ALCOHOL CONSUMPTION AND MISUSE AMONG PATIENTS WITH RESPIRATORY FAILURE BEFORE AND DURING THE COVID-19 PANDEMIC

Alcohol misuse is associated with the development of respiratory failure, including the acute respiratory distress syndrome (ARDS), with attendant poorer outcomes. Epidemiologic data suggest that alcohol consumption increased during the COVID-19 pandemic;however, consumption habits among critically ill patients are not well-described. We hypothesized that the prevalence of alcohol consumption and misuse among patients with respiratory failure would be greater during the pandemic relative to years prior. Patients with respiratory failure requiring mechanical ventilation in a single academic hospital were enrolled from 2015-19 (pre-pandemic) and 2020-21 (pandemic). Data regarding demographics and alcohol use were obtained from the medical record. Phosphatidylethanol (PEth) was measured in a subset of patients in red blood cells collected within 48 hours of intubation. Patients were characterized as follows: (1) likely drinking: patient or proxy endorsed any drinking on the Alcohol Use Disorders Identification Test (AUDIT) or in social history;or patient was admitted for an alcohol-related diagnosis;or patient had detectable blood alcohol or PEth;(2) likely alcohol misuse: PEth ≥ 250;or AUDIT-C ≥3 (women) or ≥4 (men);or AUDIT ≥5 (women) or ≥ 8 (men);or patient was admitted for an alcohol-related diagnosis;(3) likely severe alcohol misuse: PEth ≥400. Univariable statistics were utilized as appropriate (Fisher's Exact Test, Kruskal- Wallis Test). The prevalence of likely drinking in the study population (n = 195) varied over time (p <0.0001). When pandemic (n = 67) and pre-pandemic patients (n = 128) were compared, the prevalence of likely drinking (93% vs 58%, p <0.0001), likely alcohol misuse (49% vs 33%, p <0.03) and likely severe alcohol misuse (32% vs 10%, p <0.003) were all greater during the pandemic. Among likely drinking patients with available PEth data (n = 97), median PEth was significantly higher during the pandemic compared to pre-pandemic years (251 [135-702] versus 87 [16-374], p = 0.002). Among the subset of patients enrolled during the pandemic, the prevalence of likely current drinking was 100% in those with COVID-19 compared to 76% among non-COVID patients (p = 0.002), but alcohol misuse did not differ. We conclude that alcohol consumption, and alcohol misuse, have increased during the pandemic in this patient cohort. These changes in drinking patterns are concerning and warrant additional research regarding their impact on susceptibility and outcomes for respiratory failure.

IN-HOSPITAL DELIRIUM IS ASSOCIATED WITH INCREASED MORTALITY RISK IN PATIENTS WITH ADMITTED WITH COVID-19 AND A HISTORY OF ALCOHOL MISUSE

Introduction: In-hospital delirium is a risk factor for worse critical care outcomes. Alcohol misuse contributes to increased risk of critical illness and greater pneumonia severity. We sought to determine if alcohol misuse was associated with increased risk of critical illness or complications of critical illness including delirium amongst hospitalized COVID-19 patients. Methods: Retrospective study across 12 University of Colorado hospitals (March 2020-April 2021). Adults with a COVID-19 diagnosis were included. Alcohol misuse was defined by validated ICD-10 codes (F10.1∗, F10.2∗, F10.9∗). Multivariable, mixed effects logistic regression models were used to estimate effects of alcohol misuse adjusting for age, sex, body mass index, diabetes and liver disease. Results: We included 6,454 hospitalized COVID-19 patients aged 60 (SD 18 years) with 53% male. Twenty-four percent (n = 1561) required ICU admission, 14.4% (n = 927) endotracheal intubation with a median ICU stay of 6 days [IQR 2 to 15] and duration of mechanical ventilation of 10 days [IQR 5 to 18]. Ten percent of the cohort (n = 644) died in hospital. Delirium was identified in 4% (n = 254) and 257 patients (4%) had alcohol misuse. Patients with misuse were younger (52 vs 60 p < 0.01), men (77 vs 52%, p < 0.01), had more co-morbid liver disease (p-value < 0.01) and received less remdesivir (29 vs 50%, p < 0.01). There was no difference in dexamethasone or vasopressor use. In adjusted analyses, alcohol misuse was associated with 54% increased odds of ICU admission (aOR, 1.54, 95%CI 1.13 to 2.09, p < 0.01) and 46% increased odds of mechanical ventilation (aOR. 1.46, 95% CI 1.03 to 2.08, p < 0.01) compared to patients without misuse. There was a significant interaction between misuse and in-hospital delirium on in-hospital death with odds of death highest amongst patients with both alcohol misuse and identified delirium. Conclusions: Alcohol misuse was associated with increased need for critical care including ICU admission and mechanical ventilation. Delirium was an important modifiable risk factor for worse outcomes in hospitalized patients with alcohol misuse, and the odds of in-hospital death were significantly increased in the presence of bothmisuse and delirium.

The impact of the colorado state-wide lockdown on non-COVID-19 related ICU admissions and mortality

RATIONALE: The COVID-19 pandemic has rapidly become the most significant worldwide public health crisis in the modern era. Like other states around the country, the state of Colorado instituted a statewide lockdown to combat increasing case and hospitalization rates for COVID-19 throughout the state. The impact of this mandate on the ICU admission rates and outcomes of other medical problems has never been investigated. Our study aimed to determine the effects of stay-at-home orders on outcomes for other diagnoses by analyzing ICU admission rates and outcomes of patients presenting to the ICU for non-COVID related issues before, during, and after the statewide mandate. METHODS: We performed a retrospective analysis of all ICU admissions in three phases: before (2 months prior), during, and 1.5 months after the statewide lockdown (March 26 to April 27, 2020). We included all patients admitted to the University of Colorado Health System hospitals ICUs within this defined time period. A time-to-event analysis was performed with the date of index ICU stay set as time zero. Baseline characteristics were obtained. Primary outcome measures were 28-day mortality and all-time mortality. Kaplan-Meier curves were used to estimate survival probabilities, while Cox regression and multivariable logistic regression were utilized to model phase-specific mortality controlling for comorbidities, demographics, and admission diagnoses. Counts of typical ICU admission diagnoses were also analyzed to determine any changes across lockdown periods. RESULTS: 9201 total ICU admissions occurred, of which 8154 (88.6%) were non-COVID-19 related. Approximately 57.4% were male with a mean age of 60.4 years. 28-day mortality rates for non-COVID-19 ICU admissions were 475 (11.0%), 127 (13.8%), and 306 (10.5%) before, during, and after the lockdown, respectively. The increased mortality during lockdown persisted after adjustment for comorbidities and demographics (HR=1.23, 95% CI, 1.007 to 1.512, p = 0.043). Acute respiratory failure was the most common diagnosis in each time period, and increased during lockdown (p<0.001). Admissions for sepsis increased during lockdown and decreased after (p = 0.001);myocardial infarction (MI) admission decreased during lockdown but increased after (p = 0.014);and alcohol withdrawal (AW) admission increased both during and after lockdown (p < 0.001). CONCLUSIONS: For non-COVID-19 related ICU admissions, the mortality rate increased during the state-wide shutdown but decreased after shutdown, although this difference became insignificant after controlling for patient admission diagnoses. Admission diagnoses also differed with more admissions for sepsis and AW during lockdown and more admissions for MI and AW after lockdown.