Fig. 2: Projected intensive care unit (ICU) surges and days under lock-down for the optimized strategy versus four alternative strategies.
From: Design of COVID-19 staged alert systems to ensure healthcare capacity with minimal closures
Optimal is the recommended strategy. The Optimal two-stage strategy is optimized to respect ICU capacity under a two-stage alert system, and the Optimal hospital strategy respects total hospital capacity under a four-stage alert system. The Percent ICU strategy is based on France’s mitigation policy5 and the Incidence strategy is based on reopening criteria proposed by the Harvard Global Health Institute30. a The maximum daily number of COVID-19 patients in ICUs versus the number of days under the most restrictive red alert level. Each point represents the result of a single stochastic simulation under one of the five policies (indicated by color). The plot includes 300 points per policy; the vertical black line indicates the estimated COVID-19 ICU capacity of 331 beds for the Austin area. The vertical stratification of the green and purple points stems from a model assumption that stages must be in place for a minimum of 14-days before they can shift. The Optimal policy is designed to minimize the use of costly stages while having 95% of the peak-demand values within ICU capacity. On the rare occasion that hospital admissions transiently exceed the red threshold, a return to orange is often triggered as soon as the 2-week minimum passes. b The expected proportion of days spent in each stage, colored in the same manner as Fig. 1 and the 95th percentile of unmet ICU demand measured in patient-days above capacity, in gray with values indicated on the right y-axis.
