Integrating Community Health Workers into Community-Based Primary Care Practice Settings to Improve Blood Pressure Control Among South Asian Immigrants in New York City: Results from a Randomized Control Trial
See Editorial by Brook
The Million Hearts initiative aims to reduce heart attacks and strokes1 and includes strategies used by clinicians and health centers to improve hypertension control.2 Million Hearts has focused on public and private sector mobilization for a core set of objectives, which includes cardiovascular disease risk factors.3,4 However, challenges have been identified working with small, community-based primary care practices, which often have limited resources, high staff turnover, and reduced capacity to perform initiative recommendations.5,6 These practices care for a high percentage of racial and ethnic minority patients and are acutely connected with the needs and burdens of these communities.7
Blood pressure (BP) control remains suboptimal in minority groups compared with Whites, and declines in CVD mortality over the past decade have been less evident in Asian American populations.8,9 While a sizeable literature has examined the impact of hypertension interventions for Black communities,10 few have been implemented in South Asian populations.11 South Asians Americans, individuals tracing their heritage to India, Pakistan, Bangladesh or other parts of the South Asian diaspora, experience a disproportionally high burden of CVD, diabetes, and hypertension. Results from a national longitudinal study of South Asians in the United States found that 43% of men and 35% of women between 40 and 84 years of age had hypertension.12 Additionally, this group experiences distinct cultural, linguistic, and social barriers impeding access to healthcare and lifestyle behavior. A qualitative study on challenges faced by South Asian migrants observed a lack of understanding in how to access relevant self-management resources.13 Although disaggregated data among South Asian subgroups is limited,14 prior research has demonstrated that Asian Americans are less likely than the average population to have their BP checked (74% versus 83%, respectively).15 Primary care practices also face challenges in supporting the needs of South Asian patients, including ability to offer culturally- and linguistically-appropriate health coaching on chronic disease management,16 which may result in a diminished capacity to identify, manage, and refer South Asian hypertensive patients to appropriate BP management resources. Evidence-based hypertension management programs adapted to the South Asian community hold great potential to mitigate disparities for this population.17
Within the arsenal of evidence-based programs, Million Hearts has called for the use of community health workers (CHWs) as an impactful method to improve hypertension.18 CHWs, as frontline public health professionals, play an integral role in disseminating effective health interventions through their deep knowledge and connection to communities, cultivated through shared sociodemographic or cultural characteristics and trusted relationships. CHWs provide culturally-appropriate health education, assist in translation or interpretation, and advocate for individual and community health needs, providing vital social support and capacity-building to ensure sustainable health improvements in their communities.19
CHW-based interventions have demonstrated effectiveness in improving hypertension outcomes through medication adherence, appointment keeping, and implementation of lifestyle behavior modifications20 and are recommended as a strategy to improve BP outcomes in minority communities.21 Factors associated with successful implementation of CHW-based hypertension interventions include: strong provider-CHW training and collaboration, engaging CHWs with shared lived experience of their communities, and building community capacity centered around strong networks of community care, including engagement of clinical and community stakeholders.22 In the United States, CHW-based hypertension interventions have predominantly focused on low-income Black or Latino communities.22–24 However, a CHW-led randomized controlled trial among Filipino Americans observed 3.2 times greater adjusted odds of BP control for the treatment group versus the control group.3 While CHW-based interventions have been effective in improving chronic disease outcomes in South Asian American communities,25–27 no studies have investigated the impact on hypertension management. There exists a need to explore implementation of Million Hearts aims in high-risk South Asian populations.
The overall goal of IMPACT (Integrating Million Hearts for Provider and Community Transformation) was to evaluate the feasibility, adoption, and effectiveness of an integrated CHW-led health coaching and practice-level electronic health record (EHR)-based intervention to improve hypertension control among South Asian patients in New York City primary care practices.28 We examine the effect of the program on reductions in systolic BP, diastolic BP, and BP control at follow-up.
Methods
The data that support the findings of this study are available from the corresponding author, upon reasonable request.
Study Design
Project IMPACT was informed by a coalition of South Asian community-based organizations established in 2007 with expertise in the development and implementation of culturally-tailored community-clinical linkage models. The coalition provided feedback on the CHW component of the intervention, including review and adaptation of the curriculum, engagement strategies, and patient materials.28 Project IMPACT included 2 study components: an EHR intervention and a CHW intervention. Details on the EHR intervention component have been described previously.28,29 The CHW intervention utilized an unblinded randomized controlled trial design and was launched in 14 primary care practice sites 1 year after EHR intervention implementation. The practices used an average 2 full-time and 2 part-time physicians and served a majority South Asian, Medicaid population, treating 178 patients per week on average.30 CHWs were reflective of the patient population; there were 4 men and 3 women ranging from ages mid-20s to mid 50s. All spoke Bangla, Hindi, Punjabi, and/or Urdu, and received standardized CHW core competency training.31 The study was designed as an integrated intervention; CHWs assisted clinic staff with translations and facilitated medication reminders/refills and appointment scheduling; communication and activities varied by site based on clinic capacity and preferences.32 This study was approved by the institutional review boards of New York University School of Medicine and the City University of New York Graduate School of Public Health and Health Policy.
Participants
Participants were recruited from July 2017 to September 2018. A preliminary list of patients meeting criteria for study participation was identified through the EHR registry at each site. Criteria included: South Asian ethnicity using race, ethnicity, language, and surname data; age 18–85; diagnosed hypertension using International Classification of Diseases 9/10 code; and SBP≥140 or DBP≥90 mm Hg within the previous 6 months. These individuals received a letter from their physician inviting participation in the intervention, and CHWs contacted them via telephone to confirm eligibility and assess preferred language and availability to participate in education sessions. CHWs also conducted tabling at primary care practice offices, whereby interested patients in the waiting room completed a screening form and confirmed eligibility through a BP reading.
Screened patients were eligible to enroll into the intervention if they met and confirmed the following criteria: (1) self-reported ancestry from India, Bangladesh, Pakistan, Nepal, Sri Lanka, or the Indo-Caribbean; (2) age 18–85; (3) not pregnant at the time of screening; and (4) one of the following: (a) a diagnosis of hypertension using International Classification of Diseases 9/10 code in practice EHR and an uncontrolled BP reading (SBP≥140 mm Hg or DBP≥90 mm Hg) measured at screening; or (b) a diagnosis of hypertension and an uncontrolled BP reading at any office visit within 6 months before screening verified during EHR review. Eligible and interested participants were invited to complete the consent process, baseline survey, and attend the first education session.
Of 2435 individuals screened for eligibility, 2356 were identified through the EHR and mailed a recruitment letter, and 79 were identified and screened via clinic-based, in-person recruitment. A total of 303 individuals were enrolled into the study (Figure 1). Of those excluded (n=2129), 203 were ineligible, 312 declined, 388 were initially reached but never completed screening, and 1226 were not reached, had messages left, or were not contacted because recruitment goals were met. Three additional individuals were ineligible post-consent. Signed informed consent was completed before study enrollment. After participation in the first educational session, by site, individuals were randomly assigned to the treatment (n=159) or control (n=144) group by sex, using the Microsoft Excel randomization function. Spousal/family dyads were randomized into the same study arm based on the randomization of women and older individuals; there were 12 dyads in the treatment group and 8 in the control group.
Intervention
CHWs delivered a standard hypertension management curriculum adapted from the National Heart, Lung, and Blood Institute Healthy Heart, Healthy Family program, other community-based CHW curriculum materials implemented in South Asian communities, and Million Hearts Initiative materials. The intervention consisted of 5 monthly 60- to 90-minute group education sessions that used adult learning techniques and group-based activities. All materials were culturally- and linguistically-adapted for relevancy to South Asian populations. Details on the culturally-tailored aspects of the intervention have been described previously; briefly, these included: translation into South Asian languages; inclusion of specific foods, exercise, and cultural norms; and consideration of family dynamics impacting chronic disease management.28 Multiple sessions were held at primary care practice offices and community spaces to accommodate varying participant schedules. CHWs followed up with participants every 2 weeks by phone or in-person; during these contacts, CHWs engaged participants in goal-setting on hypertension control, including medication adherence, physical activity (PA), and nutrition, as identified jointly by the CHW and the patient. CHWs made referrals to community resources (eg, food pantries, mental health services) through established partnerships with a network of South Asian serving community-based organizations. Control group participants were instructed to seek care as usual after the first session and were engaged at follow-up time points for data collection. They were eligible to participate in the full intervention after study completion.
Outcome Measurements
All outcome measures and survey tools were collected for both treatment and control group participants by the CHWs at baseline, 3-months, and 6-months. BP control (SBP<140 and DBP<90 mm Hg) at 6-month follow-up was the primary outcome measure, based on 2010 JNC 8 hypertension guidelines, which were current at that time and aligned with the Healthcare Effectiveness Data and Information Set reimbursement guidelines.33 Recently recommended American Heart Association guidelines use <130/80 for BP control.34 CHWs collected BP measurements using an Omron 10 Series Upper Arm BP monitor. Three BP measurements were taken after participants were seated for 5 minutes; the average of the 2nd and 3rd measurements was used. At 6-month follow-up, EHR BP measurements filled in some missing BP readings.
Secondary clinical outcomes included SBP, DBP, and body mass index. Body mass index was calculated using weight collected by CHWs using Omron Body Composition Monitor scales and height from the EHR. Asian body mass index categorizations (underweight/normal weight: <23, overweight: 23-27.49, and obese: ≥27.5) were used.35 Behavioral measures were collected by CHWs using a survey tool at baseline and follow-up time points and included: self-reported BP management (cutting salt from diet, weight reduction, on a PA program, actively learning about healthy meals, and taking one’s own BP); PA (moderate weekly PA, vigorous weekly PA, recommended weekly PA); and Patient-Reported Outcomes Measurement Information System (PROMIS) Global scale measures (Global Physical Health, Global Mental Health, general health, and ability to carry out usual activities).
Statistical Analysis
The planned study sample size was determined based on the main outcome measure: BP control. We assumed that there would be an effect size of 8 percentage points (15% success in the intervention arm versus 7% success in the control arm). We proposed a study of 350 participants (25 at each of 14 sites) evenly randomized to the 2 groups; simulations indicated that a study of this size would have >80% power to detect the assumed difference, using a 2-sided, 0.05 level test. These calculations assumed a 20% loss to follow-up.
Percentages and means were calculated for demographic and health behavior measures. Standardized mean differences between groups were calculated for each measure; ≥10% indicated a group imbalance.
For the primary analysis, we compared BP control between the study groups at 6-months. A mixed effect generalized linear model with a logit link function assessed intervention effectiveness for controlled hypertension (Yes/No). The model included a practice-level random effect to account for clustering at the practice level, and was adjusted for age, sex, baseline SBP, and days between BP measurements. For secondary analyses, percentages of categorical variables and mean±standard deviation for continuous variables are presented for baseline and 6 months. Mixed effect linear models and mixed effect generalized linear models with a logit link assessed adjusted group differences at 6-months for continuous and binary outcomes, respectively. All models included the same adjustments as the primary outcome. Three-month BP data were not included in analyses.
A subgroup analysis was performed on individuals self-reporting a diabetes diagnosis, based on our future research interests in co-morbid diabetes and hypertension among South Asians. Percentage of BP control and mean±SD of SBP and DBP are presented for baseline and 6-month time points. Mixed effect generalized linear models (logit link function used for BP control) assessed adjusted group differences in each 6-month BP measure, using previous adjustments.
Results
Descriptive Findings
Table 1 presents baseline characteristics of randomized individuals. Approximately half of participants were women (54.1%), mean age was 56.8±11.2 years, 45.3% had <high school education, 87.6% were married/living with a partner, and 40.3% were employed. All were foreign-born, and mean years in the United States was 13.7±9.9. Although randomization achieved balance across some subject characteristics, we identified group differences using standardized mean differences for years in the United States, marital status, education, body mass index categories, taking BP medication, and current paan use (a South Asian practice of chewing betel nuts). Mean clinical BP measures were statistically (but not clinically) different for treatment and control groups: BP control (37.7% versus 42.4%), SBP (137.8 versus 138.6), DBP (86.3 versus 86.6), respectively.
| n | Total (n=303) | Intervention (n=159) | Control (n=144) | Standardized mean difference | |
|---|---|---|---|---|---|
| Sociodemographics | |||||
| Women, % | 303 | 54.1 | 54.1 | 54.2 | 0.002 |
| Age (y), mean±SD | 303 | 56.8±11.2 | 56.9±11.1 | 56.7±11.4 | 0.018 |
| Time lived in the United States (y), mean±SD | 295 | 13.7±9.9 | 14.5±10.1 | 12.7±9.7 | 0.187* |
| Married/living with partner, % | 298 | 87.6 | 89.2 | 85.8 | 0.329* |
| Education, % | 298 | 0.158* | |||
| Less than high school | 45.3 | 45.3 | 45.3 | ||
| High school/some college | 22.5 | 25.2 | 19.4 | ||
| College graduate | 32.2 | 29.6 | 35.3 | ||
| Employed, % | 298 | 40.3 | 39.7 | 40.9 | 0.022 |
| Clinical measures | |||||
| BMI, mean±SD, kg/m2 | 302 | 28.1±4.9 | 28.1±4.7 | 28.0±5.2 | 0.014 |
| BMI categories, % | 302 | 0.161* | |||
| Normal/underweight (<23.0 kg/m2) | 13.2 | 12.7 | 13.9 | ||
| Overweight (23.0–27.49 kg/m2) | 36.1 | 32.9 | 39.6 | ||
| Obese (≥27.5 kg/m2) | 50.7 | 54.4 | 46.5 | ||
| BP Control | 303 | 39.9 | 37.7 | 42.4 | 0.094 |
| SBP, mean±SD, mm Hg | 303 | 138.2±18.5 | 137.8±17.9 | 138.6±19.3 | 0.045 |
| DBP, mean±SD, mm Hg | 303 | 86.4±10.2 | 86.3±10.4 | 86.6±10.1 | 0.037 |
| Health behaviors | |||||
| Self-reported diabetes diagnosis | 303 | 52.5 | 57.2 | 52.1 | 0.088 |
| Taking BP medication, % | 302 | 98.0 | 99.4 | 96.5 | 0.203* |
| Moderate weekly PA (min), mean±SD | 295 | 97.2±155.5 | 103.4±178.0 | 90.2±125.4 | 0.085 |
| Vigorous weekly PA (min), mean±SD | 299 | 26.6±73.9 | 25.1±69.8 | 28.2±78.5 | 0.041 |
| Current cigarette smoker, % | 299 | 4.7 | 4.4 | 5.0 | 0.025 |
| Current paan use, % | 295 | 7.8 | 9.7 | 5.7 | 0.149* |
| Global Physical Health, mean±SD | 278 | 43.0±8.0 | 43.3±8.5 | 42.7±7.4 | 0.075 |
| Global Mental Health, mean±SD | 294 | 46.9±7.5 | 47.1±7.8 | 46.7±7.1 | 0.049 |
| General health, mean±SD | 303 | 2.9±0.8 | 2.9±0.8 | 3.0±0.8 | 0.050 |
| Carry out usual activities, mean±SD | 303 | 3.3±1.0 | 3.3±1.0 | 3.3±1.0 | 0.004 |
BMI indicates body mass index; BP, blood pressure; SBP, systolic blood pressure; DBP, diastolic blood pressure; PA, physical activity; SD, standard deviation; mm Hg, millimeters of mercury.
* Group imbalance in standardized mean difference (≥10%) between intervention and control for years in United States, marital status, education, BMI categories, BP medication, and paan use.
Among treatment group participants, 91% completed the follow-up survey and 86% completed all 5 sessions, while 6% completed the first session only. All completed the goal-setting behavioral action plan and 94% completed at least 1 additional goal-setting encounter. Among control group participants, 92% completed the follow-up survey.
Effect of the Intervention Versus Control Group on BP Control
Among individuals with follow-up BP data at 6-months (treatment, n=154; control, n=137), 68.2% of the treatment group and 41.6% of the control group had controlled BP (P<0.001). Individuals in the treatment group had 3.7 [95% CI, 2.1–6.5] times the odds of achieving BP control at follow-up, while adjusting for baseline SBP, sex, gender, practice, and days between BP measurements. Initially, a model was fit that also included a family-level effect for family/spousal dyads, but the estimated variance was 0 and the effect was not included in the final model. No significant differences were found for BP control in a sex-specific analysis (data not shown). Figure 2 presents BP control by intervention and control group across baseline, 3-month, and 6-month time points.
Effect of the Intervention Versus Control Group on SBP and DBP
Consistent with the main outcome, mean SBP and DBP were lower in the intervention group at 6-months compared withthe control group. At 6 months, mean SBP was 133.6±12.6 in the intervention group compared with141.1±16.6 in the control group (effect size −6.8 [95% CI, −9.5 to −4.2]); mean DBP was 82.4±7.4 in the intervention group compared with 87.5±9.0 in the control group (effect size −4.7 [95% CI −6.3 to −3.2]). Figure 2 presents BP by groups at baseline, 3-month, and 6-month time points.
Changes in BP Control, SBP, and DBP Among Individuals With Diabetes by Treatment and Control Groups
Among a subgroup analysis of individuals self-reporting diabetes (n=161), the treatment group demonstrated a decrease in SBP and DBP, and an increase in BP control, whereas the control group showed an increase in SBP and DBP and a decrease in BP control. Individuals in the treatment group had 6.3 [95% CI, 2.9–14.8] times the odds of achieving BP control at follow-up compared with individuals in the control group, while adjusting for baseline SBP, sex, gender, practice, and days between BP measurements (see Table 2).
| Intervention Group | Control Group | Effect size [95% CI]*† | |||||
|---|---|---|---|---|---|---|---|
| N | Baseline | 6 months | N | Baseline | 6 months | ||
| BMI, mean±SD, kg/m2 | 142 | 28.1±4.8 | 27.8±4.6 | 132 | 27.9±5.3 | 28.3±5.3 | −0.6 [−1.1 to −0.2] |
| BP self-management, % | |||||||
| Cutting salt from diet | 144 | 65.3 | 81.9 | 133 | 67.7 | 68.4 | 2.6 [1.4–4.9] |
| Weight reduction | 144 | 20.8 | 31.9 | 133 | 21.8 | 14.3 | 3.3 [1.8–6.5] |
| On a PA program | 144 | 14.6 | 25.7 | 133 | 11.3 | 11.3 | 3.1 [1.6–6.4] |
| Actively learning about healthy meals | 144 | 11.8 | 31.3 | 133 | 10.5 | 6.0 | 11.4 [4.8–30.9] |
| Takes own BP | 144 | 60.3 | 69.0 | 133 | 55.7 | 56.8 | 2.0 [1.0–3.9] |
| Physical activity (PA) | |||||||
| Moderate PA (min), mean±SD, weekly | 142 | 104.0±182.0 | 89.8±84.3 | 129 | 92.6±126.2 | 60.0±92.2 | 28.4 [9.4–47.3] |
| Vigorous PA (min), mean±SD, weekly | 142 | 25.7±69.3 | 8.7±29.5 | 130 | 30.1±81.4 | 10.5±30.5 | −0.8 [−7.7 to 6.2] |
| Recommended weekly PA, % | 142 | 30.0 | 30.3 | 129 | 35.8 | 25.6 | 1.4 [0.8–2.6] |
| PROMIS Global Health, mean±SD | |||||||
| Global Physical Health | 129 | 42.9±8.7 | 45.1±8.8 | 116 | 41.9±7.2 | 41.4±8.9 | 3.0 [1.4–4.6] |
| Global Mental Health | 138 | 46.4±7.6 | 49.2±8.0 | 125 | 46.0±6.7 | 46.0±7.7 | 2.9 [1.5–4.4] |
| General health | 144 | 2.9±0.8 | 3.4±0.8 | 133 | 2.9±0.8 | 3.1±0.7 | 0.3 [0.1–0.4] |
| Carry out usual activities | 143 | 3.2±1.0 | 3.4±0.9 | 133 | 3.3±1.0 | 3.1±0.9 | 0.3 [0.1–0.5] |
| BP outcomes among individuals with diabetes | |||||||
| SBP, mean±SD, mm Hg | 88 | 137.4±17.9 | 133.5±12.0 | 73 | 138.6 ±19.4 | 144.6±18.1 | −10.7 [−14.4 to −7.0] |
| DBP, mean±SD, mm Hg | 88 | 84.6±11.2 | 80.5±7.4 | 73 | 85.3±10.9 | 87.7±10.9 | −6.7 [−9.1 to −4.1] |
| BP control (<140/90 mm Hg), % | 88 | 38.6 | 70.5 | 73 | 43.8 | 38.4 | 6.3 [2.9–14.8]* |
BMI indicates body mass index; BP, blood pressure; DBP, diastolic blood pressure; mm Hg, millimeters of mercury; PA, physical activity; PROMIS, Patient-Reported Outcomes Measurement Information System; SBP, systolic blood pressure; and SD, standard deviation.
* Adjusted for baseline measure, sex, age, and practice.
† Odds ratio reported for binary outcomes; for continuous outcomes this is a difference.
Difference in Behavioral Outcomes by Treatment and Control Groups
Table 2 presents baseline and 6-month secondary outcome measures by study group, as well as differences between intervention and control groups for each measure. There was no difference in mean body mass index by study group at 6 months. All BP self-management activities were greater for the intervention group compared with the control group. At 6 months, 81.9% of the treatment group reported cutting salt from their diet, compared with 68.4% of the control group (odds ratio [OR]=2.6 [95% CI, 1.4–4.9]); 31.9% of the treatment group reported weight reduction compared with 14.3% of the control group (OR=3.3, [95% CI, 1.8–6.5]); 25.7% of the treatment group reported being on a PA program compared with 11.3% of the control group (OR=3.1, [95% CI, 1.6–6.4]); 31.3% of the treatment group reported actively learning about healthy meals compared with 6.0% of the control group (OR=11.4, [95% CI, 4.8–30.9]), and 69.0% of the treatment group reporting taking their own BP compared with 56.8% of the control group (OR=2.0, [95% CI, 1.0–3.9]). At 6-months, mean moderate PA was 89.8±84.3 in the treatment group compared with 60.0±92.2 in the control group (effect size 28.4, [95% CI, 9.4–47.3]). There was no difference in recommended weekly PA between the groups.
All PROMIS Global Health indicators were higher in the intervention group compared with the control group at 6-months. At 6-months, mean Global Physical Health was 45.1±8.8 in the intervention group compared with 41.4±8.9 in the control group (effect size 3.0, [95% CI, 1.4–4.6]); mean Global Mental Health was 49.2±8.0 in the intervention group compared with 46.0±7.7 in the control group (effect size 2.9, [95% CI, 1.5–4.4]); mean general health was 3.4±0.8 in the intervention group compared with 3.1±0.7 in the control group (effect size 0.3, [95% CI, 0.1–0.4]); and mean of carrying out usual activities was 3.4±0.9 in the treatment group compared with 3.1±0.9 in the control group (effect size 0.3, [95% CI, 0.1–0.5]).
Discussion
This study demonstrates that a culturally-adapted CHW-led intervention among hypertensive South Asians delivered in community-based primary care practice settings can achieve significant improvements in BP control after 6 months. Previous research has focused on minority groups such as African Americans and Filipinos,3,22–24 but this is the first study designed to improve BP management among South Asian Americans. This is significant because although changes in overall BP control have been examined in community-based lifestyle interventions targeting heart disease and diabetes risk factors among South Asian Americans,25–27,36 to our knowledge this is the first study to examine the impact of implementation of such models in small, community-based primary care practices. While Evidence Now has demonstrated that implementation of community-clinical linkage models in small practice settings can improve BP control,37 the engagement of primary care settings serving South Asian populations in such initiatives has been noticeably absent. Thus, our findings provide support for models to enhance BP control using CHWs in small community-based practice settings for South Asian communities.
Additionally, our findings contribute to the growing evidence-base for the effectiveness of CHW approaches BP control improvement among South Asian populations experiencing disproportionate burdens of CVD risk factors. Community-based lifestyle interventions among South Asian Americans have yielded mixed results with respect to BP outcomes: although some studies observed no significant intervention effect on BP changes,25,27 1 study observed a significant reduction in SBP.36 Our findings align with 2 recent CHW-based hypertension interventions utilizing family-based home education and training of physicians conducted among South Asians in their home country,38,39 which similarly observed significantly greater BP control in the intervention group.
The intervention group demonstrated significant improvements in BP control self-management strategies and PROMIS Global Health measures compared with the control group. While both groups showed an increase in self-reported weekly and moderate PA at follow-up, our regression analyses found that the treatment group had higher moderate weekly PA at 6-months compared with the control group. Similar intervention have shown improvements in PA-related outcomes among South Asians Americans,25,27 although 1 community-based trial observed no significant between-group differences.25 The impact of the intervention on behavioral outcomes indicates that culturally-adapted CHW models are important in improving patient activation in hypertension self-management for underserved populations facing linguistic and cultural barriers to care, aligning with previous findings.22,23 The improvements in physical, mental, and overall health measures provide important evidence on the impact of CHWs in improving well-being, particularly among immigrant and minority communities, lending support to a growing body of literature on the multi-faceted ways in which CHWs improve health.40,41
Finally, our intervention demonstrated SBP and DBP reduction and greater BP control among individuals with co-morbid diabetes and hypertension. Although evidence supports targeting BP reduction to <140/90 mm Hg among individuals with diabetes42 and clinical and lifestyle recommendations promote hypertension control for individuals with diabetes,43 there is a gap in the implementation of evidence-based strategies to address co-morbidities among minority communities that face social and cultural barriers in chronic disease management. Previous interventions among South Asian immigrants with diabetes have not examined BP outcomes or have not observed significant intervention effects. Our findings provide important support in addressing co-morbid diabetes and hypertension in South Asian and other high-risk communities (rather than addressing a singular chronic disease), through a CHW model that provides culturally-tailored health coaching on CVD risk reduction more broadly. This high-risk subgroup analysis was not an original focus of our intervention, but will be further explored in future research.27,44
Study Limitations and Strengths
The study has some limitations. First, this was an unblinded randomized controlled trial, which may lead to response bias but does not detract from the study findings, and behavioral measures were collected by the CHWs, which may lead to overestimations in both groups. Second, study-collected BP measurements were missing for 9% of individuals at 6 months and for 12.5% of individuals at 3 months; we supplemented missing 6-month BP measurements using EHR medical records for a completion rate of 96%. The EHR-obtained measurements reflect a single reading, rather than the average of 2nd and 3rd readings. Third, key study measurements were not included (eg, number of physician visits, changes in medication dosage, and nonadherence to medication), which may account for some of the BP control change. Fourth, our study was conducted in a limited English proficient, predominantly immigrant, South Asian population, and findings may not be generalizable to other minority communities. Fifth, our models did not include lifestyle changes that may be linked to BP reduction; these predictors will be evaluated in future analyses. A large number of eligible individuals could not be reached for engagement in the program; this limitation should be considered in future replication and scale-up efforts. Finally, we did not examine the sustainability of BP outcomes after 6 months or the relationship between intervention dosage and outcomes, which will be explored in future analyses.
As the first study to examine the effectiveness of a CHW intervention on hypertension management among South Asian Americans using a rigorous randomized controlled trial design, our results provide support to strategies designed to advance BP control among minority populations, particularly those seeking care in small primary care practices. We contribute to the evidence-base that CHWs are effective members of the healthcare team and can deliver lifestyle interventions in limited English proficient, immigrant communities.
Our study findings contribute empirical evidence that culturally-relevant models can be effectively implemented for hypertension management in diverse populations. Given the recent growth of the US-based South Asian population, the high burden of hypertension and CVD faced by South Asians, and the context of the ongoing global COVID-19 pandemic which has demonstrated worse outcomes for minority communities and individuals with CVD co-morbidities,45,46 integrating CHWs into clinical care through sustainable models may be particularly relevant for South Asian communities. CHW approaches also facilitate adherence to medication and health promotion strategies and address upstream factors impacting chronic disease management.41 Past studies have suggested that CHWs are a cost-effective model for chronic disease management among minority or underserved communites.47 As the demand for patient-centered approaches in clinical settings expands, CHW models have growing clinical and public health relevance in the context of hypertension management.
Leave a Reply