OBJECTIVE: Although diagnostic performance has recently improved by using new diagnostic methodologies, acute patient management is usually initiated after considering only fairly elementary findings of CSF examination. Using these early findings it is often difficult to distinguish between bacterial and aseptic (viral) meningitis. In order to help distinguish these two categories, scoring tools have been proposed that are more or less complex and validated. METHODS: The aim of this study was to establish a simple scoring tool and compare it to other available decision making systems. We retrospectively analysed all the meningitis cases from our patients at our institution and established a scoring tool for pediatric meningitis and for meningitis in adults by using categorized analysis tree methodology. RESULTS: Main categories for bacterial etiology were, leucocytosis >15 giga, CSF leucocytes count >1700 per ml, CSF neutrophil percentage >80, CSF protein >2.3g/l and glucose CSF/blood ratio <0.33 for adults and CSF leucocytes count >1800, CSF neutrophil percentage >80, CSF protein >1.2g/l and glucose CSF/blood ratio <0.3 for children. Additionally, our new scoring tool and five published ones were compared using our data and two external data sets; from these scores, three, including ours, exhibited good sensitivity and specificity. We then performed several thousand Monte Carlo simulations of both bacterial and viral meningitis for children and adults. We found that our scoring tool (Meningitest) had very high performances with positive and negative predictive values of 97% and 94%, respectively. CONCLUSION: Thus, from this analysis of five meningitis scoring systems, we believe that our new tool is simple, does not need any complex calculation and is effective in identifying bacterial vs viral meningitis in fully immunocompetent children and adults.
OBJECTIVE: Although diagnostic performance has recently improved by using new diagnostic methodologies, acute patient management is usually initiated after considering only fairly elementary findings of CSF examination. Using these early findings it is often difficult to distinguish between bacterial and aseptic (viral) meningitis. In order to help distinguish these two categories, scoring tools have been proposed that are more or less complex and validated. METHODS: The aim of this study was to establish a simple scoring tool and compare it to other available decision making systems. We retrospectively analysed all the meningitis cases from our patients at our institution and established a scoring tool for pediatric meningitis and for meningitis in adults by using categorized analysis tree methodology. RESULTS: Main categories for bacterial etiology were, leucocytosis >15 giga, CSF leucocytes count >1700 per ml, CSF neutrophil percentage >80, CSF protein >2.3g/l and glucose CSF/blood ratio <0.33 for adults and CSF leucocytes count >1800, CSF neutrophil percentage >80, CSF protein >1.2g/l and glucose CSF/blood ratio <0.3 for children. Additionally, our new scoring tool and five published ones were compared using our data and two external data sets; from these scores, three, including ours, exhibited good sensitivity and specificity. We then performed several thousand Monte Carlo simulations of both bacterial and viral meningitis for children and adults. We found that our scoring tool (Meningitest) had very high performances with positive and negative predictive values of 97% and 94%, respectively. CONCLUSION: Thus, from this analysis of five meningitis scoring systems, we believe that our new tool is simple, does not need any complex calculation and is effective in identifying bacterial vs viral meningitis in fully immunocompetent children and adults.
Authors: Nguyen T Huy; Nguyen T H Thao; Doan T N Diep; Mihoko Kikuchi; Javier Zamora; Kenji Hirayama Journal: Crit Care Date: 2010-12-31 Impact factor: 9.097
Authors: Nguyen Tien Huy; Nguyen Thanh Hong Thao; Nguyen Anh Tuan; Nguyen Tuan Khiem; Christopher C Moore; Doan Thi Ngoc Diep; Kenji Hirayama Journal: PLoS One Date: 2012-11-28 Impact factor: 3.240