Decision-tree derivation and external validation of a new clinical decision rule (DISCERN-FN) to predict the risk of severe infection during febrile neutropenia in children treated for cancer

Background

In 2017, international guidelines proposed a new treatment for febrile neutropenia in children with cancer, adjusted for the risk of serious infection through clinical decision rules (CDRs). To date, none of the proposed CDRs have performed well enough in high-income countries for use in clinical practice. Our study aimed to build and validate a new CDR (DISCERN-FN) to predict the risk of serious infection in children with febrile neutropenia.

Methods:

We conducted two prospective studies. First, a prospective derivative study included all episodes of febrile neutropenia in children (aged < 18 years) diagnosed with cancer and previously treated and admitted for an episode of febrile neutropenia, with the exception of patients already treated with antibiotics for this episode, febrile neutropenia not caused by chemotherapy, patients receiving palliative care and patients with a stem cell transplant for less than 1 year, from April 1, 2007 to December 31, 2011 from two pediatric cancer centers in France. We collected the children's medical histories, clinical and laboratory data, and analyzed their associations with serious infections. Sipina software was used to derive the CDR as a decision tree. Second, a prospective, national, external validation study was conducted in 23 centers from January 1, 2012 to May 31, 2016. The primary outcome was severe infection, defined by bacteremia, positive bacterial culture from a usually sterile site, local infection with a high potential for expansion, or an invasive fungal infection. The CDR was applied a posteriori to all episodes to evaluate sensitivity, specificity, and negative probability ratio.

findings

The lead set included 539 episodes of febrile neutropenia (270 episodes in patients with blood cancer [median age 7·5 years, IQR 3·7–11·2; 158 (59 %) boys and 112 (41%)
girls] and 269 in patients with solid tumors [median age 6·6 years, IQR 2·9–14·2;
140 (52 %) boys and 129 (48%) girls]). Significant variables introduced into the decision tree were cancer type (solid tumor versus blood cancer), age, high-risk chemotherapy, fever, C-reactive protein concentration (24-48 hours after admission), and leukocyte and platelet counts and procalcitonin. (on admission and 24-48 hours after admission). For the derivation set, CDR sensitivity was 98% (95% CI 93-100), specificity 56% (51-61) and negative probability ratio 0.04 (0.01-0.15). 1,806 febrile neutropenia episodes were analyzed in the validation set (mean age 8.1 years [SD 4·8], 1014 (56%) boys and 792 (44%) girls), of which 332 (18%, 95% CI 17-20) were associated with serious infection. For the validation set, the CDR had a sensitivity of 95% (95% CI 91–97), a specificity of 38% (36–41), and a negative probability ratio of 0.13 (0.08–0.21). ). Our CDR reduced the risk of serious infection to a post-test probability of 0.8% (95% CI 0.2–2 9) in the lead set and 2.4% (1.5–3 9) in the validation set. The validation study is registered with ClinicalTrials.gov, NCT03434795.

Interpretation

The use of our CDR significantly reduced the risk of serious infection after testing in both the derivation and validation groups, suggesting that this CDR would improve clinical practice enough to be introduced in appropriate settings.

Financing

National League Against Cancer.