Key messages

• It is not clear whether preemptive strategy affects mortality because the certainty of the evidence is very low, but it might slightly decrease the use of antifungal agents in patients with prolonged febrile neutropenia.
• It is not possible to conduct an adequate balance between benefits and risks and costs, given the high level of uncertainty associated to the existing evidence.

What is the evidence.
See evidence matrix  in Epistemonikos later

We identified three systematic reviews [4],[5],[6] including twelve primary studies [7],[8],[9],[10],[11],[12],[13],[14],[15],[16],[17],[18] from which five correspond to randomized controlled trials [9],[10],[13],[15],[18]. One of the trials was excluded from this summary since it does not compare against empirical strategy [9].

What types of patients were included

Two studies included only adults [10],[15], one study included both adults and children [18] and one study did not report the age of the patients [13].

What types of interventions were included

All studies used different criteria for starting antifungal therapy in the preemptive strategy, and different tests for early diagnosis of invasive fungal disease. Three studies used galactomannan test twice a week [10],[15],[18]; one study used nested PCR for Aspergillus [15] and one study non-nested PCR for Aspergillus and Candida [13]; only one study performed blood cultures [13]. All studies used imaging as diagnostic test; two studies performed chest computed tomography if serological tests were positive [15],[18], one study performed chest and abdomen computed tomography to all patients [13], and one study performed chest radiography followed by chest computed tomography [10]. 

What types of outcomes
were measured

To whom this evidence does and does not apply

• The evidence included in this summary is applicable to patients with hematological malignancies undergoing stem cell transplant or chemotherapy, presenting febrile neutropenia and remaining febrile despite empirical antibiotic treatment. 

• All-cause mortality and use of antifungal therapy were selected as critical outcomes for decision-making by the authors of this summary.
• We did not include the outcome invasive fungal infection because it was considered less important than the selected outcomes when choosing a treatment over the other, and also because of the susceptibility to risk of bias derived from lack of blinding.

• Because of the high level of uncertainty it is not possible to assess risk/benefit properly.
• Preemptive strategy might decrease antifungal therapy use, and also the potential associated adverse effects. However this information must be taken with caution for decision-making because of the very low certainty of the evidence about mortality.

• Physicians and patients who prefer to avoid the risk of adverse effects above the uncertain benefit of the empirical therapy, will probably choose the preemptive strategy.
• It is particularly important to inform the patient about the uncertainty associated with this decision. 

• Because of the high uncertainty it is not possible to assess cost/benefit properly.
• Preemptive strategy costs vary depending on the tests selected for invasive fungal infection diagnosis. The costs of empirical therapy also vary depending on the antifungal agent selected.
• Even if the existing evidence about the effects were more certain it would be desirable to formally assess cost-efectiveness. The characteristics of this type of decision will be different depending on where it is implemented.

Differences between this summary and other sources

• The conclusions of our summary partially agree with the systematic reviews identified. The more recent and complete review [6] concludes preemptive antifungal strategy is a reasonable alternative to empirical treatment in terms of outcomes and costs. However, our summary puts more emphasis on the associated uncertainty of the existing evidence.
• Our summary agrees with the guideline of the Infectious Diseases Society of America [1] which considers preemptive antifungal strategy as an acceptable alternative to empirical treatment in clinically stable neutropenic patients that still have fever after four to seven days of broad-spectrum antibiotics, and have no evidence of fungal infection (serological tests, chest and paranasal computed tomography without fungal infection, and absence of fungal agents such as Candida or Aspergillus on cultures). They suggest starting antifungal agents if any of the tests is positive. Most of the evidence identified in this summary was not available in 2010 when the guideline was published. 

• The probability that future studies would change the conclusions of this summary is high, considering the uncertainty of the existing evidence.
• There are no new or ongoing trials answering this question, based on a search of trials published after the search date of the more recent review identified, and a search in the the International Controlled Trials Registry Platform of World Health Organization. One of the included studies is still open, so new data might emerge [19].