{"id":30396,"date":"2025-06-19T18:19:08","date_gmt":"2025-06-19T18:19:08","guid":{"rendered":""},"modified":"-0001-11-30T00:00:00","modified_gmt":"-0001-11-30T00:00:00","slug":"evaluating-the-need-for-new-governance-structures-in-healthcare-to-oversee-ai-implementation-and-adaptive-clinical-decision-support-3748360","status":"publish","type":"post","link":"https:\/\/www.simbo.ai\/blog\/evaluating-the-need-for-new-governance-structures-in-healthcare-to-oversee-ai-implementation-and-adaptive-clinical-decision-support-3748360\/","title":{"rendered":"Evaluating the Need for New Governance Structures in Healthcare to Oversee AI Implementation and Adaptive Clinical Decision Support"},"content":{"rendered":"<p>Clinical decision support systems have traditionally provided clinicians with evidence-based recommendations or alerts to improve patient outcomes. Classical CDS tools are static, meaning that given the same input, they deliver the same output each time. Adaptive CDS represents a change from this model. These systems use machine learning algorithms to adjust and improve their recommendations based on new clinical evidence, health data, and updated interpretative methods.<\/p>\n<p>Because Adaptive CDS can change their decision logic over time, they offer a more personalized approach to patient care. This responsiveness to current data and research may improve clinical workflows and patient safety. However, it also introduces risks related to unexpected errors, bias, and transparency issues.<\/p>\n<p>The American Medical Informatics Association (AMIA) recently noted a key difference between two types of Adaptive CDS tools: <strong>Marketed Adaptive CDS<\/strong>, which are commercial products regulated by the Food and Drug Administration (FDA) under the 21st Century Cures Act, and <strong>Self-Developed Adaptive CDS<\/strong>, typically created and used internally by healthcare systems without formal regulatory oversight.<\/p>\n<p>The lack of federal regulation for Self-Developed Adaptive CDS leaves patients vulnerable to risks such as algorithmic bias or incorrect recommendations that may not be caught until harmful effects occur. Dr. Joseph Kannry, Chair of AMIA\u2019s Public Policy Committee, has stated that current policies and oversight for Adaptive CDS are inadequate. He emphasized the urgent need for new governance approaches to protect patient safety and maintain accountability.<\/p>\n<h2>The Case for New Governance Models in Healthcare Institutions<\/h2>\n<p>Healthcare organizations in the United States face growing complexity as they add AI-based tools to clinical and operational processes. The rapid increase in health data, along with advances in cloud computing and machine learning, requires a fresh look at how AI tools are managed and supervised.<\/p>\n<p>AMIA suggests creating dedicated governance bodies within healthcare institutions to oversee the development, use, and monitoring of Adaptive CDS. These groups would handle validation, maintenance, and risk assessment tailored to the distinct challenges posed by AI-driven tools. Their recommendations include two core components:<\/p>\n<ul>\n<li><strong>Institution-Level AI Governance Committees<\/strong>: responsible for local oversight, model validation, assessing transparency, and ongoing risk reviews.<\/li>\n<li><strong>Cross-Institutional Oversight Structures<\/strong>: collaborative networks or registries to track performance, share practices, and ensure consistency of standards across healthcare organizations.<\/li>\n<\/ul>\n<p>Additionally, AMIA proposes <strong>Adaptive CDS Centers of Excellence<\/strong> to promote research, testing, and dissemination of safe machine learning-based clinical decision support. These centers would help develop regulatory frameworks, conduct safety evaluations, and provide education on AI implementation.<\/p>\n<p>Patricia C. Dykes, AMIA Board Chair, has highlighted the role of informatics professionals in guiding these efforts. Without appropriate oversight, the expansion of AI risks unintended effects that could hurt patient safety and reduce confidence in these technologies.<\/p>\n<h2>Transparency and Communication Standards: Essential Pillars for AI Accountability<\/h2>\n<p>A major challenge in supervising Adaptive CDS is the limited transparency about how these systems are trained and how they make decisions. AMIA calls for strict standards requiring healthcare organizations to disclose important details about AI tools, such as:<\/p>\n<ul>\n<li>Methods used for algorithm training, including datasets, cohort selection, and possible sources of bias.<\/li>\n<li>The origin and meaning of input data, including how the data were collected and prepared before being used by the model.<\/li>\n<li>Communication standards similar to medication labeling mandated by the FDA, offering clear guidance on intended use, contraindications, adverse effects, and considerations for special patient groups.<\/li>\n<\/ul>\n<p>Without such transparency, it becomes difficult to verify the safety and effectiveness of these tools. That situation limits accountability and can impact patient care negatively.<\/p>\n<p>For medical practice administrators and IT managers, this means a clear need to require AI vendors and internal developers to provide comprehensive documentation explaining the AI\u2019s behavior, its limits, and conditions for use. Integrating these transparency measures with current electronic health record (EHR) systems and workflows will help prevent misuse or misinterpretation of AI recommendations.<\/p>\n<p><!--smbadstart--><\/p>\n<div class=\"ad-widget case-study-ad\" smbdta=\"smbadid:sc_21;nm:UneQU319I;score:0.89;kw:data-entry_0.98_insurance-extraction_0.94_ehr_0.89_sm-process_0.78_form-automation_0.72;\">\n<h4>AI Call Assistant Skips Data Entry<\/h4>\n<p>SimboConnect extracts insurance details from SMS images &#8211; auto-fills EHR fields.<\/p>\n<div class=\"client-info\">\n    <!--<span><\/span>--><br \/>\n    <a href=\"https:\/\/simbo.ai\/schedule-connect\">Speak with an Expert \u2192<\/a>\n  <\/div>\n<\/div>\n<p><!--smbadend--><\/p>\n<h2>Implications for Medical Practice Administrators, Owners, and IT Managers<\/h2>\n<p>In the U.S. healthcare system, administrators and owners are charged with maintaining operational efficiency while ensuring patient safety and compliance. Adopting AI and Adaptive CDS in clinical workflows requires them to understand complex regulatory frameworks and predict emerging governance needs.<\/p>\n<h3>Policy Gaps and Liability Considerations<\/h3>\n<p>At present, regulatory oversight for Self-Developed Adaptive CDS is incomplete, placing healthcare institutions in a difficult position. These organizations must independently confirm AI tools are safe and also consider legal risks from algorithm errors or biased outcomes. Owners and administrators should prepare governance structures and protocols to reduce these risks by:<\/p>\n<ul>\n<li>Setting up multidisciplinary AI oversight committees that include clinicians, IT specialists, legal experts, and quality managers.<\/li>\n<li>Requiring thorough risk assessments and bias evaluations before deploying AI tools.<\/li>\n<li>Conducting regular monitoring and post-deployment reviews to detect new issues or performance declines.<\/li>\n<\/ul>\n<h3>Data Management and Security<\/h3>\n<p>Adaptive CDS depends on large amounts of confidential patient data, which are updated frequently. Protecting patient privacy, ensuring HIPAA compliance, and securing data storage are critical. IT managers must use secure cloud environments with encryption, access controls, and audit logs to maintain data confidentiality and integrity.<\/p>\n<h3>Staff Training and Workflow Integration<\/h3>\n<p>Using Adaptive CDS involves changing established clinical workflows. Practice administrators should create training programs to inform providers about AI\u2019s capabilities and limits. Emphasizing clinician oversight and joint decision-making remains important. IT teams have to ensure system interoperability so AI recommendations fit seamlessly into existing electronic medical records and telehealth platforms.<\/p>\n<p><!--smbadstart--><\/p>\n<div class=\"ad-widget regular-ad\" smbdta=\"smbadid:sc_17;nm:AJerNW453;score:2.88;kw:hipaa_0.99_compliance_0.96_encryption_0.93_data-security_0.85_call-privacy_0.77;\">\n<h4>HIPAA-Compliant Voice AI Agents<\/h4>\n<p>SimboConnect AI Phone Agent encrypts every call end-to-end &#8211; zero compliance worries.<\/p>\n<p>  <a href=\"https:\/\/simbo.ai\/schedule-connect\" class=\"cta-button\">Claim Your Free Demo \u2192<\/a>\n<\/div>\n<p><!--smbadend--><\/p>\n<h2>AI and Workflow Automation: A Growing Consideration in Healthcare Operations<\/h2>\n<p>Besides clinical decision support, AI is playing a larger role in automating administrative tasks. Solutions like those from Simbo AI focus on front-office phone automation. These tools offer practical improvements to operational efficiency in medical settings.<\/p>\n<h3>Reducing Administrative Burdens with AI-Powered Phone Systems<\/h3>\n<p>Front-office duties such as patient scheduling, appointment reminders, and triage calls often require significant time and carry risks of human error or delays. AI-powered phone automation handles these repetitive tasks by:<\/p>\n<ul>\n<li>Answering patient calls around the clock without wait times.<\/li>\n<li>Gathering structured information from callers to route requests appropriately.<\/li>\n<li>Confirming appointments and managing cancellations or rescheduling in real-time.<\/li>\n<li>Responding to routine patient questions, allowing staff to focus on more complex or urgent issues.<\/li>\n<\/ul>\n<p>For practice administrators, investing in such AI answering services may reduce missed calls, improve patient satisfaction, and lower staffing costs.<\/p>\n<h3>Integration with Clinical AI Systems<\/h3>\n<p>When workflow automation links to clinical AI tools like Adaptive CDS, medical practices can improve coordination between front-office and clinical care. For example:<\/p>\n<ul>\n<li>Automated call systems can identify urgent symptoms reported by patients and alert clinicians using Adaptive CDS to prioritize care.<\/li>\n<li>Information gathered during calls can contribute to AI-based risk models for early intervention.<\/li>\n<li>Scheduling aligned with clinical recommendations can reduce no-shows and optimize resource use.<\/li>\n<\/ul>\n<p>This integration requires healthcare IT leaders to develop technical standards and interoperability protocols, supported by governance frameworks that protect data accuracy, privacy, and compliance.<\/p>\n<p><!--smbadstart--><\/p>\n<div class=\"ad-widget checklist-ad\" smbdta=\"smbadid:sc_14;nm:AOPWner28;score:0.99;kw:reminder_0.1_appointment-reminder_0.89_patient-notification_0.73;\">\n<div class=\"check-icon\">\u2713<\/div>\n<div>\n<h4>AI Call Assistant Reduces No-Shows<\/h4>\n<p>SimboConnect sends smart reminders via call\/SMS &#8211; patients never forget appointments.<\/p>\n<p>    <a href=\"https:\/\/simbo.ai\/schedule-connect\" class=\"download-btn\"> Let\u2019s Chat <\/a>\n  <\/div>\n<\/div>\n<p><!--smbadend--><\/p>\n<h2>Addressing Algorithmic Bias and Patient Safety Concerns<\/h2>\n<p>Algorithmic bias is a known risk in machine learning models trained on healthcare data that may reflect social, racial, or economic disparities. AMIA warns that without proper oversight, Adaptive CDS tools could reinforce or increase these inequities, potentially harming patient safety.<\/p>\n<p>Healthcare administrators and owners in the U.S. should:<\/p>\n<ul>\n<li>Make sure training datasets for AI tools represent the diversity of their patient populations.<\/li>\n<li>Conduct regular audits to identify biased outcomes or unequal recommendations affecting vulnerable groups.<\/li>\n<li>Involve diverse stakeholders, including patients and community representatives, in AI policy development and supervision.<\/li>\n<\/ul>\n<p>These steps not only promote fairness in clinical care but also support ethical practices and risk management.<\/p>\n<h2>Preparing for the Future of AI Governance in U.S. Healthcare<\/h2>\n<p>Integrating AI, especially Adaptive CDS, into medical practice requires a shift in governance from traditional regulatory models to more flexible, institution-led oversight, combined with national cooperative efforts.<\/p>\n<p>Medical practice managers and IT teams should prepare by:<\/p>\n<ul>\n<li>Creating internal policies for AI use that include validation processes, transparency requirements, and risk reduction measures.<\/li>\n<li>Engaging with national initiatives such as AMIA\u2019s Adaptive CDS Centers of Excellence to keep abreast of best practices.<\/li>\n<li>Partnering with technology providers, including AI vendors like Simbo AI, to ensure their systems meet governance and transparency standards.<\/li>\n<li>Supporting legislative and regulatory changes that define the federal role in supervising Self-Developed Adaptive CDS.<\/li>\n<\/ul>\n<p>Only through a combination of strong local management, cooperation between organizations, and clear communication can U.S. healthcare adopt AI safely without compromising patient care.<\/p>\n<h2>Final Thoughts<\/h2>\n<p>The growing use of AI in healthcare marks a turning point requiring action by medical practice administrators, owners, and IT managers across the country. The lack of comprehensive federal oversight for Adaptive Clinical Decision Support systems means healthcare organizations must take the lead in governance. Establishing transparency guidelines, communication standards, multidisciplinary oversight groups, and connections with national Adaptive CDS Centers of Excellence is critical for managing risks and protecting patients.<\/p>\n<p>At the same time, AI-driven front-office automation shows the wider operational benefits AI can bring in reducing administrative workloads and improving patient interactions. Combining these tools with clinical decision support oversight will be key to successful and safe AI use.<\/p>\n<p>Healthcare leaders should regard AI both as a clinical aid and a complex technology that needs careful monitoring, clear communication, and ongoing evaluation. The time to develop these governance structures is now to ensure AI improves the quality and safety of healthcare delivery.<\/p>\n<section class=\"faq-section\">\n<h2 class=\"section-title\">Frequently Asked Questions<\/h2>\n<div class=\"faq-container\">\n<details>\n<summary>What is the focus of the AMIA position paper?<\/summary>\n<div class=\"faq-content\">\n<p>The AMIA position paper focuses on the policy framework for adaptive clinical decision support (CDS) systems that utilize artificial intelligence (AI) applications in healthcare.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What is &#8216;Adaptive CDS&#8217;?<\/summary>\n<div class=\"faq-content\">\n<p>Adaptive CDS refers to clinical decision support systems that can learn and change their performance over time based on new clinical evidence and data, enabling personalized decision support.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What is the difference between Marketed ACDS and Self-Developed ACDS?<\/summary>\n<div class=\"faq-content\">\n<p>Marketed ACDS is sold to customers and is subject to FDA oversight, while Self-Developed ACDS is created in-house by healthcare systems without regulatory oversight.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What are the current gaps in the regulation of Adaptive CDS?<\/summary>\n<div class=\"faq-content\">\n<p>The existing policy landscape is inadequate, leaving patients exposed to algorithmic bias and safety issues due to gaps in federal jurisdiction.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What is the significance of transparency in Adaptive CDS?<\/summary>\n<div class=\"faq-content\">\n<p>Transparency in how Adaptive CDS is trained is crucial for accountability, requiring clear standards for training datasets, model design, and data acquisition.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What communication standards are suggested for Adaptive CDS?<\/summary>\n<div class=\"faq-content\">\n<p>The AMIA paper suggests establishing communication standards for the intended use, expected users, and operational guidance of Adaptive CDS to aid in evaluation and maintenance.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>Why is oversight essential for Adaptive CDS?<\/summary>\n<div class=\"faq-content\">\n<p>Oversight ensures that Adaptive CDS achieves safety and effectiveness by managing implementation through consistent systems and controls.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What mechanisms does AMIA propose for governance of AI in healthcare?<\/summary>\n<div class=\"faq-content\">\n<p>AMIA calls for the creation of new bodies or departments for governing AI implementation and Adaptive CDS, along with Centers of Excellence for testing and evaluation.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What is the urgency behind the AMIA&#8217;s recommendations?<\/summary>\n<div class=\"faq-content\">\n<p>The rapid advancement of AI in healthcare necessitates urgent safeguards to ensure safe and effective use of machine learning applications.<\/p>\n<\/p><\/div>\n<\/details>\n<details>\n<summary>What role does AMIA aim to play in the execution of the policy agenda?<\/summary>\n<div class=\"faq-content\">\n<p>AMIA seeks to position itself as the leading organization to execute the policy agenda for the safe and effective use of Adaptive CDS in the U.S. healthcare system.<\/p>\n<\/p><\/div>\n<\/details><\/div>\n<\/section>\n","protected":false},"excerpt":{"rendered":"<p>Clinical decision support systems have traditionally provided clinicians with evidence-based recommendations or alerts to improve patient outcomes. Classical CDS tools are static, meaning that given the same input, they deliver the same output each time. Adaptive CDS represents a change from this model. These systems use machine learning algorithms to adjust and improve their recommendations [&hellip;]<\/p>\n","protected":false},"author":6,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[],"tags":[],"class_list":["post-30396","post","type-post","status-publish","format-standard","hentry"],"acf":[],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/posts\/30396","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/users\/6"}],"replies":[{"embeddable":true,"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/comments?post=30396"}],"version-history":[{"count":0,"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/posts\/30396\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/media?parent=30396"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/categories?post=30396"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.simbo.ai\/blog\/wp-json\/wp\/v2\/tags?post=30396"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}