The Premier Healthcare Database, Special COVID-19 Release (PHD-SR), is a US, hospital-based, all-payer database used by the Centers for Disease Control and Prevention to inform COVID-19 response activities (5,6). The database contains deidentified records from >1,000 nongovernment, community, and teaching hospitals that contributed inpatient data during the analytic period. We used diagnosis codes from the International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM), listed for each hospitalization and identified hospitalizations involving fungal infections (fungal hospitalizations) and COVID-19 (COVID-19 hospitalizations) during January 1, 2019–December 31, 2021 (Appendix Table 1). We defined COVID-19–associated fungal hospitalizations as those in which both a COVID-19 and fungal infection diagnosis were listed during the same hospitalization.

We estimated annual hospitalization rates (per 10,000 population) by fungal infection type and calculated average annual percentage change during 2019–2021. For COVID-19–associated fungal hospitalizations (2020–2021 only), we calculated hospitalization rates per 10,000 COVID-19 hospitalizations. We stratified 2020–2021 fungal hospitalizations by COVID-19 association and fungal infection type and compared patient demographics, US hospital census regions and urban–rural classifications (https://www.cdc.gov/nchs/data_access/urban_rural.htm), lengths of hospital stays, intensive care unit (ICU) admissions, invasive mechanical ventilation (IMV) receipt, and in-hospital deaths. We assessed annual trends in fungal hospitalizations by using Cochran-Armitage tests and compared fungal hospitalizations according to COVID-19 status by using χ² tests.

During 2019–2021, a total of 59,212 fungal hospitalizations were identified in the PHD-SR. Rates of fungal hospitalizations (per 10,000 hospitalizations) increased from 22.3 in 2019 to 25.0 in 2020 and 26.8 in 2021 (p<0.01), representing an average annual percentage change of 8.5% (Table 1). Average annual rates of hospitalization significantly increased for each fungal infection, except for those caused by Pneumocystis spp., Cryptococcus spp., and other specified fungi (Table 1).

During 2020–2021, a total of 5,288 (13.4%) of 39,423 fungal hospitalizations were COVID-19–associated. Rates of COVID-19–associated fungal hospitalizations (per 10,000 COVID-19 hospitalizations) increased by 24.9% (43.1% to 57.4%; p<0.01). Annual rates increased significantly for COVID-19–associated fungal hospitalizations involving blastomycosis (0.2 to 0.5 [65.6% change]; p<0.01), aspergillosis (7.9 to 18.9 [58.2% change]; p<0.01), mucormycosis (0.7 to 1.1 [39.8% change]; p = 0.02), histoplasmosis (1.1 to 1.6 [32.1% change]; p = 0.03), pneumocystosis (1.9 to 2.6 [25.4% change]; p = 0.03), and other specified mycoses (1.7 to 2.5 [32.9% change]; p<0.01). Compared with non–COVID-19–associated fungal hospitalizations, COVID-19–associated fungal hospitalizations more frequently involved aspergillosis (27.8% vs. 16.9%; p<0.01), mucormycosis (1.8% vs. 1.4%; p = 0.03), and unspecified mycoses (24.3% vs. 18.5%; p<0.01) and, in general, less frequently involved other fungal infection types (Table 2).

Figure

Figure. Age distribution of patients in study of increased hospitalizations involving fungal infections during COVID-19 pandemic, United States, January 2020–December 2021. Hospitalizations for fungal infections were COVID-19–associated (n = 5,288) or...

Median patient age was 63 (interquartile range [IQR] 52–72) years for COVID-19–associated hospitalizations versus 61 (IQR 46–72) years for non–COVID-19–associated hospitalizations (p<0.01) (Figure). Compared with hospitalizations of patients with non–COVID-19–associated fungal infections, hospitalizations of patients with COVID-19–associated fungal infections more frequently involved those who were male (59.9% vs. 57.5%; p<0.01) and Hispanic/Latino (18.8% vs. 11.7%; p<0.01); occurred in hospitals located in the western United States (29.1% vs. 27.5%; p<0.01); involved longer hospital stays (21 [IQR 11–35] days vs. 9 [IQR 4–17] days; p<0.01); and involved ICU-level care (70.0% vs. 35.5%; p<0.01), IMV receipt (64.4% vs. 22.5%; p<0.01), and increased in-hospital deaths (48.5% vs. 12.3%; p<0.01) (Table 3).

Longer hospital stays, higher ICU admission rates, more IMV receipts, and more deaths were generally observed for hospitalizations caused by COVID-19–associated fungal infections than for non–COVID-19–associated fungal infections, regardless of the specific fungal pathogens involved (Appendix Tables 2, 3). COVID-19–associated fungal hospitalizations with the highest percentages of deaths involved aspergillosis (57.6%), invasive candidiasis (55.4%), mucormycosis (44.7%), and unspecified mycoses (59.0%).

Analysis of a large US healthcare services database indicated that hospitalization rates involving fungal infections increased significantly during 2019–2021, primarily driven by hospitalizations of patients with COVID-19–associated fungal infections. During 2020–2021, a total of 13.4% of fungal hospitalizations were COVID-19–associated, and COVID-19–associated fungal infections were associated with ≈2-fold increase in ICU admission rates and ≈4-fold increase in in-hospital death rates compared with non–COVID-19–associated fungal hospitalizations. Consistent with national mortality data, hospitalizations of patients with COVID-19–associated (compared with non–COVID-19–associated) fungal infections most often involved invasive candidiasis and aspergillosis and disproportionately occurred among non-White male patients in the western United States (2). Racial or ethnic disparities observed for fungal infection–associated hospitalization rates might relate to longstanding inequities in social health determinants, such as lack of access to medical care or occupational exposures, and prevalence of underlying conditions, such as diabetes, that might increase fungal and COVID-19 infection risk among certain minority groups (2,7–9). Also consistent with national mortality data, hospitalization rates for COVID-19–associated aspergillosis and mucormycosis increased from 2020 to 2021 (2), likely reflecting a greater burden of COVID-19 during 2021 than 2020 (https://gis.cdc.gov/grasp/covidnet/covid19_5.html), increased clinician awareness and testing for COVID-19–associated mold infections (10,11), and increased use of corticosteroids for COVID-19 treatment, a major risk factor for aspergillosis and mucormycosis (4). Our findings emphasize the importance of maintaining a high index of clinical suspicion for fungal infections in patients at high risk, including those with COVID-19, and the need for increased fungal disease surveillance to detect and evaluate emerging trends.

The first limitation of our study is that, although ICD-10-CM codes for COVID-19 correlate well with SARS-CoV-2 test results in PHD-SR data (12), fungal ICD-10-CM codes might be associated with underreporting, misclassification, and nonspecific coding of pathogenic fungi, particularly those causing candidemia and invasive mold disease (13–15). Second, PHD-SR data are broadly representative of US hospitals, and hospital types remained relatively consistent during the analytic period. However, data might overrepresent certain regions of the country, particularly the South, and participating hospitals can vary over time. Finally, we suspect that most COVID-19–associated fungal infections were secondary complications of COVID-19 because of the natural history of fungal disease in patients with respiratory infections (3), but we could not verify this supposition by using PHD-SR data.

Our analysis underscores the substantial burden of patient hospitalizations with fungal infections in the United States and indicates that increased hospitalizations involving fungal infections occurred during the COVID-19 pandemic. As the COVID-19 pandemic evolves, and to increase preparedness for future infectious diseases outbreaks, comprehensive public health surveillance for fungal diseases is needed to characterize disease epidemiology and guide efforts to prevent illness and death.