Practice Fusion’s Intervention Risk Management Practices

Practice Fusion, we prioritize the safety, reliability, and effectiveness of our Predictive Decision Support Interventions (PDSIs) by adhering to comprehensive Intervention Risk Management (IRM) practices. These practices are designed to meet the stringent guidelines set forth by the ASTP/ONC and align with relevant sections of the National Institute of Standards and Technology (NIST) AI Risk Management Framework. In addition, it aligns with FAVES principles of being fair, appropriate, valid, effective, and safe for all users.

Our approach begins with a risk analysis, where a cross-functional team evaluates potential risks and adverse impacts associated with the validity, reliability, robustness, fairness, intelligibility, safety, security, and privacy of our predictive models. This helps to ensure that our models are accurate, consistent, resilient, unbiased, transparent, safe, secure, and protective of patient privacy.

To address any identified risks, we implement risk mitigation strategies. These include regular updates and validation of our predictive models, ongoing monitoring and improvement of model performance, and the implementation of safeguards to reduce the risk of adverse impacts on patient care.

Governance is a key aspect of our IRM practices. We enforce policies and controls to ensure that data is acquired ethically and legally, managed with high standards of integrity, and used responsibly to support accurate and fair predictions.

To ensure that our practices remain up-to-date and effective, we review each release for Predictive Decision Support Interventions as well as conduct an annual review of our Predictive Decision Support Interventions and IRM practices. This allows us to update our documentation and practices as necessary, reflecting the latest standards and ensuring ongoing improvement.

Practice Fusion’s IRM practices help ensure that its Predictive Decision Support Interventions align to applicable laws and regulations including ASTP/ONC certification criteria and the NIST AI Risk Management Framework.