Using Analytics to Combat Sepsis


Dignity Health has now rolled this capability out to 30 hospitals, and can monitor performance across its hospital system to extend best practices in specific facilities and identify areas for improvement. It reviews key performance indicators (KPI) weekly, including nurse and provider response times, compliance with the SEP-1 bundle, and variation in patient care.

Overall, the program is aimed at initiating treatment as soon as possible for patients who present with symptoms of systemic inflammatory response syndrome, sepsis, or severe sepsis. The automated dashboard provides two levels of information: patient-specific data alerts to nurses and doctors when an individual is declining, and response time alerts monitored for timeliness of delivery-appropriate treatment protocols. Each week, doctors and nurses review a variety of dashboard metrics, including:

  • Average and/or geometric mean length of stay
  • Sepsis mortality rate
  • Timeliness of intervention/bundle compliance
  • Response to alert (dismiss or action)
  • Alert volume
  • Alert performance
  • Renal failure

Given the need for near real-time monitoring and confirmation of symptoms, a number of data sources must be incorporated into the analysis. Data sources, including electronic medical records, admission, discharge, and transfers, provide a comprehensive view of a patient’s status and risk of infection. The process of data acquisition, transformation, and standardization allows for delivery of critical information in a consumable format to help doctors and nurses react quickly. An enterprise analytics hub that incorporates clinical, cost, pharmaceutical, quality, and patient behavioral data across the care continuum forms the program’s foundation.

Since it launched in February 2015, the bio-surveillance program has tracked more than 500,000 patients, resulting in quicker decisions to provide treatment earlier, which saves lives while improving operational costs. Initial program objectives are to save a minimum of 400 lives per year and to reduce length of stay in the ICU by 10%. Additionally, the implementation of the dashboard, alerts, and feedback loops into nurse and provider performance is driving dramatic improvement; this technology has provided a level of transparency across the enterprise that directly affects patient care and costs, along with identifying improvement opportunities in medical coding.

Dignity Health has learned multiple lessons from this program and uncovered other opportunities to improve patient care and costs. Key lessons include:

  • It’s important to define, measure, and monitor agreed-upon KPIs.
  • Alerts must be directed to the right people, such as the primary nurse, charge nurse, triage nurse, and attending provider.
  • The first 24–48 hours after the last alert are critical.
  • The resulting information is configured in the workflow to ensure source data, documentation, and alerts are incorporated into operational processes.
  • A feedback loop is critical to monitor workflow performance, assign a goal owner for the KPIs, and provide nursing and provider attribution to monitor response times and patient improvement.

As a result of this project, program objectives are on track to realize dramatic improvements in the identification, diagnosis, and treatment of sepsis, and valuable lessons have been learned. Together, the use of an enterprise analytics hub, the development and monitoring of KPIs, and the documentation of workflow with a feedback loop can significantly reduce the impact of critical conditions and improve cost and operational processes. But most importantly, they can improve patient safety and outcomes, which is the ultimate measure of success.

Paula Brown leads the industry solutions, implementation, and delivery teams for SAS Health and Life Sciences solutions portfolio across pharmaceutical, healthcare, and health provider industries. Her teams are responsible for successful customer engagements that include enterprise and industry solutions, technical architecture, implementation, and advisory services. Prior to joining SAS more than 13 years ago, Brown served as president and CEO of InTechnology Corporation, a venture-backed strategy management and technology consulting company. She may be contacted at


Allen, T. L. (2013, November 14). Reducing mortality from sepsis. H&HN. Retrieved June 13, 2016, from

Butcher, L. (2016, January 13). Stepping up against sepsis. H&HN. Retrieved May 31, 2016, from

Center for Medicare and the Center for Clinical Standards and Quality. (2015, September 18). Hospital inpatient prospective payment system and long-term acute care hospital prospective payment system fiscal year 2016 final rule. Medicare Learning Network. Retrieved from

Scutti, S. (2014, March 18). Want to decrease your risk of septic shock? Get out of the ICU. Medical Daily. Retrieved June 13, 2016, from

Torio, C. M., & Moore, B. J. (2016, May). National inpatient hospital costs: The most expensive conditions by payer, 2013 #204. Retrieved June 13, 2016, from