The existing clinical decision rules (CDRs) for managing febrile neutropenia in children with cancer have not optimally performed so far. A new study has built and validated a new CDR (DISCERN-FN) for predicting the risk of severe infection in children with febrile neutropenia (FN). Results of this study were published in the journal The Lancet Child and Adolescent health.
The project involved two prospective studies. The first study from France enrolled children (below 18 years) diagnosed with cancer and with episodes of FN. The study did not include patients already treated with antibiotics for FN, receiving palliative care, FN not induced by chemotherapy, and with stem cell allograft for less than a year. Medical history, clinical and laboratory data were used to analyse the association of FN with severe infection.
From Jan 1, 2012, to May 31, 2016, a second prospective external study was done across 23 centres. The primary outcome of the study was set as severe infection defined by bacteraemia, local infection (with a high potential of extension) or invasive fungal infection. All episodes were analysed to evaluate the sensitivity, specificity, and negative likelihood ratio of the CDR.
A total of 539 FN episodes were included in the derivation set with 270 episodes in patients with blood cancer (median age 7.5 years, IQR 3.7–11.2; 158 boys and 112 girls) and 269 episodes in patients with solid tumours (median age 6.6 years, IQR 2.9–14.2; 140 boys and 129 girls). The variables introduced in the decision tree included age, high-risk chemotherapy, cancer type, level of fever, C-reactive protein concentration, leucocyte, platelet counts and procalcitonin. The derivation set showed a high CDR sensitivity (98%), specificity (56%), and a negative likelihood ratio of 0.04 (0.01-0.15).
The validation set included 1806 FN episodes with a mean age of 8.1 years (1014 boys and 792 girls, of which 332 were linked with severe infection). The validation set showed a high CDR sensitivity (95%), specificity (38%), and a negative likelihood ratio of 0.13 (0.08-0.21).
The new CDR reduced the risk of severe infection with a probability of 0.8% in derivation and 2.4% in the validation set.
The new CDR demonstrates its efficacy in reducing the risk of infection in both derivation and validation groups. It should be implemented in appropriate settings for improving clinical practice.
Reference
Top
