Ebola Virus Disease, Democratic Republic of the Congo

•Nonparametric Survival Analyses,  Actuarial Forecasts, and Prediction Limits

•Data are case reports, cures (gueries), and deaths for population and health workers: similar to SIR model with both deaths and cures

–Some missing data: ignored, imputed, or truncated inadmissible counts

–Minor alterations, usually count reductions

•Methods: max. likelihood, least squares, and bootstrap

–Alterations and missing data mess up estimators, but nonparametric max. likelihood (npmle) and nonparametric least squares (nplse) estimators are pretty robust and agreeable

•Transient Markov chain model of deaths and cures agrees too

•Actuarial forecasts, including standard deviation of forecast,

–Account for covariance of actuarial rate estimators and extrapolations but not change of trend or variance of case reports

•Prediction (tolerance) upper limits accounting for estimator uncertainty and extrapolations

Please see EVD20189.PDF at link below.

Ebola Virus Maladie République démocratique du Congo

Survie des Analyses et prévisions

•Résumé : Les Analyses de survie non paramétrique, prévisions actuarielles et prédiction limites

•Données sont des agents de santé, décès, gueries, rapports de cas certaines données manquantes :

•Méthodes: maximum de vraisemblance, moindres carrés, et bootstrap

–Modifications mineures de chefs d’accusation irrecevable ignorés, imputée ou tronqué, généralement compter des réductions