Expert: AI Must Augment, Not Replace, Clinicians
In a recent podcast, critical care specialist Dr. Amel Havkic argued that AI is most effective as "augmented intelligence" that supports, rather than supplants, clinical judgment. He stressed the need for informatics professionals who can help innovators build solutions that are approved by clinicians and can be scaled safely. This requires bridging the gap between bedside medicine and MedTech strategy.
- The American Nurses Credentialing Center (ANCC) offers the Informatics Nursing Certification (RN-BC), which requires a current RN license, a bachelor's degree, and specific practice or academic experience in informatics. For example, one eligibility pathway is a minimum of 2,000 hours in informatics nursing within the last three years. - AI is being leveraged in intensive care for predictive analytics to forecast events like sepsis, pressure injuries, and delirium, enabling earlier intervention. These models often use data from large ICU databases such as MIMIC-III/IV and eICU and can identify subtle trends in patient data that may not be immediately apparent to human observers. - A major source of frustration for ICU nurses is the fragmented nature of IT systems, which often lack interoperability between different departments (e.g., physician and nursing systems). This leads to redundant data entry, searching for information across multiple platforms, and can compromise patient safety when information is not readily available. - Health Level Seven (HL7) and Fast Healthcare Interoperability Resources (FHIR) are key standards for exchanging health information. FHIR, the more modern standard, uses web-based technologies (RESTful APIs) to define modular components of data called "resources," making it easier for different systems and applications, including mobile apps, to share data. - EHR optimization projects can significantly reduce the documentation burden on nurses. One UCHealth project focusing on Epic flowsheets cut documentation time for acute care nurses by 18 minutes per 12-hour shift, saving over 64,800 hours annually across the system. This was achieved by eliminating unnecessary documentation options and redesigning workflows to hide irrelevant information. - The 21st Century Cures Act, implemented through rules from the ONC and CMS, mandates greater interoperability and patient access to their health data. These regulations require healthcare providers and IT vendors to adopt standardized APIs, like FHIR, to prevent "information blocking" and allow patients to access their records through third-party applications. - A foundational understanding of data science is increasingly important in nursing informatics, encompassing skills in statistical analysis, data visualization, and machine learning. Nurses play a critical role in the data lifecycle, from capturing data in the EHR to interpreting data science reports to improve patient care and operational efficiency. - Common usability issues reported by ICU staff with clinical information systems include unclear presentation of medication information, redundant data entry requirements, scattered information across different sections, and slow system speed. A European survey found that due to poor IT system usability, nurses spend over an hour per workday on tasks like unnecessary clicking and navigating, which could be reallocated to patient care.
