Tasty Morsels of Critical Care 049 | Candida

26 Jul

Welcome back to the tasty morsels of critical care podcast.

Today we’ll look at everyone’s favourite yeast – Candida. Firstly, remember the distinction between yeasts and moulds. Yeasts, like Candida species are single celled critters whereas moulds like aspergillus are multicellular.

Candidiasis is common in ICU patients. It grows commonly in our BAL samples, it grows in our urine samples. It’s often the only +ve micro we turn up because the patients are mostly picked in meropenem and linezolid and the candida is all that is left or becomes selected out. It’s important to note that these BAL and urine samples of candida are common and generally reflect colonisation rather than infection and in general they shouldn’t be treated. That being said knowing what’s there in the event of the patient getting super sick might be somewhat helpful in guiding therapy.

the LITFL post describes candida as invasive if:

  • grown in blood cultures
  • found in a sterile cavity
  • cultured from 2 non contiguous sites (so for example you could argue that candida in sputum and urine fills this criteria though in reality I find we rarely treat this type of growth unless they’re super sick)
  • the species is a non commensal form of candida
  • culture from wound or burn biopsy

Risk factors for invasive candidiasis can be summed up as “being proper sick” but they can be somewhat more nuanced to include things like invasive lines, some kind of intra abdominal perforation or major burns or some horrible immunosuppression.

We make the diagnosis here using a variety of tools. The easiest being when it pops up in blood cultures on a sick patient or commonly it can grow from drain fluid in some abdominal disaster, particularly the higher up the GI tract the disaster occurs. Having a tissue diagnosis with a named species with sensitivities makes treatment fairly straightforward.

There are a variety of named scoring systems that can help you estimate risk of an invasive candidal infection that rejoice in the names like the “Nebraska criteria”, the “Candida score” and the “Paphitou” score. However I don’t see anyone using them in clinical practice.

More commonly we’re likely to be faced with a deteriorating septic patient with RF for candidiasis and we empirically start an antifungal. The issue now comes with stopping or supporting the ongoing use of the antifungal. Here tests like beta-d-glucan probably have a role. BDG is a cell wall component of all fungi so it certainly is not specific for candida but if it’s through the roof in a sick patient then coverage with an antifungal is probably reasonable. On the other hand if it’s undetectable and the indication for the antifungal was a bit soft in the first place then maybe it’s time to stop. The sensitivity is somewhere in the 75-85% range so I would not use it to rule out the super sick, super high risk patient.

When it comes to treatment we have to consider source control. If it’s coming from a belly full of collections then you need to drain the collections. Once source control is out of the way we now face a variety of choices. Ultimately if we have grown a candida then we will eventually get sensitivities and we can tailor therapy appropriately but all that takes time and we need to start something before then.

In general we turn to the echinocandins, drugs like anidulafungin and caspofungin. These are work horses in the ICU with a low side effect profile and importantly cover the vast majority of candida likely to grow. They also have nice action against biofilm embedded bugs which is very handy given the number of line related candida infections we might see. Fluconazole is of course an option but it struggles when it’s something other than candida albicans hence I see it more often used as a deescalation agent. The other agents like amphotericin and voriconazole will work in general for candida but are generally not needed.

In terms of guidelines, the 2012 European microbiology guidelines have Caspo or Anidulafungin as first line for documented invasive candidiasis. The IDSA in 2016 suggests fluconazole in stable patients and then caspo or anidulafungin if unstable.

Finally there is often great anxiety and hand wringing regarding the eye balls when someone grows a candida in the blood. Every time we grow candida in the blood we need to twist the arm of some poor eye doctor to come and do an exam for us. Indeed candida end ophthalmitis is a well described and devastating metastatic complication of candida but the eye balls have the useful characteristic of being easily accessible to clinical examination. If the eyes are red and nasty then be worried. If not then maybe you’re OK. This is supported by a 2019 JAMA Ophthalmology SRMA that suggested that a routine exam was useful in 3/7000 patients and made the gentle suggestion that maybe the IDSA should rethink their guidance please and thank you.

If you do happen to find candida in the eye ball then you need to rethink your antifungal strategy as our beloved echinocandins just don’t penetrate that well and you should be reaching for fluconazole, or flucytosine or indeed some kind of needle injection of stuff into the eye.


Oh’s Manual 73


Breazzano MP, Day HR Jr, Bloch KC, Tanaka S, Cherney EF, Sternberg P Jr, Donahue SP, Bond JB 3rd. Utility of Ophthalmologic Screening for Patients With Candida Bloodstream Infections: A Systematic Review. JAMA Ophthalmol. 2019 Jun 1;137(6):698-710. doi: 10.1001/jamaophthalmol.2019.0733. PMID: 30998819.

IDSA 2016 Guidance

And I missed it in the prep but of course IBCC has an excellent post that covers a lot more than I squeezed in here and is honestly, compulsory reading

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.