Candida Bloodstream Infections in 10 U.S. Surveillance Areas, 2017–2021

Candidemia — a bloodstream infection caused by Candida yeasts — is one of the most common health care-associated bloodstream infections in the United States, and it brings high costs, serious illness and high death rates (all-cause mortality around 25%). Candida normally lives in the gut and on the skin; it can cause surface (mucocutaneous) or invasive infection, and from the blood it can spread to organs such as the abdomen, bones, eyes, heart, kidneys or lungs.

Why it happens

Candidemia can follow a breach in the body's skin and mucosal barriers — including the gut, which is why abdominal surgery is a risk factor. Critical illness raises the risk through stress on the immune system and through devices such as central venous catheters that give the yeast a way in. Other risk factors include cancer, hemodialysis, diabetes, immune-suppressing drugs including corticosteroids, total parenteral nutrition and antibiotics; injection drug use has more recently emerged as an important risk for candidemia outside health care.

Candida albicans has historically caused most cases, but other species — C. glabrata, C. parapsilosis, C. tropicalis and C. krusei — have become more common, and they more often resist antifungal drugs and may be deadlier. Candida auris, highly transmissible and often multidrug-resistant, has caused health care outbreaks.

The surveillance

Candidemia is not nationally notifiable, so CDC tracks it through its Emerging Infections Program, a partnership with state health departments and academic partners. Surveillance began at two sites in 2008 and now covers 10 sites — parts of California, Colorado, Georgia, Maryland, Minnesota, New Mexico, New York, Oregon and Tennessee, plus all of Connecticut from 2019 — about 21.5 million people, 7% of the U.S. population in 2019.

A case is a blood culture positive for any Candida species in a resident of a surveillance area; further positive cultures within 30 days count as the same case. Laboratories report cases, surveillance officers review medical records, and isolates go to CDC for species identification and antifungal susceptibility testing.

Results

7,381 cases in 6,235 patients were identified in 2017–2021, an overall incidence of 7.4 per 100,000 people — lowest in 2019 (6.8) and highest in 2021 (7.9), and slightly below the 8.7 recorded in 2012–2016.

Annual candidemia incidence, overall and by age group, 2017–2021. CDC figure.

Who was affected (2021 unless noted):

• Incidence was highest among adults 65 and older and next highest among infants under 1 (8.0).
• Males (8.7) had higher incidence than females (7.0).
• Black patients (12.8) had higher incidence than non-Black patients (5.6).
• By site, incidence was highest in Maryland (14.5), Tennessee (10.1) and Georgia (10.0), and lowest in Oregon (4.8).

Annual candidemia incidence by surveillance site, 2017–2021. CDC figure.

Risk factors across all cases: 80.7% had recently received antibiotics, 63.7% had a central venous catheter, 36.2% had diabetes (rising from 32.1% in 2017 to 38.0% in 2021), 29.2% had recent surgery, 25.6% chronic kidney disease, 24.3% cancer, 22.1% chronic lung disease, 16.6% liver disease, 19.6% total parenteral nutrition, and 9.0% a history of injection drug use (down from 10.7% to 7.1%).

Changes during the COVID-19 years:

• Cases with health care onset rose from 52.2% (2017) to 58.0% (2021), and those with an ICU stay before diagnosis from 38.3% to 44.9%.
• Cases with a positive SARS-CoV-2 test within 90 days rose from 10.4% (2020) to 17.7% (2021).
• Cases with another, non-Candida organism in the blood in the preceding days rose from 17.8% to 28.7%.

Treatment and outcomes:

• Echinocandins (such as micafungin) were used in 49.8% of cases and azoles in 47.7%. Untreated cases rose from 18.0% in 2017 to 35.4% in 2020 (30.1% in 2021) — most often because the patient died before the culture result came back.
• Complications included endocarditis (4.0%), abscess (2.4%), osteomyelitis (1.4%), septic embolus (1.0%) and endophthalmitis or chorioretinitis (0.9%).
• 55.1% of cases involved an ICU stay around the time of diagnosis; the median hospital stay was 16 days.
• In-hospital death occurred in 32.6% of cases overall, reaching 36.1% in 2021.

Species and resistance: C. albicans caused 37.1% of cases, C. glabrata 30.4%, C. parapsilosis 13.5% and C. tropicalis 6.1%; C. glabrata led in California and Maryland, and in New Mexico its share rose from 15.4% to 48.9%. C. auris accounted for 0.4%. Of 6,576 isolates tested, 5.6% resisted fluconazole and under 1% resisted echinocandins or several drugs, with no change over time.

What it means

The higher death rate and the rise in ICU-linked and health care-onset cases in 2020–2021 likely reflect the COVID-19 pandemic — strained health systems and more critically ill patients susceptible to candidemia. The higher incidence among older adults, males and Black patients matches earlier findings; the growing share of older and Black patients may reflect COVID-19's heavier toll on those groups. Local differences — incidence roughly three times higher in Maryland than Oregon — show why local data matter.

Infectious Diseases Society of America guidelines recommend echinocandins as the preferred first treatment for most patients, yet only about half of cases received them — partly because some patients died before diagnosis, and possibly because of low awareness.

The authors recommend strict measures to prevent health care-associated bloodstream infections, vigilance for candidemia as a complication of critical illness, and continued surveillance of at-risk groups and resistance.

Limitations: only 10 areas were covered; data collected changed over the years; chart reviews may vary and under-record risk factors; and blood cultures miss some infections.

Sources

• Jenkins EN, Gold JAW, Benedict K, et al. "Population-Based Active Surveillance for Culture-Confirmed Candidemia — 10 Sites, United States, 2017–2021." MMWR Surveillance Summaries 74(4), Centers for Disease Control and Prevention.
• The report gives two different starting years for the rise in in-hospital deaths (2019 in its summary, 2017 in its discussion) and two different time frames for the incidence among adults 65 and older; this page gives neither.