作者: A. M. Baart , W. L. A. M. de Kort , K. G. M. Moons , Y. Vergouwe
DOI: 10.1111/J.1423-0410.2010.01382.X
关键词:
摘要: Background and Objectives Each year, a relevant proportion of whole blood donors is deferred from donation because low haemoglobin (Hb) levels. Such temporary deferrals are demoralizing, may never return for donation. Reliable predictions Hb levels guide the decision whether can be invited next In this study, prediction model was developed risk levels. Materials Methods Individual data 5191 were analysed; 143 had level. Eleven candidate predictors considered in logistic regression models to predict The performance studied with receiver operating characteristic (ROC) curve. Internal validity assessed bootstrap procedure. Results Strong sex, seasonality, level measured at previous visit, difference between two visits, time since deferral total number donations past 2 years. validation showed an area under ROC curve 0·87. Conclusion provides accurate discrimination appropriate valuable determine donation, or some interventions such as postponement invitation warranted. Potentially, could decrease donor