Canidate Shantrel S, Gracy Hannah R, McIntosh Sean, Liu Yiyang, Fisk-Hoffman Rebecca, Rich Shannon, Mavian Carla, Cook Robert L, Prosperi Mattia, Salemi Marco
Department of Epidemiology, College of Public Health and Health Professions and College of Medicine, University of Florida, Gainesville, FL.
Department of Human, Development and Family, Science, College of Human Sciences, Auburn University, Auburn, AL.
Proc (IEEE Int Conf Healthc Inform). 2025 Jun;2025:588-597. doi: 10.1109/ichi64645.2025.00072. Epub 2025 Jul 22.
Developing and validating novel molecular HIV surveillance (MHS) tools capable of predicting the growth and trajectory of localized outbreaks driven by specific transmission clusters is key to the . This study explored stakeholders' perspectives on HIV prevention and treatment regarding a developing deep-learning framework, and its ability to predict HIV transmission cluster trajectories and inform decision-making on HIV prevention and treatment scale-up approaches in Florida. We conducted five virtual focus group discussions with 16 clinical health professionals and state and local public health personnel. Focus group discussions were audio-recorded, transcribed using Zoom transcription, and manually coded using a reflexive thematic analysis approach. Overall, participants reported a high level of acceptability for using MHS tools. However, when exploring their perspectives on using the DeepDynaForecast tool,participants discussed their acceptance criteria, including key features that the DeepDynaForecast tool should have and the need to determine the data types the tool should generate to meet their needs and be deemed acceptable. Before implementation, participants felt the tool should undergo extensive software testing, followed by end-users receiving comprehensive training and the developers determining how the DeepDynaForecast tool could integrate with existing MHS tools. Likewise, participants discussed using the data generated by DeepDynaForecast to increase HIV prevention, education, outreach activities, and mobilization efforts in communities where the most HIV diagnoses occur, as well as increase behavioral change communication efforts. Participants also expressed concerns about HIV-related stigma, a potentially dangerous unintended consequence of using existing and new MHS tools. Current MHS tools have helped inform and evaluate HIV prevention and treatment efforts in the US. A novel MHS tool such as DeepDynaForecast may be critical to achieving the Ending the HIV Epidemic (EHE) goals and curbing the spread of HIV in Florida and in the US.
开发并验证能够预测由特定传播集群驱动的局部疫情的增长和轨迹的新型分子HIV监测(MHS)工具是[此处原文缺失关键信息]的关键。本研究探讨了利益相关者对HIV预防和治疗的看法,涉及一个正在开发的深度学习框架,以及其预测HIV传播集群轨迹并为佛罗里达州HIV预防和治疗扩大规模方法的决策提供信息的能力。我们与16名临床卫生专业人员以及州和地方公共卫生人员进行了五次虚拟焦点小组讨论。焦点小组讨论进行了录音,使用Zoom转录进行转录,并采用反思性主题分析方法进行手动编码。总体而言,参与者报告对使用MHS工具的接受程度很高。然而,在探讨他们对使用DeepDynaForecast工具的看法时,参与者讨论了他们的接受标准,包括DeepDynaForecast工具应具备的关键特征,以及确定该工具应生成何种数据类型以满足他们的需求并被认为可接受的必要性。在实施之前,参与者认为该工具应进行广泛的软件测试,随后终端用户接受全面培训,并且开发者确定DeepDynaForecast工具如何与现有的MHS工具集成。同样,参与者讨论了使用DeepDynaForecast生成的数据来加强在HIV诊断最多的社区的HIV预防、教育、外展活动和动员工作,以及加强行为改变沟通工作。参与者还对与HIV相关的耻辱感表示担忧,这是使用现有和新的MHS工具可能产生的潜在危险的意外后果。当前MHS工具已有助于为美国的HIV预防和治疗工作提供信息并进行评估。像DeepDynaForecast这样的新型MHS工具对于实现“终结HIV流行”(EHE)目标以及遏制HIV在佛罗里达州和美国的传播可能至关重要。