"Low-risk" prediction rule for pediatric oncology patients presenting with fever and neutropenia

Abstract

Purpose: To prospectively derive and validate a clinical prediction rule to allow a more tailored approach to the management of pediatric oncology outpatients presenting with fever and neutropenia.

Patients and methods: The clinical prediction rule was derived over a 1-year period and then validated over the following 8 months in a new set of fever and neutropenia episodes. Patients were excluded if they presented with comorbidity or an abnormal chest x-ray (CXR).

Results: Significant bacterial infection (SBI; defined as any blood or urine culture positive for bacteria, interstitial or lobar consolidation on CXR, or unexpected death from infection) was documented in 43 of the 227 episodes. Multivariate analysis found four significant factors: bone marrow disease, general appearance unwell on initial examination, monocyte count less than 0.1 x 10(9)/L, and peak oral or oral equivalent temperature greater than 39 degrees C. Only the monocyte count contributed to determining a low-risk group, excluding SBI with 84% sensitivity (95% confidence interval [CI], 61% to 100%), 42% specificity (95% CI, 38% to 46%), and a negative predictive value of 92% (95% CI, 76% to 100%). If the monocyte count was >/= 0.1 x 10(9)/L at the time of presentation (low risk), the incidences of SBI and bacteremia were 8% and 5%, respectively, versus 25% and 17% in the high-risk group. When validated in a new population of 136 episodes of fever and neutropenia, the incidences of SBI and bacteremia in the low-risk group were 12% and 5%, respectively, and 25% and 19% in the high-risk group.

Conclusion: Pediatric oncology outpatients with fever and neutropenia who present with an initial monocyte count of >/= 0.1 x 10(9)/L and do not have comorbidity or an abnormal CXR at the time of presentation are at lower risk for SBI and can be considered for less aggressive initial therapy.