-
CHIME: COVID-19 Hospital Impact Model for Epidemic
- An open source contribution by Penn Medicine to the healthcare community
- Using epidemiological patterning, it predicts a theoretical number of people festering the virus which can help with capacity planning.
As we get ready for the extra demands that the COVID-19 pandemic will position on our hospital system, our functioning heads need up-to-date forecasts of what surplus resources may be entailed in the future. Cognisant approximations of the number of patients who will require hospitalization, mechanical ventilation, ICU infrastructure, etc. in the very near future will significantly enhance the promptness of responses and mitigation strategies.
To this end, the Predictive Healthcare team at Penn Medicine has created a tool that leverages SIR modeling to aid hospitals with capacity planning around COVID-19.
CHIME lets hospitals to input information about their patients and adjust suppositions around the coverage and conduct of COVID-19. It runs a standard SIR model to predicts the number of new admissions each day, accompanied by the day-to-day hospital survey. These predictions can be used to develop worst- and best-case pictures to support capacity formulations.
The most significant feature in this model is the doubling time. It describes how quickly a disease spreads. Familiarities with other geographical contexts hint that the doubling time may span from 3 to 13 days or more. The US, according to Our World in Data, has a four-day doubling rate. It does not paint a promising picture in the fight towards Corona. Dr. Liz Specht guesstimated even at a doubling rate of six, “all hospital beds in the US will be filled by about May 10.”
You may enter your doubling time estimate in an online version of CHIME and observe what happens.
