Patrick Sean Sullivan

Background: State-level estimates of hepatitis C virus (HCV) prevalence are needed to guide resource allocation and to provide a baseline against which to measure progress toward elimination goals. Population prevalence is not directly observable, requiring estimation from indirect data sources. Methods: We adapted a Bayesian spatial integrated abundance model to estimate state-level HCV prevalence during 2017–2020 across 48 states and the District of Columbia. The model integrated six HCV-related outcomes: acute and chronic surveillance cases, HCV-related deaths, observations of HCV in [MarketScan] administrative claims data, diagnoses of HCV in Medicaid recipients, and treatment with direct-acting antivirals in Medicaid recipients. Estimates were anchored to a national prevalence estimate, and the model accounted for data source–specific selection, heterogeneity in HCV risk factors across states, and geospatial correlation. Results: Estimated average prevalence was 1.32% (95% credible interval [CrI]: 0.94%–1.80%), corresponding to 3.33 million (95% CrI: 2.37–4.65 million) adults with HCV infection across 48 states and DC. State estimates ranged from 0.74% in North Dakota to 2.22% in Oklahoma (median state-specific prevalence, 1.27%), with higher prevalence concentrated in South Central states, Appalachia, and the West, and the lower prevalence in the upper Midwest, Southeast, and New England. Ten states accounted for 56% of estimated infections. Estimates were stable in sensitivity analyses (most differences <1 percentage point). Conclusions: These estimates quantify state-level HCV burden prior to the 2021 federal expansion of HCV surveillance funding, providing a baseline for monitoring elimination progress. Geographic variation indicates that resource needs will differ by jurisdiction.

BACKGROUND HIV pre-exposure prophylaxis (PrEP) is highly effective in preventing HIV infection when taken as prescribed, but individual and structural barriers persist and limit access to and uptake of this effective preventive medicine. Following the launch of the End the HIV Epidemic (EHE) initiative in the US in 2019, a variety of HIV prevention programs have been implemented, ranging from telemedicine PrEP prescribing to messaging campaigns to facilitate PrEP use. However, limited information exists regarding the types and characteristics of the programs implemented. OBJECTIVE Describe (1) HIV prevention efforts, with a focus on PrEP promotion programs, that were implemented by health departments between 2016-2025 in the US, (2) the distribution and characteristics of funding sources by program types, and (3) program-specific characteristics. METHODS We developed a survey to assess PrEP promotion programs. The survey was refined based on feedback from an advisory board comprising volunteers from 14 health jurisdictions. We identified eight PrEP promotion programs implemented between 2016-2025 - (1) PrEP Navigation, (2) PrEP Outreach and Education, (3) Messaging Campaigns, (4) Training for Clinicians and Clinical Staff, (5) Telemedicine PrEP, (6) PrEP Assistance Program, (7) Safety-Net Clinics, and (8) Optimization of Electronic Medical Record (EMR)/Electronic Health Record (EHR). Survey questions included whether a jurisdiction had such a program in place, program start and end periods, funding source, period of funding coverage, and program-specific contextual questions. The survey was sent to all state health departments, the District of Columbia, and seven local health jurisdictions (n=58) via email. The survey was launched on May 14, 2025, and data collection concluded on Apr 30, 2026. RESULTS 47/58 (81%) health jurisdictions completed the survey. PrEP navigation (79%) was the most common program and EHR changes (19%) was the least common. Among the implemented programs, the proportion reported as ongoing at the time of the survey ranged between 58%-94%. The proportion of health jurisdictions that received federal funding to support each of the programs ranged from 53%-100%, state funding ranged from 32%-59%, and other funding ranged from 18%-32%. The characteristics identified were specific to each program. For example, PrEP navigation programs (n=37) included services such as linkage to a PrEP provider (100%), PrEP counseling (95%), insurance/benefit navigation (84%), and communication and engagement support (65%). The proportion of health jurisdictions implementing each program increased from 2%–23% in 2016 to 17%–70% in 2025. CONCLUSIONS A variety of PrEP promotion programs were increasingly implemented from 2016-2025 across the nation, with the majority of programs active at the time of the survey. Programs were supported by a range of funding sources. Government grant opportunities played an important role in enabling program implementation. Future research should evaluate the effectiveness of these programs and their impact on PrEP uptake and retention in PrEP care. CLINICALTRIAL NA INTERNATIONAL REGISTERED REPORT RR2-10.2196/80911

Background: Chatbots have the potential to reduce barriers to preexposure prophylaxis (PrEP), including lack of awareness, misconceptions, and stigma, by providing anonymous and continuous support. However, in the context of PrEP, chatbots are still nascent; they lack personalized informational expertise, peer experiential expertise, and human-like emotional support to facilitate future PrEP uptake. These personalized and relatable forms of support are crucial for increasing engagement, influencing health decisions, and fostering resilience and well-being. Objective: In this paper, we describe the iterative development and evaluation plans of a retrieval-augmented generation (RAG) chatbot for providing personalized information, peer experiential expertise, and human-like emotional support to PrEP candidates. Methods: We used an iterative design process consisting of 2 phases: prototype conceptualization and iterative chatbot development. In the conceptualization phase, we identified real-world PrEP needs and designed a functional dialogue flow diagram for PrEP support. Chatbot development included developing 2 components: a query preprocessor and a RAG module. The preprocessor uses the Segment Any Text (SAT) tool (developed by Markus Frohmann, Igor Sterner, Ivan Vulić, Benjamin Minixhofer, and Markus Schedl) for query segmentation and a Gemma 2 fine-tuned support classifier to identify informational, emotional, and contextual data from real-world queries. To implement the RAG module, we used Sentence-Bidirectional Encoder Representations from Transformers (SBERT) embeddings with cosine similarity, and performed topic matching to identify topically relevant documents based on the query topic to support document retrieval. Extensive prompt engineering was used to guide the large language model (LLM), Gemini-2.0-flash, in generating tailored responses. We conducted 10 rounds of internal evaluations to assess and improve the chatbot responses based on 10 criteria: clarity, accuracy, actionability, relevancy, information detail, tailored information, comprehensiveness, language suitability, tone, and empathy. Finally, we conducted a blinded comparative study and used a linear mixed model to validate the chatbot against a general LLM and real-world user responses. Results: We developed a RAG chatbot and iteratively refined it based on the internal evaluation feedback. Prompt engineering is essential in guiding the LLM to generate responses tailored to information, experiential, and emotional user needs. We found that prompt effectiveness varied with task complexity; this was likely due to LLM sensitivity to the structure of prompts and to linguistic variability. Prompt decomposition and segmenting prompt instructions helped improve comprehensiveness and relevancy for complex and long queries. The linear mixed model analysis revealed that while both the general LLM and RAG were preferred over user responses, the general LLM maintained a superior edge over the RAG chatbot across most criteria, with the exceptions of empathy and tone. Conclusions: Our RAG chatbot leverages social media data to provide personalized information, peer experiences, and human-like emotional support; these elements are essential for addressing PrEP misconceptions and promoting self-efficacy. Further analysis incorporating expert and user feedback will be conducted to help validate and improve the chatbot’s potential.

OBJECTIVES: Effective multidirectional and longitudinal communication is essential to patient- and family-centered care (PFCC) in the ICU. While communication technologies have the potential to facilitate effective communication, their utilization, especially during the COVID-19 pandemic, has been variable. This scoping review answers the following research questions: What communication tools, strategies, and programs have been used in ICUs? What are their reported effects on PFCC and multidirectional communication? To what extent do these studies consider equity, diversity, and inclusion principles? DATA SOURCES: The scoping review followed Arksey and O’Malley’s (2005) framework and adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews. We conducted a search covering the period from 2007 to 2024, focusing on the concepts of communication, ICU, and PFCC in OVID Medline, EBSCO CINAHL, and OVID Embase. Relevant studies were screened and reported in this review. STUDY SELECTION: Searches identified 3280 unique sources, which were screened at a three-stage asynchronous review. A total of 131 articles were included in this review. DATA EXTRACTION: Communication tools, strategies, and programs were categorized into technology-based and non–technology-based interventions. Technology-based interventions included eight subcategories such as data-enabled devices, electrolarynx, and eye trackers. Non–technology-based strategies included 13 subcategories such as family meetings, diaries, and dedicated personnel. DATA SYNTHESIS: Positive outcomes were reported in 82% of studies for both technology-based and non–technology-based communication tools, strategies, and programs on PFCC. CONCLUSION: Most technology and non–technology communication tools, strategies, and programs positively impact communication and PFCC in the ICU. However, only 4% of studies focused on equity, diversity, and inclusion in the evaluation, design, or development of their intervention, indicating a need for further research in this area.

Objective: This review aims to identify the contribution of user experience features and underlying technical features to sustained engagement in unguided chatbots for improving health-related behaviors. Materials and Methods: Following PRISMA-2020 guidelines, we conducted a systematic review, searching PubMed, ACM, APA PsycINFO, Cochrane, Web of Science, and IEEE Xplore from June to September 2022 and updated in April 2025. Data was analyzed via Synthesis without Meta-Analysis (SWiM), to understand the relationship between user engagement overall and individual experience metrics. Results: Customizable avatars and flexible input interactions may enhance overall user engagement. Conversely, pre-scripted content that lacks personalization and emotional support negatively impacts user satisfaction and adherence to health interventions. Other features contributing to sustained engagement are in-app technical assistance, user learning features, and crisis support systems. A strong positive correlation (r = 0.808, n = 16) was observed between user satisfaction and engagement, specifically for satisfaction dimensions including need fulfillment (r = 0.872, n = 6), willingness to recommend chatbot (r = 0.817, n = 4) and user enjoyment (r = 0.971, n = 3) in SWiM analysis. The limited application of large language models and retrieval augmented generation techniques may constrain the quality of support available to users and overall sustained engagement. Conclusion: Effective unguided chatbot design requires an emphasis on interactive educational elements, in-app technical assistance and crisis support, and personalized content. This can be achieved with high context awareness, input understanding, and quality content generation. Our findings suggest that user satisfaction is a primary driver of sustained engagement, though further research is needed to validate individual user satisfaction features for sustained engagement.

Background: Accurate estimates of the population sizes of men who have sex with men (MSM) are essential for evaluating HIV-related interventions. We aimed to estimate MSM populations at county and state levels using recent data from the decennial census and the National Health and Nutrition Examination Survey. Methods: We used 2020 Decennial Census data to calculate a weight for each U.S. county, reflecting the proportion of male-male partner households relative to counties of similar urbanicity based on the 2013 National Center for Health Statistics Urban–Rural Classification Scheme. We applied these weights to urbanicity-stratified estimates of the prevalence of adult men who reported sex with a man in the past 12 months, derived from National Health and Nutrition Examination Survey (2015–2020). Multiplying these percentages by adult male populations produced county estimates, which were aggregated to state levels. Results: We estimated approximately 2.1 million MSM in the United States, representing 1.7% of adult males. State-level estimates ranged from 0.3% in Wyoming (n = 675) to 3.7% in the District of Columbia (n = 9709). California had the largest MSM population (n = 318 612; 2.1%), followed by Florida (n = 191 199; 2.3%) and Texas (n = 173 751; 1.6%). At the county level, Los Angeles County, CA, had the largest MSM population (n = 82 513; 2.1%), followed by Cook County, IL (n = 39 580; 2.0%), and Broward County, FL (n = 34 321; 4.7%). Broward County, FL (4.7%), and San Francisco County, CA (4.6%), had the highest proportions relative to their male populations. Conclusions: County- and state-level MSM estimates provide crucial denominators for calculating disease rates and informing public health interventions.

Background: Antiretroviral therapy (ART) has become a cornerstone of human immunodeficiency virus (HIV) management. However, a challenge in HIV care and policy is ensuring individuals remain engaged in care and on treatment over time. Discontinuation of ART is common for various reasons, and prolonged treatment interruptions can lead to worse health outcomes at the individual level and increased HIV transmissions at the public health level. Objective: A cost-consequence analysis was conducted to evaluate the economic and public health impact of reducing the interval to ART restart among people with HIV (PWH) who have disengaged from care. Methods: A state transition disease model was developed to calculate the economic benefits from improving treatment restart patterns from a United States healthcare payer perspective. Two hypothetical cohorts of 1000 PWH who discontinued ART were compared: a standard-of-care cohort where restart occurs 32 weeks after discontinuation, and a comparator cohort exploring the impact of reducing the time between ART discontinuation and restart to 12 weeks. Individuals were assigned to CD4-related health states, and rates of viral suppression were considered. Four outcomes, ART costs, CD4 health state costs, CD4-related mortality, and new HIV transmissions were calculated over a three-year time horizon. Cost savings from averted HIV cases were valued based on the lifetime excess healthcare costs for a PWH. Results: Increasing the proportion of individuals restarting ART and reducing time to restart was estimated to avert 88 HIV transmissions. This corresponds to a number needed to treat, defined as the number of PWH who would need to experience the earlier restart pattern of the comparator cohort, of 11 to avoid one new transmission, and $101 083 857 lifetime cost savings. Cost savings attributable to improved CD4 counts in the cohort were also found. Conclusion: Enhancing ART restart patterns improves health and provides considerable cost savings by improving individuals’ CD4 counts and reducing new HIV transmissions from people who are viremic. Effective policies to promote care engagement and treatment adherence are predicted to improve the health of PWH and reduce new HIV cases.

Background: Transgender women are disproportionately affected by HIV in the United States. Evidence shows that HIV self-testing increases awareness of HIV status, preventing transmission. Culturally appropriate strategies are needed to provide HIV self-tests (HIVSTs) to transgender women. Objective: This paper aims to describe an adaptation for transgender women of a comparative effectiveness trial of an intervention (Know@Home website) on the use of HIVSTs, counseling, and linkage of participants to appropriate services initially designed to reach Black or African American and Hispanic or Latino men who have sex with men (Implementation of Rapid HIV Self-Testing Among MSM Project). Methods: The ADAPT-ITT (Assessment, Decision, Adaptation, Production, Topical Experts, Integration, Training, and Testing) framework guided the adaptation. Project staff worked with consultants and a community advisory board to revise all study materials (ie, intervention web content, marketing text and images, and language used in the surveys) to be tailored to, and inclusive of, health topics relevant to transgender women. We conducted a pilot study with transgender women recruited through online venues. Participants were randomly assigned 2:1 to the intervention arm or control arm. All participants were mailed 2 HIVSTs, the OraQuick In-Home HIV Test, and had access to a standard resource list including information on how to find a nearby testing location, HIV prevention, and sexual health. Only intervention arm participants had access to the study website, “Know@Home.org.” The Know@Home website allowed participants in the intervention arm to order sexually transmitted infection test kits, condoms, and lubricant and provided HIV prevention information, service locators, and risk assessment tools. Adaptations to the intervention included modifying language, images, and links to outside resources to be transgender specific or gender neutral. Online surveys were adapted by removing inappropriate gender terms and replacing them with culturally appropriate terms for gender identity, genitalia, and condom use. Four months after completing the baseline (enrollment) survey, all participants were asked to complete a final follow-up survey. Upon completing the 4-month survey, participants were mailed a dried blood spot collection kit to be returned by mail for the laboratory testing of the sample. The results of all HIVSTs could be reported in real time in an online survey. Video health counseling was available, upon request, for all participants during study participation. Results: Enrollment into the pilot study began on April 22, 2021, and concluded on July 7, 2021, yielding a total of 102 transgender women participants, with data collection completed in January 2022. Conclusions: By using the ADAPT-ITT model, we produced materials to be culturally and linguistically appropriate for transgender women. The findings of this study have the potential to inform future research studies among transgender women and underscore the importance of involving subject matter experts and community members in the development of tailored materials.