Using Natural Language Processing to Identify Social Support Needs Among African American and Hispanic Family Caregivers for Persons with Dementia (Preprint)

Author:

Yoon SunmooORCID,Davis Nicole J,Moise Nathalie,Kronish Ian M.,Alcantara Carmela,Broadwell Peter,Nho Joo Young,Mittelman Mary

Abstract

BACKGROUND

Older African and Hispanic Americans are disproportionately more likely than older White Americans to have dementia.

OBJECTIVE

To assess social support needs among African American and Hispanic dementia caregivers to establish a foundation that can inform designs for culturally sensitive social support interventions.

METHODS

Design: This mixed-methods study applied natural language processing (NLP) on qualitative data. Setting: Academic medical center in New York Participants: Volunteer sample of seven experts participated together in the series of in-depth expert panels. The four dementia caregivers were recruited from the dementia caregiver registry, as well as via flyers posted within and outside of the institution. Health professionals were recruited from Columbia University Irving Medical Center. Eligible caregivers included those who had been a primary family dementia caregiver for at least three years, spoke English or Spanish, and used text messaging. Eligible professionals included those with at least 5 years of experience and publication on dementia caregiving. An eligible health communication specialist included one with a degree in health literacy.

RESULTS

A total of 72 topics (38 needs, 34 strategies, 2,322 terms) were identified through the NLP process: 1) 21 topics regarding disease, medication and communication from the information support corpus (coherence: 0.470, 38.77%); 2) 20 topics regarding care coordination and social services from the instrumental support corpus (coherence: 0.513, 23.97%); 3) 21 topics regarding psychological distress and copying from the emotional support corpus (coherence: 0.509, 21.25%); 4) 9 topics regarding self-evaluations on caregiving from the appraisal support corpus (coherence: 0.461, 15.99%).

CONCLUSIONS

This study is among the first to use NLP on qualitative data gathered from expert panels. New findings on unmet needs include for providers to disclose disease prognosis, to understand cultural norms on seizure treatment as a result of side effect of donepezil, and to familiarize with culturally sensitive terminologies and routine medications for dementia to reduce distrust.

CLINICALTRIAL

NCT03865498

Publisher

JMIR Publications Inc.

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3