Jul 2026· Journal of Medical Internet Research· Vol 28· 0 citations· 73 references
Medicine
TL;DR
By extending a sociocognitive framework through empirical grounding in HIV clinical practice, this study offers design insights for developing AI-powered CDSS that are trustworthy, context-aware, and capable of supporting actionable decision-making in HIV care settings and beyond.
Abstract
Abstract Background AI-powered clinical decision support systems (CDSS) have shown promise in improving prediction, monitoring, and treatment optimization across clinical domains, including HIV care. However, translating AI outputs derived from electronic health records into clinically meaningful, trustworthy, and actionable decision support remains challenging, underscoring the need for more human-centered and socioecologically grounded CDSS design. Objective This study aimed to explore how we can effectively translate the outputs of machine learning models based on HIV electronic health records into a real AI-powered CDSS for HIV care. Using the human-in-the-loop method, we engaged a set of stakeholders, including HIV physicians, nurse practitioners, infectious disease pharmacists, social workers, and case managers. Stakeholders interacted with an AI-powered CDSS prototype to identify barriers and challenges to adoption, as well as to inform a more holistic and context-aware AI-powered CDSS design. Methods We conducted a field study at Prisma Health in South Carolina that included pre- and postsurveys, interactive usability testing sessions, think-alouds, and in-depth interviews with 16 clinicians providing HIV care between March and September 2025. We analyzed survey responses using descriptive statistics, and then transcribed and analyzed think-aloud and interview data using an etic and emic approach. Results Clinicians identified multiple challenges and design considerations for AI-powered HIV CDSS, demonstrating that clinician-AI interaction is inherently sociotechnical and embedded across multiple socioecological levels. While clinicians relied on familiar clinical indicators as cognitive anchors for interpreting AI predictions, they emphasized that social determinants of health were central to their own risk assessment and clinical decision-making. Additionally, clinicians' trust in AI is conditional and develops over time, with explainability and actionability emerging as critical factors for translating predictions into meaningful clinical interventions. Conclusions Findings highlight the need to move beyond technically accurate predictions toward AI-powered CDSS designs that align with clinicians’ cognitive practices and socioecological realities of HIV care. By extending a sociocognitive framework through empirical grounding in HIV clinical practice, this study offers design insights for developing AI-powered CDSS that are trustworthy, context-aware, and capable of supporting actionable decision-making in HIV care settings and beyond.
AI-integrated Clinical Decision Support Systems (AI-CDSS) consolidate patient data with medical expertise to aid healthcare practitioners in deciding their approach. They promote guideline compliance, enhance treatment safety, and lead to more accurate diagnoses. This narrative review examines the potential role and emerging applications of AI-CDSS in strengthening primary healthcare systems by reviewing existing literature and evidence to evaluate current implementations, benefits, and limitations of AI-CDSS in the local healthcare context, with a particular focus on gaps to be addressed in the Pakistani healthcare landscape. Currently, although Pakistan is only utilizing AI-CDSS at a small scale and limited to tertiary care settings, these efforts encompass multiple domains such as drug safety, diagnostic help, and treatment planning. Findings have shown positive results and advocate for implementing AI-CDSS systems on a larger scale, which can reduce the burden on the healthcare system in Pakistan. Unfortunately, multiple concerns, such as user adoption and warning fatigue, must be dealt with before implementing a large-scale application of this system. Our research proposes a framework for the sustainable integration of AI-CDSS into Pakistan's healthcare system, with a focus on scaling solutions to primary care environments such as Basic Health Units and rural health centers. Overall, AI-CDSS holds significant potential to enhance accessibility, quality, and capacity of primary healthcare in Pakistan, provided that existing systemic and operational barriers are effectively addressed.
Huma Ahsan, F. Qazi, Amna Amir Jalal et al.· Liaquat National Journal of...· 0 citations
Background: Artificial Intelligence (AI) has emerged as one of the most transformative technologies in modern healthcare, significantly influencing clinical decision-making processes. AI-driven systems, including machine learning algorithms, deep learning models, natural language processing, and clinical decision support systems, have demonstrated the ability to analyze vast amounts of healthcare data, identify patterns, predict outcomes, and support evidence-based clinical decisions. As healthcare systems face increasing patient complexity, workforce shortages, and demands for improved quality of care, AI offers opportunities to enhance diagnostic accuracy, treatment planning, risk prediction, and healthcare efficiency. However, concerns regarding algorithm transparency, ethical accountability, data privacy, bias, and professional acceptance continue to challenge its widespread adoption.
Objectives
To systematically review the evidence regarding the role of artificial intelligence in clinical decision-making across healthcare settings.
To identify the benefits and effectiveness of AI-supported clinical decision-making systems in improving patient outcomes.
To examine challenges, ethical considerations, and barriers associated with AI implementation in clinical decision support.
To evaluate implications for healthcare professionals, particularly nurses and physicians, in AI-assisted clinical practice.
Methods: This systematic review was conducted according to Joanna Briggs Institute (JBI) methodology and reported following PRISMA 2020 guidelines. Electronic databases including PubMed, Scopus, Web of Science, CINAHL, ScienceDirect, and Google Scholar were searched for peer-reviewed studies published between 2019 and 2025. The review question was structured using the PICOT framework. Methodological quality was assessed using appropriate JBI Critical Appraisal Tools. Due to heterogeneity among study designs and outcomes, findings were synthesized narratively.
Results: Thirty-five studies met the inclusion criteria. Evidence demonstrated that AI-assisted clinical decision-making significantly improved diagnostic accuracy, risk stratification, treatment planning, predictive analytics, medication management, and workflow efficiency. AI systems were particularly effective in radiology, oncology, cardiology, intensive care, and chronic disease management. However, challenges related to explainability, algorithmic bias, legal liability, data privacy, workforce adaptation, and ethical governance were consistently reported.
Conclusion: Artificial intelligence has substantial potential to strengthen clinical decision-making and improve healthcare outcomes. Successful integration requires robust regulatory frameworks, transparent algorithms, ethical governance, clinician training, and preservation of patient-centered care principles. AI should function as a supportive tool that enhances rather than replaces human clinical judgment.
Arun James, Dr. Jomon Thomas, Dr. Deepika Verma et al.· PAIN, JOINTS, SPINE· 0 citations
Background Preventable adverse drug events (ADEs) remain a major source of hospital morbidity, mortality, and healthcare costs worldwide. Clinical decision support systems (CDSS) integrated into electronic health records (EHRs) were developed to reduce unsafe prescribing, yet evidence of their real-world effectiveness remains mixed. The emergence of artificial intelligence (AI) and machine learning (ML) offers new opportunities to enhance medication safety but also introduces risks such as algorithmic bias, technology-induced error, and reduced clinician vigilance. Objectives This narrative review critically examines: (1) evidence for the effectiveness of CDSS in reducing preventable ADEs; (2) human factors influencing interactions between clinicians and AI-enabled safety tools; and (3) conceptual, methodological, and governance challenges affecting the safe implementation of digital health technologies. Methods A structured narrative review was conducted using the SANRA framework and reported in accordance with PRISMA-ScR guidance where applicable. Searches of PubMed/MEDLINE, CINAHL, Embase, Scopus, and IEEE Xplore covered literature published between January 2015 and March 2024, supplemented by seminal earlier studies. Following eligibility screening, 75 studies were included in a thematic synthesis and quality appraisal using established risk-of-bias tools. Results Five themes emerged: the transition from passive to adaptive decision support; AI's dual role as both a safety enhancer and a source of new risks; persistent alert fatigue; the often-overlooked contribution of nursing vigilance; and gaps in equity, governance, and technology-induced error research. From these findings, we propose the Clinical Safety Intelligence Loop (CSIL), a conceptual framework that positions AI within a sociotechnical system while embedding equity, governance, and continuous feedback as core components. Conclusion Achieving medication safety improvements requires moving beyond technology-focused solutions toward systems-level approaches integrating AI, clinician cognition, organizational culture, and governance. The CSIL offers a useful framework for guiding this transformation, although further empirical validation is needed.
M. Alruwaili, U. Paul-Chima, Ugwu Chinyere Nneoma· Frontiers in Digital Health· 0 citations
BACKGROUND
The advent of health technologies associated with artificial intelligence (AI) is deemed a transformative shift in the delivery of medical care within our lifetime. Nevertheless, there is a communication gap between intelligent models and clinical experts. Transitioning towards Human-Centered AI can serve as a means to bridge this gap.
METHODS
This study introduces a human-centered rule extraction model based on the Non-Dominated Sorting Genetic Algorithm (NSGA-II), designed to enhance the interpretability and clinical utility of diagnostic tools in healthcare. This model autonomously generates diagnostic rules, adjusts threshold values for variables, and involves clinical experts in evaluating the rules, thereby ensuring the relevance and applicability of extracted rules in real-world settings.
RESULTS
Experiments on the WBC, WDBC, and Pima datasets showed that the proposed model outperformed state-of-the-art rule extraction methods in the literature in terms of predictive value accuracy (PVA) and support. On the subset of data covered by the extracted rules, it achieved accuracy comparable to traditional black-box methods without sacrificing interpretability. The extracted rules were clinically evaluated by 13 domain physicians, with all approved rules achieving a content validity index (CVI) of at least 0.85. Additionally, the model provides multiple high-performance alternative diagnostic rules per class, giving clinicians practical flexibility.
CONCLUSIONS
Our approach emphasizes the importance of multidisciplinary collaboration between AI specialists and healthcare professionals, aiming to build trust in AI-driven diagnostic systems through transparency and clinical validation.
F. Ahouz, M. Sohrabi, A. Golabpour· BMC Medical Informatics and...· 0 citations
This perspective aims to evaluate the integration of AI and machine learning into pediatric emergency departments (PEDs) across three core domains: clinical decision support, stakeholder engagement, and medical education.
Lorenzo Gasparini, Nicola Gobbi, D. Zama et al.· Pediatric Reports· 0 citations
This article shows how the Technology Acceptance Model and Technology Trust are combined in the Unified Theory of Acceptance and Use of Technology (UTAUT) and how the Perceived Empathy Model (PEM) and Perceived Clarity of Accountability Model (PCA) are extensions of these models.