Article | May 21, 2023
AI-enabled, home-based cardiac care by nurses – abstract presented at the ESC Heart Failure Congress 2023. AI point-of-care ultrasound can improve access, enabling novice nurses to perform cardiac screening in patients’ homes.
Key Findings
- In this prospective feasibility pilot in Tunisia, 7 novice nurses with 1 day of training performed AI-supported POCUS at patients' homes, with 5 of 7 achieving a minimum standard to participate, demonstrating that brief training is sufficient for task-shifting cardiac screening to non-specialist nurses.
- The AUC for detecting cardiac dysfunction (LVEF below 50% or LAVI above 34 mL/m2) was 0.86 (95% CI 0.77-0.96) with a specificity of 81%, establishing that home-based AI-POCUS by nurses can meaningfully screen for clinically significant abnormalities.
- Specificity was 89% for reduced LVEF and 85% for elevated LAVI individually, and the AUC of nurse-led AI-POCUS was significantly higher than NT-proBNP at a cut-off of greater than 125 pg/mL (P = 0.04), outperforming a standard blood biomarker used for the same purpose.
- The study provides the first prospective evidence that novice nurse-led home-based AI echo can detect cardiac dysfunction with diagnostic performance comparable to a clinic visit with a senior physician, directly addressing the access gap to echocardiography in low- and middle-income countries.
- These results, published in European Heart Journal Digital Health, provide the foundation for scaling home-based cardiac monitoring in resource-constrained settings where clinic attendance is a persistent barrier to early heart failure diagnosis.