An AI-Based Whole-Person Care Summarization Tool for Care Management Providers

To provide effective care and support to patients with complex medical and social needs, providers must have synthesized whole-person insights on their health-related social needs (HRSN). Often, the only sources of these insights are large numbers of unstructured data coming from a range of sources, which providers struggle to synthesize. Care team members at Pair Team, a medical group that delivers a high-needs care management program for Medicaid beneficiaries in California, routinely review lengthy patient charts comprising demographics, medical and behavioral history, HRSN data, care team interaction history, and care plan information. Pair Team developed and deployed an AI Patient Summary tool to synthesize these diverse information sources. The tool exhibits high accuracy and minimal demographic bias. In full deployment across the practice, it generates approximately 300 summaries daily and has decreased the time spent reviewing a patient chart. This framework provides a scalable model for organizations seeking to safely utilize AI for data extraction to enable providers of complex care management to deliver care more effectively.

Authors

Institutions

Publication Details

Journal
NEJM Catalyst
Published
2026-09-16
DOI
https://doi.org/10.1056/cat.26.0025
Primary Topic
Machine Learning in Healthcare
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
article

An AI-Based Whole-Person Care Summarization Tool for Care Management Providers

Shefali Kumar, Jessie Juusola, Nathan Favini, Neil Batlivala et al.
NEJM Catalyst
Machine Learning in Healthcare
article

An AI-Based Whole-Person Care Summarization Tool for Care Management Providers

Shefali Kumar, Jessie Juusola, Nathan Favini, Neil Batlivala, Kejia Zhu, Jacob Mulligan
article en

Abstract

To provide effective care and support to patients with complex medical and social needs, providers must have synthesized whole-person insights on their health-related social needs (HRSN). Often, the only sources of these insights are large numbers of unstructured data coming from a range of sources, which providers struggle to synthesize. Care team members at Pair Team, a medical group that delivers a high-needs care management program for Medicaid beneficiaries in California, routinely review lengthy patient charts comprising demographics, medical and behavioral history, HRSN data, care team interaction history, and care plan information. Pair Team developed and deployed an AI Patient Summary tool to synthesize these diverse information sources. The tool exhibits high accuracy and minimal demographic bias. In full deployment across the practice, it generates approximately 300 summaries daily and has decreased the time spent reviewing a patient chart. This framework provides a scalable model for organizations seeking to safely utilize AI for data extraction to enable providers of complex care management to deliver care more effectively.

NEJM CatalystVol. 7(10)
National Society of Professional Engineers (US), University of California System (US), Office of the Chief Scientist (IL), Health Outcomes Solutions (United States) (US), Berkeley Public Health Division (US), Carnegie Mellon University (US), University of California, Berkeley (US)
Openalex Percentile: Top 9%
Machine Learning in Healthcare
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

An AI-Based Whole-Person Care Summarization Tool for Care Management Providers — Shefali Kumar, Jessie Juusola, et al. · NEJM Catalyst (2026) | TGRS Research Map | TGRS