New hospital-based data contradicts HUD estimates on homelessness

Authored by today.uic.edu and submitted by savvas_lampridis

Researchers at the University of Illinois at Chicago report that in Illinois hospital visits associated with homelessness have tripled since 2011.

Their findings, which are published in the American Journal of Public Health, also show that beginning in 2016, annual conservative estimates of homelessness using hospital-based data have exceeded similar estimates from the U.S. Department of Housing and Urban Development, or HUD.

“The data suggests that homelessness may be increasing, despite official HUD estimates showing a substantial decrease in homelessness,” said Dana Madigan, a doctoral student at UIC and first author of the study.

Using discharge records from 2011 to 2018, Madigan and her colleagues identified visits to hospital emergency departments in Illinois in which patients indicated they were affected by a lack of housing or homelessness.

Their analysis included more than 200 hospitals, representing 96.5% of all patient admissions statewide in this period.

By matching patient date of birth, gender and race/ethnicity, they estimated the number of unique homeless individuals that visited hospitals in Illinois. They compared their estimates with HUD’s annual estimates of homelessness based on Point-in-Time counts for Illinois, taken on a single night in January.

The most conservative estimates of homelessness using hospital data of unique emergency department visitors increased annually from 6,613 in 2011 to 15,815 in 2018. HUD estimates showed some year-over-year increases, but an overall 24% decrease from 14,009 in 2011 to 10,643 in 2018.

“At the very least, this suggests that state and national estimates of homelessness will be more accurate if they also take into account readily available information from hospitals,” Madigan said.

Lee Friedman, senior author of the study, said that without accurate data, states will not get adequate resources to address homelessness.

“Homelessness is an ‘invisible’ condition that is undoubtedly hard to measure, but that doesn’t mean state and federal agencies get a pass on doing a poor job,” said Friedman, UIC associate professor of environmental and occupational health sciences at the School of Public Health. “HUD’s numbers, which are the primary driver of public policy, may be seriously flawed, and this study shows that hospital data, which is available in every state, is a feasible data system to incorporate into estimates of homelessness to improve accuracy.”

Linda Forst, UIC professor of environmental and occupations health sciences, is a co-author of the study, which was supported in part by the National Institute for Occupational Safety and Health Illinois Education and Research Center (T42OH008672).

c0y0t3_sly on January 19th, 2020 at 19:12 UTC »

HUD depends very heavily on a one day "census" of the homeless for these numbers. It's actually happening next week, and if you want to see how this sausage gets made you can almost certainly volunteer in your home town to help conduct it. Short answer: trying to use a field census to count homelessness is a fool's errand anyway, and they absolutely should be doing it using distributed community data sources like this.

EbilSmurfs on January 19th, 2020 at 18:12 UTC »

I'm starting to get worried how one should handle US Gov. data. The Social Costs of CO2 are up to 10x higher than the EPA claims, and the US is lacking proper filling of many science based posts including NASA.

So what should one do when dealing with the US Government and it's science? It seems clear to me that at the moment the US Government is pretty, well, bad at science. When, or even should, we go back to trusting it? This is concerning, and I would like to see a way forward where we go back to having decent science published by the US Government.

discomuffin94 on January 19th, 2020 at 17:53 UTC »

It's very easy to scew statistics in your favor. Never trust them unless the data set and method are explicitly detailed