Robust outbreak surveillance of epidemics in Sweden
ABSTRACT Outbreak detection is of interest in connection with several diseases and syndromes. The aim is to detect the progressive increase in the incidence as soon as possible after the onset of the outbreak. A semiparametric method is applied to Swedish data on tularaemia and influenza. The method is constructed to detect a change from a constant level to a monotonically increasing incidence. If seasonal effects are present, the residuals from a model incorporating these can be used. The properties of the method are evaluated by application to Swedish data on tularaemia and influenza and by simulations. The suggested method is compared with subjective judgments as well as with other algorithms. The conclusion is that the method works well. A user-friendly computer program is described.
- SourceAvailable from: Anette Hulth01/2012; 3(1). DOI:10.4172/2157-7420.1000108
- [Show abstract] [Hide abstract]
ABSTRACT: Public health surveillance aims at lessening disease burden, e.g., in case of infectious diseases by timely recognizing emerging outbreaks. Seen from a statistical perspective, this implies the use of appropriate methods for monitoring time series of aggregated case reports. This paper presents the tools for such automatic aberration detection offered by the R package surveillance. We introduce the functionality for the visualization, mod-elling and monitoring of surveillance time series. With respect to modelling we focus on univariate time series modelling based on generalized linear models (GLMs), multivariate GLMs, generalized additive models and generalized additive models for location, shape and scale. This ranges from illustrating implementational improvements and extensions of the well-known Farrington algorithm, e.g., by spline-modelling or by treating it in a Bayesian context. Furthermore, we look at categorical time series and address overdisper-sion using beta-binomial or Dirichlet-Multinomial modelling. With respect to monitoring we consider detectors based on either a Shewhart-like single timepoint comparison between the observed count and the predictive distribution or by likelihood-ratio based cumulative sum methods. Finally, we illustrate how surveillance can support aberration detection in practice by integrating it into the monitoring workflow of a public health institution. Al-together, the present article shows how well surveillance can support automatic aberration detection in a public health surveillance context.
- [Show abstract] [Hide abstract]
ABSTRACT: Multivariate control charts are of interest in industrial production as they enable the joint monitoring of several components. Recently, there has been an increased interest also in other areas such as the detection of bioterrorism, transac-tion strategies in finance and the surveillance of outbreaks of epidemic diseases based on spatial information. Multivariate counterparts to the univariate Shewhart, EWMA, and CUSUM methods have earlier been proposed. General approaches to multivariate surveillance are reviewed. The challenges of evaluating multivariate surveillance meth-ods are of special concern. Optimality is usually hard to derive, and even to define, in multivariate problems. This is true also for multivariate surveillance. Multivariate on-line surveillance problems can be complex. The sufficiency principle can be of great use to clarify the structure of some problems. Here it is used to discuss metrics for evaluation of multivariate surveillance. It is demonstrated that the sufficiency principle allows important reductions of some classes of multivariate surveillance problems. This is used to determine optimal methods.