Underserved communities at risk of heart failure: effectiveness of community-based recruitment strategies and burden of unrecognised cardiovascular disease

Background
Models to predict incident heart failure (HF) are often not representative of socioeconomically deprived and ethnically diverse populations, limiting generalisability.[1-3] The burden of unrecognised cardiovascular disease (CVD) predisposing to HF in underserved communities is not well described.

Purpose
We designed a prospective community-based pilot study to predict incident HF in primary care (EARLY-HF). We describe baseline cohort characteristics, assess the effectiveness of recruitment strategies for engaging socioeconomically deprived and ethnically diverse populations, and quantify the burden of unrecognised CVD in a community setting.

Methods
To date EARLY-HF has enrolled 550 adults aged ≥50 years with two or more cardiometabolic risk factors (diabetes, hypertension, ischaemic heart disease, obesity, atrial fibrillation, chronic obstructive pulmonary disease, or chronic kidney disease stage 3) and no prior diagnosis of HF. Participants were recruited from community settings via: (i) targeted identification and text messaging using electronic health records in eight primary care centres serving deprived and ethnically diverse populations; (ii) a four-week targeted social media campaign on Meta platforms using posters and recruitment videos co-developed through patient and public involvement workshops (Figure 1A); (iii) ten community outreach events; (iv) direct referral from family or friends. All participants attended a baseline study visit including ECG, blood sampling including NT-proBNP, and imaging.

Results
Baseline characteristics are shown in Figure 1B. The cohort included 185 (34%) participants from the most deprived Index of Multiple Deprivation (IMD) quintile and 97 (18%) participants of non-White ethnicity. Text messaging through primary care and social media were the most effective recruitment strategies, enrolling 444 (81%) and 68 (12%) participants respectively, while maintaining high representation from deprived and non-White populations (Figure 2). Over four weeks, the social media campaign generated 15,207 advertisement clicks and 1,012 expressions of interest at a cost of £5.93 per sign-up. Community outreach events enrolled 5 (1%) participants and were resource intensive.

A substantial burden of previously unrecognised CVD was identified: 125 (23%) participants had clinically actionable cardiac abnormalities, including 46 (9%) with previously unrecognised myocardial infarction on cardiovascular magnetic resonance imaging. 331 (60%) participants had poorly controlled blood pressure (≥140/90 mmHg) and 162 (30%) had elevated NT-proBNP (≥125 pg/ml).

Conclusions
We show that primary care-embedded and digital recruitment strategies can effectively engage underserved communities at risk of HF, at scale and at low cost. This population has a very high burden of unrecognised cardiovascular disease.