Hospitalization Linked With Substance Use and Psychiatric Disorders in COVID-19

Hospitalization is more likely for patients with COVID-19 who have psychiatric disorders and substance abuse disorders.

A greater likelihood of hospitalization exists for patients with COVID-19 and with both psychiatric disorder (PD) and substance use disorder (SUD) compared with individuals with COVID-19 and with neither disorder or with either disorder alone, according to study findings published in the Journal of the American Medical Association Psychiatry.

Investigators sought to evaluate the association of comorbid PD and SUD with the probability of hospitalization among patients with COVID-19 in the emergency department (ED). The primary endpoint was any hospitalization determined by billing codes .

They conducted a retrospective cross-sectional study that included more than 1.2 million patients with COVID-19 treated in 1 of 970 US EDs and inpatient hospitals from April 2020 to August 2021. Primary exposures were any past diagnosis of PD (posttraumatic stress disorder (PTSD), major depression or other mood disorder, bipolar disorder, anxiety, attention-deficit/hyperactivity disorder [ADHD], or schizophrenia) and/or SUD (cocaine, sedatives, cannabis, alcohol, stimulants, opioids or other substances). Investigators calculated differences in the probability of hospitalization to evaluate its association with both SUD and PD compared with SUD alone, PD alone, or neither condition. Nicotine dependence was treated as a comorbid covariate vs a primary exposure variable. Independent variables consisted of sex, race and ethnicity, age, primary insurance, and the US Census region of the hospital.

Investigators analyzed discharge data using the Premier PINC AI Healthcare Data Special Release COVID-19 (PHD-SR COVID-19) edition (release date, September 14, 2021), a 48-state, all-payer longitudinal electronic health record database that included 1,274,219 adult patients with COVID-19 at least 18 years of age (mean age, 54.6±19.1 years; 52.4% women) and found 4.6% had an SUD (mean age, 50.1 years; 38.3% women), 18.6% had a PD (mean age 59.0 years; 62.3% women), and 2.3% had both (mean age, 50.4 years; 46.9% women).

Substance use disorders appear to have a greater association than psychiatric disorders with the probability of hospitalization.

There were 567,766 (44.6%) of patients hospitalized directly from the ED. Hospitalized patients vs those not admitted were more likely men (52.4%), older (mean age 63.6 vs 47.4 years), insured by Medicare, living in urban areas, or living in the Northeast census region. Among the hospitalized patients, about three-fourths had a chronic disease vs one-third of patients with COVID-19 who were discharged from an ED.

Alcohol (2.1%), cannabis (1.3%), opioids (1.0%), and poly (≥2) SUDs (0.9%) were the most common SUDs. Anxiety (12.9%), major depression (9.8%), poly (≥2) PDs (6.4%), and schizophrenia (1.4%) were the most prevalent PDs. Among patients with ADHD, PTSD, other mood disorders, or schizophrenia, prevalence of SUD exceeded 21%. Hospitalized patients vs those discharged from the ED were more likely to have a SUD (6.1% vs 3.4%), PD (26.5% vs 12.2%), or both SUD and PD (3.1% vs 1.6%). They found SUD vs PD appeared to have a stronger association with the likelihood of hospitalization.

Investigators noted the probability of hospitalization of those with both SUD and PD was higher by a weighted mean of 20 percentage points (range, 6-36; interquartile range [IQR], 16-25) than those with neither condition, by 12 percentage points (range, -4 to 31; IQR, 8-16) than those with PD alone, and by 4 percentage points (range, -7 to 15; IQR, -2 to 7) than those with SUD alone, based on specific SUD-PD pairs (Q <.05 [P value adjusted for false discovery rate]).

Study limitations include the nature of retrospective cross-sectional design, not designed to be a probability sample, social desirability bias, under-coding, about one-third of patients with secondary discharge of COVID-19 inflating hospitalization numbers, misclassification by including SUD inactive condition or excluding SUD with an undetermined degree of activity, and asymptomatic patients with COVID-19 not identified nor were those treated out-of-network or in a non-ED setting. Additionally, patients were only included if they sought emergency care in 48 states, a shortage of hospital beds may have postponed hospitalization, and a lack of accurate determination of vaccine attenuation though a vaccine was not available during the initial 8 months of the study.

Among “emergency department patients with COVID-19, patients with both psychiatric disorder and substance use disorder had a greater probability of hospitalization, compared with those with psychiatric disorder alone, substance use disorder alone, or neither disorder,” investigators concluded. They added, “Substance use disorders appear to have a greater association than psychiatric disorders with the probability of hospitalization.” They suggest screening and evaluation for SUDs and PDs are necessary to better estimate the likelihood of hospitalization among patients in the ED with COVID-19 and coexisting SUD and PD.


Schieber LZ, Dunphy C, Schieber RA, Lopes-Cardozo B, Moonesinghe R, Guy GP Jr. Hospitalization associated with comorbid psychiatric and substance use disorders among adults with COVID-19 treated in US emergency departments from April 2020 to August 2021. JAMA Psychiatry. Published online February 15, 2023. doi:10.1001/jamapsychiatry.2022.5047