Cargando…
Mortality-associated clinical risk factors in patients with febrile neutropenia: A retrospective study
OBJECTIVES: To determine the predictors for 30-day all-cause mortality in patients with febrile neutropenia (FN) and develop a prediction score. METHODS: The electronic medical records of patients undergoing chemotherapy with FN between 2018 and 2019 were reviewed. Multivariate logistic regression w...
Autores principales: | , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Elsevier
2021
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9216401/ https://www.ncbi.nlm.nih.gov/pubmed/35757817 http://dx.doi.org/10.1016/j.ijregi.2021.09.002 |
Sumario: | OBJECTIVES: To determine the predictors for 30-day all-cause mortality in patients with febrile neutropenia (FN) and develop a prediction score. METHODS: The electronic medical records of patients undergoing chemotherapy with FN between 2018 and 2019 were reviewed. Multivariate logistic regression was performed to identify factors associated with 30-day all-cause mortality to develop a parsimonious model. A prediction score was developed from the model's coefficients of each predictor. RESULTS: There were 273 FN episodes in 153 patients. The overall mortality rate was 12.5%. Pre-existing cardiovascular disease (OR 22.45), alteration of consciousness on admission (OR 18.50), anemia (OR 4.33), acute kidney injury (AKI) (OR 13.15), causative pathogen identified (OR 8.68), intensive care unit admission (OR 0.13), septic shock (OR 18.72), and the need for mechanical ventilation (OR 22.65) were associated with mortality. After exploring confounding effects between factors, septic shock, anemia, AKI, and the need for mechanical ventilation were selected to develop the prediction score which provided good sensitivity (87.88%) and specificity (90.91%) with an area under the ROC curve of 0.8939. CONCLUSIONS: Septic shock, anemia, AKI, and the need for mechanical ventilation were associated with FN mortality. Our prediction score is effective in discriminating high and low-risk patients for mortality. |
---|