Hospital Beds and ICUs Vary Across the U.S.
Using GraphXR to identify vulnerable regions and populations amidst pandemic.
As the surge of COVID19 overwhelms metropolitan areas such as New York, the question of sufficient healthcare resources emerges across the US. On April 9th, New York peaked with 1,036 deaths in a day and surpassed their supply of hospital beds and ICUs. Left to build makeshift hospitals in Central Park, the same question is posed to other less-equipped regions of the U.S. — many of them more economically or demographically vulnerable. Where do these disparities of resources lie and which populations are at most risk?
Visualizing the distribution of hospital beds in all 50 states is a first step towards answering this question. Below, a 3-D scatter plot depicts a state’s percentage of at-risk individuals aged 60+ (x-axis) against the state’s count of hospital beds per 1000 individuals (y-axis) and the state’s total population (z-axis). For each state, the size of the node is scaled to the total number of hospital beds while the color is mapped to the total number of at-risk individuals (brown being the most).
With data extrapolated from the American Hospital Association (AHA) and the 2018 US Census through the Kaiser Family Foundation, the largest population of at-risk individuals is unveiled to be in California (represented by the brown node). This makes sense given that California is the most populated state with around 39 million people. However, its at-risk population is only 19% of the state’s total population as compared to Florida and Maine with around 26%. A major concern, however, is the concentration of hospital beds available. California carries 1.8 hospital beds for every 1000 individuals, the third-lowest in the country after Oregon and Washington. While social distancing has slowed down the infection rate in California thus far, relaxing these orders too soon could easily result in a surge of cases that would overwhelm the number of hospital beds available.
Northwest states like Oregon and Washington are similar to California in that they report the lowest concentrations of hospital beds available, making their at-risk populations more vulnerable. On the opposite coast, Vermont, Maine, and Florida have the highest densities of at-risk individuals, more than 25%. Although Vermont and Maine have drastically smaller populations compared to Florida with 20 million, these small states carry a higher concentration of vulnerable residents that would likely need hospitalization. It’s also important to note that Florida is the 3rd most populated state in the U.S. with the 2nd highest density of at-risk individuals. As Florida relaxes its social distancing orders by reopening beaches, a potential surge in its cases would fall disproportionately on its at-risk population.
Total hospital beds per state is a broad overview of where vulnerable populations lie in relation to healthcare resources. There is quite a bit of variance on the county lines as well, especially in terms of ICU capacity. Assuming hospital beds are not converted into ICUs, the next visualization captures ICU density county-by-county across each state.
Current data on the number of ICUs and hospital beds is not easy to obtain. For this reason, health system researchers like the Harvard Global Health Institute (HGHI) are making it their mission to launch regional capacity estimates for local governance to take initiative on pandemic readiness. Transparency on where the resources are located is crucial in the race against time.
With data aggregated from Kaiser Health News analysis along with the 2019 US Census lat/lng coordinates, we can map the density of ICUs by county. Above, this visualization reveals that about 53% of U.S. counties (shown in red) have 0 ICUs while 47% of counties report more than 1 ICU and 19% with more than 20 ICUs.
It may be obvious that metropolitan areas like Los Angeles have some of the highest numbers of ICUs in the country. However, the question becomes about what percentage of the population can be hospitalized before those resources are exhausted. To try to answer this, projections are made by the U.S. hospital markets, also known as Hospital Referral Regions (HRR), through localized data sets, discerning the number of ICUs and hospital beds needed if such a surge occurred. This localized data with a clear depiction of ICU density across county lines better defines the lack of preparedness in a majority of the U.S.
In this fight against the exponential spread of COVID-19 through social distancing and shelter-in-place measures, there are several potential outcomes that can play out. Visibility around ICUs and hospital beds is crucial to how healthcare facilities allocate resources strategically while taking note of their at-risk populations and facilities shortages across the board. Visualizing data at the state and county level helps to expose the lack of resources that healthcare professionals face. By exposing these gaps, hopefully, more informed and timely decisions can be made to drive public health legislation, equip workers on the front lines, and save lives across the U.S.
