Impact of the COVID-19 Pandemic on Emergency Department Visits

United States, January 1, 2019-May 30, 2020

Kathleen P. Hartnett, PhD; Aaron Kite-Powell, MS; Jourdan DeVies, MS; Michael A. Coletta, MPH; Tegan K. Boehmer, PhD; Jennifer Adjemian, PhD; Adi V. Gundlapalli, MD, PhD

Disclosures

Morbidity and Mortality Weekly Report. 2020;69(23):699-704. 

In This Article

Abstract and Introduction

Introduction

On March 13, 2020, the United States declared a national emergency to combat coronavirus disease 2019 (COVID-19). As the number of persons hospitalized with COVID-19 increased, early reports from Austria,[1] Hong Kong,[2] Italy,[3] and California[4] suggested sharp drops in the numbers of persons seeking emergency medical care for other reasons. To quantify the effect of COVID-19 on U.S. emergency department (ED) visits, CDC compared the volume of ED visits during four weeks early in the pandemic March 29–April 25, 2020 (weeks 14 to 17; the early pandemic period) to that during March 31–April 27, 2019 (the comparison period). During the early pandemic period, the total number of U.S. ED visits was 42% lower than during the same period a year earlier, with the largest declines in visits in persons aged ≤14 years, females, and the Northeast region. Health messages that reinforce the importance of immediately seeking care for symptoms of serious conditions, such as myocardial infarction, are needed. To minimize SARS-CoV-2, the virus that causes COVID-19, transmission risk and address public concerns about visiting the ED during the pandemic, CDC recommends continued use of virtual visits and triage help lines and adherence to CDC infection control guidance.

To assess trends in ED visits during the pandemic, CDC analyzed data from the National Syndromic Surveillance Program (NSSP), a collaborative network developed and maintained by CDC, state and local health departments, and academic and private sector health partners to collect electronic health data in real time. The national data in NSSP includes ED visits from a subset of hospitals in 47 states (all but Hawaii, South Dakota, and Wyoming), capturing approximately 73% of ED visits in the United States able to be analyzed at the national level. During the most recent week, 3,552 EDs reported data. Total ED visit volume, as well as patient age, sex, region, and reason for visit were analyzed.

Weekly number of ED visits were examined during January 1, 2019–May 30, 2020. In addition, ED visits during two 4-week periods were compared using mean differences and ratios. The change in mean visits per week during the early pandemic period and the comparison period was calculated as the mean difference in total visits in a diagnostic category between the two periods, divided by 4 weeks ([visits in diagnostic category {early pandemic period} – visits in diagnostic category {comparison period}]/4). The visit prevalence ratio (PR) was calculated for each diagnostic category as the proportion of ED visits during the early pandemic period divided by the proportion of visits during the comparison period ([visits in category {early pandemic period}/all visits {early pandemic period}]/[visits in category {comparison period}/all visits {comparison period}]). All analyses were conducted using R software (version 3.6.0; R Foundation).

Reason for visit was analyzed using a subset of records that had at least one specific, billable International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) code. In addition to Hawaii, South Dakota, and Wyoming, four states (Florida, Louisiana, New York outside New York City, and Oklahoma), two California counties reporting to the NSSP (Santa Cruz and Solano), and the District of Columbia were also excluded from the diagnostic code analysis because they did not report diagnostic codes during both periods or had differences in completeness of codes between 2019 and 2020. Among eligible visits for the diagnostic code analysis, 20.3% without a valid ICD-10-CM code were excluded. ED visits were categorized using the Clinical Classifications Software Refined tool (version 2020.2; Healthcare Cost and Utilization Project), which combines ICD-10-CM codes into clinically meaningful groups.[5] A visit with multiple ICD-10-CM codes could be included in multiple categories; for example, a visit by a patient with diabetes and hypertension would be included in the category for diabetes and the category for hypertension. Because COVID-19 is not yet classified in this tool, a custom category, defined as any visit with the ICD-10-CM code for confirmed COVID-19 diagnosis (U07.1), was created.[6] The analysis was limited to the top 200 diagnostic categories during each period.

The lowest number of visits reported to NSSP occurred during April 12–18, 2020 (week 16). Although visits have increased since the nadir, the most recent complete week (May 24–30, week 22) remained 26% below the corresponding week in 2019 (Figure 1). The number of ED visits decreased 42%, from a mean of 2,099,734 per week during March 31–April 27, 2019, to a mean of 1,220,211 per week during the early pandemic period of March 29–April 25, 2020. Visits declined for every age group (Figure 2), with the largest proportional declines in visits by children aged ≤10 years (72%) and 11–14 years (71%). Declines in ED visits varied by U.S. Department of Health and Human Services region,* with the largest declines in the Northeast (Region 1, 49%) and in the region that includes New Jersey and New York (Region 2, 48%) (Figure 2). Visits declined 37% among males and 45% among females across all NSSP EDs between the comparison and early pandemic periods.

Figure 1.

Weekly number of emergency department (ED) visits — National Syndromic Surveillance Program, United States,* January 1, 2019– May 30, 2020
*Hawaii, South Dakota, and Wyoming are not included.
Vertical lines indicate the beginning and end of the 4-week coronavirus disease 2019 (COVID-19) early pandemic period (March 29–April 25, 2020) and the comparison period (March 31–April 27, 2019).

Figure 2.

Emergency department (ED) visits, by age group (A) and U.S. Department of Health and Human Services (HHS) region* (B) — National Syndromic Surveillance Program, United States, March 31–April 27, 2019 (comparison period) and March 29–April 25, 2020 (early pandemic period)
*Region 1: Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont; Region 2: New Jersey and New York; Region 3: Delaware, District of Columbia, Maryland, Pennsylvania, Virginia, and West Virginia; Region 4: Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South Carolina, and Tennessee; Region 5: Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin; Region 6: Arkansas, Louisiana, New Mexico, Oklahoma, and Texas; Region 7: Iowa, Kansas, Missouri, and Nebraska; Region 8: Colorado, Montana, North Dakota, and Utah; Region 9: Arizona, California, and Nevada; Region 10: Alaska, Idaho, Oregon, and Washington.
Hawaii, South Dakota, and Wyoming are not included.

Among all ages, an increase of >100 mean visits per week from the comparison period to the early pandemic period occurred in eight of the top 200 diagnostic categories (Table). These included 1) exposure, encounters, screening, or contact with infectious disease (mean increase 18,834 visits per week); 2) COVID-19 (17,774); 3) other general signs and symptoms (4,532); 4) pneumonia not caused by tuberculosis (3,911); 5) other specified and unspecified lower respiratory disease (1,506); 6) respiratory failure, insufficiency, or arrest (776); 7) cardiac arrest and ventricular fibrillation (472); and 8) socioeconomic or psychosocial factors (354). The largest declines were in visits for abdominal pain and other digestive or abdomen signs and symptoms (–66,456), musculoskeletal pain excluding low back pain (–52,150), essential hypertension (–45,184), nausea and vomiting (–38,536), other specified upper respiratory infections (–36,189), sprains and strains (–33,709), and superficial injuries (–30,918). Visits for nonspecific chest pain were also among the top 20 diagnostic categories for which visits decreased (–24,258). Although not in the top 20 declining diagnoses, visits for acute myocardial infarction also declined (–1,156).

During the early pandemic period, the proportion of ED visits for exposure, encounters, screening, or contact with infectious disease compared with total visits was nearly four times as large as during the comparison period (Table) (prevalence ratio [PR] = 3.79, 95% confidence interval [CI] = 3.76–3.83). The other diagnostic categories with the highest proportions of visits during the early pandemic compared with the comparison period were other specified and unspecified lower respiratory disease, which did not include influenza, pneumonia, asthma, or bronchitis (PR = 1.99; 95% CI = 1.96–2.02), cardiac arrest and ventricular fibrillation (PR = 1.98; 95% CI = 1.93–2.03), and pneumonia not caused by tuberculosis (PR = 1.91; 95% CI = 1.90–1.93). Diagnostic categories that were recorded less commonly during the early pandemic period included influenza (PR = 0.16; 95% CI = 0.15–0.16), no immunization or underimmunization (PR = 0.28; 95% CI = 0.27–0.30), otitis media (PR = 0.35; 95% CI = 0.34–0.36), and neoplasm-related encounters (PR = 0.40; 95% CI = 0.39–0.42).

In the 2019 comparison period, 12% of all ED visits were in children aged ≤10 years old, compared with 6% during the early pandemic period. Among children aged ≤10 years, the largest declines were in visits for influenza (97% decrease), otitis media (85%), other specified upper respiratory conditions (84%), nausea and vomiting (84%), asthma (84%), viral infection (79%), respiratory signs and symptoms (78%), abdominal pain and other digestive or abdomen symptoms (78%), and fever (72%). Mean weekly visits with confirmed COVID-19 diagnoses and screening for infectious disease during the early pandemic period were lower among children than among adults. Among all ages, the diagnostic categories with the largest changes (abdominal pain and other digestive or abdomen signs and symptoms, musculoskeletal pain, and essential hypertension) were the same in males and females, but declines in those categories were larger in females than males. Females also had large declines in visits for urinary tract infections (–19,833 mean weekly visits).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....