Raktim Mukhopadhyay

Large language model (LLM)–based systems have tremendous potential to improve patient-centered health care, especially for medically underserved populations. However, realizing this potential requires careful consideration of the socio-technical contexts in which LLMs are used. The design of such systems should consider the vulnerabilities of any medically underserved group that the system intends to support, and provide trustworthy evidence for its use. This viewpoint reports the lessons learned from our experience and research with the CREATE (Culture, Respect, Education, Advancement, Trust, and Expertise) framework for engaging multiple stakeholders to guide the integration of LLM-based systems for improved health care delivery. We propose an extension of the traditional hierarchy of evidence model and identify key junction points in clinical workflows at which LLM-based systems can potentially be used to improve patient care. The key messages can be summarized as follows: (1) users of AI technologies, such as LLM-based tools, must be aware of the strengths, limitations, and impact of these technologies on health care delivery workflows and on the quality of generated evidence on which health care recommendations are based; (2) users must also be aware of the impact of data quality on AI outputs and on the evidence generated from the use of these systems; (3) the evidence pyramid provides a framework that facilitates the evaluation of the output generated by AI systems; (4) the CREATE framework facilitates the engagement of a variety of stakeholders. We propose concrete approaches for its validation and implementation; (5) any system designed to support people with opioid use disorder (OUD) needs to consider the overall lack of trust and stigmatizing experiences of this population with the health care system, and we discuss various aspects of the evaluation process necessary to build trust; (6) to discuss research directions that need to be addressed before LLM-based systems can be integrated usefully in patient health care.

UNSTRUCTURED Large language model-based systems (LLMs) have tremendous potential to improve patient-centered healthcare, especially for medically underserved populations. However, realizing this potential requires careful consideration of the socio-technical contexts in which LLMs are used. The design of such systems should consider the vulnerabilities of any medically underserved group that the system intends to support and provide trustworthy evidence for its use. In this work, we study the adoption of LLM-based systems as socio-technical systems from the diffusion of innovations (DOI) theory perspective. We extend the CREATE (Culture, Respect, Education, Advancement, Trust and Expertise) framework to guide integration of LLM-based systems for improved healthcare delivery. We also propose an extension of the traditional hierarchy of evidence model appropriate for our context and discuss risks associated with the use of such systems. Furthermore, we identify key junction points in clinical workflows at which LLM-based systems can potentially be used to improve patient care. In addition to general considerations associated with the development of any patient-oriented LLM-based system, any system designed to support people with opioid use disorder (OUD) needs to consider the overall lack of trust and stigmatizing experiences of this population with the healthcare system. We therefore discuss various aspects of the evaluation process necessary to build trust.

Even though several publicly accessible pharmacovigilance databases are available, extracting data from them is a technically challenging process. Existing tools typically focus on a single database. We present SurVigilance, an open-source tool that streamlines the process of retrieving safety data from seven major pharmacovigilance databases. SurVigilance provides a graphical user interface as well as functions for programmatic access, thus enabling integration into existing research workflows. SurVigilance utilizes a modular architecture to provide access to the heterogeneous sources. By reducing the technical barriers to accessing safety data, SurVigilance aims to facilitate pharmacovigilance research.

Background: People with opioid use disorder (OUD) have the highest rates of hepatitis C virus (HCV) infection. Despite the availability of curative HCV medication, people with OUD have limited health care access largely due to stigma. In a recent, pragmatic, randomized controlled trial (RCT), we compared a facilitated telemedicine intervention for HCV treatment integrated into opioid treatment programs (OTPs) with off-site referral. Facilitated telemedicine is bidirectional videoconferencing between a remote provider and a patient, supported by a case manager who facilitates the telemedicine encounter. The case manager schedules telemedicine visits, provides appointment reminders, and operates the digital equipment. Among 602 participants in the RCT, 90% (n=262) were cured through facilitated telemedicine and 39% (n=123) were cured through off-site referral. In this work, a multidisciplinary group of investigators, who directed the RCT, conducted a workshop, “Advancing Viral Hepatitis Screening and Treatment in Opioid Treatment Settings – Models & Resources,” at the American Association for Treatment of Opioid Dependence Conference in May 2024 to disseminate knowledge of facilitated telemedicine, including implementation considerations. We highlighted facilitated telemedicine as a patient-centered, sociotechnical, pragmatic health care delivery model for underserved populations. Objective: This study aimed to identify lessons learned to successfully overcome challenges of facilitated telemedicine implementation for HCV treatment integrated into OTPs. Methods: We partnered with the National Alliance of State and Territorial AIDS Directors in planning the workshop. The workshop consisted of 7 presentations on topics related to facilitated telemedicine implementation. The workshop was recorded and transcribed by Zoom (Zoom Communications). The transcripts served as data for the thematic analysis. The transcripts were interpreted to elucidate patterns of meanings and nuances derived from each presentation. In an iterative process, preliminary findings were compared and coalesced into themes. Verbatim quotes from the workshop were highlighted to support the themes. Results: We developed 3 themes. First, patient-centered care promotes HCV treatment for underserved populations through facilitated telemedicine. Case managers leveraged the destigmatizing environment of the OTP to build trust with patients, promoting an HCV cure through facilitated telemedicine. Second, sociotechnical approaches expand health care access for people with OUD. To be effective, facilitated telemedicine integrates 2 necessary components, a social aspect and a technical aspect. Third, facilitated telemedicine supports pragmatic research emphasizing people with OUD. Pragmatic research of facilitated telemedicine is needed to assess sustainability and scaling considerations beyond OTPs. Overall, we found that facilitated telemedicine overcame the digital divide, promoting access to digital technology, internet provision, and digital literacy. Conclusions: Facilitated telemedicine incorporates both a technical and a social component. The technical component largely addresses geographical challenges, while the social component addresses temporal (ie, care coordination) issues, promotes trust, and largely assuages patients’ concerns related to HCV treatment. The patient-centered, sociotechnical intervention can address the digital divide, thereby increasing health care access.

A suite of empirical Bayes methods to use in pharmacovigilance. Contains various model fitting and post-processing functions. For more details see Tan et al. (2025) , Koenker and Mizera (2014) and Efron (2016) .

BACKGROUND People with opioid use disorder (OUD) have the highest rates of hepatitis C virus (HCV) infection. Despite the availability of curative HCV medication, people with OUD have limited healthcare access largely due to stigma. In a recent, pragmatic, randomized controlled trial (RCT), we compared a facilitated telemedicine intervention for HCV treatment integrated into opioid treatment programs to offsite referral. Facilitated telemedicine is bidirectional videoconferencing between a remote provider and a patient, supported by a case manager who facilitates the telemedicine encounter. In the RCT, we found that facilitated telemedicine overcame the digital divide, promoting access to digital technology, internet provision, and digital literacy. In this work, a multidisciplinary group of investigators, who directed the RCT, conducted a workshop entitled “Advancing Viral Hepatitis Screening and Treatment in Opioid Treatment Settings – Models & Resources”. We sought to disseminate knowledge of facilitated telemedicine at the American Association for Treatment of Opioid Dependence Conference in May 2024. OBJECTIVE We identified lessons learned to successfully overcome challenges of facilitated telemedicine implementation for HCV treatment integrated into opioid treatment programs. METHODS We partnered with the National Alliance of State & Territorial AIDS Directors in planning the workshop. The workshop consisted of seven presentations on topics related to facilitated telemedicine implementation. The workshop was recorded and transcribed by Zoom. The transcripts served as data for the thematic analysis. The transcripts were interpreted to elucidate overall meanings and nuances derived from each presentation. In an iterative process, preliminary findings were compared until consensus was reached. Preliminary findings were aggregated and coalesced into themes. Verbatim quotes from the workshop were highlighted to support the themes. RESULTS We developed three themes: (1) Patient-centered care promotes HCV treatment for underserved populations through facilitated telemedicine. Case managers leveraged the destigmatizing environment of the opioid treatment program to build trust with patients, promoting an HCV cure through facilitated telemedicine. (2) Sociotechnical approaches expand healthcare access for people with OUD. To be effective, facilitated telemedicine integrates two necessary components, a social aspect and a technical aspect. (3) Facilitated telemedicine supports pragmatic research emphasizing people with OUD. A patient advisory committee, peer pipeline, and learning lunches were permissive to overcoming the digital divide. Additional pragmatic trials of facilitated telemedicine are needed to assess sustainability and scaling considerations beyond methadone treatment programs. CONCLUSIONS Facilitated telemedicine incorporates both a technical and a social component. The technical component largely addresses geographical challenges while the social addresses temporal (i.e., care coordination) issues, promotes trust, and largely assuages patients’ concerns related to HCV treatment. The patient-centered, sociotechnical intervention can satisfactorily address the digital divide thereby increasing healthcare access. CLINICALTRIAL ClinicalTrials.gov Identifier: NCT02933970