(Press-News.org) As artificial intelligence is rapidly developing and becoming a growing presence in healthcare communication, a new study addresses a concern that large language models (LLMs) can reinforce harmful stereotypes by using stigmatizing language. The study from researchers at Mass General Brigham found that more than 35% of responses in answers related to alcohol- and substance use-related conditions contained stigmatizing language. But the researchers also highlight that targeted prompts can be used to substantially reduce stigmatizing language in the LLMs’ answers. Results are published in The Journal of Addiction Medicine.
“Using patient-centered language can build trust and improve patient engagement and outcomes. It tells patients we care about them and want to help,” said corresponding author Wei Zhang, MD, PhD, an assistant professor of Medicine in the Division of Gastroenterology at Mass General Hospital, a founding member of the Mass General Brigham healthcare system. “Stigmatizing language, even through LLMs, may make patients feel judged and could cause a loss of trust in clinicians.”
LLM responses are generated from everyday language, which often includes biased or harmful language towards patients. Prompt engineering is a process of strategically crafting input instructions to guide model outputs towards non-stigmatizing language and can be used to train LLMs to employ more inclusive language for patients. This study showed that employing prompt engineering within LLMs reduced the likelihood of stigmatizing language by 88%.
For their study, the authors tested 14 LLMs on 60 generated clinically relevant prompts related to alcohol use disorder (AUD), alcohol-associated liver disease (ALD), and substance use disorder (SUD). Mass General Brigham physicians then assessed the responses for stigmatizing language using guidelines from the National Institute on Drug Abuse and the National Institute on Alcohol Abuse and Alcoholism (both organizations’ official names still contain outdated and stigmatizing terminology).
Their results indicated that 35.4% of responses from LLMs without prompt engineering contained stigmatizing language, in comparison to 6.3% of LLMs with prompt engineering. Additionally, results indicated that longer responses are associated with a higher likelihood of stigmatizing language in comparison to shorter responses. The effect was seen across all 14 models tested, although some models were more likely than others to use stigmatizing terms.
Future directions include developing chatbots that avoid stigmatizing language to improve patient engagement and outcomes. The authors advise clinicians to proofread LLM-generated content to avoid stigmatizing language before using it in patient interactions and to offer alternative, patient-centered language options. The authors note that future research should involve patients and family members with lived experience to refine definitions and lexicons of stigmatizing language, ensuring LLM outputs align with the needs of those most affected. This study reinforces the need to prioritize language in patient care as LLMs become increasingly used in healthcare communication.
Authorship: In addition to Zhang, Mass General Brigham authors include Yichen Wang, Kelly Hsu, Christopher Brokus, Yuting Huang, Nneka Ufere, Sarah Wakeman, and James Zou.
Disclosures: None.
Funding: This study was funded by grants from the May Center Clinic for Digital Health in partnership with the Mayo Clinic Office of Equity, Inclusion, and Diversity and Dalio Philanthropies.
Paper cited: Wang, Y. et. al. “Stigmatizing Language in Large Language Models for Alcohol and Substance Use
Disorders: A Multi-Model Evaluation and Prompt Engineering Approach” DOI: 10.1097/ADM.0000000000001536
###
About Mass General Brigham
Mass General Brigham is an integrated academic health care system, uniting great minds to solve the hardest problems in medicine for our communities and the world. Mass General Brigham connects a full continuum of care across a system of academic medical centers, community and specialty hospitals, a health insurance plan, physician networks, community health centers, home care, and long-term care services. Mass General Brigham is a nonprofit organization committed to patient care, research, teaching, and service to the community. In addition, Mass General Brigham is one of the nation’s leading biomedical research organizations with several Harvard Medical School teaching hospitals. For more information, please visit massgeneralbrigham.org.
END
A study just published on Eurosurveillance has found evidence of substantial lifetime exposure to hepatitis B and C viruses in Ukraine in a 2021 nationwide, representative sample of the population, with findings also suggesting significant ongoing circulation of these viruses.
Ahead of World Hepatitis Day 2025, viral hepatitis remains a major global health concern, and is one of the priority infectious diseases under the United Nations’ Sustainable Development Goal 3. [1][2] The World Health Organization (WHO) Global Hepatitis ...
The National Biofoundry Project Team at the Korea Research Institute of Bioscience and Biotechnology (KRIBB), led by Dr. Haseong Kim, has spearheaded an international joint research effort (including institutions from Korea, the U.S., the U.K., Singapore, and others—10 in total) to create a new standard framework that simplifies and enhances the accuracy and efficiency of synthetic biology research. This framework is anticipated to serve as an international standard for biofoundries—automated laboratories in synthetic biology.
Driven by advancements in deep-tech fields such as synthetic biology ...
Use of billing codes in big data sets to find diagnoses can result in up to two-thirds of cases being mistakenly identified, new UCLA-led research finds.
Databases frequently used for medical research such as those for the Centers for Medicare & Medicaid Services or the National Inpatient Survey typically rely on ambulatory billing codes to identify diseases or medical procedures, but their accuracy is rarely verified in publications that rely on this data, the researchers write in a report published in the peer-reviewed journal ...
Hospitalizations and emergency department (ED) visits can be distressing and costly for nursing home residents – especially those who are severely impaired or terminally ill. Despite their vulnerability, these individuals are frequently transferred to hospitals, even though up to 40% of such transfers over the past 25 years are considered potentially avoidable by health care professionals.
These unnecessary transfers not only cause distress and discomfort for residents and families but also lead to hospital-acquired complications and added costs for the health care system. In the United States, hospital transfers from nursing homes significantly ...
Schizophrenia is a serious mental health condition that affects thoughts, moods, perceptions, and behaviors. Affected individuals experience positive symptoms like delusions and hallucinations, and negative symptoms like social withdrawal, cognitive deficits, disorganized thoughts and speech, and a decreased experience of pleasure. While schizophrenia is caused by a complex interplay of genetic, environmental, and lifestyle factors, precise mechanisms remain elusive.
Animal models provide valuable insights into the neurobiological mechanisms that underpin schizophrenia. However, conventional behavioral assessments ...
Chemotherapy-induced nausea and vomiting are among the most distressing side effects of anti-cancer treatment, particularly for those receiving highly emetogenic regimens such as anthracycline plus cyclophosphamide combinations. This major side effect compromises a patient’s quality of life and willingness to continue therapy. Therefore, there is a crucial need to devise an effective antiemetic management approach for optimizing cancer care and patient well-being.
Against this backdrop, a new study, led by Professor Mitsue ...
Immune checkpoint inhibitors (ICIs), a powerful form of immunotherapy, have revolutionized cancer treatment by unleashing the body’s own immune system to fight tumors. These compounds target the programmed cell death-ligand 1 (PD-L1), a surface protein typically found on tumor cells, which enables the tumors to avoid recognition by immune T cells. By disrupting PD-L1’s function with specially tailored antibodies, ICI-based strategies have brought hope to countless patients with cancer. However, ...
Self-disclosure, or the process of conveying one’s details to others verbally, is crucial for communication. Self-disclosure includes expressing personal information, thoughts, and feelings. It encompasses self-expression and clarification, social validation and control, as well as relationship development, and is closely related to reciprocity, intimacy, trust, interactional enjoyment, and satisfaction.
In recent years, technological advancements have paved the way for new forms of communication, including video-conferencing and embodied virtual ...
In a first-of-its-kind experiment tracing evolution across 25 generations, scientists have discovered that marine copepods—the tiny crustaceans at the heart of the ocean food web—rely on a largely unknown biological toolkit to survive the stresses of climate change.
Published July 15, 2025, in the Proceedings of the National Academy of Sciences, the study reveals that it’s not only genetic changes (permanent alterations to DNA) that help these animals adapt to warming and acidifying ocean conditions. In addition, little-known epigenetic changes (temporary “on/off” chemical modifications to parts ...
The accumulation of misfolded proteins in the brain is central to the progression of neurodegenerative diseases like Huntington’s, Alzheimer’s and Parkinson’s. But to the human eye, proteins that are destined to form harmful aggregates don’t look any different than normal proteins. The formation of such aggregates also tends to happen randomly and relatively rapidly – on the scale of minutes. The ability to identify and characterize protein aggregates is essential for understanding and fighting neurodegenerative diseases.
Now, using deep learning, ...
