Finding Answers to Sepsis in Intelligent Data and Technology

Sepsis is a critical condition for health clinicians and system administrators across the globe. It occurs when bacteria enter the bloodstream and can arise from seemingly common infections such as those of the urinary tract, skin, or even a dental abscess. When left untreated or not treated in a timely manner, the infection can cause internal organs like the heart, kidney, and lungs to malfunction.  

The World Health Organization (WHO) reports that one in five deaths worldwide are caused by sepsis — which means 11 million people die of sepsis each year. Those at highest risk are the very young, the elderly, and people with serious injuries or medical conditions such as diabetes, AIDS, cancer, or liver disease — but no one is immune to sepsis.

The sepsis statistics in the U.S. alone are alarming. According to the Center for Disease Control (CDC), one out of every three patients who dies in a hospital each year has been diagnosed with sepsis — that’s at least 1.7 million adults diagnosed and nearly 270,000 of them dying of sepsis.  

A costly condition

Treating sepsis involves a prolonged stay in the intensive care unit and complex therapies that drive higher costs and have greater risks. The most recent statistics list sepsis as by far the most expensive condition to treat in U.S. hospitals, costing nearly $62 billion for inpatient and skilled nursing facilities (SNF) alone.

People with sepsis are two to three times more likely to be readmitted to the hospital than people with many other conditions including heart failure, pneumonia, and chronic obstructive pulmonary disease. Because of a host of pre-existing underlying conditions, readmissions due to sepsis are also more expensive than readmissions due to any of these other conditions.

This is all outside the context of COVID-19. The effects of the coronavirus are being evaluated as another level of complication in identifying and treating sepsis patients. Critically ill COVID patients also share a diagnosis of Sepsis-3 due to the inflammatory response of the virus. 

Insights through analytics

Sepsis is clearly a problem that the medical and scientific communities have not adequately been able to wrap their heads around. The inability to save lives and reduce costs related to sepsis has not been for a lack of significant research by the CDC and others. 

With the huge numbers of patients affected every day by this illness comes important data. And inside that data is where the answers to sepsis lie — along with the ability to identify patterns early to either prevent a septic condition or intervene and prevent a negative outcome after a sepsis diagnosis. 

Many healthcare providers and their partners are taking the right approach, analyzing patient data as it is occurring — looking at real-time lab values, oxygen levels and hydration, which in turn empower clinicians to intervene quickly and stop the spread of infection. But given the complexities of sepsis, this has been challenging and we still have a long way to go in creating changes that will positively impact our sepsis population.

Retrospective data banks

To make significant strides combatting sepsis, a retrospective view of cases will help healthcare providers better understand what worked and what didn’t— and build a collaborative way to change practice and treatment patterns for better outcomes. 

Developing a baseline of what is occurring at the provider level at a healthcare facility can help clinicians identify not only treatment patterns that may be failing, but also challenge process areas that may not have been previously recognized. Given the years of data available, a historical view of past performance with a deep dive into key metrics such as ventilator use, ICU length of stay and mortality can spotlight potential negative consequences and provide a closer look at compliance or turnaround times. 

Investigating case management and planning workflows can help reveal factors affecting the patient’s risk of readmission — for example analyzing patient discharge locations to identify poor-performing long-term or skilled nursing facilities with a high bounce-back rate. 

Creating a “sepsis data bank” enables predictive models to be developed for identifying potential sepsis cases prior to onset. By preventing further mortalities and reducing overall costs, facilities can prescriptively identify patients who fall into high-risk categories and empower clinicians to intercede before the person ever enters the hospital. 

Pivotal technology tools

Gaining a retrospective view through an easily consumable online dashboard or scorecard where performance and improvement can easily be tracked over time is pivotal. Month-over-month and year-over-year comparison measures are also crucial as are baselines. The ability to add a spark-line representing when significant changes were implemented provides feedback to clinicians on how well they are doing, and creates a necessary feedback loop into their sepsis data bank to expand the breadth of data and improve the accuracy of models. 

Regardless of the approach, the analysis of sepsis will remain a priority for health systems and providers across the U.S. and globe for the foreseeable future. Defining what works best for your organization is key to understanding not only how you are currently dealing with the problem, but the steps you need to take to improve the lives of your patients.

Conduent Advanced Analytics Solutions
Conduent's DataVision Platform helps healthcare providers address the challenges handling sepsis within their facilities.  The technology provides the ability to evaluate the patient’s sepsis journey from diagnosis through post-discharge and establish key metrics to assist in reviewing hospital standards and strategies being employed to address sepsis. Using retrospective data in combination with prospective statistical models, the platform displays readmissions, length of stay and mortality in an intuitive, actionable dashboard aimed at increasing visibility and insight into to future sepsis patients and preventing negative outcomes.   

About the Author

Anna Daly

Anna Daly is VP of Data Analytics and Innovation for Conduent Healthcare where she leads the predictive analytics and cloud technology solutions for the healthcare provider group. Her focus centers on the expansion of clinical data management and business intelligence strategies to help clients understand the value of the clinical data and how this asset can drive change across the healthcare continuum. Anna holds a Master’s in Nursing Informatics from Vanderbilt University and is active in the special needs community as well as bringing awareness to end of life issues and the hospice mission.

Follow on Linkedin More Content by Anna Daly
Previous Flipbook
e-Book: Healthcare Analytics
e-Book: Healthcare Analytics

Next Article
Reassessing your CHNA and Implementation Strategies for COVID-19
Reassessing your CHNA and Implementation Strategies for COVID-19