annual census is 46,000 patients. Adult patients represent (≥18 years) 80% of all ED visits. The demographics of the patient population reflect those of the primary service area: 35% African American, 35% Hispanic, 20% Caucasian, and an increasing number of Asians and Native Americans.

Our team has demonstrated the ability to recruit patients that identify as part of a racial minority group. [21] The pilot trial AHEAD2 was conducted in the same ED and enrolled 150 participants over 10 months. Out of the 150 participants, 70% identified as Black and 17.3% identified as Latinx. Similar participation is expected in the TOUCHED trial. The goal is to reach the recruitment target in 4 years. We utilize several approaches to achieve a high level of subjects from minority groups. First, our RAs receive specific training to provide culturally sensitive approaches to recruitment. Second, many of our RAs identify as being from minority groups and are generally well accepted by potential participants.

Inclusion and Exclusion Criteria

Inclusion criteria: 1) At least two recorded BP readings between BP > 140/90 mm Hg and ≤ 180/110 mm Hg from the current ED visit prior to discharge, 2) verbal fluency in English or Spanish, and 3) age 18–75 years. Patients with, or without, cardiovascular disease will be eligible for the trial.

Exclusion criteria: 1) Unable to verbalize comprehension of study or impaired decision making, 2) Lives outside the Chicago area or plans to move in the next year, 3) Currently pregnant or expecting to be pregnant in the next year.

Recruitment

Once patients have been evaluated, treated, and identified for discharge from the ED, they are screened against the study eligibility criteria. It is important for patients to be identified for discharge to ensure they have completed their clinical care and to maximize use of time for consent and initial assessments before physically leaving the ED. The tracking board in the ED displays current BP readings, other vitals, chief complaint, and discharge status which can be used to evaluate eligibility. Patients meeting inclusion criteria are approached by trained RAs for study participation and the study is explained in detail. If the patient agrees to participate, they sign written informed consent.

Baseline BP Determination

Once eligibility is established, BP is measured following the American Heart Association Guidelines and approaches used in clinical trials.[22, 23] Participants are instructed to sit upright, with back supported and feet uncrossed for 5 minutes before and during the first BP reading is measured. The first reading is discarded to maintain accuracy. Two additional BP values are measured each separated by 1 minute. These two values are averaged to generate the research measurement. If the 2 BP values differ by >10 mm Hg, then a fourth BP is measured and the two closest values from the 2^nd, 3^rd, and 4^th BP measurements are averaged to generate the research measurement. The instrument used to measure BP is the Welch Allyn BP monitor which is used in the ED. It has been validated and is calibrated annually.[24]

Baseline Assessments

All study participants complete assessments designed to evaluate current HTN knowledge, medication adherence, and determination of risk of cardiovascular complications at baseline and 6-months post-enrollment. These assessments include:

1. Hypertension Knowledge Survey:[25] The HTN knowledge survey is a 9-item, validated tool developed to assess hypertension knowledge in low literacy patient populations. The scale assesses respondents’ knowledge in defining HTN, lifestyle, and behaviors that may affect BP levels, and the long-term consequences of HTN. The survey has been validated in an urban population that included a high proportion of Black and Latino patients. Scores are categorized into tertiles that indicate low (≤7), medium (8), or high (9) levels of HTN knowledge.

2. Modified Morisky Scale Survey:[26] The modified Morisky scale is a validated 4-item instrument to assess self-reported patient adherence related to antihypertensive medication. The modified Morisky scale provides a total score with a range of 0 to 4, with higher scores indicating lower adherence to medication. The scores of the modified Morisky scale can be classified as low adherence (3–4), medium adherence (1–2) and high adherence (0) based on its criterion validity with BP control. The appropriate licensing and training to use the instrument was obtained.

3. Atherosclerotic Cardiovascular Disease (ASCVD) Risk Assessment:[27] Using ASCVD Risk Score calculators, participants have their risk score calculated by entering their age, tobacco use status (Y/N), SBP, Cholesterol/HDL, and BP treatment (Y/N), and history of diabetes (Y/N). Point of care cholesterol/HDL measurement is obtained by the research assistant using the Polymer Technology Systems CardioChek PA Analyzer, which analyzes up to 4 cholesterol types with a single finger stick of blood and is CLIA-waived and FDA approved.

4. Primary Care Provider Referral Process: All patients indicating they do not have a primary care provider are referred to Mile Square Health Center. It is a federally qualified health center and UI Health partner. The patient’s contact information is transmitted via a secure intranet tool directly to the appointment facilitator. The appointment facilitator contacts the participants directly to secure appointments within 48–72 hours at the nearest FQHC facility.

Randomization

After consent and baseline data collection, participants are assigned to one of the two arms (1: usual care, 2: intervention) using a stratified block randomization scheme. Two stratifying variables were used, sex (M, F), and baseline BP level (140/90 mmHg ≤ BP ≥160/100 mmHg, BP≥160/100 mmHg). Random blocks of size 6 or 9 are first chosen within each stratum followed by random allocation in one of the two treatment groups. This process ensures that the study arms remain balanced throughout recruitment. A random allocation table was first created using SAS PROC PLAN, which was then fed into the RedCap data management program to determine treatment allotment.

E2 Intervention

Participants randomized to the intervention arm (Arm 2) receive additional educational and empowerment resources to assist in the management of HTN. The customized video is bi-lingual and culturally sensitive.

1. Visual Echocardiogram Image Clips: Since HTN is a “silent killer”, images of underlying organ damage from HTN can show participants the severity of the disease. We use a series of visual images of age- and gender-matched echocardiograms of people with and without damage as a tool to educate and motivate patients to change their behavior to improve their BP. The video is customized to include simple explanations of possible long-term impact of uncontrolled HTN to accompany the imagery. Additionally, the video is bi-lingual and culturally sensitive. We have found that the real time visualization of cardiac ultrasound images with active discussion of findings is a significant patient motivator and empowerment tool, and was a significant factor in the success of a previous pilot study.[20]

2. Hypertension Educational Video: The 5-minute video educates participants in Arm 2 about high BP, how it is diagnosed, and the importance of medication and lifestyle changes in preventing secondary complications. The lifestyle changes that are discussed are reduction in salt intake, reduction in alcohol consumption, cessation of smoking, and increase in physical activity. The customized video is available in both English and Spanish and is culturally sensitive.

3. Smartphone-enabled Blood Pressure Monitoring Kit and Text Messages: Regular self-monitoring of BP is recommended as an important component of HTN management.[2, 4, 5] Home BP monitoring has been shown to improve adherence to medications, induce healthy lifestyle changes, and aid in optimizing of treatment.[2, 4, 5] All participants randomized to the intervention group receive a smartphone-enabled BP monitoring kit comprised of an FDA-cleared, wireless, self-inflating, BP monitor (Withings BP-801) and the Health Mate^™ mobile app which is compatible with both iPhone (iOS 5.0 or higher) and Android devices (4.0 or higher). Participants without a smartphone or unable to download the Health Mate app on their phone due to lack of space or compatibility are loaned a smartphone (iPhone 7S) for the duration of the study (6 months). The Health Mate app automatically launches when the participants places the cuff on their arm and turns on the monitor to begin the BP measurement. All timestamped BP readings automatically sync with the app, which creates an easy-to-understand chart of all the measurements and provides participants with instant color-coded feedback based on AHA recommendations for hypertension. Synced data are automatically uploaded from the mobile app to the iCardia server used in this study for remote data collection. iCardia is a previously validated and secure research platform hosted in a HIPAA-compliant environment at the University of Illinois at Chicago. It has been used in multiple NIH-funded studies to facilitate remote collection of patient-generated health data from various commercially available connected health devices such as wearable activity trackers, blood pressure monitors and smart weight scales.[28–31]

   Participants are shown how to use the BP monitor and app by RAs and asked to demonstrate understanding of use prior to leaving the ED. Participants are asked to check their BP at home at least weekly. The RAs set up automatic reminders on each participant’s mobile app that prompt them to measure their BP at a time of their choice. In case of missed measurements for seven consecutive days, participants receive a text-message through the iCardia platform for a maximum of four consecutive days or until they complete a BP measurement. In case of missed measurements for 14 consecutive days, participants are contacted by phone by the RAs. All data is remotely monitored by the coordinators through the iCardia platform.

   Participants are informed that the Health Mate app is not a telemonitoring app (i.e. the data are not actively monitored by a clinical care team). In the event they experience a hypertensive crisis (>180/110 mm Hg) they must contact a physician immediately.

   Participants allocated to the intervention group also receive a program of behavior change text messages (TMs) targeting BP self-monitoring adherence and medication adherence. TMs were developed by our research team in English and also translated in Spanish for the purposes of this study. TMs are grounded in a range of underlying theoretical frameworks such as Health Belief Model, Theory of Reasoned Action, Theory of Planned Behavior, and Social-Cognitive Theory. Behavior change techniques include the provision of educational information about the importance of medication intake and adherence BP self-monitoring, encouragement, information about the consequences of uncontrolled BP, intention formation, monitoring self-behavior, barrier identification, prompting and cues, and personal reminders. Participants receive 4 TMs per week via the iCardia platform at a time of their preference.

4. Post-Acute Care Hypertension Transition Consultation (PACHT-c): All participants randomized to the E2 intervention receive a focused consultation with either a clinical pharmacist or an advance practice nurse (APN). To ensure consistency throughout the trial, the PACHT-c intervention is standardized and remains interchangeable between the clinical pharmacists and APNs. A training manual was developed to individually train each clinical pharmacist and APN on the project. During the consultation, the pharmacist/APN reviews the screening assessments, and general principles of BP control including nutrition, exercise, and smoking cessation, if applicable. BP is managed according to the current published guidelines available regarding initiation of first-line antihypertensive medications.[32] Patients with BP >140/90 mmHg are either initiated on antihypertensive medications, or current antihypertensives are titrated by the pharmacist/APN during the consultation, if appropriate. All PACHT-c clinic notes are entered into REDCap and in the EHR, which is shared between both the ED and our affiliate FQHC. During the COVID pandemic, the clinical pharmacists and nurse practitioners have remained physically present in the ED. However, modifications were recently made to include the use of telehealth consultations to limit patient interaction and prospectively plan for potential interruptions of in person availability (quarantine due to potential COVID exposure, COVID child care issues. The telehealth consultation was recently trialed with 2 patients randomized to arm 2. The clinical pharmacist was physically present in the ED (in case there was an issue) but completed the consultations virtually without issue.

Follow-up

All participants are asked to return to the hospital for research BP measurements at 3- and 6-months after their enrollment. All participants are contacted via phone call at 1-, 3- and 6-months post-enrollment to verify contact information and schedule study follow-up appointments at 3- and 6-months. In light of the COVID outbreak, we have implemented all institution required safety procedures for all in-person interactions and trained staff accordingly.[33]A video was filmed featuring the PI, assuring participants it is safe to return to the hospital for the visit and demonstrating the procedure. The video is sent via text to participants via iCardia. Of those participants that were due to return post-COVID outbreak, 80% returned to complete their follow-up visit, which is line with the retention rate pre-outbreak. A text message is sent out via iCardia the day before as a reminder of their follow up appointment. Return follow-up hospital visits are brief, limited to 15 minutes or less and at no charge to the study participants.

In the event a participant is not reached, RAs follow a defined retention algorithm which includes text messages, mailings, and/or telephone calls to secondary contact person provided during recruitment. To maximize retention efforts, participants are reimbursed for any in-person data collection ($50 for their 3-month follow up and $50 for their 6-month follow-up post-randomization). A rideshare, public transit, and parking vouchers, are also provided as needed. In addition, participants receive a welcome card, birthday card, and quarterly newsletters to their mailing address as an additional means of communication during their participation in the study.

Data Collection and Management

Data are collected at enrollment (baseline), 3-months and 6-months post-randomization. All data collection takes place in a private room in the ED by RAs. At the 3-month follow-up, only repeat BP measurements are taken. At 6-month, RAs conduct repeat BP measurements, the Modified Morisky survey, and an exit interview to provide feedback about the study. The exit interview contains questions on participant satisfaction with all study components including the BP monitor and mobile health app for those in the intervention group.

Research assistants and outcome assessors are blinded to the initial study arm assignment for follow-up appointments. All data are collected electronically using a pre-installed REDCap app on a study iPad. To maintain confidentiality and security of data, participants are assigned a study ID. The study ID functions as the main identifier, access to personal identifiers (such as first, last name, birthdate) is restricted post-enrollment. Data analysis is conducted using de-identified data.

Electronic data entry is frequently monitored by the study coordinators to identify any data errors or omissions, and immediately addressed with study personnel. Written data such as consent forms and primary care referral forms are kept in a locked filing cabinet in a locked room only accessible to key study personnel.

Outcomes

The primary outcome is the average systolic BP measurement at 6 months versus baseline. Secondary outcomes include average diastolic BP at 6 months, average systolic and diastolic BP at 3 months, change in cardiovascular risk score at 6 months, and change in medication adherence at 6 months.

Statistical Considerations

Sample size and power calculations

This study will recruit a total of 770 participants across the 4-year recruitment span. The primary hypothesis of the study is that the difference in BPs from baseline to month 6 is different between the two treatment arms. Testing this hypothesis amounts to testing difference in mean BP between the two treatment groups at 6 months, due to randomization (since the mean BP at baseline is the same across groups). Thus, sample size estimation is determined by comparing the mean BP at 6 months using a two-sample, two-sided, 5% significance t-test with 80% and 85% statistical power. Sample size is calculated first to detect a pre-determined effect size (difference in means divided by standard deviation) and then converted to BP units, namely mmHg.

Data from our pilot study resulted in estimates of DBP standard deviation of 12 mmHg and SPB standard deviation of 21 mmHg.[21] For example, the required sample size to detect an effect size of 0.25 (0.25*12 = 3 mmHg in DBP, and 0.25*21 = 5.25 mmHg in SBP) is 504 for 80% power and 576 for 85% power, using a two-sided, two-sample 5% significance t-test. An attrition-adjustment to the calculated sample size give 673 and 769 patients for 80% and 85%, respectively.

Table 1 illustrates the effect size in relation to BP standard deviation and sample size. The bolded N indicates the sample size after the assumed 25% attrition rate. The highlighted line indicates the selected sample size for this trial which is a total of 770 with 85% power. Additional power is achieved when mean BP comparisons are adjusted for baseline BP determinations, even assuming a modest correlation between 6-month and baseline measurements.[34]

Table 1.

Sample Size Calculation

Effect Size			80% Power		85% Power
Standardized	mmHg*		Sample Size		Sample Size
	DBP	SBP	N	N**	N	N**
0.20	2.40	4.20	786	1049	898	1198
0.25	3.00	5.25	504	673	576	769
0.30	3.60	6.30	350	467	400	534
0.35	4.20	7.35	258	345	294	393
0.40	4.80	8.40	198	265	178	238
0.45	5.40	9.45	156	209	178	238
0.50	6.00	10.5	126	169	144	193

^*Effect size is defined as difference in means divided by standard deviation. From our pilot study: DBP Standard deviation 12 mmHg; SBP Standard Deviation 21 mmHg.

^**25% attrition is assumed and thus required sample size, N**, is obtained by dividing calculated sample size, N, by 0.75.

Data analysis

Descriptive statistics will be used to assess completeness of study data, normality of outcome measures, and potential covariates as well as to identify potential covariate imbalances between study arms. For the primary analysis of each study hypothesis we will adopt an “intention to treat”, (ITT) modality. We will apply the independent samples t-test and Chi-Squared test to compare key measures between treatment arms to ascertain balancing of important measures that may have an influence on the outcome variables such as baseline BP, age, race and sex.” Mean BPs will be compared across groups using a two-sample, two-sided, 5% significance t-test. The analysis will then be extended to gain statistical power by adjusting the comparison by baseline BP levels. Key variables (age, sex, race/ethnicity) will be compared across treatment groups and if different will be used in a multivariate mixed effects model.

It is possible that data may not be missing at random. For subjects who drop out of the study before their outcome data can be obtained, we propose to use a multiple imputation method to impute their outcome. This multiple imputation will be implemented using a model based on SBP and DBP at baseline and the 6 or 12 months for all subjects with observed data. We will use five separate imputations, and will average the parameters across all five imputations for the final analysis.

Aim 1:

Evaluate the effectiveness of an ED based intervention (Arm 2) on the primary outcome of mean SBP differences at 6-months post-intervention compared to usual care (Arm 1).

We hypothesize that the mean change in SBP at 6 months after randomization will be higher in the intervention arm compared to usual care. Linear regression models will be used to compare change in continuous BP between treatment arms at 6 months; mixed effects linear regression models will be used to combine the two-time points into a single model.

Aim 2:

Evaluate the effectiveness of an ED based intervention (Arm 2) on the secondary outcome of mean SBP and DBP differences at 3 months and mean DBP differences at 6 months post-intervention compared to usual care (Arm 1).

We hypothesize that the mean change in SBP and DBP at 3 months and mean DBP at 6 months after randomization will be higher in the intervention arm compared to usual care. Linear regression models will be used to compare change in continuous BP between treatment arms at 3 months and DBP at 6 months; mixed effects linear regression models will be used to combine the two-time points into a single model.

Aim 3:

Examine if the intervention (Arm 2) reduces racial disparities. Specifically, if the intervention is as effective in racial minorities as non-minorities. This will be an exploratory analysis and will primarily focus on the changes in cardiovascular risk score/profile.

Change in the cardiovascular risk score at 3 and 6 months will be analyzed to assess racial disparity by testing the interaction between a binary indicator of treatment group and a binary indicator of minority versus non-minority. The distribution of change scores will be examined to determine whether linear regression is appropriate; if not, ordinal categories will be defined, and proportional odds logistic regression will be used. Models will be fit at each of 3 and 6 months, and mixed-effects models will be used to combine the two-time points into a single model.

An intent-to-treat analysis will be implemented to accommodate missing data. Furthermore, missing data will be handled by using a mixed-effects model approach, which assumes that the data is missing at random and hence will not bias the analysis. If we determine that data may not be randomly missing and thus non-ignorable then two well accepted approaches to handle this situation will be implemented, namely selection models and pattern mixture models.

Sampling balance for the key predictors will be checked and adjusted for in the analysis, and their mediating and moderating effects will be examined if needed. We address power for Aims 1 and 2 for unadjusted tests. Adjustment for covariates will decrease power slightly by using up degrees of freedom, but the adjustment will also create more accurate estimates, so the expected power will be similar.

Study Team Structure

The principle investigator is a board-certified emergency medicine physician. The project team consists of five co-investigators: a board-certified emergency medicine physician certified in ultrasonography, a clinical pharmacist, a biostatistician with clinical trials experience, a biomedical and health informatics researcher with expertise in digital health interventions, a cardiovascular disease/minority health disparities researcher, as well as a full time project manager, a part-time project coordinator, and fifteen research assistants. Additionally, the study has two consultants: a clinical pharmacist and a statistician with significant experience in epidemiologic studies, as well as an internal advisory committee in which its three members have expertise in clinical research methods, emergency medicine, and technology utilization in research. Internal advisory members are not actively involved in the study but are available to answer study-related questions as they arise.

The Data Safety and Monitoring Board (DSMB) is composed of 7 members from institutions not affiliated with the study institution. DSMB members share expertise in community health, cardiology, HTN management, preventive medicine, and statistical analysis. DSMB members assemble to ensure the safety of the subjects and the validity and integrity of the data generated. The DSMB convenes bi-quarterly to review and provide feedback on study protocol and performance and must recommend approval of the study before protocol is implemented or continued.

Research Assistant Training Protocols

Undergraduate and medical students were recruited and trained as research assistants (RAs) for the study. Additionally, a dedicated non-student research assistant was hired to facilitate evening and overnight participant recruitment.

RAs were trained in-person by the project coordinator/manager in small groups of 3 or 4. The comprehensive training consisted of human subject research training, data entry via REDCap, study design, patient eligibility criteria, study protocol and use of the hospital Electronic Health Records (EHR) system used to identify potential participants. A manual of operations (MOP) was written and distributed to all RAs during training.

Figure 2 illustrates the training and assessments the RAs must complete prior to independent recruitment. The human subject research training is facilitated by the university’s Office of the Protection of Research Subjects which utilizes the CITI Program’s online modules. Certificates of completion were obtained and sent to the project manager before being added as research personnel on the study’s Institutional Review Board application. The training on study protocol includes study design, patient eligibility criteria, use of EHR system, and REDCap data entry follows the human subject research training. RAs are then assessed via a written examination before beginning clinical shadowing with project coordinator/manager and senior RAs. Clinical shadowing ceased once a RA observed senior members enroll an Arm 1 and Arm 2 participant, and the RA was observed by the project coordinators to complete multiple enrollments without error.

Figure 2:

Volunteer Training

Progress to date

Participant recruitment began in February 2019. From February 19, 2019 to March 12, 2020 there were 107 participants enrolled. Recruitment was temporarily suspended on March 13, 2020-September 11, 2020 due to the COVID-19 global pandemic. Recruitment was scheduled to end in August 2022 but will likely be extended due to the temporary suspension.

Discussion

The NIH-funded TOUCHED study is a two-arm randomized controlled trial to compare an ED initiated education and empowerment intervention in predominately minority participants with asymptomatic uncontrolled HTN versus usual care, on a single primary outcome of mean systolic BP differences between arms at 6-months. The trial addresses the role that EDs can play in secondary cardiovascular prevention in high risk populations being discharged from an emergency department with asymptomatic uncontrolled HTN through an ED initiated education/empowerment intervention.

The TOUCHED trial builds upon previous NIH-funded pilot trials from the same principle investigator. In 2014, a 1-year pilot study entitled “Longitudinal Study of Asymptomatic Hypertensive Patients with Subclinical Heart Disease in an Urban Emergency Department Population: A Pilot Study” was funded by the National Center for Advancing Translational Sciences.[20] A small cohort of patients (N=10) presenting to the ED with HTN was selected for a brief educational HTN intervention. The intervention included a bedside echocardiogram, a brief HTN education session, and counsel on the reversibility of the echocardiogram findings. They were asked to return 9 months post-intervention for a BP measurement. The results suggested EDs could serve as a point of screening and intervention for HTN patients.

In 2015, the NHLBI funded “A Hypertension Emergency Department Intervention Aimed at Decreasing Disparities (AHEAD2)” as a 1 year feasibility and acceptability trial.[21] The aim was to develop a feasible ED-initiated intervention for HTN patients with a focus on reducing HTN disparities for underrepresent populations. The Intervention depended on the study arm and may have included preprinted discharge instructions, a HTN education video, BP monitoring kit, limited bedside echocardiogram, and/or a pharmacist consultation within 48–72 of ED discharge. The results (n = 150) showed increased BP control among participants randomized to the intervention group. The pilot demonstrated the feasibility of implementing an ED initiated HTN intervention and the role EDs can play in secondary cardiovascular prevention through education and empowerment. The pilot stood as the foundation for the present TOUCHED trial with some changes to improve the efficacy and streamline the enrollment process.

The TOUCHED study is the first multi-year, fully powered randomized control trial funded by the NHLBI addressing asymptomatic uncontrolled HTN in an urban emergency department setting. Risk assessment coupled with a streamlined education and empowerment component appropriate for an ED setting is not currently standard for patients with asymptomatic uncontrolled HTN being discharged from the ED. Recent emergency medicine literature suggests that if evidence-based guidelines were developed that demonstrate effectiveness for a brief and cost-effective intervention for high-risk patients, it would result in greater adherence with referral guidelines and heightened awareness of secondary prevention interventions available to ED physicians. This study addresses the evidence gaps in the options available to asymptomatic uncontrolled hypertensive patients being discharged from the ED and engages a high-risk population. We believe the intervention could be portable to other ED settings with similar at – risk patient populations.

Highlights.

• Uncontrolled hypertension may be as high as 46% in emergency departments (ED)

• The ED is well positioned to decrease health disparities in hypertension control

• This is a randomized controlled trial in 770 adults with uncontrolled hypertension

• Subjects will receive either usual care or the intervention including self-monitorin

• The trial will establish guidelines for high risk patients with hypertension in the ED