Recognition of heart failure with micro electro-mechanical sensors using commercially available smartphone, the REFLECS study

Abstract Background Heart failure (HF) is the leading cause of hospitalization in people over the age of 65 years. More recently, cardiac motion sensor technology has emerged as a promising technology to detect HF. Methods In this multicenter study, we examined whether accelerometer and gyroscope signals from motion sensors collected using commercially available smartphones can classify HF status. Participants were enrolled from Finland and the United States. Participants hospitalized with acute decompensated HF were assessed in the acute state and re-assessed in the stabilized state prior to discharge. Outpatient participants were assessed in the stable state. In a pre-specified pooled data analysis, state specific algorithms to detect HF were first derived using logistic regression and validated using boostrap aggregation method (10 repeats) followed by sensitivity analysis in participants with HF with reduced or preserved ejection fractions, participants in atrial fibrillation and according to age, sex or body mass index. Results A total of 217 HF participants had a total of 474 assesment: acute state = 174; stabilized state = 128; stable state = 172. 786 controls were included. The mean age of all participants with HF was 67 years, mean ejection fraction (EF) 39% and mean NT-proBNP 5778 ng/l. 74% had HF with reduced EF (HFrEF) and 38% had atrial fibrillation. Across all three HF states, the algorithm had an AUC of 0.95, sensitivity of 85%, specificity 90%, and an accuracy of 89%. The performance of the algorithm was not affected by HF state, ejection fraction, age, sex, body mass index (BMI) or presence of atrial fibrillation. Conclusion A simple commercially available smartphone based assessment of cardiac function with motion sensors shows distinct features in participants in decompensated, stabilized and stable HF states in comparison to controls.Heart failure detection ROC curves