The “missing ingredient” to get ahead of outbreaks
Like many, I too hoped that we had turned the final corner on the COVID-19 pandemic. But the surge of the Delta variant and continued vaccine hesitancy means we are increasingly likely to see a twindemic — that is, simultaneous COVID and seasonal flu — later this year.
If there is indeed a twindemic, early, actionable information about what’s going around will be more important than ever. If you know your area’s risk level, you can take precautions. If you fall ill but know what’s going around, you can take the right steps to address your symptoms and get better faster.
Back in 2012, we set out to capture the missing ingredient from all existing public health data sets that would allow us to give people the information they need to stay well or avoid illness: real time, geographically-precise indicators of illness. While public health data including doctor’s visits and lab results tell part of the story, we need to talk to people far earlier in the process, as soon as they feel sick, if we are to get ahead of an outbreak.
Today, Kinsa’s system allows us to access unique data the healthcare system misses entirely:
- Data from mildly symptomatic individuals — well before they see a doctor or go to the lab.
- Data on how fast an illness spreads in the home or in school — a good indication of how fast it will spread in the community.
- Data from underserved communities — these families are often larger and underinsured, seeking care late or not at all.
This unique, early data is Kinsa’s secret sauce, allowing us to detect emerging epidemics and understand seasonal illness trends weeks ahead of other systems. With years of historical data combined with specific, hyper-local, real-time symptoms trends, we can produce a long-lead forecast for the entire season as early as November 1.
This is powerful for two reasons: first, it gives local health systems ample time to prepare. Second, it serves as a baseline of “expected” illness levels. When we see illness levels outside the expected range for a given time and place, that’s a problem worth investigating because it could mean the emergence of something new. We also see how often certain symptoms occur, how quickly they tend to worsen, and how fast they spread — key indicators that help us characterize a potential epidemic.
The ability to identify outbreaks early is only useful if we use that knowledge to curb transmission and keep communities healthier. Fortunately, Kinsa does that too.
Our school health program FLUency is 5,000 schools nationwide monitors local, school, and grade illness levels, triggering alerts to parents, school nurses and administrators, guiding the most vulnerable to preventative actions. The program has shown exceptionally high engagement levels (65% of parents, teachers, administrators and school nurses engage on a weekly level), and has driven a 27% decrease in illness-based school absenteeism. Fifty percent increased hand washing and 97% of teachers disinfect their classrooms more frequently after participating, showing that knowledge of what’s going around influences behavior, keeping us all healthier.
These results encapsulate the benefits of an early warning system for spreading illness. It’s not rocket science, but it is innovative data science that can change the way we plan for and respond to infectious illness — and keep families and communities healthier.