PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
January 14, 2026European Heart Journal - Digital HealthOpen Access

Remote and noninvasive monitoring of childhood cancer survivors for elevated NT-proBNP using Apple Watch ECG

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Cardiotoxic childhood cancer treatments place survivors at lifelong cardiovascular risk, and whether an existing ECG-AI model could detect elevated NT-proBNP using a single-lead Apple Watch ECG remained untested.

Does an ECG-AI model applied to a single-lead Apple Watch ECG accurately detect elevated NT-proBNP in adult survivors of childhood cancer?

Population

580 adult survivors of childhood cancer in the SJLIFE cohort

Comparison

ECG-AI model applied to Apple Watch ECG vs laboratory NT-proBNP

Design

Cohort diagnostic validation study

Key result

The ECG-AI model achieved an AUC of 0.84 for identifying elevated NT-proBNP levels (>300 pg/mL) in childhood cancer survivors using Apple Watch ECG.

Authors

KBKatie BarrIKI KarabayirDMD Mulrooney

Discussion

Loading...

Member takes

Overview

May support wearable ECG-AI screening for elevated NT-proBNP in survivors; leaves open prospective validation before clinical use.

Key Points

  • The aim is to evaluate whether an existing ECG-AI model can detect elevated NT-proBNP levels in childhood cancer survivors by utilizing a single-lead Apple Watch ECG.
  • Employed an existing ECG-AI model previously developed for standard clinical ECGs.
  • Collected paired single-lead Apple Watch ECGs and NT-proBNP data from the St. Jude Lifetime Cohort.
  • Applied the model without fine-tuning on the Apple Watch ECGs to predict NT-proBNP levels.
  • Calculated performance metrics, including area under the receiver operating characteristic curve (AUC) for the binary outcome.
  • Identified 2.8% of participants had elevated NT-proBNP levels (>300 pg/mL).
  • Achieved an AUC of 0.84 for detecting elevated NT-proBNP levels, indicating high accuracy in prediction.
  • Participants included 580 survivors with diverse demographics, primarily White and female.

Structured PICO

Does an ECG-AI model applied to a single-lead Apple Watch ECG accurately detect elevated NT-proBNP in adult survivors of childhood cancer?

P
Population
580 adult survivors of childhood cancer from the St. Jude Lifetime Cohort, mean age 37, 49.5% female, 82.2% White.
I
Intervention
ECG-AI model applied to single-lead Apple Watch ECG
O
Outcome
Detection of elevated NT-proBNP (>300 pg/mL)surrogate

An existing ECG-AI model applied to single-lead Apple Watch ECGs can accurately detect elevated NT-proBNP in adult survivors of childhood cancer, offering a potential non-invasive screening tool.

Cite This Study

Barr et al. (2026) studied this question. The ECG-AI model achieved an AUC of 0.84 for identifying elevated NT-proBNP levels (>300 pg/mL) in childhood cancer survivors using Apple Watch ECG.

synapsesocial.com/papers/696719c1c0d1e3cfbfce92d6https://doi.org/10.1093/ehjdh/ztaf143.084
View Full Paper
Ask AI
Bookmark
Share