Yongjie Sha

BACKGROUND There are increasing interests in developing AI tools to identify and address individual-level social determinants of health in both health care and human service settings. These activities are part of social care integration, which at the individual level involves identifying individuals with social risks (awareness) and connecting them with relevant social care resources (assistance). Social care providers such as community health workers and social workers are deemed critical stakeholders in both settings. Chatbots have shown feasibility and acceptability for social risk screening in emergency departments and primary care centers. However, we do not know if a screening chatbot is worth developing for the safety net health care and human service settings, where there are visitors with greater social needs and less organizational resources. OBJECTIVE The study aims to investigate the perceived value proposition of an AI-based chatbot for social risk screening from the perspectives of social care providers in safety net health care and human service organizations. Providers’ perceived value propositions of other AI-based applications for social care integration were also examined. METHODS We conducted semi-structured interviews with 19 social care providers who have experience with awareness and/or assistance from 16 safety net health care and human service organizations in Michigan. Interview questions focused on their experiences and challenges regarding awareness and assistance when applicable. A simulated screening chatbot based on ChatGPT-4o was also used to solicit their feedback on the technology. The nonadoption, abandonment, scale-up, spread, and sustainability (NASSS) framework was used to guide data analysis. Interview transcripts were first coded deductively, then inductively. RESULTS Social care providers perceived the screening chatbot as offering limited value. This is mainly because many participants engaged in assistance activities and they noted addressing social needs is a multi-step process requiring follow-up that screening chatbots do not provide. In addition, they valued cultivating trust as many patients/clients have high social needs and lack trust in the health care system, and felt chatbots present new challenges for maintaining essential trust and care quality with clients/patients. Instead of a screening chatbot, we identified that technologies to reduce documentation burden could improve providers’ efficiency and potentially increase time spent with patients/clients. This is because social risks and needs documentation generates administrative burden for providers. We also found that technologies to improve referral accuracy and engagement could improve providers’ effectiveness, as study participants have overall limited access to technologies that typically support referral-related activities. CONCLUSIONS Social care providers in the safety net preferred AI-based applications for addressing documentation burden and social needs assistance rather than for social risk screening. Future strategies to develop AI tools for social care integration should align with social care providers’ professional values and focus on equity-centered care.

This study investigated the acceptability of using digital health, including telemedicine (e.g., video calls on a computer or smartphone), mHealth (a mobile app, text messaging, and social media messages), and artificial intelligence-enabled chatbots to communicate health information and facilitate access and uptake of sexual and reproductive healthcare among lesbian, bisexual, and queer (LBQ+) women of color (WOC) in the United States.

We conducted semi-structured interviews with 20 members who engaged in community-based effort to reverse overdoses using a smartphone-based app in an Eastern United States (U.S.) city. Drawing from feminist ethics of care, we identify how the caring practices of community members extend from administering a medical intervention to building trust and support between the care receivers and caregivers in the case of opioid overdose response. Contrary to the predominant patient-centered care paradigm, we emphasize community-centered care, which acknowledges the resistance of individuals and attends to reallocating caring responsibility and building relationships within the community. Our results highlight how trust intersects with social ecologies of care in the highly stigmatized context of opioid overdose and that trustful and less hierarchical relationships are critical sources of care for groups experiencing marginalization. We discuss applying harm reduction principles in designing health technologies for substance use disorders. We also discuss research and design opportunities for community-centered design for marginalized individuals and community caregivers.

Background: Social network approaches to testing allow individuals (indexes) to distribute tests to social networks (alters). This quasi-experimental study compared two social network-based testing strategies in promoting human immunodeficiency virus (HIV) testing among Chinese gay, bisexual, and other men who have sex with men (GBMSM). Methods: GBMSM aged ≥18 years were recruited from Guangzhou, China. From May to September 2019, indexes could distribute blood-based HIV/syphilis dual self-testing kits to people within their social network. Indexes recruited from October 2019 to January 2020 could send HIV testing cards to their social networks for free facility-based tests. Alters were encouraged to upload a photo verification of test results. Indexes and alters received incentives during both periods. Results: There were 245 participants who were assessed for eligibility and 208/245 (84.9%) were eligible. 106 and 102 indexes were recruited in the secondary distribution and testing card arms respectively. 154/208 (74.0%) completed follow up at 1 month. 92 indexes in the secondary distribution arm self-reported distributions to 179 unique alters, and 62 in the testing card arm to 26 unique alters. An average of 1.95 (standard deviation [s.d.] = 1.90) HIV/syphilis dual self-tests and 0.42 (s.d. = 0.78) HIV testing cards were distributed, generating a risk difference of 1.53 (95% confidence interval [CI] 1.09, 1.96). Indexes self-identifying as gay (P = 0.007) or having previously tested (P = 0.02) distributed more tests. Secondary distribution cost less per alter tested (USD120 vs USD9408). Conclusions: Secondary distribution engaged more GBMSM to distribute tests and reached more GBMSM to test compared to referral cards, suggesting advantage in facilitating testing uptake among Chinese GBMSM.

Secondary distribution of HIV self-test is promising to increase testing uptake while the facilitators and barriers of secondary distribution remain unclear. In-depth interviews were conducted with 22 MSM who had participated in secondary distribution of HIVST in southern China. Data were thematically analyzed to capture participants' motivations, procedures, and challenges when accessing and distributing self-tests. MSM in China are willing to distribute HIVST to members in their social network, but their decision-making is impacted by their ability to broach sexual health conversations, evaluations of the recipients, and perceived trustworthiness of the self-test. Our study suggested that several strategies, including creating a friendly environment for sexual health conversations and establishing nationwide policies related to quality assured self-tests and standardized self-testing protocols, may be helpful to advance this approach in China.

Background: Digital network–based methods may enhance peer distribution of HIV self-testing (HIVST) kits, but interventions that can optimize this approach are needed. We aimed to assess whether monetary incentives and peer referral could improve a secondary distribution program for HIVST among men who have sex with men (MSM) in China. Methods and findings: Between October 21, 2019 and September 14, 2020, a 3-arm randomized controlled, single-blinded trial was conducted online among 309 individuals (defined as index participants) who were assigned male at birth, aged 18 years or older, ever had male-to-male sex, willing to order HIVST kits online, and consented to take surveys online. We randomly assigned index participants into one of the 3 arms: (1) standard secondary distribution (control) group (n = 102); (2) secondary distribution with monetary incentives (SD-M) group (n = 103); and (3) secondary distribution with monetary incentives plus peer referral (SD-M-PR) group (n = 104). Index participants in 3 groups were encouraged to order HIVST kits online and distribute to members within their social networks. Members who received kits directly from index participants or through peer referral links from index MSM were defined as alters. Index participants in the 2 intervention groups could receive a fixed incentive ($3 USD) online for the verified test result uploaded to the digital platform by each unique alter. Index participants in the SD-M-PR group could additionally have a personalized peer referral link for alters to order kits online. Both index participants and alters needed to pay a refundable deposit ($15 USD) for ordering a kit. All index participants were assigned an online 3-month follow-up survey after ordering kits. The primary outcomes were the mean number of alters motivated by index participants in each arm and the mean number of newly tested alters motivated by index participants in each arm. These were assessed using zero-inflated negative binomial regression to determine the group differences in the mean number of alters and the mean number of newly tested alters motivated by index participants. Analyses were performed on an intention-to-treat basis. We also conducted an economic evaluation using microcosting from a health provider perspective with a 3-month time horizon. The mean number of unique tested alters motivated by index participants was 0.57 ± 0.96 (mean ± standard deviation [SD]) in the control group, compared with 0.98 ± 1.38 in the SD-M group (mean difference [MD] = 0.41),and 1.78 ± 2.05 in the SD-M-PR group (MD = 1.21). The mean number of newly tested alters motivated by index participants was 0.16 ± 0.39 (mean ± SD) in the control group, compared with 0.41 ± 0.73 in the SD-M group (MD = 0.25) and 0.57 ± 0.91 in the SD-M-PR group (MD = 0.41), respectively. Results indicated that index participants in intervention arms were more likely to motivate unique tested alters (control versus SD-M: incidence rate ratio [IRR = 2.98, 95% CI = 1.82 to 4.89, p-value < 0.001; control versus SD-M-PR: IRR = 3.26, 95% CI = 2.29 to 4.63, p-value < 0.001) and newly tested alters (control versus SD-M: IRR = 4.22, 95% CI = 1.93 to 9.23, p-value < 0.001; control versus SD-M-PR: IRR = 3.49, 95% CI = 1.92 to 6.37, p-value < 0.001) to conduct HIVST. The proportion of newly tested testers among alters was 28% in the control group, 42% in the SD-M group, and 32% in the SD-M-PR group. A total of 18 testers (3 index participants and 15 alters) tested as HIV positive, and the HIV reactive rates for alters were similar between the 3 groups. The total costs were $19,485.97 for 794 testers, including 450 index participants and 344 alter testers. Overall, the average cost per tester was $24.54, and the average cost per alter tester was $56.65. Monetary incentives alone (SD-M group) were more cost-effective than monetary incentives with peer referral (SD-M-PR group) on average in terms of alters tested and newly tested alters, despite SD-M-PR having larger effects. Compared to the control group, the cost for one more alter tester in the SD-M group was $14.90 and $16.61 in the SD-M-PR group. For newly tested alters, the cost of one more alter in the SD-M group was $24.65 and $49.07 in the SD-M-PR group. No study-related adverse events were reported during the study. Limitations include the digital network approach might neglect individuals who lack internet access. Conclusions: Monetary incentives alone and the combined intervention of monetary incentives and peer referral can promote the secondary distribution of HIVST among MSM. Monetary incentives can also expand HIV testing by encouraging first-time testing through secondary distribution by MSM. This social network–based digital approach can be expanded to other public health research, especially in the era of the Coronavirus Disease 2019 (COVID-19). Trial registration: Chinese Clinical Trial Registry (ChiCTR) ChiCTR1900025433

Background: The COVID-19 outbreak was designated a global pandemic on March 11, 2020. The relationship between vaping and contracting COVID-19 is unclear, and information on the internet is conflicting. There is some scientific evidence that vaping cannabidiol (CBD), an active ingredient in cannabis that is obtained from the hemp plant, or other substances is associated with more severe manifestations of COVID-19. However, there is also inaccurate information that vaping can aid COVID-19 treatment, as well as expert opinion that CBD, possibly administered through vaping, can mitigate COVID-19 symptoms. Thus, it is necessary to study the spread of inaccurate information to better understand how to promote scientific knowledge and curb inaccurate information, which is critical to the health of vapers. Inaccurate information about vaping and COVID-19 may affect COVID-19 treatment outcomes. Objective: Using structural topic modeling, we aimed to map temporal trends in the web-based vaping narrative (a large data set comprising web-based vaping chatter from several sources) to indicate how the narrative changed from before to during the COVID-19 pandemic. Methods: We obtained data using a textual query that scanned a data pool of approximately 200,000 different domains (4,027,172 documents and 361,100,284 words) such as public internet forums, blogs, and social media, from August 1, 2019, to April 21, 2020. We then used structural topic modeling to understand changes in word prevalence and semantic structures within topics around vaping before and after December 31, 2019, when COVID-19 was reported to the World Health Organization. Results: Broadly, the web-based vaping narrative can be organized into the following groups or archetypes: harms from vaping; Vaping Regulation; Vaping as Harm Reduction or Treatment; and Vaping Lifestyle. Three archetypes were observed prior to the emergence of COVID-19; however, four archetypes were identified post–COVID-19 (Vaping as Harm Reduction or Treatment was the additional archetype). A topic related to CBD product preference emerged after COVID-19 was first reported, which may be related to the use of CBD by vapers as a COVID-19 treatment. Conclusions: Our main finding is the emergence of a vape-administered CBD treatment narrative around COVID-19 when comparing the web-based vaping narratives before and during the COVID-19 pandemic. These results are key to understanding how vapers respond to inaccurate information about COVID-19, optimizing treatment of vapers who contract COVID-19, and possibly minimizing instances of inaccurate information. The findings have implications for the management of COVID-19 among vapers and the monitoring of web-based content pertinent to tobacco to develop targeted interventions to manage COVID-19 among vapers.

BACKGROUND The COVID-19 outbreak was designated a global pandemic on March 11, 2020. The relationship between vaping and contracting COVID-19 is unclear, and information on the internet is conflicting. There is some scientific evidence that vaping cannabidiol (CBD), an active ingredient in cannabis that is obtained from the hemp plant, or other substances is associated with more severe manifestations of COVID-19. However, there is also inaccurate information that vaping can aid COVID-19 treatment, as well as expert opinion that CBD, possibly administered through vaping, can mitigate COVID-19 symptoms. Thus, it is necessary to study the spread of inaccurate information to better understand how to promote scientific knowledge and curb inaccurate information, which is critical to the health of vapers. Inaccurate information about vaping and COVID-19 may affect COVID-19 treatment outcomes. OBJECTIVE Using structural topic modeling, we aimed to map temporal trends in the web-based vaping narrative (a large data set comprising web-based vaping chatter from several sources) to indicate how the narrative changed from before to during the COVID-19 pandemic. METHODS We obtained data using a textual query that scanned a data pool of approximately 200,000 different domains (4,027,172 documents and 361,100,284 words) such as public internet forums, blogs, and social media, from August 1, 2019, to April 21, 2020. We then used structural topic modeling to understand changes in word prevalence and semantic structures within topics around vaping before and after December 31, 2019, when COVID-19 was reported to the World Health Organization. RESULTS Broadly, the web-based vaping narrative can be organized into the following groups or archetypes: harms from vaping; Vaping Regulation; Vaping as Harm Reduction or Treatment; and Vaping Lifestyle. Three archetypes were observed prior to the emergence of COVID-19; however, four archetypes were identified post–COVID-19 (Vaping as Harm Reduction or Treatment was the additional archetype). A topic related to CBD product preference emerged after COVID-19 was first reported, which may be related to the use of CBD by vapers as a COVID-19 treatment. CONCLUSIONS Our main finding is the emergence of a vape-administered CBD treatment narrative around COVID-19 when comparing the web-based vaping narratives before and during the COVID-19 pandemic. These results are key to understanding how vapers respond to inaccurate information about COVID-19, optimizing treatment of vapers who contract COVID-19, and possibly minimizing instances of inaccurate information. The findings have implications for the management of COVID-19 among vapers and the monitoring of web-based content pertinent to tobacco to develop targeted interventions to manage COVID-19 among vapers.