Analytics and augmented intelligence (AI) offer a scalable promise to healthcare: It is possible to reduce health disparities, reduce waste, and improve patient safety, where one project’s success can springboard additional massive systemwide healthcare improvements.
On May 12, Farhana Nakhooda, SVP of Healthcare & Life Sciences, Asia Pacific, at Health Catalyst, spoke on this very topic at the AI World Congress (AWC). Nakhooda presented among other leaders in the AI space, each presenter outlining different innovative strategies to improve healthcare.
During her presentation, Nakhooda revealed different health systems that achieved great improvements through AI, with the promise of a larger, lasting impact.
One example Nakhooda referenced during her presentation at AWC was the success seen by ChristianaCare when conducting a systematic equity analysis.
ChristianaCare’s use of AI to provide more equitable care in the face of COVID-19, and the larger impact in reducing health disparity across its system, shows the scalable promise of AI. To ensure practices help reduce—not exacerbate—health disparities, ChristianaCare established health equity as a strategic priority, pursuing it with the same commitment and focus it would give to other domains of quality.
Using AI, it conducted systematic equity analysis to guide interventions. ChristianaCare was able to develop predictive tools to start to understand disparity across different clinical areas. One of the areas that they looked at was in COVID-19 testing.
ChristianaCare standardized the collection of personal characteristic data across its hundreds of registration points and mapped historical data to the current standard. The organization used Health Catalyst’s analytics applications as the organization’s single “source of truth” for personal characteristic data.
The organization also utilized 18 months of historical data from its data platform, master person data that conforms to the enterprise data standard, and AI to perform an equity analysis.
ChristianaCare created the ability to evaluate nine different measures over seven conditions for six equity dimensions, including age, race, ethnicity, gender, language, and zip code. The organization performed single- and multivariable analyses to perform a full range of analyses on each outcome and equity dimension, visualizing the data in a dashboard and creating a unique statistic for measuring, comparing, and tracking health equity.
As administrators learned more from the AI’s equity analysis, they changed some processes, and increased access to care through innovative clinics that combined virtual and primary care. By considering factors such as race, ethnicity, language, and zip codes, they’re able to take a clearer look at how to advance health equity and improve quality.
During her presentation, Nakhooda also offered insights on how the use of AI can make significant reductions to healthcare waste, as seen with the University of Pittsburgh Medical Center (UPMC), a system of 30 hospitals in Pittsburgh. The health system employed an analytics application to get a better understanding of where waste was occurring across its departments.
Healthcare waste is a result of a lack of understanding about how much it costs to provide patient care, preventable errors, unnecessary treatments, and a lack of knowledge about how those costs compare with the outcomes achieved, according to the Institute of Medicine Roundtable on Value and Science-Driven Healthcare. This is a problem affecting healthcare systems around the globe as the report details 30 to 50 percent of all healthcare resource expenditures are quality-associated waste.
Like many other healthcare systems, waste was a problem at UPMC. Spiraling costs and a changing environment required the health system to identify and address waste, reduce total cost of care, take on more risk, and transition from payment for volume to payment for value to create long-term sustainability.
The health system had been employing traditional financial analytics in their hospital, using relative value units (RVU) and ratio of costs-to-charge (RCC). These methods did not provide the level of detail and accuracy necessary to inform decisions that would enable UPMC to overcome industry threats to its sustainability.
To remedy this, UPMC implemented activity-based costing to deliver detailed and actionable cost data. By utilizing PowerCosting™, an application that leverages Health Catalyst’s analytics platform and activity-based costing models to deliver accurate and actionable data. The application delivers detailed and actionable cost data across the analytics environment, and supports service line reporting, contract modeling, and clinical process improvement.
By using PowerCosting™, UPMC was able to move away from an antiquated way of financial analytics and start looking at clinical variation and cost variation across their clinical community, resulting in a $3 million cost savings over two years.
UPMC also had increased visibility into cost variation and drivers of inefficiency in the operating room. This insight led to reduced readmission complications and improved patient satisfaction.
Many people look at AI as something very sophisticated and overly complicated, but UPMC’s application of the technology in this basic area is transformational and scalable.
Improving Patient Safety
Nakhooda also shared how Allina Health, a not-for-profit healthcare system of 12 hospitals and 90 clinics serving patients throughout Minnesota and western Wisconsin, used analytics and AI to make significant improvements to patient safety.
Allina Health wanted to reduce the number of patients that had severe sepsis and septic shock, limit surgical site infections and pressure injuries, and improve the mortality rate for these patients.
Despite a rapid process improvement project focused on the early identification of sepsis, sepsis mortality rates remained higher than desired. By turning to Health Catalyst to replace its burdensome manual review process, Allina Health was able to identify opportunities for improvement and develop evidence-based processes for sepsis identification and treatment.
Using analytics and AI, Allina Health was able to receive over $1 million in sepsis cost savings. They achieved a 30 percent reduction in the mortality rate of severe sepsis and septic shock. They saw an 80 percent relative reduction in elective colorectal surgical site infections.
From an operational perspective, they saw an 18 percent reduction in length of stay (LOS) of patients with severe sepsis and septic shock, and a 19 percent reduction in systemwide LOS.
Further, Allina Health found 216 more cases of pressure injuries – helping to identify where the organization should conduct a root cause analysis.
These combined insights led to a total cost savings of $125 million each year.
Building on Solutions
AI can improve current quality of care, remove inefficiencies, and bolster patient safety, but these insights can also pave the way for wider solutions across health systems and set a stage for an informed healthcare future.
As Nakhooda mentioned in her presentation at AWC, the promise of analytics and AI is that the data analysis shows you where you are, where you should be, and how to make improvements to bridge that gap in one small project that can build a better overall health system.