SPOT- AI to Spot Sepsis at HCA Healthcare

Background

Hospital Corporation of America (HCA) Healthcare was founded in 1968 in Tennesse. As a for profit hospital group with over 38,000 physicians, 2000 healthcare sites around the US and even in the UK, and 31.2 million patient encounters every year, HCA Healthcare stands as the largest for profit health system in the United States. The organization has been at the forefront for patient healthcare not just because of their large volume but because of their top performance and healthcare outcomes (eg. the maternal mortality rate at HCA Healthcare is half of the national average).

This is not to say that HCA Healthcare has not had issues around ethical behavior and mismanagement in the past. 2003 concluded a decade long federal investigation around HCA Healthcare’s fraudulent billing to Medicare that resulted in what could be argued as the costliest health-care fraud investigation- a payment of over $2 billion.

Since the settled fraud investigation, HCA got back on solid financial footing with a new focus on patient first. This paired with new governance guidelines “that the inspector general of the Health and Human Services Department described as unprecedented in scope and level of detail.” The company’s brand took a serious blow but the new guidelines and third party audits, focused business development in key areas, and renewed focus on innovative healthcare operations helped sustain and grow the company up to today.

 

Sepsis Prediction and Optimization of Therapy (SPOT)

HCA Healthcare is “using artificial intelligence (AI) and algorithms based on patient vital signs, labs, nursing reports, and other data to help identify sepsis approximately 18 hours earlier than the best clinicians.” Sepsis is important because diagnosis is hard to detect early on and per Dr. Jonathan Perlin, HCA’s CMO, once “you realize there’s a fire, it’s out of control.” SPOT is able to identify potential cases 18 hours faster than the top clinicians. From 2017 to 2018 SPOT was able to reduce mortality by an additional 23% (initial improvements not attributed to SPOT were 38% from 2003-2018). SPOT is an innovative AI and algorithm that helps HCA provide better outcomes for their patients and cost save but they would not have been able to develop this technology without a few key factors.

 

AI is highly dependent on the quality and quantity of information inputs. HCA Healthcare can monopolize on its huge data set that covers a wide demographic of patients. (For instance, Medicare and Medicaid make up ~40% of their payer mix, managed care and private payers make up ~45%, and self pay ~5$.) In addition to the number of patients, HCA healthcare also has this vast amount of data from earlier on than many of the new challengers in the market. This is crucial because there are changed that should be accounted for in terms of time. Protocol and/or medical interventions that were standard five years ago may no longer be appropriate. Even more nuanced, implementation of a new protocol/intervention could result in data that should not be considered standard (as is with any change management). These influxes and changes are mitigated in AI and machine learning algorithms by omission or by having enough volume and complexity in the algorithm. In addition to HCA Healthcare’s data, other organizations have also partnered to use SPOT in preventing sepsis related deaths and complications.

 

SPOT is able to provide continuous monitoring that physicians and healthcare providers do not usually have ability to. This level of monitoring is insightful because it is able to perceive data in real time while checking and predicting patient outcomes. There is a level of limitation with current medical systems that creates lag points between data being available, centralized, processed (in terms of determining a healthcare plan of action), and execution. SPOT removes these friction points and highlights at risk patients for providers to focus on. Importantly, SPOT is embedded in HCA Healthcare’s existing EMR, providing minimal adaption resistance. Because of this integration across HCA Healthcare’s network, implementing SPOT across all these networks would provide economies of scale in cost savings.

 

Key facts about Sepsis

  • About 270,000 people in the US die from the infection each year
  • It costs Medicare more than $6 billion in 2015
  • More people die from sepsis than from breast cancer, prostate cancer, and AIDS combined
  • It is the 11th leading cause of death in the U.S
  • The 9th leading cause of death in all hospitals
  • The 3rd leading cause of death in all intensive care units (ICU)
  • The leading cause of death in non-cardiac ICUs

 

Points of Consideration

 

Opportunity

While cost savings are still being evaluated for SPOT, HCA Healthcare has the opportunity to use this tool as a value creation source beyond the brand and healthcare outcomes of its own patients. HCA Healthcare is in a position to partner with other hospital groups (especially those that leverage the same EMR). This could be an additional source of revenue or data collection that would further SPOTs accuracy or provide insights outside of HCA Healthcare’s scope. Getting FDA approval and/or regulatory buy in for their product would allow for other hospitals and physicians to further buy into SPOTs offering. Lastly, SPOT has high sensitivity so false positives do occur. Qualified feedback from healthcare professionals still find the alert helpful because it usually is an indicator of other issues that need to be managed, but as sepsis specific tool, this aspect would need to be further analyzed and improved upon.

 

Competitors

This opportunity is highly dependent on the availability of alternative and the cost structure that is put into place for this and other products. There are other companies and hospital groups that have been looking towards creating their own version of SPOT but there are also companies that are not using AI to tackle sepsis. One example is T2Biosystems.

 

T2Biosystems created an FDA approved bacteria panel identifies:

  • 50% of all bacterial bloodstream infections1
  • 70% of all blood culture species in the emergency department2
  • 90%of deadly ESKAPE pathogens3

The results com back within 3-5 hours which could be a potential threat to SPOTs offering.

 

Conclusion

 

HCA healthcare has leveraged its AI, SPOT, in order to save lives and hopefully reduce costs. SPOT is able to capitalize on the quantity and quality of data that HCA Healthcare has alongside the possibility of running the digital component 24/7. The result is a decreased mortality rate from sepsis of 23%, with thousands of lives saved. There is a possibility to leverage SPOT as a product to other hospitals but considerations around sensitivity, competitive products, and pricing could prove challenging.

 

References

https://money.cnn.com/magazines/fortune/fortune_archive/2004/02/09/360103/index.htm

https://marketrealist.com/2014/12/exploring-payer-mix-trends-at-hca-holdings/

https://www.t2biosystems.com/products-technology/t2bacteria-panel/

https://hcahealthcare.com/util/forms/press-kit/HCA-presskit-fact-sheet-a.pdf

https://hcatodayblog.com/2018/09/10/spot-how-hca-is-sniffing-out-sepsis-early/

https://www.t2biosystems.com/products-technology/t2bacteria-panel/

https://www.modernhealthcare.com/care-delivery/hca-uses-predictive-analytics-spot-sepsis-early

Previous:

Data Analytics drives down health insurance costs for small businesses

Leave a comment