Effect of a cardiovascular post discharge sms support program on healthcare utilisation - the hearthealth program

Abstract Background Cardiovascular disease (CVD) is a leading cause of mortality worldwide[1]. Targeting modifiable risk factors is key to primary and secondary prevention of CVD but many barriers to good risk factor control exist[2]. SMS message programs have shown to be an efficient modality to deliver information to improve knowledge and risk factor profiles for CVD patients[3]; however, their impact on healthcare utilisation has not been thoroughly examined. The HeartHealth program is a text message post-discharge support program delivered to CVD patients. It comprised regular semi-personalised text messages providing CVD information, advice, and support to patients over a 6-month period. Purpose This study aimed to examine the impact of the Heart Health program on healthcare service utilisation. Methods Patients ≥18 years old who recently attended cardiology clinics or were discharged from a cardiology unit, between April 2020 and April 2022 were invited to the program. We compared healthcare utilisation rates in patients who enrolled into the HeartHealth text program (intervention group) to contemporaneous patients who did not enrol in the program (control group). The primary outcome was the incidence rate of all-cause healthcare utilisation events (composite of ED presentations and/or hospital admissions for any medical cause) in the 6 months post-discharge. The secondary outcomes were incidence rates of CVD events and unplanned CVD events. All analyses were adjusted for demographic and co-morbidity variables. Results 11542 patients were included in analysis (intervention: 4324, control: 7218), mean age 58.8 years, 60.7% were male, and 78.3% with English as their preferred language. HeartHealth Intervention participants had lower rates of the primary outcome of healthcare service events compared to control participants (mean number of events 0.43 intervention vs 0.52 control, adjusted rate ratio [RR], 0.91 [95% CI, 0.83-0.99], p=0.033, Table 1). The intervention was significantly more effective in older patients compared to younger patients (RR, 0.82, [95% CI, 0.72-0.93] in patients aged ≥65 years, compared to patients