1 October 2011 in Infection Control & Hospital Epidemiology
Objective.: Employees of long-term care facilities (LTCFs) who have contact with residents should be vaccinated against influenza annually to reduce influenza incidence among residents. This investigation estimated the magnitude of the benefit of this recommendation.Methods.: The New Mexico Department of Health implemented active surveillance in all of its 75 LTCFs during influenza seasons 2006-2007 and 2007-2008. Information about the number of laboratory-confirmed cases of influenza and the proportion vaccinated of both residents and direct-care employees in each facility was collected monthly. LTCFs reporting at least 1 case of influenza (defined alternately by laboratory confirmation or symptoms of influenza-like illness [ILI]) among residents were compared with LTCFs reporting no cases of influenza. Regression modeling was used to obtain adjusted odds ratios (aORs) and 95% confidence intervals (CIs) for the association between employee vaccination coverage and the occurrence of influenza outbreaks. Covariates included vaccination coverage among residents, the staff-to-resident ratio, and the proportion of filled beds.Results.: Seventeen influenza outbreaks were reported during this 2-year period of surveillance. Eleven of these were laboratory confirmed (n = 21 residents) and 6 were defined by ILI (n = 40 residents). Mean influenza vaccination coverage among direct-care employees was 51% in facilities reporting outbreaks and 60% in facilities not reporting outbreaks (P = .12). Increased vaccination coverage among direct-care employees was associated with fewer reported outbreaks of laboratory-confirmed influenza (aOR, 0.97 [95% CI, 0.95-0.99]) and ILI (aOR, 0.98 [95% CI, 0.96-1.00]).Conclusions.: High vaccination coverage among direct-care employees helps to prevent influenza in LTCFs.
17 September 2014 in Computational Statistics
10 March 2016 in Journal of Statistical Computation and Simulation
This article discusses two asymmetrization methods, Azzalini's representation and beta generation, to generate asymmetric bimodal models including two novel beta-generated models. The practical utility of these models is assessed with nine data sets from different fields of applied sciences. Besides this tutorial assessment, some methodological contributions are made: a random number generator for the asymmetric Rathie–Swamee model is developed (generators for the other models are already known and briefly described) and a new likelihood ratio test of unimodality is compared via simulations with other available tests. Several tools have been used to quantify and test for bimodality and assess goodness of fit including Bayesian information criterion, measures of agreement with the empirical distribution and the Kolmogorov–Smirnoff test. In the nine case studies, the results favoured models derived from Azzalini's asymmetrization, but no single model provided a best fit across the applications considered. In only two cases the normal mixture was selected as best model. Parameter estimation has been done by likelihood maximization. Numerical optimization must be performed with care since local optima are often present. We concluded that the models considered are flexible enough to fit different bimodal shapes and that the tools studied should be used with care and attention to detail.
29 June 2020 in Quality and Reliability Engineering International
27 November 2017 in Communications in Statistics - Theory and Methods
JAJoanlise Marco de Leon Andrade
Joanlise Marco de Leon Andrade
Maximum likelihood (ML) estimation of relative risks via log-binomial regression requires a restricted parameter space. Computation via non linear programming is simple to implement and has high convergence rate. We show that the optimization problem is well posed (convex domain and convex objective) and provide a variance formula along with a methodology for obtaining standard errors and prediction intervals which account for estimates on the boundary of the parameter space. We performed simulations under several scenarios already used in the literature in order to assess the performance of ML and of two other common estimation methods.
1 January 2021 in Cadernos de Saude Publica
Using data collected by the Brazilian National Household Sample Survey - COVID-19 (PNAD-COVID19) and semi-Bayesian modelling developed by Wu et al., we have estimated the effect of underreporting of COVID-19 cases in Brazil as of December 2020. The total number of infected individuals is about 3 to 8 times the number of cases reported, depending on the state. Confirmed cases are at 3.1% of the total population and our estimate of total cases is at almost 15% of the approximately 212 million Brazilians as of 2020. The method we adopted from Wu et al., with slight modifications in prior specifications, applies bias corrections to account for incomplete testing and imperfect test accuracy. Our estimates, which are comparable to results obtained by Wu et al. for the United States, indicate that projections from compartmental models (such as SEIR models) tend to overestimate the number of infections and that there is considerable regional heterogeneity (results are presented by state).