Demonstrating the use of high-volume electronic medical claims data to monitor local and regional influenza activity in the US.

INTRODUCTION: Fine-grained influenza surveillance data are lacking in the US, hampering our ability to monitor disease spread at a local scale. Here we evaluate the performances of high-volume electronic medical claims data to assess local and regional influenza activity. MATERIAL AND METHODS: We used electronic medical claims data compiled by IMS Health in 480 US locations to create weekly regional influenza-like-illness (ILI) time series during 2003-2010. IMS Health captured 62% of US outpatient visits in 2009. We studied the performances of IMS-ILI indicators against reference influenza surveillance datasets, including CDC-ILI outpatient and laboratory-confirmed influenza data. We estimated correlation in weekly incidences, peak timing and seasonal intensity across datasets, stratified by 10 regions and four age groups (<5, 5-29, 30-59, and 60+ years). To test IMS-Health performances at the city level, we compared IMS-ILI indicators to syndromic surveillance data for New York City. We also used control data on laboratory-confirmed Respiratory Syncytial Virus (RSV) activity to test the specificity of IMS-ILI for influenza surveillance. RESULTS: Regional IMS-ILI indicators were highly synchronous with CDC's reference influenza surveillance data (Pearson correlation coefficients rho≥0.89; range across regions, 0.80-0.97, P<0.001). Seasonal intensity estimates were weakly correlated across datasets in all age data (rho≤0.52), moderately correlated among adults (rho≥0.64) and uncorrelated among school-age children. IMS-ILI indicators were more correlated with reference influenza data than control RSV indicators (rho = 0.93 with influenza v. rho = 0.33 with RSV, P<0.05). City-level IMS-ILI indicators were highly consistent with reference syndromic data (rho≥0.86). CONCLUSION: Medical claims-based ILI indicators accurately capture weekly fluctuations in influenza activity in all US regions during inter-pandemic and pandemic seasons, and can be broken down by age groups and fine geographical areas. Medical claims data provide more reliable and fine-grained indicators of influenza activity than other high-volume electronic algorithms and should be used to augment existing influenza surveillance systems.

Spatial Transmission of 2009 Pandemic Influenza in the US.

The 2009 H1N1 influenza pandemic provides a unique opportunity for detailed examination of the spatial dynamics of an emerging pathogen. In the US, the pandemic was characterized by substantial geographical heterogeneity: the 2009 spring wave was limited mainly to northeastern cities while the larger fall wave affected the whole country. Here we use finely resolved spatial and temporal influenza disease data based on electronic medical claims to explore the spread of the fall pandemic wave across 271 US cities and associated suburban areas. We document a clear spatial pattern in the timing of onset of the fall wave, starting in southeastern cities and spreading outwards over a period of three months. We use mechanistic models to tease apart the external factors associated with the timing of the fall wave arrival: differential seeding events linked to demographic factors, school opening dates, absolute humidity, prior immunity from the spring wave, spatial diffusion, and their interactions. Although the onset of the fall wave was correlated with school openings as previously reported, models including spatial spread alone resulted in better fit. The best model had a combination of the two. Absolute humidity or prior exposure during the spring wave did not improve the fit and population size only played a weak role. In conclusion, the protracted spread of pandemic influenza in fall 2009 in the US was dominated by short-distance spatial spread partially catalysed by school openings rather than long-distance transmission events. This is in contrast to the rapid hierarchical transmission patterns previously described for seasonal influenza. The findings underline the critical role that school-age children play in facilitating the geographic spread of pandemic influenza and highlight the need for further information on the movement and mixing patterns of this age group.

Estimating the timing of mother-to-child transmission of the human immunodeficiency virus type 1 using a viral molecular evolution model.

BACKGROUND: Mother-to-child transmission (MTCT) is responsible for most pediatric HIV-1 infections worldwide. It can occur during pregnancy, labor, or breastfeeding. Numerous studies have used coalescent and molecular clock methods to understand the epidemic history of HIV-1, but the timing of vertical transmission has not been studied using these methods. Taking advantage of the constant accumulation of HIV genetic variation over time and using longitudinally sampled viral sequences, we used a coalescent approach to investigate the timing of MTCT. MATERIALS AND METHODS: Six-hundred and twenty-two clonal env sequences from the RNA and DNA viral population were longitudinally sampled from nine HIV-1 infected mother-and-child pairs [range: 277-1034 days]. For each transmission pair, timing of MTCT was determined using a coalescent-based model within a Bayesian statistical framework. Results were compared with available estimates of MTCT timing obtained with the classic biomedical approach based on serial HIV DNA detection by PCR assays. RESULTS: Four children were infected during pregnancy, whereas the remaining five children were infected at time of delivery. For eight out of nine pairs, results were consistent with the transmission periods assessed by standard PCR-based assay. The discordance in the remaining case was likely confused by co-infection, with simultaneous introduction of multiple maternal viral variants at the time of delivery. CONCLUSIONS: The study provided the opportunity to validate the Bayesian coalescent approach that determines the timing of MTCT of HIV-1. It illustrates the power of population genetics approaches to reliably estimate the timing of transmission events and deepens our knowledge about the dynamics of viral evolution in HIV-infected children, accounting for the complexity of multiple transmission events.

Impact of birth seasonality on dynamics of acute immunizing infections in Sub-Saharan Africa.

We analyze the impact of birth seasonality (seasonal oscillations in the birth rate) on the dynamics of acute, immunizing childhood infectious diseases. Previous research has explored the effect of human birth seasonality on infectious disease dynamics using parameters appropriate for the developed world. We build on this work by including in our analysis an extended range of baseline birth rates and amplitudes, which correspond to developing world settings. Additionally, our analysis accounts for seasonal forcing both in births and contact rates. We focus in particular on the dynamics of measles. In the absence of seasonal transmission rates or stochastic forcing, for typical measles epidemiological parameters, birth seasonality induces either annual or biennial epidemics. Changes in the magnitude of the birth fluctuations (birth amplitude) can induce significant changes in the size of the epidemic peaks, but have little impact on timing of disease epidemics within the year. In contrast, changes to the birth seasonality phase (location of the peak in birth amplitude within the year) significantly influence the timing of the epidemics. In the presence of seasonality in contact rates, at relatively low birth rates (20 per 1000), birth amplitude has little impact on the dynamics but does have an impact on the magnitude and timing of the epidemics. However, as the mean birth rate increases, both birth amplitude and phase play an important role in driving the dynamics of the epidemic. There are stronger effects at higher birth rates.

Linking time-varying symptomatology and intensity of infectiousness to patterns of norovirus transmission.

BACKGROUND: Norovirus (NoV) transmission may be impacted by changes in symptom intensity. Sudden onset of vomiting, which may cause an initial period of hyper-infectiousness, often marks the beginning of symptoms. This is often followed by: a 1-3 day period of milder symptoms, environmental contamination following vomiting, and post-symptomatic shedding that may result in transmission at progressively lower rates. Existing models have not included time-varying infectiousness, though representing these features could add utility to models of NoV transmission. METHODS: We address this by comparing the fit of three models (Models 1-3) of NoV infection to household transmission data from a 2009 point-source outbreak of GII.12 norovirus in North Carolina. Model 1 is an SEIR compartmental model, modified to allow Gamma-distributed sojourn times in the latent and infectious classes, where symptomatic cases are uniformly infectious over time. Model 2 assumes infectiousness decays exponentially as a function of time since onset, while Model 3 is discontinuous, with a spike concentrating 50% of transmissibility at onset. We use Bayesian data augmentation techniques to estimate transmission parameters for each model, and compare their goodness of fit using qualitative and quantitative model comparison. We also assess the robustness of our findings to asymptomatic infections. RESULTS: We find that Model 3 (initial spike in shedding) best explains the household transmission data, using both quantitative and qualitative model comparisons. We also show that these results are robust to the presence of asymptomatic infections. CONCLUSIONS: Explicitly representing explosive NoV infectiousness at onset should be considered when developing models and interventions to interrupt and prevent outbreaks of norovirus in the community. The methods presented here are generally applicable to the transmission of pathogens that exhibit large variation in transmissibility over an infection.

The role of weather on the relation between influenza and influenza-like illness.

Influenza epidemics, enabled by viral antigenic drift, occur invariably each winter in temperate climates. However, attempts to correlate the magnitude of virus change and epidemic size have been unsatisfactory. The incidence of influenza is not typically measured directly, but rather derived from the incidence of influenza-like illness (ILI), a clinical syndrome. Weather factors have been shown to influence the manifestation of influenza-like symptoms. We fitted an influenza transmission model to time series of influenza-like illness as monitored from 2003 to 2010 by two independent symptomatic surveillance systems (Influenzanet and EISN) in three European countries. By assuming that seasonality only acts upon the manifestation of symptoms, the model shows a significant correlation between the absolute humidity and temperature at the time of infection, and the proportion of influenza infections fulfilling the clinical ILI case definition, the so-called ILI factor. When a weather-dependent ILI factor is included in the model, the epidemic size of influenza-like illness becomes dependent not only on the susceptibility of the population at the beginning of the epidemic season but also on the weather conditions during which the epidemic unfolds. The combination reduces season-to-season variation in epidemic size and, interestingly, leads to a non-monotonic trend whereby the largest ILI epidemic occurs for moderate initial susceptibility.

The role of seasonality and import in a minimalistic multi-strain dengue model capturing differences between primary and secondary infections: complex dynamics and its implications for data analysis.

In many countries in Asia and South-America dengue fever (DF) and dengue hemorrhagic fever (DHF) has become a substantial public health concern leading to serious social-economic costs. Mathematical models describing the transmission of dengue viruses have focussed on the so-called antibody-dependent enhancement (ADE) effect and temporary cross-immunity trying to explain the irregular behavior of dengue epidemics by analyzing available data. However, no systematic investigation of the possible dynamical structures has been performed so far. Our study focuses on a seasonally forced (non-autonomous) model with temporary cross-immunity and possible secondary infection, motivated by dengue fever epidemiology. The notion of at least two different strains is needed in a minimalistic model to describe differences between primary infections, often asymptomatic, and secondary infection, associated with the severe form of the disease. We extend the previously studied non-seasonal (autonomous) model by adding seasonal forcing, mimicking the vectorial dynamics, and a low import of infected individuals, which is realistic in the dynamics of dengue fever epidemics. A comparative study between three different scenarios (non-seasonal, low seasonal and high seasonal with a low import of infected individuals) is performed. The extended models show complex dynamics and qualitatively a good agreement between empirical DHF monitoring data and the obtained model simulation. We discuss the role of seasonal forcing and the import of infected individuals in such systems, the biological relevance and its implications for the analysis of the available dengue data. At the moment only such minimalistic models have a chance to be qualitatively understood well and eventually tested against existing data. The simplicity of the model (low number of parameters and state variables) offer a promising perspective on parameter values inference from the DHF case notifications.

Explaining rapid reinfections in multiple-wave influenza outbreaks: Tristan da Cunha 1971 epidemic as a case study.

Influenza usually spreads through the human population in multiple-wave outbreaks. Successive reinfection of individuals over a short time interval has been explicitly reported during past pandemics. However, the causes of rapid reinfection and the role of reinfection in driving multiple-wave outbreaks remain poorly understood. To investigate these issues, we focus on a two-wave influenza A/H3N2 epidemic that occurred on the remote island of Tristan da Cunha in 1971. Over 59 days, 273 (96%) of 284 islanders experienced at least one attack and 92 (32%) experienced two attacks. We formulate six mathematical models invoking a variety of antigenic and immunological reinfection mechanisms. Using a maximum-likelihood analysis to confront model predictions with the reported incidence time series, we demonstrate that only two mechanisms can be retained: some hosts with either a delayed or deficient humoral immune response to the primary influenza infection were reinfected by the same strain, thus initiating the second epidemic wave. Both mechanisms are supported by previous empirical studies and may arise from a combination of genetic and ecological causes. We advocate that a better understanding and account of heterogeneity in the human immune response are essential to analysis of multiple-wave influenza outbreaks and pandemic planning.

Parameter Estimation in Epidemiology: from Simple to Complex Dynamics.

We revisit the parameter estimation framework for population biological dynamical systems, and apply it to calibrate various models in epidemiology with empirical time series, namely influenza and dengue fever. When it comes to more complex models like multi-strain dynamics to describe the virus-host interaction in dengue fever, even most recently developed parameter estimation techniques, like maximum likelihood iterated filtering, come to their computational limits. However, the first results of parameter estimation with data on dengue fever from Thailand indicate a subtle interplay between stochasticity and deterministic skeleton. The deterministic system on its own already displays complex dynamics up to deterministic chaos and coexistence of multiple attractors.

Influenza A gradual and epochal evolution: insights from simple models.

The recurrence of influenza A epidemics has originally been explained by a "continuous antigenic drift" scenario. Recently, it has been shown that if genetic drift is gradual, the evolution of influenza A main antigen, the haemagglutinin, is punctuated. As a consequence, it has been suggested that influenza A dynamics at the population level should be approximated by a serial model. Here, simple models are used to test whether a serial model requires gradual antigenic drift within groups of strains with the same antigenic properties (antigenic clusters). We compare the effect of status based and history based frameworks and the influence of reduced susceptibility and infectivity assumptions on the transient dynamics of antigenic clusters. Our results reveal that the replacement of a resident antigenic cluster by a mutant cluster, as observed in data, is reproduced only by the status based model integrating the reduced infectivity assumption. This combination of assumptions is useful to overcome the otherwise extremely high model dimensionality of models incorporating many strains, but relies on a biological hypothesis not obviously satisfied. Our findings finally suggest the dynamical importance of gradual antigenic drift even in the presence of punctuated immune escape. A more regular renewal of susceptible pool than the one implemented in a serial model should be part of a minimal theory for influenza at the population level.