Today we released Brief 7, which looks at available hospital capacity and the “Runway” we have in North Carolina hospitals. tl; dr: At current trends, we will start to hit capacity of ICU in just over 4 weeks and about 6 weeks for overall hospital beds. #ncpol https://twitter.com/nccovid/status/1336328835868938240

As we ( @PharmDJD and @A_McKethan ) have in previous briefs, we use a pretty simple method for this analysis. Future Use = Current Use * (1+growth)^time. The appeal of this is that there is not much to it -- it uses recent trends, current data, and the exponential function.

We do 5 estimates: one for “how long until we hit capacity at current growth?”, and then four on “how fast would growth have to be to hit capacity at 4 / 8/ 12/ 16 weeks?” We do it for {statewide + 7 regions} x {acute, ICU} = 16 different settings.

(We know that most hospitalization models use the preparedness regions. We like the PHRST better for a number of reasons: they are county-based (rather than point-based) and partition the state).

One caveat: our analysis assumes all else equal – namely, that hospitals cannot expand runway through (e.g.) cutting elective procedures, expanding capacity, etc. There are a number of safety valves that can be used. Of course, it also assumes underlying trends don’t get worse.

We find that, given our current growth rates, we have about 6.8 weeks of runway for acute beds. This would put us around January 23. For ICU beds, we project capacity being reached around January 7.

There are regional differences. Charlotte and Asheville regions have shorter runways for both, while Wilmington has longer runways. There are regional variations in case growth, which portends future growth in hospitalizations. https://twitter.com/gmarkholmes/status/1330939366185840640?s=20

Of course, these forecasts are not destiny; we can extend our runway and give us time for the vaccines to make their way through the public. While the next week or so is largely determined (today’s case is next week’s hosp.), we can move the needle in the medium term. -30-