Ruijun Ji1, Haipeng Shen1, Yuesong Pan1, Wanliang Du1, Penglian Wang1, Gaifen Liu1, Yilong Wang1, Hao Li1, Xingquan Zhao1, Yongjun Wang2. 1. From the Tiantan Comprehensive Stroke Center, Tiantan Hospital, Capital Medical University, Beijing, China (R.J., Y.P., W.D., P.W., G.L., Yilong Wang, H.L., X.Z., Yongjun Wang); China National Clinical Research Center for Neurological Diseases, Beijing, China (R.J., Y.P., W.D., P.W., G.L., Yilong Wang, H.L., X.Z., Yongjun Wang); and Department of Statistics and Operation Research, University of North Carolina, Chapel Hill (H.S.). 2. From the Tiantan Comprehensive Stroke Center, Tiantan Hospital, Capital Medical University, Beijing, China (R.J., Y.P., W.D., P.W., G.L., Yilong Wang, H.L., X.Z., Yongjun Wang); China National Clinical Research Center for Neurological Diseases, Beijing, China (R.J., Y.P., W.D., P.W., G.L., Yilong Wang, H.L., X.Z., Yongjun Wang); and Department of Statistics and Operation Research, University of North Carolina, Chapel Hill (H.S.). yongjunwang1962@gmail.com.
Abstract
BACKGROUND AND PURPOSE: We aimed to develop a risk score (intracerebral hemorrhage-associated pneumonia score, ICH-APS) for predicting hospital-acquired stroke-associated pneumonia (SAP) after ICH. METHODS: The ICH-APS was developed based on the China National Stroke Registry (CNSR), in which eligible patients were randomly divided into derivation (60%) and validation (40%) cohorts. Variables routinely collected at presentation were used for predicting SAP after ICH. For testing the added value of hematoma volume measure, we separately developed 2 models with (ICH-APS-B) and without (ICH-APS-A) hematoma volume included. Multivariable logistic regression was performed to identify independent predictors. The area under the receiver operating characteristic curve (AUROC), Hosmer-Lemeshow goodness-of-fit test, and integrated discrimination index were used to assess model discrimination, calibration, and reclassification, respectively. RESULTS: The SAP was 16.4% and 17.7% in the overall derivation (n=2998) and validation (n=2000) cohorts, respectively. A 23-point ICH-APS-A was developed based on a set of predictors and showed good discrimination in the overall derivation (AUROC, 0.75; 95% confidence interval, 0.72-0.77) and validation (AUROC, 0.76; 95% confidence interval, 0.71-0.79) cohorts. The ICH-APS-A was more sensitive for patients with length of stay >48 hours (AUROC, 0.78; 95% confidence interval, 0.75-0.81) than those with length of stay <48 hours (AUROC, 0.64; 95% confidence interval, 0.55-0.73). The ICH-APS-A was well calibrated (Hosmer-Lemeshow test) in the derivation (P=0.20) and validation (P=0.66) cohorts. Similarly, a 26-point ICH-APS-B was established. The ICH-APS-A and ICH-APS-B were not significantly different in discrimination and reclassification for SAP after ICH. CONCLUSION: The ICH-APSs are valid risk scores for predicting SAP after ICH, especially for patients with length of stay >48 hours.
BACKGROUND AND PURPOSE: We aimed to develop a risk score (intracerebral hemorrhage-associated pneumonia score, ICH-APS) for predicting hospital-acquired stroke-associated pneumonia (SAP) after ICH. METHODS: The ICH-APS was developed based on the China National Stroke Registry (CNSR), in which eligible patients were randomly divided into derivation (60%) and validation (40%) cohorts. Variables routinely collected at presentation were used for predicting SAP after ICH. For testing the added value of hematoma volume measure, we separately developed 2 models with (ICH-APS-B) and without (ICH-APS-A) hematoma volume included. Multivariable logistic regression was performed to identify independent predictors. The area under the receiver operating characteristic curve (AUROC), Hosmer-Lemeshow goodness-of-fit test, and integrated discrimination index were used to assess model discrimination, calibration, and reclassification, respectively. RESULTS: The SAP was 16.4% and 17.7% in the overall derivation (n=2998) and validation (n=2000) cohorts, respectively. A 23-point ICH-APS-A was developed based on a set of predictors and showed good discrimination in the overall derivation (AUROC, 0.75; 95% confidence interval, 0.72-0.77) and validation (AUROC, 0.76; 95% confidence interval, 0.71-0.79) cohorts. The ICH-APS-A was more sensitive for patients with length of stay >48 hours (AUROC, 0.78; 95% confidence interval, 0.75-0.81) than those with length of stay <48 hours (AUROC, 0.64; 95% confidence interval, 0.55-0.73). The ICH-APS-A was well calibrated (Hosmer-Lemeshow test) in the derivation (P=0.20) and validation (P=0.66) cohorts. Similarly, a 26-point ICH-APS-B was established. The ICH-APS-A and ICH-APS-B were not significantly different in discrimination and reclassification for SAP after ICH. CONCLUSION: The ICH-APSs are valid risk scores for predicting SAP after ICH, especially for patients with length of stay >48 hours.
Authors: Sandro Marini; Andrea Morotti; Umme K Lena; Joshua N Goldstein; Steven M Greenberg; Jonathan Rosand; Christopher D Anderson Journal: Neurocrit Care Date: 2018-02 Impact factor: 3.210
Authors: Amit K Kishore; Andy Vail; Benjamin D Bray; Angel Chamorro; Mario Di Napoli; Lalit Kalra; Peter Langhorne; Joan Montaner; Christine Roffe; Anthony G Rudd; Pippa J Tyrrell; Diederik van de Beek; Mark Woodhead; Andreas Meisel; Craig J Smith Journal: Eur Stroke J Date: 2016-06-01
Authors: Willeke F Westendorp; Jan-Dirk Vermeij; Nina A Hilkens; Matthijs C Brouwer; Ale Algra; H Bart van der Worp; Diederik Wj Dippel; Diederik van de Beek; Pual J Nederkoorn Journal: Eur Stroke J Date: 2018-03-08