Could you predict harm before it happens?

The next generation of patient safety and health care quality improvement will include not only the retrospective and concurrent analytic capabilities but also prospective: the capability to predict harm events before they occur.

Our harm predictive analytics - currently in research and development at clients - aspires to do just that, by applying clinical and data science techniques to continuously evaluate, enhance, and improve our underlying predictive modeling. Through highly accurate predictive algorithms and a frontline-focused user experience, Pascal’s approach optimizes for alert fatigue avoidance and seeks to empower staff to intervene at the appropriate points in the workflow based on evolving clinical risk.

Our innovative method for predicting all-cause harm includes:

  • Advanced data science - We go beyond many predictive models that use static multi-variable regressions by implementing an algorithmic approach that can be improved over time to increase accuracy. This model uses statistical analysis, machine learning, and other advanced methods of data science. Our technology is capable of running clinical data through thousands of potential predictors in real-time in order to assess risk.

  • Academic collaboration - Based on consultation across Pascal’s National Collaborative, our team across our Clinical Services and Applied Science groups have developed a standardized harm classification coding approach, which considers the whole view of each patient, rather than traditional binary classifications.  Getting definition and classifications questions resolved well is critical for predicting the outcomes that we want to achieve.
  • Clinical practice - Our solution incorporates clinical use cases benefit from the combined experience across many of the field’s leading health systems in enhancing how our clients use our safety management system and information provided for action and improvement. 
  • Unique data set - Supported by our Pascal PRIME cloud service, our harm predictive analytics benefit from Small Data - a very large clinically confirmed data set - as well as hundreds of millions of clinical events in our Big Data set - unified, standardized, and normalized across the field’s only real-time PSO client community. 

We deliver real and measurable results because we combine the right clinical methodology and the right technology to support scale and cost-effective execution.

Our methodology is designed to create enduring change across healthcare organizations, everywhere from the front line to the executive offices. It's been developed by our team of experts in safety, quality and process improvement, in concert with Collaborative partners. Our technology platform helps them execute on their vision. That's our sweet spot--bringing together the methodology and the technology needed to achieve results.


We have developed a results-oriented and clinically rich application model through more than 100 years of combined experience creating safer and more reliable healthcare delivery.

Learn More.



Pascal HealthBench supports all of our applications as the first patient safety SaaS platform in healthcare, and is used by many of the world's leading healthcare delivery systems.

Learn More.



Our solutions are developed to lead to measurable impact where it matters, for the hospital and the patient. We look forward to opportunities to measure that impact with you.

Learn More.