An Incidental Yeast in the Catheter Bag
A single patient, found to have candiduria on a routine post-operative culture. The disagreement isn't about whether her diabetes raised her risk — it's about whether an incidental, asymptomatic finding meets any of the three narrow reasons the guideline actually recommends treating it.
Yolanda P., a 66-year-old woman who still works two mornings a week at the bakery she has owned for three decades, fractured her hip stepping off a curb she has stepped off a thousand times before, and is now two days out from surgical repair. She has type 2 diabetes, managed with metformin and a basal insulin she started reluctantly last year after her A1c crept above 8%, and an indwelling catheter placed for the surgery that nobody has yet had a reason to reconsider. A routine post-operative urinalysis and culture, sent as part of standard surgical monitoring rather than for any specific complaint, comes back growing Candida albicans, greater than 10,000 colony-forming units per milliliter — no fever, no dysuria the catheter would make hard to report anyway, no cloudy or foul urine, nothing that distinguishes this culture from the asymptomatic bacteriuria any long-term catheter eventually produces, except that this time the organism is a yeast rather than a bacterium, and yeast in urine tends to prompt a reflexive urge to treat that bacteriuria alone rarely does.
The 2016 IDSA candidiasis guideline, led by Pappas et al., is direct about this exact scenario: asymptomatic candiduria does not require treatment except in three specific circumstances — in patients who will undergo urologic manipulation, in neutropenic patients, and in very-low-birth-weight infants under 1500 grams. Yolanda is none of these. What the guideline does recommend for her, and recommends strongly, is the thing that has nothing to do with a drug: eliminating the predisposing factor, which in her case is a catheter with no remaining indication. Her diabetes raises her baseline risk of candiduria occurring in the first place, a real and well-documented association, but risk of occurrence is not the same question the guideline is answering, which is whether treating an asymptomatic finding, once it has already occurred, changes anything for the better.
On post-operative rounds
She's diabetic, and diabetic patients with candiduria are more likely to progress to something worse than a non-diabetic patient would be. I'd rather start fluconazole now than wait to see if this becomes a real infection.
Pappas et al.'s 2016 IDSA guideline names exactly three groups in whom treating asymptomatic candiduria is recommended — patients about to undergo urologic manipulation, neutropenic patients, and very-low-birth-weight infants under 1500 grams. Diabetes elevates her risk of developing candiduria, which she already has, but it isn't one of the three reasons the guideline treats it once it's there. And the guideline's first recommendation here isn't a drug at all — it's a strong recommendation to eliminate the predisposing factor wherever feasible, which for Yolanda is a catheter nobody has a remaining surgical reason to keep in at day two.
Wanting to close off a plausible complication is understandable, but the guideline was written specifically because that instinct, applied broadly, doesn't reduce complications — it mostly just increases unnecessary antifungal exposure.
I'd add the cost side explicitly: fluconazole exposure in patients who don't need it is part of what has shifted candiduria isolates over time toward more fluconazole-resistant non-albicans species. It's not just that treating Yolanda doesn't help Yolanda — it's that treating patients like Yolanda routinely makes the drug less useful for the patients who actually need it later.
Agreed: catheter removed same day, no antifungal started, repeat urinalysis only if symptoms (fever, dysuria, flank pain) develop.
Not treated as a disagreement requiring further discussion, but recorded: the orthopedic hospitalist's instinct to treat proactively in a diabetic patient was accepted by all three voices as a reasonable read of general risk, just not the specific question the guideline is answering — a distinction the hospitalist agreed was clarifying rather than contested further.