Infectious Disease Compendium


As always, the 2016 IDSA guidelines are the best summary and/or a great ambien substitute (PubMed).


A yeast that includes the human pathogens C. albicans, C. auris (PubMed)(may be misidentified as Candida haemulonii, Candida famata, Saccharomyces cerevisiae, and Rhodotorula glutinis), C. dubliniensis (Review), C. glabrata, C. guilliermondii, C. haemulonii, C. krusei, C. kyfir (PubMed), C. lusitaniae, C. parapsilosis, C. palmioleophila, C. pseudotropicalis, C. rugosa, C. tropicalis. There are other Candida that cause the occasional infection and resultant case report in the extremely unlucky patient.

It is NOT pronounced Can-dee-da. Got it? Its Can-da-da.

Blood cultures stink for diagnosing Candidemia, best bet is blood cultures, 1,3 B-D-Glucan and blood PCR all at once (PubMed).

The 1,3 B-D-Glucan assay: positive if > 80 picogram/ml. False positives with hemodialysis, those who receive albumin or IgG. The higher the result, the worse the outcome and a falling titer is a good sign; > 416 is the cut off for ominousness (PubMed).

As in all tests there are false positives and if slow to fall look for a deep infection (Pubmed).

Epidemiologic Risks

Part of the normal human flora. Up to a third of cases in the ICU may be due to unsuspected nosocomial clusters/transmission (PubMed), often the case with C. parapsilosis (PubMed). The classic risks for bloodstream infections are central lines, TPN, broad spectrum antibiotics, neutropenia and a major surgical procedure. The more places you grow Candida (wounds, mouth, urine) the more likely it will invade the blood.  The risks are (From NEJM Review):

  • Critical illness, with particular risk among patients with long-term ICU stay
  • Abdominal surgery, with particular risk among patients who have anastomotic leakage or have had repeat laparotomies
  • Acute necrotizing pancreatitis
  • Hematologic malignant disease
  • Solid-organ transplantationSolid-organ tumors
  • Neonates, particularly those with low birth weight, and preterm infants
  • Use of broad-spectrum antibiotics
  • Presence of central vascular catheter, total parenteral nutrition
  • Hemodialysis
  • Glucocorticoid use or chemotherapy for cancer
  • Candida colonization, particularly if multifocal (colonization index >0.5 or corrected colonization index >0.4)

IVDA is a risk for Candidemia, at least in the US NE (PubMed). It was also an issure in the 1970's. The more things change the more they stay the same.

While C. albicans is most common in the West, in Wayampi Amerindians in remote areas it is uncommon (PubMed).

People routinely give piddly ass doses of fluconazole, and it drives the risk for developing fungemia due to fluconazole-non-susceptible Candida (PubMed).

CARD9 Deficiency causes spontaneous CNS Candidiasis (PubMed).

C. auris is the emerging Candida of the 21st century, new since 2009 and often resistant. As of 2018, most of the US cases are in New York, New Jersey, California, Illinois and Florida. But I am sure it will spread. Usually occurs in patients with prior antifungal exposure.

C. rugosa is common in Indian trauma patients.


- Thrush/Esophagitis: occurs in the immuno-incompetent especially AIDS (CD4 < 200), cancer patients and patients on steroids. Can be brought on by antibiotics in normal people.

- Vaginitis: usually in normal hosts, can be a harbinger of HIV disease and diabetes.

- Cystitis: almost always associated with instrumentation (a foley not a flute).

- Line Related fungemia: the big question you have to ask yourself is: where, if anywhere, did that Candida go? Depending on the study 15-30% of patients will have fungus balls in their eyes. If the time to positivity for candida is < 30 hours, it is more likely a catheter infection as the source of the Candidemia (PubMed).

But does it mean that everyone with Candidemia need a fundascopic exam as the guidelines say? Probably not. Those who complain of eye problems certainly and perhaps those who cannot complain (Pubmed). But in one study only 8% had eye involvement, it was asymptomatic and did not impact outcomes. So eye exam should likely best be done in those who have symptoms (PubMed).

The general rule of thumb with infections is the sooner the correct therapy and the sooner the source is controlled (in this case, pulling the catheter) the better the patient will do. So pull the CVC, in some studies it improves survival for Candida (PubMed).

If the time to positivity for Candida is < 30 hours, it is more likely a catheter infection as the source of the Candidemia (PubMed), you remember that if S. aureus has a time to positivity of < 24 hours, it is more likely and endovascular infection. However, a catheter tip for Candida is only rarely associated with fungemia: about 2.7 % of cases (PubMed).

- Pneumonia. Nope. Candida almost never causes pneumonia and can almost always be ignored, even in a BAL. "We concede, by the present data, that it is convincingly proven that Candida species are at most a very rare cause of pneumonia (PubMed)".; 0.07% of positive BAL's will actually be the real deal.

- Disseminated: the big risk factors are broad spectrum antibiotics, TPN, central venous catheter, neutropenia and major surgical procedures. Beware: blood cultures can be negative or delayed. Use the nifty Candida score to see if preemptive Candida therapy is warranted.

- Meningitis: usually postop/trauma.

- Endocarditis: often heroin users or a complication of central lines, especially right sided.  At least 4.2% of candademias will have endocarditis (PubMed).

Empiric therapy in high risk patients with persisting fever on antibiotics and no other reasons, may want to treat even if the blood cultures are negative. Look in the eyes and check a beta glucan. Mortality goes up every 24 hours you delay appropriate therapy. Prophylaxis is not that beneficial in the non-neutropenic.

Manifestations can be Candida endophthalmitis (occurs in 10 (PubMed) to 15 to 30%) or skin lesions (ecthyma gangrenosum or folliculitis). The strain in the urine and the strain in the blood match only 48% of the time (PubMed); so the urine is often not the source of Candidemia. Consider fundascopic exam on all patients with PROVEN candidemia, start therapy within 24 hrs (duh) and document clearance of blood cultures after starting therapy.

Serial (1-3)-Beta-D-glucan has good diagnostic parameters in cancer patients (PubMed).

- Peritonitis/abscess. Do not ignore Candida in the peritoneum, which is common after gastric perforation. Drain it as you would a bacterial abscess and treat it if you want a good outcome (Pubmed). As the terminator said, Drain da pus if you want to live (PubMed).

- Chronic Mucocutaneous: a disease associated with a congenital inability to recognize candida or thymomas. It is due to the STAT1 mutation (PubMed) in an autosomal dominant form of the disease as well as inherited interleukin 12 receptor β1 deficiency (PubMed).


(see IDSO website for complete treatment guidelines)

Most important for fungemia/sepsis: Start an echinocandin first, NOT an azole, and pull the damn catheter (PubMed). It was this approach that allowed a surgical service of all things to have better outcomes than a medical service. How humiliatin' (PubMed).

When starting an antifungal, consider prior exposure. Fluconazole and to a lesser extent echinocandins may breed resistance, especially in C. glabrata (PubMed) and C. parapsilosis.

Empiric treatment of disseminated Candida with an echinocandin is associated with decreased hospital mortality and increased clinical success (PubMed).

C. albicans is the only Candida that is reliably susceptible to fluconazole; pending identification you should start with caspofungin or one of its cousins. However, C. glabrata, C. albicans, C. krusei, C. parapsilosis and C. tropicalis can be resistant to the echinocandins.

C. auris: don't use azoles except isavuconazole . C. auris can be resistant to all three classes of antifungals. Use (Pubmed) amphotericin B or echinocandins (caspofungin or micafungin (perhaps the best (PubMed)) or anidulafungin)(PubMed). In one study for C. auris isavuconazole was the most active agent (MIC90 = 0.125 mg/L) followed by posaconazole (MIC90 = 0.5 mg/L) (PubMed) (PubMed). You really do need susceptibility testing to guide therapy.

C. rugosa, 16% are resistant to fluconazole but susceptible to voriconazole, amphotericin and flucytosine.

Rare Candida can have odd susceptibilities: "C. guilliermondii strains exhibited high rates of azole MICs above ECVs (fluconazole, 17%; voriconazole and posaconazole, 24%). The 2 species that commonly were positive for caspofungin MICs above ECVs were C. kefyr (82% vs. 17% among other species; p<0.001) and C. lusitaniae (21%) " (PubMed).

- Thrush/Esophagitis: Clotrimazole, nystatin, amphotericin OR fluconazole all work just fine in most patients, especially if C. albicans. If it doesn't respond worry re: right diagnosis, resistant yeast or poor absorption (a real issue with itraconazole if the patient has no gastric acid). If resistant to fluconazole or itraconazole, posaconazole works 75% of the time (PubMed). In African AIDS patients, 750 mg fluconazole single dose worked as well as a two week course (PubMed).

- Vaginitis: there are a billion topical agents that work as does po fluconazole 150 mg po x 1. For recurrent disease, 150 mg fluconazole a week therapy works (PubMed). The use of probiotics (as compared, I suppose to the amateur ones) is still not clear (PubMed).

- Cystitis: Probably not a cause of fever and symptoms although it may represent a marker/risk for disseminated disease, especially in the ICU. Treat? Probably a waste of time if the foley is left in and self-limited if removed. Amphotericin bladder wash (but is a cruel and unusual punishment) was equally efficacious in achieving overall cure, and resulted in greater clearance of candiduria compared to fluconazole. Amphotericin bladder is equal fluconazole for the treatment of candiduria and may be preferred over fluconazole in patients with renal dysfunction (Pubmed) although not pleasant for the patient.

A real problem as best treatment is the least likely to occur: remove the foley. Fluconazole is the one imidazole that has reasonable urinary levels: 400 mg a day for 14 d. Can also try amphotericin irrigations: 50 mg of amphotericin B in 1 liter of sterile water and infused at 40 ml/hour, but will be passe in the next guidelines. Flucytosine 25 mg/kg qid for 7–10 days. Even though urinary levels are diddley, a small case series suggests that caspofungin has efficacy (PubMed). I high risk patients, treat cystitis like dissemination. Changing the catheter alone cures the infection 10% of the time, removing the catheter (alone) cures 50% of infections.

- Catheter related fungemia: the sicker the host the more aggressive the therapy. At a minimum a 14 day course of relatively high dose (800 load, then 400 mg a day) of fluconazole, the really ill patient should get amphotericin B or an echinocandin (caspofungin or micafungin or anidulafungin). And pull the catheter. Always (PubMed).

- Endocarditis, which can be a complication of both catheters and IVDA, is best treated with valve resection and a long course of ? amphotericin ? fluconazole (PubMed) ? other. While classically an absolute indication for valve removal, I have cured two cases in patients who were deemed not surgical candidates and this has been supported in the literature (PubMed).

For prosthetic valve endocarditis, start with lipid amphotericin not an echinocandin (PubMed). Improves survival. And, oddly, surgery did not have better outcomes than medical therapy alone, although small numbers and retrospective.

- Osteomyelitis - debride and 6 to 12 months of some sort of therapy or other. It is not like there are randomized trials for this infection (PubMed) (PubMed).

- Disseminated Candida/fungemia

Always start with caspofungin or micafungin or anidulafungin when known or suspected; survival is better (PubMed). And pull the CVC (PubMed). The problem is Candida is slow to grow in blood culture and for every hour it takes for the blood cultures turn positive, mortality goes up 4% (PubMed). So if you have a high index of suspicion for Candida in blood, start therapy.  After 5 plus days of an echinocandin, if repeat blood cultures are negative and the patient is doing well then change to po fluconazole, at least 400 mg a day.

Note the 'at least' in the last paragraph. For example, survival increases with fluconazole dose for C. glabrata. 400 mg is the minimum dose (PubMed).

May want to follow beta-glucan levels as a surrogate for response (PubMed).

Neutropenic: An echinocandin (caspofungin or micafungin or anidulafungin), perhaps preferred (PubMed), or lipid amphotericin B 3–5 mg/ kg daily.

Candida loves to go to the back of the eye, and if there is fungal endophthalmitis, may want to avoid echinocandins as penetration into the eye is negligible. "Fluconazole, voriconazole, and flucytosine achieve therapeutic intravitreal concentrations, whereas the echinocandins and all formulations of amphotericin do not. Most experience has accumulated with fluconazole. There is less experience with voriconazole, but there are data on the efficacy and safety of intravitreal injection of this agent. Flucytosine should be used in combination with amphotericin and not as sole therapy (PubMed)".

How long to treat for fungemia? Treat with the echinocandinin until better, perhaps as little as 4 days for C. albicans (PubMed), then high dose fluconazole. A minimum of 14 days if the goal is to prevent late onset chorioretinitis (PubMed).

A study using historical comparisons suggested better outcomes with echinocandins (PubMed).

Alternatives: fluconazole 800-mg (12-mg/kg) loading dose, then 400 mg (6 mg/kg) daily; or voriconazole 400 mg (6 mg/kg) bid for 2 doses then 200 mg (3 mg/kg ) bid. An echinocandin (caspofungin or micafungin or anidulafungin) or lipid amphotericin B is preferred for most patients. Fluconazole is recommended for patients without recent azole exposure.

Non neutropenic: Fluconazole 800-mg (12-mg/kg) loading dose, then 400 mg (6 mg/kg) daily or an echinocandin (caspofungin or micafungin or anidulafungin).

Alternatives: Lipid amphotericin B 3–5 mg/kg daily; or amphotericin B 0.5–1 mg/kg daily; or voriconazole 400 mg (6 mg/kg) bid for 2 doses, then 200 mg (3 mg/kg) bid.

Use the nifty Candida score to see if preemptive Candida therapy is warranted. However, prophylactic fluconazole in high risk ICU patients did not prevent infections or improve outcomes (PubMed).

As adjunctive therapy of chronic disseminated candidiasis aka &hepatosplenic candidiasis, prednisone at a dose of 0.5–0.8 mg/kg leads to rapid resolution of symptoms, suggesting in part a Immune Reconstitution Syndrome (PubMed).

All the Echinocandins are the same for killing Candida (PubMed). Make sure that if you are going to treat disseminated candida with fluconazole have a minimum starting dose of 400 - 800 mg qd. Once the patient has responded to IV therapy, it is reasonable to change to po fluconazole (C. albicans) or voriconazole (C. krusei).

If C. glabrata, avoid fluconazole or voriconazole unless you have sensitivities back. And C. glabrata is increasingly resistant to echinocandins (PubMed). For example "Of 146 isolates, 30 (20.5%) were resistant to fluconazole, 15 (10.3%) to caspofungin, and 10 (6.8%) to multiple drugs (9 caspofungin- resistant isolates were also resistant to fluconazole, 1 to amphotericin B) (PubMed)."

For C. parapsilosis, there is reduced susceptibility to echinocandins. It may make little clinical impact; patients do fine (PubMed).

In the NEUTROPENIC use either lipid amphotericin B or caspofungin or micafungin or anidulafungin.


In most studies in adults there is no benefit in fluconazole in preventing ICU disease but there is a literature to suggest probiotics are effective in decreasing mucocutaneous colonization with Candida, perhaps decreasing the risk of invasive disease (Pubmed).

Candida in the sputum is always to be ignored, Candida pneumonia is rare, you can cut off one of my hands you can still count the number of cases with fingers to spare. Save the middle finger for optimal driving.

Delay in treating fungemia increases mortality (PubMed), although not all studies show benefit from the initial choice of therapy (PubMed).

Only Candida albicans can be reliably treated with fluconazole/itraconazole.

C. krusei is often resistant to fluconazole/itraconazole and can develop resistance to echinocandins on therapy (PubMed).

C. lusitaniae can be resistant to amphotericin. Really.

C. parapsilosis may be selected with the use of caspofungin (PubMed).

C. haemulonii and C. pseudohaemulonii: treat with caspofungin or micafungin (PubMed), amphotericin and azoles are unreliable.

Curious Cases

Relevant links to my Medscape blog

Answer these questions three

Look. Up in the sky. It's a bird. It's a plane. It's

Fungus Ball

The Bugs are more complicated than we suspect.

25 Years, Ain't Never Seen This Bug Before


The Real Deal?

Right the first time

Getting close to the EMR

Didn't See That Coming. Again.

Doug Flutie has nothing on me

Things fall apart

Tricky Yeast

A Ticket to the Fungal Ball

Swing and a Miss


I Don't Believe the Diagnosis

The Yeast Tells All

I see it as often as a short-period comet

Somewhere Endovascular

Working with minimal data


Candidemia: Still not really knowing what to do.

Last Ditch Effort

Close. But no cigar.

In Cultures I Trust

The more things change, the more things are different

Trying to Catch Up

What is worse: Candida on a valve or in the esophagus

I wish the monkey preferred bananas

Twists and Turns

Famata is not famata

Third Time to be a Charm?

Roman Names

Blind Pigs and Acorns

Last Update: 04/30/18.