Where do Sepsis Predictive Models Fall Down?

A leading EHR provider took heat for their Sepsis model only detecting 7% of cases missed by clinicians.  Today we explore an article by Angelique Russell on why that is unacceptable and totally understandable.

FTA
The alarms you see on television exist only for the sickest patients at best. There is some data supporting the cost effectiveness of transitioning to continuous monitoring for all patients, but progress towards this goal is often hampered by competing financial priorities.  Neither human nor machine can predict risk on null data:

But Epic is not relying on such an algorithm to detect sepsis from labs and vital signs, it is relying on the medical billing code for sepsis, which has an even looser definition subject to varying payer guidelines. When a patient is deteriorating, the "acuity level" or resources necessary to care for that patient increase, so it is necessary that hospitals bill higher amounts for higher acuity care. But in practice, this can result upcoding.

---

This article is worth checking out.  From the front lines of care from a practicing data scientist. 

I love how she closes.

Despite what you might have heard about magical unicorns, data science is a team sport. In a clinical setting, it requires guidance from nurses, doctors, clinical quality specialists, medical device experts, EHR experts, data engineers, and data scientists. More than that, it requires leadership to ensure effective collaboration and proper prioritization of patient outcomes above all other end points.

#healthcare #healthIT #cio #cmio #datascience #chime #himss

A leading EHR provider took heat for their Sepsis model only detecting 7% of cases missed by clinicians.  Today we explore an article by Angelique Russell on why that is unacceptable and totally understandable.

FTA
The alarms you see on television exist only for the sickest patients at best. There is some data supporting the cost effectiveness of transitioning to continuous monitoring for all patients, but progress towards this goal is often hampered by competing financial priorities.  Neither human nor machine can predict risk on null data:

But Epic is not relying on such an algorithm to detect sepsis from labs and vital signs, it is relying on the medical billing code for sepsis, which has an even looser definition subject to varying payer guidelines. When a patient is deteriorating, the "acuity level" or resources necessary to care for that patient increase, so it is necessary that hospitals bill higher amounts for higher acuity care. But in practice, this can result upcoding.

---

This article is worth checking out.  From the front lines of care from a practicing data scientist. 

I love how she closes.

Despite what you might have heard about magical unicorns, data science is a team sport. In a clinical setting, it requires guidance from nurses, doctors, clinical quality specialists, medical device experts, EHR experts, data engineers, and data scientists. More than that, it requires leadership to ensure effective collaboration and proper prioritization of patient outcomes above all other end points.

#healthcare #healthIT #cio #cmio #datascience #chime #himss