Informing the Development of a Chatbot for Pharmacist Needs in HIV Care
Bibliographic record
Abstract
Background: Chatbots -"conversational agents" -are computer programs designed to mimic conversations with human users.Recent advances in artificial intelligence (AI) algorithms and their application in digital healthcare have made AI-based chatbots increasingly valuable.Chatbots have become progressively more popular in various industries, including healthcare, as they offer numerous benefits to patients and healthcare providers alike.Chatbots can help healthcare providers with tasks such as retrieving health-and medication-related information, keeping medical records, scheduling appointments, and triaging patient concerns.For instance, pharmacists need up-to-date HIV-and antiretroviral therapy (ART)-related information when counseling people with HIV (PWH).However, with thousands of academic articles on prescription and nonprescription medications published each year, they find it challenging to stay current on every aspect of HIV care.MARVIN, an artificial intelligence-based chatbot developed to assist PWH in taking ART, initially developed in 2020, could be adapted to respond to this need.Aim: The aim of this thesis was to generate evidence on the needs of pharmacists to inform the adaptation of the MARVIN chatbot for improving services provided to PWH in Québec, Canada.Methods: This thesis consists of two manuscripts.The first manuscript provides a comprehensive review of the roles and benefits chatbots offer in healthcare and the populations they serve.The second manuscript entails a preliminary report that summarizes the design, conduct and results of an online survey to assess the HIV care needs of pharmacists in Québec, Canada, deployed between December 2022 and May 2023.For the first manuscript, a rapid review was performed, and data extracted on chatbot roles, users, and benefits were synthesized using content analysis.For the second manuscript (preliminary report), an online needs assessment questionnaire based on the Knowledge, Attitudes and Practices (KAP) model was administered to pharmacists, and response distributions descriptively analyzed to elicit central tendencies and variations with regard to knowledge, attitudes and practices.Results: The findings from the rapid review and the needs assessment questionnaire provide valuable insights that inform the recommendations and implications for the adaptation and development of MARVIN-Pharma in HIV care.Key conclusions based on the rapid review suggest that adapting an AI chatbot to meet the specific needs of pharmacists in HIV care can significantly enhance support for PWH.By providing remote consultation and treatment advice, facilitating medication management, offering educational 6 resources, and streamlining administrative tasks, the chatbot can empower pharmacists to deliver more effective and personalized care, ultimately improving health outcomes for PWH.Key conclusions based on the pharmacist needs assessment questionnaire for MARVIN-Pharma include the necessity for MARVIN-Pharma to address pharmacists' knowledge gaps in HIV treatment; provide targeted educational and training materials; integrate patient information needs; support various HIV care services; and address barriers facing pharmacists in HIV care.Userfriendly design, continuous support, and gathering user feedback are essential for enhancing the usability and effectiveness of MARVIN-Pharma, as well as connecting it with a constantly updated, reliable HIV information resource.Conclusion: These findings and recommendations provide insights for the content configuration and development of MARVIN-Pharma, guiding its design and development to meet the needs of pharmacists in HIV care as an innovative, practical tool.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.006 | 0.023 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.001 | 0.003 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.026 | 0.004 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".