Demographic Trends in Emergency Department Visits for Psychiatric Concerns During the COVID-19 Pandemic.

Objective: To describe changes in emergency department (ED) psychiatric visits during the pandemic in both rural and nonrural regions in the United States. Methods: This cohort study was performed across 22 EDs in the Midwest and Southern United States from January 1, 2019 to April 22, 2021. Prevalence of psychiatric visits before and after the COVID-19 pandemic, defined as starting on March 1, 2020, were compared. Psychiatric and nonpsychiatric visits were defined on the basis of primary clinician-assigned diagnosis. The primary end point was average daily visits normalized to the average daily visit count before the pandemic, labeled as relative mean daily visits (RMDVs). Results: Psychiatric visits decreased by 9% [RMDVs, 0.91; 95% confidence interval (CI), 0.89-0.93] during the pandemic period, whereas nonpsychiatric visits decreased by 17% (RMDVs, 0.83; 95% CI, 0.81-0.84). Black patients were the only demographic group with a significant increase in psychiatric visits during the pandemic (RMDVs, 1.12; 95% CI, 1.04-1.19). Periods of outbreaks of psychiatric emergencies were identified in most demographic groups, including among male and pediatric patients. However, the outbreaks detected among Black patients sustained the longest at 6 months. Unlike older adults who experienced outbreaks in the spring and fall of 2020, outbreaks among pediatric patients were detected later in 2021. Conclusion: In this multisite study, total ED visits declined during the pandemic; however, psychiatric visits declined less than nonpsychiatric visits. Black patients experienced a greater increase in psychiatric emergencies than other demographic groups. There is also a concern for increasing outbreaks of pediatric psychiatric visits as the pandemic progresses.

Practical development and operationalization of a 12-hour hospital census prediction algorithm.

Hospital census prediction has well-described implications for efficient hospital resource utilization, and recent issues with hospital crowding due to CoVID-19 have emphasized the importance of this task. Our team has been leading an institutional effort to develop machine-learning models that can predict hospital census 12 hours into the future. We describe our efforts at developing accurate empirical models for this task. Ultimately, with limited resources and time, we were able to develop simple yet useful models for 12-hour census prediction and design a dashboard application to display this output to our hospital's decision-makers. Specifically, we found that linear models with ElasticNet regularization performed well for this task with relative 95% error of +/- 3.4% and that this work could be completed in approximately 7 months.

Cardiac Abnormalities in COVID-19 and Relationship to Outcome.

OBJECTIVE: To characterize the clinical and transthoracic echocardiographic features and 30-day outcomes of hospitalized patients with coronavirus disease 2019 (COVID-19). METHODS: Retrospective cohort study that included consecutive inpatients with COVID-19 infection who underwent clinically indicated transthoracic echocardiography at 10 sites in the Mayo Clinic Health System between March 10 and August 5, 2020. Echocardiography was performed at bedside by cardiac sonographers according to an abbreviated protocol. Echocardiographic results, demographic characteristics, laboratory findings, and clinical outcomes were analyzed. RESULTS: There were 179 patients, aged 59.8±16.9 years and 111 (62%) men; events within 30 days occurred in 70 (39%) patients, including prolonged hospitalization in 43 (24%) and death in 27 (15%). Echocardiographic abnormalities included left ventricular ejection fraction less than 50% in 29 (16%), regional wall motion abnormalities in 26 (15%), and right ventricular systolic pressure (RVSP) of 35 or greater mm Hg in 44 (44%) of 101 in whom it was measured. Myocardial injury, defined as the presence of significant troponin level elevation accompanied by new ventricular dysfunction or electrocardiographic abnormalities, was present in 13 (7%). Prior echocardiography was available in 36 (20%) patients and pre-existing abnormalities were seen in 28 (78%) of these. In a multivariable age-adjusted model, area under the curve of 0.81, prior cardiovascular disease, troponin level, D-dimer level, and RVSP were related to events at 30 days. CONCLUSION: Bedside Doppler assessment of RVSP appears promising for short-term risk stratification in hospitalized patients with COVID-19 infection undergoing clinically indicated echocardiography. Pre-existing echocardiographic abnormalities were common; caution should be exercised in attributing such abnormalities to the COVID-19 infection in this comorbid patient population.