The document discusses weaknesses in current approaches to patient safety and how data-driven tools can help address these weaknesses. It describes how health systems currently take a fragmented approach to safety that focuses on specific metrics rather than reducing all-causes of harm. New tools using integrated data, machine learning, and predictive analytics allow a more proactive approach by identifying risks, recommending interventions, and enabling improvements across the care continuum. The document advocates for a sociotechnical approach combining improvements to safety culture, processes, and technology.