PURPOSE: Febrile neutropenia is a serious complication among cancer patients receiving myelosuppressive chemotherapy. Patient-specific risk factors, chemotherapy-related and disease-related characteristics can affect the clinical outcome and management of febrile neutropenia. Although many factors have been identified, they vary among different patient populations. We identified clinically-relevant febrile neutropenia predictors in Asian cancer patients through visualization of these factors. METHODS: A single-centered, retrospective study was conducted from May to July 2011 at a local cancer center. Demographics and risk factor data were collated from electronic health records and four cancer registries. Data were summarized using descriptive statistics. Additionally, potential febrile neutropenia predictors were identified using categorical principal component and multiple correspondence analyses. RESULTS: A total of 583 patients were analyzed. Majority was females (79%), Chinese (75%) and diagnosed with breast cancers (60%). Six risk factors were identified as potential predictors: types of cancer (16.9-19.8% of variance), chemotherapy regimen (anthracycline-based 11.8-12.9%, taxane-based 8.1%), liver function tests (alanine transaminase 8.6%, alkaline phosphatase 4.0%), renal function tests (serum creatinine 3.1%), prior granulocyte colony stimulating factor use (5.6%) and diabetes mellitus (6.6-6.9%). In terms of cancer types, lymphomas were more predictive than breast cancers. CONCLUSION: From our knowledge, this is the first study that has identified clinically-relevant febrile neutropenia predictors in Asian cancer patients through visualization of their risk factors. The use of these predictors to identify patients at risk for adverse reactions, such as FN, can allow clinicians to optimize prophylactic granulocyte colony stimulating factor usage in these patients.
PURPOSE:Febrile neutropenia is a serious complication among cancerpatients receiving myelosuppressive chemotherapy. Patient-specific risk factors, chemotherapy-related and disease-related characteristics can affect the clinical outcome and management of febrile neutropenia. Although many factors have been identified, they vary among different patient populations. We identified clinically-relevant febrile neutropenia predictors in Asian cancerpatients through visualization of these factors. METHODS: A single-centered, retrospective study was conducted from May to July 2011 at a local cancer center. Demographics and risk factor data were collated from electronic health records and four cancer registries. Data were summarized using descriptive statistics. Additionally, potential febrile neutropenia predictors were identified using categorical principal component and multiple correspondence analyses. RESULTS: A total of 583 patients were analyzed. Majority was females (79%), Chinese (75%) and diagnosed with breast cancers (60%). Six risk factors were identified as potential predictors: types of cancer (16.9-19.8% of variance), chemotherapy regimen (anthracycline-based 11.8-12.9%, taxane-based 8.1%), liver function tests (alanine transaminase 8.6%, alkaline phosphatase 4.0%), renal function tests (serum creatinine 3.1%), prior granulocyte colony stimulating factor use (5.6%) and diabetes mellitus (6.6-6.9%). In terms of cancer types, lymphomas were more predictive than breast cancers. CONCLUSION: From our knowledge, this is the first study that has identified clinically-relevant febrile neutropenia predictors in Asian cancerpatients through visualization of their risk factors. The use of these predictors to identify patients at risk for adverse reactions, such as FN, can allow clinicians to optimize prophylactic granulocyte colony stimulating factor usage in these patients.
Entities:
Keywords:
Clinical predictors; data reduction techniques; febrile neutropenia; multiple correspondence analysis; principal component analysis