Current hospitalizations have been one of the most helpful and reliable metrics for understanding the pandemic’s effects across the country. This metric is not subject to the same kinds of backlogs and lags in case, testing, and death data reporting, which makes it indispensable for understanding outbreaks in near real time. But like all the metrics in our dataset compiled from states, it has definitional nuances and intricacies that can affect its interpretation.
Today, as part of our wind-down process, we are releasing a one-time snapshot of our research into states’ current hospitalization definitions as of March 22, 2021. As we have previously reported, US states and territories use inconsistent data definitions for currently hospitalized COVID-19 patients. As a result, we’ve had to maintain structured notes—which we internally call “annotations”—on variations in definitions of hospitalization information across states in order to support our data entry and analysis.
Recently, we released our annotations on state case, death and test metrics. Hospitalization annotations played a similar role in supporting our decisions about how to capture state data and undergirding our research into federal hospitalizations datasets. We hope it will be useful to data users and researchers.
You can access the annotations themselves on our Data Annotations page. Read on to learn more about what kinds of variation we saw in COVID-19 current hospitalization reporting.
One of the major variations we observed in state-level data was whether states tracked patients with confirmed COVID-19, suspected COVID-19, or both.
According to the Centers for Disease Control and Prevention (CDC) and the Council of State and Territorial Epidemiologists (CSTE), a confirmed positive patient is a person who has laboratory-confirmed COVID-19 using a molecular amplification test.
Per guidance from the US Department of Health and Human Services (HHS) to hospitals and laboratories, a patient with suspected COVID-19 is defined as “a person who is being managed as though he/she has COVID-19 because of signs and symptoms suggestive of COVID-19 as described by CDC’s Guidance but does not have a laboratory positive COVID-19 test result.” This includes patients who have not been tested, are awaiting test results, and those who test negative but still exhibit signs and symptoms of COVID-19.
While some states report all currently hospitalized patients with confirmed and suspected cases of COVID-19, others have more restrictive definitions limiting their counts to laboratory-confirmed cases. There are also some states that use varying terminology for a subset of patients who do not yet have confirmed cases, but are hospitalized patients under investigation (PUI), pending test results, or presumed positive COVID-19 cases.
Data definitions were not available for many states. For example, Florida’s COVID-19 hospitalization dashboard lists that the count is for patients with a primary diagnosis of COVID-19. Our research suggests that Florida is counting all positive patients with COVID-19 symptoms, but the case types included in this remain unclear without the state providing a data definition. Another example is New Mexico. Their dashboard reports the number of “current hospitalizations,” but it does not include a data definition for what patients are included in this count.
Finally, some states list unclear data definitions. For example, North Dakota reports patients that are “hospitalized due to COVID” and the definition “includes people who are hospitalized because of COVID.” Even the eight states that report positive patients do not list a definition for what case types (confirmed vs. probable) and/or test types (PCR vs. antigen) are used to classify these positive patients.
You can find out how a state reports their current COVID-19 hospitalizations in the “State Subgroup Labels” column of our hospitalization annotations. This field lists what the state calls the subgroups it includes in its current hospitalizations metrics. If the annotation is listed as “unclear,” this indicates a data definition was either unavailable or was missing information. “Not reported” indicates that a state does not report current COVID-19 hospitalizations.
Another major source of variation in hospitalizations definitions is whether states track adult patients, pediatric patients, or both. The vast majority of states are unclear about the populations in which they are tracking current hospitalizations. The “Population” column of the annotations table lists what information we do have about the population included in a state’s currently hospitalized COVID-19 metric.
Our research also reviewed whether states were reporting their positive patients separately from patients who had not yet tested positive or whether these patients were lumped together in a topline current hospitalization metric. We found that of the 23 jurisdictions that report both metrics, 15 provide separate counts of each subset of patients and eight lump the patients into one topline metric.
To determine whether a state lumps their hospitalization metrics, review the “Cases Reporting” column in the annotations table. The cell will include the word “lumped” if the state-reported metric is lumped together. If the cell lists “unclear,” this means the data definition for the state either wasn’t provided or was unclear.
HHS hospitalization data, unlike state data, is standardized and complete
When using state-reported data, it can be tedious to make comparisons between states with varying hospitalization data definitions. Without the ability to separate out equivalent metrics, such as confirmed patients, the task becomes nearly impossible. And state reporting can change quickly—meaning our research will likely fall out of date soon.
Fortunately, the HHS now publishes state and facility-level hospitalization datasets disaggregated by confirmed and suspected adult and pediatric patients. You can learn more about how to use the federal data in the hospitalizations installment of our federal COVID-19 data 101 series.
Special thanks to Hannah Hoffman, Daniel Lin, and Kara Schechtman for additional research support.
More “Hospitalization and Death Data” posts
Looking back on a year of collecting COVID-19 data, here’s a summary of the tools we automated to make our data entry smoother and why we ultimately relied on manual data collection.
As The COVID Tracking Project comes to a close, here’s a summary of how states reported data on the five major COVID-19 metrics we tracked—tests, cases, deaths, hospitalizations, and recoveries—and how reporting complexities shaped the data.
During the worst parts of the COVID-19 pandemic, the United States struggled to keep up with COVID-19 death counts.