COVID-19 Patients With Longer Hospital Stays May Have Better Psychiatric Outcomes

Patients who had been previously diagnosed with psychiatric disorders as well as women reported the highest increase in psychopathology dimension scores at a 1-month follow-up after hospital treatment, researchers discovered in a cross-sectional cohort study published in Brain, Behavior, and Immunity. Longer hospitalization inversely correlated with scores on several psychopathology dimension questionnaires, the study authors also found. About half (55.7%) of patients scored in at least 1 psychopathological dimension’s clinical ranges at the follow-up.

From April 6 to June 9, the researchers investigated the psychopathological impact of COVID-19 on adult survivors at 1-month follow-up by screening 402 adult subjects (265 males and 140 women, aged 18 to 87 years) who survived COVID-19 for psychiatric symptoms at a hospital in Milan. The patients had each undergone evaluation –clinical evaluation, electrocardiogram, hemogasanalysis, and hematological analysis – at the emergency department and then were either admitted for severe pneumonia (300 patients, for a hospital stay of 15.31 ± 10.32 days) or managed at home (102). The researchers performed a psychiatric assessment of each patient 31.29 ± 15.7 days after the date of discharge from treatment at the hospital or 28.56 ± 11.73 days after departure from the emergency department for at-home care.

In the psychiatric assessments, the researchers conducted a clinical interview and collected sociodemographic and clinical data from the patient and used self-report questionnaires to measure current psychopathology. They used the Impact of Events Scale-Revised (IES-R), PTSD Checklist for DSM-5 (PCL-5), Zung Self-Rating Depression Scale (ZSDS), 13-item Beck’s Depression Inventory (BDI-13), State-Trait Anxiety Inventory form Y (STAI-Y), Medical Outcomes Study Sleep Scale (MOS-SS), Women’s Health Initiative Insomnia Rating Scale (WHIIRS), and Obsessive-Compulsive Inventory (OCI). They considered the scores to be in the pathological range if they were higher than the following cutoff scores: IES-R ≥ 33; PCL-5 ≥ 33; ZSDS index ≥ 50; BDI-13 ≥ 9; STAI-state ≥ 40; STAI-trait ≥ 40; WHIIRS ≥ 9; OCI ≥ 21.

They performed statistical analysis to explore the effects of sex, hospitalization, and previous history of psychiatric illness on symptoms’ severity using statistical analysis and analyzed the impact these factors would have on self-report scores.

The subjects’ self-ratings indicated 55.7% of them were in the clinical range of at least 1 of the following: PTSD (according to the IES-R or PCL-5), depression (according to the ZSDS or BDI-13), anxiety (according to the STAI-Y) and obsessive-compulsive (OC) (according to the OCI), the researchers said. Many of the subjects scored in the clinical range of multiple dimensions: 36.8% in 2, 20.6% in 3, and 10% in 4. Nearly 5% (4.5%) scored on the BDI suicide item.

The researchers found that women more frequently reported higher scores on the IES-R (χ2 = 54.98, P = < .001), the PCL-5 (χ2 = 17.91, P = < .001), the ZSDS (χ2 = 45.45, P < .001), the BDI (χ2 = 15.13, P < .001), clinical state anxiety (χ2 = 42.15, P = < .001), clinical trait anxiety (χ2 = 36.11, P = < 0.001) and sleep disturbances (χ2 = 15.70, P = < .001).

With univariate testing, they also found that the patients who had previously been diagnosed with a psychiatric disorder (36 with major depressive disorder, 28 with generalized anxiety disorder, 20 with panic attack disorder, 5 with bipolar disorder, 5 with social phobia, 3 with eating disorders and 4 with other disorders) underwent a more statistically significant impact on mental health (IES-R: β = 0.274, F = 25.86, degrees of freedom 1,273, P < .0001; PCL-5: β = 0.267, F = 24.11, P < .0001; ZSDS: β = 0.237, F = 19.25, P < .0001; BDI: β = 0.221, F = 14.74, P = .0002; STAI-Y: β = 0.238, F = 19.39, P < .0001; OCI: β = 0.190, F = 9.98, P = .0018; IRS: β = 0.123, F = 4.29, P = .0393; MOS: β = 0.099, F = 0.96, P = .329).

Patients who stayed longer at the hospital more frequently reported lower scores on the PCL-5 (r = −0.15, P = .019), ZSDS (r = −0.16, P = .009), BDI-13 (r = −0.13, P = .036), STAI-Y state (r = −0.18, P = .003), and OCI (r = −0.12, P = .044), and older patients tended to have lower scores on BDI (r = −0.12, P = .018) and MOS (r = −0.18, P = .001).

The researchers said that the inherent inability to interpret the cross-sectional study’s data collection for causality is a limitation of the study.

“With regard to the risk factor related to psychopathology, consistently with previous epidemiological studies, we have found that [women], and patients with positive previous psychiatric diagnoses, suffered more in all psychopathological dimensions,” the researchers said. “Moreover, outpatients showed increased anxiety and sleep disturbances, while the duration of hospitalization inversely correlated with PTSD, depression, anxiety, and OC symptomatology. Also considering the worse severity of COVID-19 in hospitalized patients, this observation suggests that less healthcare support could have increased the social isolation and loneliness typical of COVID-19 pandemics, thus inducing more psychopathology after remission.”


Mazza MG, De Lorenzo R, Conte C, et al. Anxiety and depression in COVID-19 survivors: Role of inflammatory and clinical predictors. Brain, Behavior, and Immunity. 2020;89(October):594-600. doi: 10.1016/j.bbi.2020.07.037