Prevalence of Alcohol Use Characterized by Phosphatidylethanol in Patients With Respiratory Failure Before and During the COVID-19 Pandemic.

Background: Alcohol misuse is overlooked frequently in hospitalized patients, but is common among patients with pneumonia and acute hypoxic respiratory failure. Investigations in hospitalized patients rely heavily on self-report surveys or chart abstraction, which lack sensitivity. Therefore, our understanding of the prevalence of alcohol misuse before and during the COVID-19 pandemic is limited.

Research Question: In critically ill patients with respiratory failure, did the proportion of patients with alcohol misuse, defined by the direct biomarker phosphatidylethanol, vary over a period including the COVID-19 pandemic?

Study Design And Methods: Patients with acute hypoxic respiratory failure receiving mechanical ventilation were enrolled prospectively from 2015 through 2019 (before the pandemic) and from 2020 through 2022 (during the pandemic). Alcohol use data, including Alcohol Use Disorders Identification Test (AUDIT)-C scores, were collected from electronic health records, and phosphatidylethanol presence was assessed at ICU admission. The relationship between clinical variables and phosphatidylethanol values was examined using multivariable ordinal regression. Dichotomized phosphatidylethanol values (≥ 25 ng/mL) defining alcohol misuse were compared with AUDIT-C scores signifying misuse before and during the pandemic, and correlations between log-transformed phosphatidylethanol levels and AUDIT-C scores were evaluated and compared by era. Multiple imputation by chained equations was used to handle missing phosphatidylethanol data.

Results: Compared with patients enrolled before the pandemic (n = 144), patients in the pandemic cohort (n = 92) included a substantially higher proportion with phosphatidylethanol-defined alcohol misuse (38% vs 90%; < .001). In adjusted models, absence of diabetes, positive results for COVID-19, and enrollment during the pandemic each were associated with higher phosphatidylethanol values. The correlation between health care worker-recorded AUDIT-C score and phosphatidylethanol level was significantly lower during the pandemic.

Interpretation: The higher prevalence of phosphatidylethanol-defined alcohol misuse during the pandemic suggests that alcohol consumption increased during this period, identifying alcohol misuse as a potential risk factor for severe COVID-19-associated respiratory failure. Results also suggest that AUDIT-C score may be less useful in characterizing alcohol consumption during high clinical capacity.