What is the Candida score and how should empirical antifungal therapy be managed in a non‑neutropenic critically‑ill patient with a score of three or higher?

Candida Score and Empirical Antifungal Management

What is the Candida Score?

The Candida score is a validated clinical prediction tool that assigns points based on four risk factors: multifocal Candida colonization (1 point), surgery (1 point), total parenteral nutrition (1 point), and severe sepsis (2 points), with a score ≥3 indicating high risk for invasive candidiasis and warranting empirical antifungal therapy. 1, 2

The score was developed to identify non-neutropenic ICU patients at highest risk for invasive Candida infections, with a cutoff of 2.5 demonstrating 81% sensitivity and 74% specificity. 1 Importantly, the rate of invasive candidiasis correlates directly with increasing scores: 0% with scores ≤3,17.6% with score of 4, and 50% with score of 5. 2

Empirical Antifungal Therapy for Score ≥3

First-Line Treatment Selection

Start an echinocandin immediately in critically ill patients with Candida score ≥3—this is the preferred empirical therapy with strong guideline support. 1, 3, 4 The IDSA recommends the following equivalent options:

• Micafungin 100 mg IV daily (no loading dose required) 3, 4
• Anidulafungin 200 mg loading dose, then 100 mg daily 1, 4
• Caspofungin 70 mg loading dose, then 50 mg daily 1, 4

Echinocandins are strongly preferred over fluconazole in this population because critically ill patients with score ≥3 typically have moderate-to-severe illness, recent broad-spectrum antibiotic exposure, and higher risk of azole-resistant species like C. glabrata or C. krusei. 1

Alternative Therapy

Fluconazole (800 mg loading dose, then 400 mg daily) may be considered only if the patient is not critically ill, has no recent azole exposure, and is unlikely to harbor fluconazole-resistant species. 1 However, given that a score ≥3 typically indicates moderate-to-severe illness, echinocandins remain the safer choice. 3, 4

Essential Adjunctive Measures

Mandatory Source Control

Remove all central venous catheters immediately in non-neutropenic patients—this is not optional. 1, 3, 5 Catheter removal is as critical as antifungal therapy itself and significantly impacts outcomes. 1, 3

Diagnostic Follow-Up

• Obtain blood cultures before starting therapy and repeat every 48-72 hours until clearance is documented. 1
• Perform dilated funduscopic examination within the first week to rule out endophthalmitis. 1, 5
• Consider biomarker testing (β-D-glucan, mannan/anti-mannan) to support diagnosis, though these remain adjunctive tools. 1

Duration and Discontinuation Strategy

For Proven Invasive Candidiasis

Continue echinocandin therapy for minimum 2 weeks after documented bloodstream clearance AND complete resolution of attributable signs and symptoms. 1, 3, 4 This is non-negotiable to prevent relapse. 3, 4

For Empirical Therapy Without Proven Infection

Reassess at 4-5 days: 3, 5

• If clinical improvement occurs without other explanation and biomarkers remain positive, continue therapy as "probable" invasive candidiasis. 6
• If no clinical response, negative cultures, and negative biomarkers with high negative predictive value, discontinue antifungals. 3, 5
• If candidemia is subsequently proven, treat for full 2 weeks post-clearance. 3, 4

Critical Pitfalls to Avoid

Never delay echinocandin initiation while awaiting culture results in patients with septic shock and score ≥3—mortality approaches 100% without timely treatment. 3, 4 Time from sepsis onset to appropriate antifungal therapy directly correlates with mortality, with delays increasing death rates by up to 50%. 4, 6

Do not use fluconazole as first-line therapy in critically ill patients with score ≥3, even without recent azole exposure—echinocandins demonstrate superior outcomes in this population. 3, 4

Do not discontinue therapy prematurely before completing the full 2-week course after blood culture clearance—this leads to relapse and treatment failure. 3, 4

Do not rely on colonization alone without calculating the Candida score—colonization has low positive predictive value, and empirical therapy based solely on colonization leads to antifungal overuse. 5, 7

Performance Characteristics

The Candida score demonstrates excellent negative predictive value (82.1%), making it highly reliable for ruling out invasive candidiasis when score is <3. 6, 2 However, positive predictive value is more modest (69.2%), meaning some patients with score ≥3 will not have invasive candidiasis. 6 This balance justifies empirical therapy given the catastrophic consequences of delayed treatment versus the manageable risks of short-term echinocandin exposure. 6, 2