Big Data is watching a million fireflies

I watched Mercy Health’s John Conroy talk today about how they manage one of the most deadly issues facing hospitals right now: Sepsis. It is a systemic infection that isn’t the reason the patient entered the hospital.

He described what they do as standing in a field in the summertime at dusk and seeing flashes of light from thousands of fireflies all around. The flashes are random but consistent, spread across all of the open space.

Sepsis can happen anywhere and takes combinations of real-time data to spot…like watching every one of those fireflies and every flash and being able to understand, anticipate and act on critical information from it.

Sepsis is every healthcare network’s nightmare. It can happen to anyone at any time, and as I said, anywhere in the hospital. Sadly, many doctors in practice today learned to watch for sepsis after their first septic patient dies. The industry-wide mortality rate is 65%. Mercy has significantly lowered their mortality rate from Sepsis and that’s a wonderful thing.

Mercy has a system that monitors every patient for sepsis by seeing every piece of information immediately, no matter where it happens across all of their facilities. They watch heart rate, blood pressure, temperature, demographic background, treatment condition, and progress. It is around 3,000 pieces of information per second that has to be pushed, pulled, digested, pattern matched and watched for temporal (time sensitive) trends. When the system spots a negative trend, they dispatch doctors and nurses immediately as Sepsis can only be treated and the patient saved by immediate action (also in combination).

Even after the alarms are shut off, they look at what happened and constantly update the rules that drive the system. They keep getting better results.

They’ve become so good at what they do that they sell their services (at cost, they’re not-for-profit) to other facilities through Mercy SafeWatch. Big Data saving lives.

Google+

Christopher Taylor