Predictive assessment of a non-linear random effects model for multivariate time series of infectious disease counts

Infectious disease counts from surveillance systems are typically observed in several administrative geographical areas. In this paper, a non-linear model for the analysis of such multiple time series of counts is discussed. To account for heterogeneous incidence levels or varying transmission of a pathogen across regions, region-specific and possibly spatially correlated random effects are introduced. Inference is based on penalized likelihood methodology for mixed models. Since the use of classical model choice criteria such as AIC or BIC can be problematic in the presence of random effects, models are compared by means of one-step-ahead predictions and proper scoring rules. In a case study, the model is applied to monthly counts of meningococcal disease cases in 94 departments of France (excluding Corsica) and weekly counts of influenza cases in 140 administrative districts of Southern Germany. The predictive performance improves if existing heterogeneity is accounted for by random effects.     

A spatio-temporal absorbing state model for disease and syndromic surveillance

Reliable surveillance models are an important tool in public health because they aid in mitigating disease outbreaks, identify where and when disease outbreaks occur, and predict future occurrences. Although many statistical models have been devised for surveillance purposes, none are able to simultaneously achieve the important practical goals of good sensitivity and specificity, proper use of covariate information, inclusion of spatio-temporal dynamics, and transparent support to decision-makers. In an effort to achieve these goals, this paper proposes a spatio-temporal conditional autoregressive hidden Markov model with an absorbing state. The model performs well in both a large simulation study and in an application to influenza/pneumonia fatality data.     

Using Social Media to Perform Local Influenza Surveillance in an Inner-City Hospital: A Retrospective Observational Study

Background

Public health officials and policy makers in the United States expend significant resources at the national, state, county, and city levels to measure the rate of influenza infection. These individuals rely on influenza infection rate information to make important decisions during the course of an influenza season driving vaccination campaigns, clinical guidelines, and medical staffing. Web and social media data sources have emerged as attractive alternatives to supplement existing practices. While traditional surveillance methods take 1-2 weeks, and significant labor, to produce an infection estimate in each locale, web and social media data are available in near real-time for a broad range of locations.

Objective

The objective of this study was to analyze the efficacy of flu surveillance from combining data from the websites Google Flu Trends and HealthTweets at the local level. We considered both emergency department influenza-like illness cases and laboratory-confirmed influenza cases for a single hospital in the City of Baltimore.

Methods

This was a retrospective observational study comparing estimates of influenza activity of Google Flu Trends and Twitter to actual counts of individuals with laboratory-confirmed influenza, and counts of individuals presenting to the emergency department with influenza-like illness cases. Data were collected from November 20, 2011 through March 16, 2014. Each parameter was evaluated on the municipal, regional, and national scale. We examined the utility of social media data for tracking actual influenza infection at the municipal, state, and national levels. Specifically, we compared the efficacy of Twitter and Google Flu Trends data.

Results

We found that municipal-level Twitter data was more effective than regional and national data when tracking actual influenza infection rates in a Baltimore inner-city hospital. When combined, national-level Twitter and Google Flu Trends data outperformed each data source individually. In addition, influenza-like illness data at all levels of geographic granularity were best predicted by national Google Flu Trends data.

Conclusions

In order to overcome sensitivity to transient events, such as the news cycle, the best-fitting Google Flu Trends model relies on a 4-week moving average, suggesting that it may also be sacrificing sensitivity to transient fluctuations in influenza infection to achieve predictive power. Implications for influenza forecasting are discussed in this report.     

Modeling the epidemic waves of AH1N1/09 influenza around the world

The 2009 swine flu pandemic was a global outbreak of a new strain of H1N1 influenza virus and there are more than 14,000 confirmed deaths worldwide. The aim of this paper is to propose new mathematical models to study different dynamics of H1N1 influenza virus spread in selected regions around the world. Spatial and temporal elements are included in these models to reproduce the dynamics of AH1N1/09 virus. Different models are used since H1N1 influenza virus spread in regions with different contact structures are not the same. We rely on time series notifications of individuals to estimate some of the parameters of the models. We find that, in order to reproduce the time series data and the spread of the disease, it is convenient to suggest spatio-temporal models. Regions with only one wave are modeled with the classical SEIR model and regions with multiple waves using models with spatio-temporal elements. These results help to explain and understand about potential mechanisms behind the spread of AH1N1 influenza virus in different regions around the world.     

Shared surveillance: meningococcal disease vs influenza

OBJECTIVE: To analyse the association between the behavior of meningococcal disease and influenza, using for this purpose population statistics for Spain for the period of 1964 to 1997. METHODS: Ecological study of the incidence of meningococcal disease and influenza in Spain from 1964 to 1997, inclusive. The study used weekly statistical data for these diseases supplied by the Compulsory Disease Reporting System (Enfermedades de Declaración Obligatoria, EDO). The deterministic component of the meningococcal disease and influenza series was studied by means of spectral analysis based on the Fast Fourier Transformation, and the non-deterministic component was studied using the ARIMA model. The Box-Jenkins method was used for pre-bleaching the series, and a cross-correlation was subsequently established between the residuals in order to detect the presence of any significant correlations between the meningococcal disease and influenza series. RESULTS: During the period from 1964 to 1997, the week that showed, on average, the greatest number of cases for the season was week 7 in the case of meningococcal disease and week 6 in the case of influenza. Spectral analysis of the meningococcal disease and influenza series clearly demonstrated the annual periodicity of both series, and periodicity of nearly 11 years for meningococcal disease and slightly over 10 years for influenza. When cross-correlation is established after prebleaching the series, positive correlations are obtained in the results of lags 0, 1, 2, and 3. Introducing influenza as an exogenous variable in the multivariate model of meningococcal disease corroborates these results. There was a statistically significant relationship between the two processes during the same week and with a three-week lapse. CONCLUSIONS: By means of a methodology not previously applied to this subject, and by the use of prolonged time-span, country-comprehensive population statistics (which includes several epidemics waves), an association was shown to exist between meningococcal disease and influenza. This suggests the need for the surveillance of the two processes in an interrelated manner.     

Diagnosing influenza: the value of clinical clues and laboratory tests.

OBJECTIVE: Our goal was to determine the utility of clinical clues, white blood cell (WBC) and differential counts, and a rapid antigen test for differentiating influenza from coexistent infectious diseases during influenza epidemics. STUDY DESIGN: Data were collected during 3 consecutive influenza outbreaks over a 2-year period. The information collected included date of onset, symptoms, vaccine status, WBC and differential counts, ZstatFlu test (ZymeTx, Oklahoma City, Ok), and influenza culture. Using culture positivity as the criterion for influenza diagnosis, we compared cases with noncases on each variable independently and by logistic regression. Receiver operating characteristic curves were plotted for WBC count, ZstatFlu, and their combination in an effort to determine the most useful diagnostic strategy. POPULATION: We included consecutive patients presenting to a family practice office with fever, cough, sore throat, myalgia, and/or headache during flu season. OUTCOMES MEASURED: The outcomes were sensitivity, specificity, and other measures of test accuracy. RESULTS: Culture-positive cases could not be reliably distinguished from those that were culture negative using symptoms or vaccination status. Both WBC count and ZstatFlu results discriminated fairly well, and their combination did somewhat better. Differential counts were not helpful. WBC counts above 8000 were associated with a low probability of influenza. The sensitivity and specificity of the ZstatFlu were 65% and 83%, respectively. CONCLUSIONS: Our data suggest that symptoms and vaccine status do not reliably identify patients with influenza. Use of WBC counts and the ZstatFlu test can be helpful. The sequence, combination, and criteria for use of these tests depend on tradeoffs between undertreatment of influenza cases and the overtreatment of noninfluenza cases, and the cost and benefit projections for individual patients.     

贝叶斯时空模型在传染病数据时空分析中的研究进展

传染病数据属于典型的时空数据,贝叶斯时空模型在传染病数据时空分析中有其独特优势,它可以综合分析传染病时空数据中蕴含的时间和空间信息。本文综述了贝叶斯时空模型的基本思想、常用模型类型、在传染病领域的应用现况、优势及注意事项。     

Influenza and Pneumococcal Vaccination Demand Responses to Changes in Infectious Disease Mortality

Objective. To test the hypothesis that individuals are more likely to receive a vaccination against influenza or pneumonia as the perceived disease threat increases.
Data Sources. This study uses two different national datasets. Individual-level information about the vaccination rates of 38,768 elderly persons are from the Behavioral Risk Factor Surveillance System, 1993–1998. Information on the combined influenza and pneumonia state mortality rates are measured from the Compressed Mortality File.
Study Design. Using both cross-sectional and state fixed-effects panel data estimators, we model an individual's probability of having an influenza or pneumococcal vaccination as a function of the lagged state mortality rate. Multiyear lags are specified in order to estimate the duration of the effect of disease mortality on individual vaccination behavior.
Principal Findings. Results support our hypothesis that influenza vaccination behavior responds positively to disease mortality, even after a one-year lag. We further find that cross-sectional estimators used in previous work yield downward-biased estimates, although even for our preferred panel data models, the estimated effects are small.
Conclusions. The findings indicate that behavioral demand responses can help to limit infectious disease epidemics, and suggest further research on how public awareness campaigns can mediate this disease threat responsiveness behavior.     

Modulation models for seasonal time series and incidence tables

We model monthly disease counts on an age-time grid using the two-dimensional varying-coefficient Poisson regression. Since the marginal profile of counts shows a very strong and varying annual cycle, sine and cosine regressors model periodicity, but their coefficients are allowed to vary smoothly over the age and time plane. The coefficient surfaces are estimated using a relatively large tensor product B-spline basis. Smoothness is tuned using difference penalties on the rows and columns of the tensor product coefficients. Heavy over-dispersion occurs, making it impossible to use Akaike's information criterion or Bayesian information criterion based on a Poisson likelihood. It is handled by selective weighting of part of the data and by the use of extended quasi-likelihood. Very efficient computation is achieved with fast array algorithms. The model is applied to monthly deaths due to respiratory diseases, for U.S. females during 1959-1998 and for ages 51-100.     

Real‐time surveillance for abnormal events: the case of influenza outbreaks

This paper introduces a method of surveillance using deviations from probabilistic forecasts. Realised observations are compared with probabilistic forecasts, and the “deviation” metric is based on low probability events. If an alert is declared, the algorithm continues to monitor until an all‐clear is announced. Specifically, this article addresses the problem of syndromic surveillance for influenza (flu) with the intention of detecting outbreaks, due to new strains of viruses, over and above the normal seasonal pattern. The syndrome is hospital admissions for flu‐like illness, and hence, the data are low counts. In accordance with the count properties of the observations, an integer‐valued autoregressive process is used to model flu occurrences. Monte Carlo evidence suggests the method works well in stylised but somewhat realistic situations. An application to real flu data indicates that the ideas may have promise. The model estimated on a short run of training data did not declare false alarms when used with new observations deemed in control, ex post. The model easily detected the 2009 H1N1 outbreak. Copyright © 2016 John Wiley & Sons, Ltd.     

Exploring spatio-temporal patterns of mortality using mixed effects models

A linear mixed effects (LME) model previously used for a spatial analysis of mortality data for a single time period is extended to include time trends and spatio-temporal interactions. This model includes functions of age and time period that can account for increasing and decreasing death rates over time and age, and a change-point of rates at a predetermined age. A geographic hierarchy is included that provides both regional and small area age-specific rate estimates, stabilizing rates based on small numbers of deaths by sharing information within a region. The proposed log-linear analysis of rates allows the use of commercially available software for parameter estimation, and provides an estimator of overdispersion directly as the residual variance. Because of concerns about the accuracy of small area rate estimates when there are many instances of no observed deaths, we consider potential sources of error, focusing particularly on the similarity of likelihood inferences using the LME model for rates as compared to an exact Poisson-normal mixed effects model for counts. The proposed LME model is applied to breast cancer deaths which occurred among white women during 1979-1996. For this example, application of diagnostics for multiparameter likelihood comparisons suggests a restriction of age to a minimum of either 25 or 35, depending on whether small area rate estimates are required. Investigation into a convergence problem led to the discovery that the changes in breast cancer geographic patterns over time are related more to urbanization than to region, as previously thought. Published in 2000 by John Wiley & Sons, Ltd.     

A cluster model for space-time disease counts

Modelling disease clustering over space and time can be helpful in providing indications of possible exposures and planning corresponding public health practices. Though a considerable number of studies focus on modelling spatio-temporal patterns of disease, most of them do not directly model a spatio-temporal clustering structure and could be ineffective for detecting clusters. In this paper, we extend a purely spatial cluster model to accommodate space-time clustering. Inference is performed in a Bayesian framework using reversible jump Markov chain Monte Carlo. This idea is illustrated using data on female breast cancer mortality from Japan. A hierarchical parametric space-time model for mapping disease is used for comparison.     

An integer-valued time series model for multivariate surveillance

In recent days, different types of surveillance data are becoming available for public health purposes. In most cases, several variables are monitored and events of different types are reported. As the amount of surveillance data increases, statistical methods that can effectively address multivariate surveillance scenarios are demanded. Even though research activity in this field is increasing rapidly in recent years, only a few approaches have simultaneously addressed the integer-valued property of the data and its correlation (both time correlation and cross-correlation) structure. In this article, we suggest a multivariate integer-valued autoregressive model that allows for both serial and cross-correlations between the series and can easily accommodate overdispersion and covariate information. Moreover, its structure implies a natural decomposition into an endemic and an epidemic component, a common distinction in dynamic models for infectious disease counts. Detection of disease outbreaks is achieved through the comparison of surveillance data with one-step-ahead predictions obtained after fitting the suggested model to a set of clean historical data. The performance of the suggested model is illustrated on a trivariate series of syndromic surveillance data collected during Athens 2004 Olympic Games.     

Development of a model to aid in reconstruction of historical silica dust exposures in the taconite industry

The ability to reconstruct employee exposure histories would be a valuable research tool for the evaluation of occupation as a factor in disease. In many cases, however, historical environmental data are available but have not been used to compute past exposures because of differences in sampling methods. This paper describes a quantitative model to convert historical environmental data (from taconite mine and mill operations) into a form consistent with current sampling methods and results and, therefore, will enable past exposure histories to be used. (Past exposure histories are to be determined in an epidemiological study.) In this study, parallel sampling results from the environmental data base were used to obtain a coefficient for the conversion of impinger-particle counts (old sampling method) to filter-respirable mass sampling results (new sampling method). Parameters in the model were estimated using multiple regression techniques. Results show that a consistent ratio exists between impinger-particle counts and filter-respirable mass concentrations for samples collected at the same locations.     

The Role of Subway Travel in an Influenza Epidemic: A New York City Simulation

The interactions of people using public transportation in large metropolitan areas may help spread an influenza epidemic. An agent-based model computer simulation of New York City’s (NYC’s) five boroughs was developed that incorporated subway ridership into a Susceptible–Exposed–Infected–Recovered disease model framework. The model contains a total of 7,847,465 virtual people. Each person resides in one of the five boroughs of NYC and has a set of socio-demographic characteristics and daily behaviors that include age, sex, employment status, income, occupation, and household location and membership. The model simulates the interactions of subway riders with their workplaces, schools, households, and community activities. It was calibrated using historical data from the 1957–1958 influenza pandemics and from NYC travel surveys. The surveys were necessary to enable inclusion of subway riders into the model. The model results estimate that if influenza did occur in NYC with the characteristics of the 1957–1958 pandemic, 4% of transmissions would occur on the subway. This suggests that interventions targeted at subway riders would be relatively ineffective in containing the epidemic. A number of hypothetical examples demonstrate this feature. This information could prove useful to public health officials planning responses to epidemics.     

Cost-effectiveness and value of information analyses of neuraminidase inhibitors for the treatment of influenza

OBJECTIVES: To assess the cost-effectiveness of alternative strategies for the treatment of suspected influenza in otherwise healthy adults and to identify future research priorities using value of information analysis. METHODS: A decision model was used to estimate the costs and effects, in terms of quality-adjusted life-years (QALYs) of amantadine, zanamivir, and oseltamivir for the treatment of influenza in otherwise healthy adults using data predominantly from meta-analysis of randomized controlled trials. Probabilistic sensitivity analysis using Monte Carlo simulation was conducted. The expected value of perfect information for the entire model and for individual parameters was calculated. RESULTS: Based on mean costs and effects, zanamivir is dominated by oseltamivir. The incremental cost-effectiveness ratio for amantadine (compared with no treatment) is pound 11,000 and pound 44,000 for oseltamivir (compared with amantadine). The probability that amantadine is cost-effective at a willingness to pay of pound 30,000 per QALY is 0.74, falling to 0.49 at pound 20,000 per QALY. Global expected value of perfect information (EVPI) is pound 2 m over 15 years if a willingness to pay threshold of pound 30,000 per QALY is assumed rising to pound 9.6 m at pound 45,000 per QALY. EVPI for only one parameter exceeds pound 500,0000 at pound 30,000 per QALY: the quality of life for untreated influenza. CONCLUSIONS: At traditionally accepted values of willingness to pay for health benefits, it is unlikely that additional research would be an efficient use of scarce resources. The only exception to this would be to examine the health-related quality of life impact of influenza in an untreated patient group. If a higher threshold value were acceptable, there are a small group of parameters that may warrant further investigation. These would, however, require comparative, potentially expensive, research studies.     

Influenza surveillance.

The main objectives of influenza surveillance are: to measure the impact of the disease by collection and analysis of epidemiological information on morbidity and mortality, and to anticipate future epidemics and pandemics by the collection and analysis of influenza viruses. The World Health Organization''s influenza programme is based on the collaboration of 98 national influenza centres in 70 countries and the 2 WHO Collaborating Centres in Atlanta and London.