COVID Behind Bars or on the Streets Linked to Hospitalization

Incarcerated and homeless individuals presenting to the emergency department required more and longer hospital care for COVID-19 than other infected patients, researchers found in a cross-sectional study.

Comparing emergency department and hospital discharge records among COVID-19 patients, incarceration and homelessness were linked to a higher rate of admission related to COVID-19 (63.5% and 64.5% vs 49.7% in the general population; P<0.001 for both), reported Martha Montgomery, MD, MHS, of the CDC in Atlanta, and colleagues.

Once admitted, patients experiencing incarceration or homelessness also had longer lengths of stay (mean 9 and 11 days vs 8) and more readmissions within 30 days of discharge (5.9% and 8.5% vs 4.6%), compared to general COVID-19 patients, the authors wrote in JAMA Network Open.

Incarcerated patients were more likely to need invasive mechanical ventilation (18.9% vs 14.2%, adjusted rate ratio 1.16, 95% CI 1.04-1.30) and to die (14.2% vs 13.6%; aRR 1.28, 95% CI 1.11-1.47) than other COVID-19 patients.

“Increased lengths of stay and readmission rates highlight the complex factors outside of COVID-19 illness with which PEI [people experiencing incarceration] and PEH [people experiencing homelessness] must contend and support the expansion of medical respite facilities,” the researchers concluded.

“When COVID-19 vaccines became available, advocating and promoting the inmates getting vaccination was stressed,” said Edward Levine, MD, of the Ohio State University Wexner Medical Center in Columbus, which collaborated on COVID-19 protocols for the congregate settings of the Ohio Department of Rehabilitation and Corrections.

Incarcerated individuals receive medical care by law, whereas homeless individuals lack regular healthcare and resources that may distinguish disease burden among them, creating gaps in understanding, the researchers noted.

“This study adds key information on COVID hospitalizations to the now large body of evidence that our pandemic response is failing people exposed to incarceration and homelessness,” said Seth Prins, PhD, of Columbia University in New York City.

For their study, Montgomery and colleagues examined electronic national data from the Premier Healthcare Database on 6,088 homeless and 2,170 incarcerated adults who were seen in the emergency department (ED) or were hospitalized for COVID-19 during April 1, 2020 to June 30, 2021. Use of primary or secondary ICD-10 diagnostic billing codes identified COVID-19 patients across 892 hospitals. Another 624,470 COVID-19 patients seen in the ED or hospitalized for COVID-19 who were not incarcerated or homeless were included as a comparator population.

Most incarcerated patients in the study were male (86%), with a mean age of 51; homeless patients were similar (72% male; mean age 50). Incarcerated and homeless patients frequently had underlying hypertension (49% and 46%), heart disease (27% and 39%), and neurological or musculoskeletal conditions (26% and 35%). Both had fewer in-hospital complications of pneumonia (70% vs 51%) and respiratory failure (52% vs 32%), compared to general COVID-19 patients (78% and 56%, respectively).

Interestingly, homeless patients hospitalized with COVID-19 had less invasive mechanical ventilation use (10%; aRR 0.64, 95% CI 0.58-0.70) and mortality (5.4%, aRR 0.53, 95% CI 0.47-0.59) than the general population.

Factors associated with higher admission rates from the emergency department in both incarcerated and homeless individuals were being male (81% and 70%), younger (median age 56 vs 55), and non-Hispanic Black (25% and 28%, respectively).

“Health systems, public health, and homeless service providers can work together to expand medical respite care options for people experiencing homelessness to reduce and shorten hospitalizations,” Montgomery told MedPage Today in an email relayed through CDC media relations.

There’s more to learn at the population-level, Montgomery acknowledged.

The analysis had several limitations, according to the researchers. Underlying medical conditions were likely underestimated. And, the study relied on medical billing and admission codes to identify patients, which may limit generalization to the broader population and introduce bias.