As hospitals scramble to prepare, they are asking the question, just how bad could it get? As I was trying to find answers to help inform our local hospitals, I struck gold with the COVID-19 Hospital Impact Model for Epidemics (CHIME), an online tool developed by the Predictive Healthcare team at UPenn Medicine.
The tool helps project the expected number of daily hospital admissions, intensive care unit (ICU) admissions, and admitted patients in need of a ventilator, as well as admitted patient totals, based on a number of assumptions about the community, hospital system, and spread of the virus. It is not predictive, but rather a tool that could be helpful in planning for the anticipated demand on the local hospital system.
I ran an analysis for Small City, USA. My model assumptions were as follows:
Model Assumptions
1 known case
1 known hospitalization
6 day case doubling time
5% case hospitalization rate
2% case ICU rate
1% case ventilated rate
7 day hospital length of stay (LOS)
9 day ICU LOS
10 day ventilator LOS
50% hospital market share (we have two main hospitals)
65,000 population
The key consideration here is case doubling time. A 6-day case doubling time simulates the spread of COVID-19 unchecked, without any mitigation efforts like social distancing. In other words, we are modeling the worst case scenario, given what we know and where we are today. Here are the results.
Without mitigation, COVID-19 hospital admissions peak at 56 days (8 weeks) with 18 new admissions per day, including 6 ICU patients, and 3 on ventilators (Figure 1). Because average length of stay is 7 days for each admitted patient and longer for those needing intensive care, the projected total number of admissions on day 56 paints a dire picture.
At its peak, COVID-19 patients would account for 122 hospital beds in 8-weeks time, with 62 patients in the intensive care unit, and 34 on ventilators. It is unclear whether our hospitals currently have, or will have, the beds, ventilators, supplies, and personnel needed to treat this number of patients (Figure 2). And what does the rest of Small City look like?
At around 9 weeks, without social distancing measures, we would see more than 17,000 cases at the peak of the COVID-19 outbreak. That is nearly 1 in 4 Small City residents (Figure 3).
Social Distancing is the Difference Maker
Now back to that oh-so important case doubling time. If we can add just 2 days to the amount of time it takes for the number of cases to double, from 6 to 8 days, best achieved though social distancing measures, we can push the peak out by 2 weeks and reduce the number of daily hospital admissions by more than 40% (Figure 4).
And the more social distancing the better. If we can reduce social contact by 25%, we see a 65% drop in daily admissions (Figure 5), and the total number of community cases drops from 17,000 to just over 10,000 at peak (Figure 6).
There is a reason why schools are closed, businesses are closed, and we are being asked told to isolate at home. Social distancing works. Without it, we risk overloading our hospitals and putting COVID-19 patients – and healthcare workers – in a very dangerous place.
For more on how social distancing works to slow the spread of COVID-19, check out this Washington Post animation.
