BACKGROUND: The prognosis of febrile neutropenia (FN) in childhood cancer has been considerably improved by the intensification of treatment, including systematic hospitalization and broad-spectrum antibiotics. As only few children present with a severe bacterial infection (SBI), clinical decision rules have been developed to distinguish those at risk for SBI. The aim of this study was to evaluate the reproducibility of six clinical decision rules proposed in the literature and to compare their performance. METHODS: This retrospective two-center cohort study included all episodes of chemotherapy-induced FN in children admitted between January 2005 and December 2006. Each rule was applied to our patients. Their sensitivity (Se) and specificity (Sp) were calculated and compared with the authors' results, to assess reproducibility. The most predictive rule was defined in advance as that yielding 100% Se, the highest Sp, and the greatest simplicity for bedside application. RESULTS: Three hundred seventy-seven episodes of FN in 167 patients were collected; 64 episodes were associated with SBI, including 36 with bacteremia. Four of the six rules were reproducible, but none were able to be validated. The most predictive rule for bacteremia had 96% Se (95% confidence interval (CI): 79-99%) and 25% Sp (95% CI: 19-33%), and the most predictive rule for SBI had 95% Se (95% CI: 87-98%), but no power of discrimination (Sp = 5%, 95% CI: 3-8%). CONCLUSION: This study emphasizes the difficulty in identifying standardized decision rules in the management of a condition with numerous clinical variables like FN. Copyright 2010 Wiley-Liss, Inc.
BACKGROUND: The prognosis of febrile neutropenia (FN) in childhood cancer has been considerably improved by the intensification of treatment, including systematic hospitalization and broad-spectrum antibiotics. As only few children present with a severe bacterial infection (SBI), clinical decision rules have been developed to distinguish those at risk for SBI. The aim of this study was to evaluate the reproducibility of six clinical decision rules proposed in the literature and to compare their performance. METHODS: This retrospective two-center cohort study included all episodes of chemotherapy-induced FN in children admitted between January 2005 and December 2006. Each rule was applied to our patients. Their sensitivity (Se) and specificity (Sp) were calculated and compared with the authors' results, to assess reproducibility. The most predictive rule was defined in advance as that yielding 100% Se, the highest Sp, and the greatest simplicity for bedside application. RESULTS: Three hundred seventy-seven episodes of FN in 167 patients were collected; 64 episodes were associated with SBI, including 36 with bacteremia. Four of the six rules were reproducible, but none were able to be validated. The most predictive rule for bacteremia had 96% Se (95% confidence interval (CI): 79-99%) and 25% Sp (95% CI: 19-33%), and the most predictive rule for SBI had 95% Se (95% CI: 87-98%), but no power of discrimination (Sp = 5%, 95% CI: 3-8%). CONCLUSION: This study emphasizes the difficulty in identifying standardized decision rules in the management of a condition with numerous clinical variables like FN. Copyright 2010 Wiley-Liss, Inc.
Authors: Adam J Esbenshade; M Cecilia Di Pentima; Zhiguo Zhao; Ayumi Shintani; Jennifer C Esbenshade; Monique E Simpson; Kathleen C Montgomery; Robert B Lindell; Haerin Lee; Ato Wallace; Kelly L Garcia; Karel G M Moons; Debra L Friedman Journal: Pediatr Blood Cancer Date: 2014-10-18 Impact factor: 3.167
Authors: Gabrielle M Haeusler; Karin A Thursky; Francoise Mechinaud; Franz E Babl; Richard De Abreu Lourenco; Monica A Slavin; Robert Phillips Journal: Br J Cancer Date: 2017-06-13 Impact factor: 7.640
Authors: Annina N von Allmen; Maxime G Zermatten; Kurt Leibundgut; Philipp Agyeman; Roland A Ammann Journal: Sci Data Date: 2018-03-13 Impact factor: 6.444
Authors: Hilde T van der Galiën; Erik A H Loeffen; Karin G E Miedema; Wim J E Tissing Journal: Support Care Cancer Date: 2018-05-19 Impact factor: 3.603
Authors: Maxime G Zermatten; Christa Koenig; Annina von Allmen; Philipp Agyeman; Roland A Ammann Journal: Sci Data Date: 2019-01-15 Impact factor: 6.444