# Addressing Disparities in Out-of-Hospital Cardiac Arrest: Utilization of Health-Related Social Needs and Predictive Analytics to Improve Clinical Outcomes

> **NIH NIH K08** · ICAHN SCHOOL OF MEDICINE AT MOUNT SINAI · 2024 · $167,383

## Abstract

PROJECT SUMMARY
 The overarching goal of this K08 research project is for Dr. Ethan Abbott, principal investigator (PI), to
establish himself as an independent physician-scientist whose research addresses healthcare-related
disparities and improves survival for out-of-hospital cardiac arrest (OHCA) patients. He has prepared with
assistance of his multi-disciplinary mentorship team a compelling and innovative research project with
matching career development training components that enables him to conduct this research, prepare and
submit his subsequent R01-supported project from the preliminary data, and launch his research career.
 Dr. Abbott’s K08 research project aims to improve OHCA survival and clinical outcomes by identifying
important individual-level health-related social needs (HRSN) domains to improve prediction of 30-day survival
after OHCA and survival to hospital discharge. Despite the importance of individual-level HRSN in health
outcomes, current OHCA predictive models only account for clinical variables, resulting in significant limitations
towards advancing health equity and improving care for patients. Use of data science techniques, such as
natural language processing (NLP) and large language models (LLMs) to identify and extract HRSN for
inclusion in predictive models, could lead to interventions that decrease OHCA mortality.
 The specific aims of Dr. Abbott’s K08 research project are to: (1) Create a baseline predictive model to
identify key pre-hospital, patient-level, hospital-level and clinical predictors of 30-day survival after OHCA and
survival to hospital discharge; (2) Evaluate the efficacy of NLP and LLMs to extract individual-level HRSN for
the OHCA cohort; and (3) Determine if inclusion of individual-level HRSN increase performance of the baseline
predictive model in predicting 30-day post-OHCA survival and survival to hospital discharge, and if the HRSN-
inclusive model performs better than the prior models NULL-PLEASE and CAST. He will rigorously develop the
models using mediation analyses, multivariable regression, and machine learning algorithms.
 Dr. Abbott’s career development plan builds on his experience as an emergency medicine physician and
junior faculty researcher to develop and acquire new skills and expertise in: (1) formal mediation analysis; (2)
application of NLP algorithms and LLMs for electronic health record information (EHR) extraction, particularly
for HRSN; (3) predictive analytics for clinical outcomes using machine learning algorithms; and (4) research
independence through professional development activities, including committee leadership positions,
grantsmanship, dissemination of research findings. The results of this K08 will generate preliminary data to
form the basis of Dr. Abbott’s subsequent R01 application submission. His R01 study will externally validate a
predictive model for OHCA and extraction using LLMs for individual-level HRSN. The public health importance
of this work is th...

## Key facts

- **NIH application ID:** 10985877
- **Project number:** 1K08HL169980-01A1
- **Recipient organization:** ICAHN SCHOOL OF MEDICINE AT MOUNT SINAI
- **Principal Investigator:** Ethan Abbott
- **Activity code:** K08 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $167,383
- **Award type:** 1
- **Project period:** 2024-09-01 → 2029-08-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10985877

## Citation

> US National Institutes of Health, RePORTER application 10985877, Addressing Disparities in Out-of-Hospital Cardiac Arrest: Utilization of Health-Related Social Needs and Predictive Analytics to Improve Clinical Outcomes (1K08HL169980-01A1). Retrieved via AI Analytics 2026-05-23 from https://api.ai-analytics.org/grant/nih/10985877. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*