With delta surging, many hospitals around the country are again under strain. The federal government regularly releases detailed hospital-level data which show how many inpatient and ICU beds are in use on a weekly basis — and what portion of them are occupied by COVID-19 patients.

Using an analysis from the University of Minnesota's COVID-19 Hospitalization Tracking Project, NPR has created a tool that allows you to see how your local hospital and your county overall are faring.

Look up the hospitals where you live to find out how many beds are filled with COVID-19 patients, and the average for hospitals in your county.

The ratio of COVID-19 hospitalizations to total beds gives a picture of how much strain a hospital is under.

That ratio is concerning when it rises above 10%, hospital capacity experts told NPR. Anything above 20% represents "extreme stress" for the hospital, according to a framework developed by the Institute for Health Metrics and Evaluation at the University of Washington.

ICU capacity is another important indicator of hospital strain. Those thresholds are higher: When 30% or more of ICU beds are filled by COVID-19 patients, it suggests a hospital is operating under "high stress." When 60% or more are full, it's considered at "extreme stress," according to IHME.

This federal hospitalization data, which is available going back to August 2020, also provides a historical perspective on the stress level over time in each state. See how your state is faring now:

All of these indicators vary depending on the size of a given hospital and its resources. The number of hospitals reporting their data also varies each week, though more have started reporting since the federal government started collecting this data.

Methodology

Data used for this story come from the Department of Health and Human Services and University of Minnesota COVID-19 Hospitalization Tracking Project.

Hospital stress levels reflect the percent of inpatient beds occupied by COVID-19 patients at reporting hospitals. Stress levels are based on a framework developed by the Institute for Health Metrics and Evaluation at the University of Washington.

Averages calculated by summing the seven-day average for each individual hospital, then dividing by the number of hospitals reporting data in a county. "No data" may indicate a sample size of fewer than four patients or hospitals that did not report statistics to the Department of Health and Human Services.

Hospitals in Louisiana are not included in the data for the week of Aug. 27-Sept. 2 due to possible reporting anomalies related to Hurricane Ida.

Copyright NPR 2021.