Comparing barriers to employee assistance program utilization in Canada and the United States using natural language processing and machine learning

Employee Assistance Program (EAP)s provide critical mental health support, yet utilization remains low in both Canada and the United States. Although qualitative studies have examined general barriers to EAP use, few have compared country-specific perceptions. This study explores barriers to EAP usage and explores potential differences in barriers between Canadian and American workers using natural language processing (NLP). This mixed-methods study, involving data transformation, included 30 semi-structured interviews with EAP-eligible employees (aged 18-65 years) in Canada and the US who had not previously used EAP services. Interviews were analyzed thematically using an inductive approach in alignment with Braun and Clarke's Thematic Analysis. The qualitative portion of this study was reported in accordance with the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines. In Phase 2 of the study, a BERT-based NLP model was trained to classify qualitative quotes by country. In phase 1, three core themes were observed: (1) Structural, informational and systematic barriers influencing EAP accessibility and coverage; (2) Reluctance to use EAP services due to psychosocial and perceptual barriers; and (3) Service quality, scope and cultural relevance to those with multiple social identities. While most barriers were common across countries, US participants reported financial concerns exclusively. Comparatively, in the NLP analysis, financial accessibility and comprehensive support through diverse EAP offerings were important in classifying American quotes, while stigma, vulnerability, and fear of workplace judgement and perceived quality and cultural relevance of EAPs were important in classifying Canadian quotes. Although many barriers to using EAPs are shared, financial and perceptual differences exist between Canadian and American workers. EAPs may benefit from implementing country-specific strategies. This could include creating strategies reducing stigma and quality concerns in Canada, while addressing concerns about cost and alternative options in the United States. Further studies should focus on generalizability and explore sector-specific and longitudinal trends in EAP perceptions.

Authors

Institutions

Publication Details

Journal
PLOS mental health.
Published
2026-04-03
DOI
https://doi.org/10.1371/journal.pmen.0000589
Primary Topic
Workplace Health and Well-being
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Comparing barriers to employee assistance program utilization in Canada and the United States using natural language processing and machine learning

Charlene Choi, Donia Obeidat, Kishana Balakrishnar, Raihana Premji et al.
PLOS mental health.
Workplace Health and Well-being
article

Comparing barriers to employee assistance program utilization in Canada and the United States using natural language processing and machine learning

Charlene Choi, Donia Obeidat, Kishana Balakrishnar, Raihana Premji, Behdin Nowrouzi-Kia, Javier Mencia Ledo, Ali Bani-Fatemi, Hong Ki Chloe Lau, Paula Allen, Marilyn Grand’Maison, Allison Kelly, Sana Siddiqui
article en

Abstract

Employee Assistance Program (EAP)s provide critical mental health support, yet utilization remains low in both Canada and the United States. Although qualitative studies have examined general barriers to EAP use, few have compared country-specific perceptions. This study explores barriers to EAP usage and explores potential differences in barriers between Canadian and American workers using natural language processing (NLP). This mixed-methods study, involving data transformation, included 30 semi-structured interviews with EAP-eligible employees (aged 18-65 years) in Canada and the US who had not previously used EAP services. Interviews were analyzed thematically using an inductive approach in alignment with Braun and Clarke's Thematic Analysis. The qualitative portion of this study was reported in accordance with the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines. In Phase 2 of the study, a BERT-based NLP model was trained to classify qualitative quotes by country. In phase 1, three core themes were observed: (1) Structural, informational and systematic barriers influencing EAP accessibility and coverage; (2) Reluctance to use EAP services due to psychosocial and perceptual barriers; and (3) Service quality, scope and cultural relevance to those with multiple social identities. While most barriers were common across countries, US participants reported financial concerns exclusively. Comparatively, in the NLP analysis, financial accessibility and comprehensive support through diverse EAP offerings were important in classifying American quotes, while stigma, vulnerability, and fear of workplace judgement and perceived quality and cultural relevance of EAPs were important in classifying Canadian quotes. Although many barriers to using EAPs are shared, financial and perceptual differences exist between Canadian and American workers. EAPs may benefit from implementing country-specific strategies. This could include creating strategies reducing stigma and quality concerns in Canada, while addressing concerns about cost and alternative options in the United States. Further studies should focus on generalizability and explore sector-specific and longitudinal trends in EAP perceptions.

PLOS mental health.Vol. 3(4)
University Health Network (CA), Centre for Addiction and Mental Health (CA), Jordan University of Science and Technology (JO), University of Toronto (CA), Toronto Rehabilitation Institute (CA), Telus (Canada) (CA), Mental Health Research Canada (CA), Krembil Foundation (CA), Krembil Research Institute
Mitacs
Quality Education
Openalex Percentile: Top 30%
Workplace Health and Well-being
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.