Promise to Practice: Navigating the Future of CDSS

By William Toth, West Region Director of Operations, AFMED, United States Air Force

The integration of Clinical Decision Support Systems (CDSS) into Electronic Health Records (EHRs) was anticipated to revolutionize healthcare delivery, offering promises of enhanced patient safety, improved clinical outcomes, and streamlined care processes through evidence-based recommendations. Despite notable successes, the journey from promise to practice has been challenged by various obstacles, hindering widespread adoption and adding burdens to clinical staff.

CDSS systems enhance clinical decision-making by delivering timely, patient-specific data at the point of care, aiming to better health outcomes and healthcare efficiency. By filtering data, automating tasks like issue flagging, and recommending guideline-based treatments, they improve patient safety, guideline adherence, reduce mortality, and increase care cost-effectiveness. Additionally, CDSS boosts healthcare team collaboration, communication and fosters patient involvement in care.

The successful adoption of CDSS hinges on building trust among clinical users, addressing barriers to adoption, and leveraging AI responsibly.

CDSS adoption has been hindered by early versions’ lack of transparency, insufficient training, interoperability challenges, and doubts about information accuracy and relevance. Clinicians worry about CDSS undermining their judgment or offering advice that doesn’t align with their expertise, experiences, or patient preferences. Alert fatigue from too many redundant and low-priority alerts has also fueled skepticism and decreased attention to critical notifications. Moreover, poor integration into clinical workflows has prompted inefficiencies and clinician workarounds, detracting from CDSS’s intended benefits.

Integrating Artificial Intelligence (AI) into CDSS holds significant potential to transform healthcare decision-making. By leveraging the capabilities of AI to analyze complex datasets, healthcare providers can benefit from more accurate predictions and personalized recommendations tailored to individual patient needs. This advancement could lead to improved patient outcomes, more efficient use of resources, and optimized treatment plans. However, the incorporation of AI into healthcare also introduces a set of regulatory, ethical, and procedural challenges. Issues such as data bias, discrimination, and the need for transparency have emerged in other industries utilizing AI and must be meticulously addressed within the healthcare sector.

By reducing cognitive burden and screen time, the introduction of AI into CDSS has the paradoxical potential of returning humanity to medicine; allowing clinician teams more time dedicated to patient interaction and less to administrative documentation. Much as the initial deployment of CDSS required adherence to the principles of the “5 Rights”, the intersection of AI with CDSS will require an expansion of these ideologies to account for the growing role that AI will play. New “Rights” such as the “Right Transparency,” “Right Ethical Use,” “Right Autonomy,” “Right Feedback & Learning,” “Right Integration,” and “Right Security” must be adopted to ensure that CDSS with AI capabilities meets the needs of clinical staff and patients without unexpectedly increasing the overall burden of efficacious care delivery. 

Transparency and responsible AI use are paramount in building trust among clinical staff and addressing regulatory and ethical concerns. CDSS recommendations must be accurate, relevant, and presented in a user-friendly manner. Clinicians need to understand how the system generates its recommendations and feel confident in the evidence supporting them. Healthcare organizations must develop transparent and explainable AI systems that clinicians can trust, while also addressing the broader ethical implications of AI in healthcare. Efforts to promote accountability and shared decision-making between clinical staff and CDSS are essential to ensure the safe and effective use of these tools while ensuring any ethical implications for bias in the decision-making can further bolster or degrade trust.

Efforts to encourage CDSS adoption must prioritize meeting clinical staff needs by involving them in development and implementation, along with comprehensive training for system familiarity. Adoption and adherence are optimized by smoothly integrating into existing workflows. Clinical informatics professionals, such as trained and experienced physicians and nurses, should lead in creating specialty-specific CDSS. Successful healthcare organizations will focus on workflow integration to ease clinician burnout, decrease cognitive load, and enhance patient interaction. Additionally, leveraging AI to analyze clinical staff usage patterns of CDSS can offer feedback for system improvement and create new value streams. To promote a culture of innovation and improvement, organizations should motivate CDS usage with policies and incentives that support and reward experimentation and ongoing enhancement.

As healthcare organizations implement more advanced CDSS with AI, the security imperative to safeguard patient data and ensure compliance with regulatory standards will increase. How AI tools are used and the implications on both the security and ethical use of patient data will have huge implications on the successful implementation of any CDSS system using AI. Staff training and understanding of data use and security protocols will become essential to mitigate risks associated with data breaches or unauthorized access. By prioritizing security, healthcare organizations can instill confidence among clinicians and patients in using CDSS while upholding the highest standards of privacy and confidentiality.

Integrating AI into clinical decision support heralds a new era of personalized, data-driven healthcare. While AI offers immense potential to enhance diagnostic accuracy, optimize treatment outcomes, and improve efficiency within healthcare systems; adoption also presents challenges such as bias, interoperability, and clinician resistance. Successful adoption of CDSS hinges on building trust among clinical users, addressing barriers to adoption, and leveraging AI responsibly. By prioritizing clinician engagement and education, designing user-friendly systems, and fostering a culture of innovation and iteration, healthcare organizations can harness the full potential of CDSS to improve patient outcomes and reduce clinician burnout. As the healthcare landscape continues to evolve, it is imperative that we refine these systems to meet the evolving needs of healthcare providers and patients alike.