SORMAS-Stats: A Novel Statistical Application for Real-time Estimation of Epidemiological Parameters of Communicable Diseases using Contact-tracing Data. (Preprint)

BACKGROUND The Surveillance Outbreak Response Management and Analysis System (SORMAS) contains a management module to support countries in epidemic response. It consists of documentation, linkage and follow-up of cases, contacts, and events. To allow SORMAS users to visualise, compute key surveillance indicators and estimate epidemiological parameters from such a network data in real time, we developed the SORMAS Statistics (SORMAS-Stats) application. OBJECTIVE The aim of this study is to describe the key visualisations, surveillance indicators and epidemiological parameters implemented in the SORMAS-Stats application, and illustrate the application of SORMAS-Stats to COVID-19 outbreak response. METHODS Based on literature review and user requests, we included the following visualisation and estimation of parameters in SORMAS-Stats: transmission network diagram, serial interval (SI), time-varying reproduction number (Rt), dispersion parameter (k) and additional surveillance indicators presented in graphs and tables. We estimated SI by fitting a lognormal, gamma, and Weibull distributions to the observed distribution of the number of days between symptoms onset dates of infector-infectee pairs. We estimated k by fitting a negative binomial distribution to the observed number of infectees per infector. We applied the Markov Chain Monte Carlo approach and estimated Rt using the incidence data and the observed SI data, computed from the transmission network data. RESULTS Using COVID-19 contact tracing data of confirmed cases reported between July 31, and October 29, 2021 in Bourgogne-Franche-Comté region of France, we constructed a network diagram containing 63570 nodes comprising 1.75% (1115/63570) events, 19.59% (12452/63570) case persons, and 78.66% (50003/63570) exposed persons, 1238 infector-infectee pairs, 3860 transmission chains with 24.69% (953/3860) having events as the index infector. The distribution with best fit to the observed SI data was lognormal distribution with mean 4.32 days (95% CI, 4.10–4.53 days). We estimated the dispersion parameter, k of 21.11 (95% CI, 7.57–34.66) and a reproductive number, R of 0.9 (95% CI, 0.58–0.60). The weekly estimated Rt values ranged from 0.80 to 1.61. CONCLUSIONS We provide an application for real-time estimation of epidemiological parameters, which are essential for informing outbreak response strategies. These estimates are commensurate with findings from previous studies. SORMAS-Stats application would greatly assist public health authorities in the regions using SORMAS or similar applications by providing extensive visualisations and computation of surveillance indicators.

Crosswind integrated concentration for various dispersion parameter systems

LFkkuh; Lrj izdh.kZu ds fy, xkSlh;u fiPNd ekWMy ¼Gaussian Plume Model½ dk O;kid :i ls iz;ksx fd;k tkrk gSA vuqizLFk iou dh dqy lkanzrk Kkr djus ds fy, xkSlh;u lw= ¼QkWewyk½ dks laxfBr fd;k gSA vuqizLFk iou dh dqy lkanzrk dh x.kuk djus ds fy, izdh.kZu izkpyksa dh fHkUu&fHkUu iz.kkfy;ksa dk mi;ksx fd;k x;k gSA lrg Lrj esa Å¡pkbZ ds vuqlkj iou xfr dh fHkUurk dk o.kZu djus ds fy, ykxfjFehd foaM izksQkby dk mi;ksx fd;k x;k gSA blesa NksM+h tkus okyh izHkkoh Å¡pkbZ dks /;ku esa j[kk x;k gSA fHkUu fHkUu izdh.kZu izkpy iz.kkfy;ksa ds fy, iwokZuqekfur lkanzrkvksa vkSj dksisugsxu ds folj.k iz;ksx ls izkIr fd, x, izsf{kr vk¡dM+ksa dh rqyuk djus ds fy, lkaf[;dh; ifjekiksa dk mi;ksx fd;k x;k gSA The Gaussian plume model is the most widely used model for local scale dispersion. The Gaussian formula has been integrated to obtain the crosswind-integrated concentration. Different systems of dispersion parameters are used to calculate the crosswind integrated concentration. A logarithmic wind profile is used to describe the variation of wind speed with height in the surface layer. The effective release height was taken into consideration. Statistical measures are utilized in the comparison between the predicted concentrations for different dispersion parameter systems and the observed concentrations data obtained from Copenhagen diffusion experiment.

Investigation of model calibration issues in the safety performance assessment of Ontario highways

A set of Safety Performance Function (SPFs) commonly known as accident prediction models, were developed for evaluating the safety of Highway segments under the jurisdiction of Ministry of Transportation, Ontario (MTO). A generalized linear modeling approach was used in which negative binomial regression models were delevoped separately for total accidents and for three severity types (Property Damage Only accidents, Fatal and Injury accidents) as a function of traffic volume AADT. The SPFs were calibrated from 100m homogenous segments as well as for variable length continuous segments that are homogeneous with respect to measured traffic and geometric characteristics. For the models calibrated for Rural 2-Lane Kings Highways, the variables that had significant effects on accident occurrence were the terrain, shoulder width and segment lenght. It was observed that the disperson parameter of the negative binomial districution is large for 100m segments and smaller for longer segments. Further investigation of the dispersion parameter for Rural 2-Lane Kings Highways showed that the models calibrated with a separate dispersion parameter for each site depending on the segment length performed better that the model calibrated considering fixed dispersion parameter for all sites. For Rural 2-Lane Kings Highways, a model was calibrated with trend considering each year as a separate observation. The GEE (Generalized Estimating Equation) procedure was use to develop these models since it incorporated the temporal correlation that exists in repeated measurements. Results showed that integration of time trend and temporal correlation in the model improves the model fit.

Investigation of model calibration issues in the safety performance assessment of Ontario highways

A set of Safety Performance Function (SPFs) commonly known as accident prediction models, were developed for evaluating the safety of Highway segments under the jurisdiction of Ministry of Transportation, Ontario (MTO). A generalized linear modeling approach was used in which negative binomial regression models were delevoped separately for total accidents and for three severity types (Property Damage Only accidents, Fatal and Injury accidents) as a function of traffic volume AADT. The SPFs were calibrated from 100m homogenous segments as well as for variable length continuous segments that are homogeneous with respect to measured traffic and geometric characteristics. For the models calibrated for Rural 2-Lane Kings Highways, the variables that had significant effects on accident occurrence were the terrain, shoulder width and segment lenght. It was observed that the disperson parameter of the negative binomial districution is large for 100m segments and smaller for longer segments. Further investigation of the dispersion parameter for Rural 2-Lane Kings Highways showed that the models calibrated with a separate dispersion parameter for each site depending on the segment length performed better that the model calibrated considering fixed dispersion parameter for all sites. For Rural 2-Lane Kings Highways, a model was calibrated with trend considering each year as a separate observation. The GEE (Generalized Estimating Equation) procedure was use to develop these models since it incorporated the temporal correlation that exists in repeated measurements. Results showed that integration of time trend and temporal correlation in the model improves the model fit.

Flat F-Manifolds, F-CohFTs, and Integrable Hierarchies

We define the double ramification hierarchy associated to an F-cohomological field theory and use this construction to prove that the principal hierarchy of any semisimple (homogeneous) flat F-manifold possesses a (homogeneous) integrable dispersive deformation at all orders in the dispersion parameter. The proof is based on the reconstruction of an F-CohFT starting from a semisimple flat F-manifold and additional data in genus 1, obtained in our previous work. Our construction of these dispersive deformations is quite explicit and we compute several examples. In particular, we provide a complete classification of rank 1 hierarchies of DR type at the order 9 approximation in the dispersion parameter and of homogeneous DR hierarchies associated with all 2-dimensional homogeneous flat F-manifolds at genus 1 approximation.

Overdispersion in COVID-19 increases the effectiveness of limiting nonrepetitive contacts for transmission control

Increasing evidence indicates that superspreading plays a dominant role in COVID-19 transmission. Recent estimates suggest that the dispersion parameter k for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is on the order of 0.1, which corresponds to about 10% of cases being the source of 80% of infections. To investigate how overdispersion might affect the outcome of various mitigation strategies, we developed an agent-based model with a social network that allows transmission through contact in three sectors: “close” (a small, unchanging group of mutual contacts as might be found in a household), “regular” (a larger, unchanging group as might be found in a workplace or school), and “random” (drawn from the entire model population and not repeated regularly). We assigned individual infectivity from a gamma distribution with dispersion parameter k. We found that when k was low (i.e., greater heterogeneity, more superspreading events), reducing random sector contacts had a far greater impact on the epidemic trajectory than did reducing regular contacts; when k was high (i.e., less heterogeneity, no superspreading events), that difference disappeared. These results suggest that overdispersion of COVID-19 transmission gives the virus an Achilles’ heel: Reducing contacts between people who do not regularly meet would substantially reduce the pandemic, while reducing repeated contacts in defined social groups would be less effective.

Inference of genome 3D architecture by modeling overdispersion of Hi-C data

We address the challenge of inferring a consensus 3D model of genome architecture from Hi-C data. Existing approaches most often rely on a two step algorithm: first convert the contact counts into distances, then optimize an objective function akin to multidimensional scaling (MDS) to infer a 3D model. Other approaches use a maximum likelihood approach, modeling the contact counts between two loci as a Poisson random variable whose intensity is a decreasing function of the distance between them. However, a Poisson model of contact counts implies that the variance of the data is equal to the mean, a relationship that is often too restrictive to properly model count data.We first confirm the presence of overdispersion in several real Hi-C data sets, and we show that the overdispersion arises even in simulated data sets. We then propose a new model, called Pastis-NB, where we replace the Poisson model of contact counts by a negative binomial one, which is parametrized by a mean and a separate dispersion parameter. The dispersion parameter allows the variance to be adjusted independently from the mean, thus better modeling overdispersed data. We compare the results of Pastis-NB to those of several previously published algorithms: three MDS-based methods (ShRec3D, ChromSDE, and Pastis-MDS) and a statistical methods based on a Poisson model of the data (Pastis-PM). We show that the negative binomial inference yields more accurate structures on simulated data, and more robust structures than other models across real Hi-C replicates and across different resolutions.A Python implementation of Pastis-NB is available at https://github.com/hiclib/pastis under the BSD licenseSupplementary information is available at https://nellev.github.io/pastisnb/

An open-access database of infectious disease transmission trees to explore superspreader epidemiology

Historically, emerging and re-emerging infectious diseases have caused large, deadly, and expensive multi-national outbreaks. Often outbreak investigations aim to identify who infected whom by reconstructing the outbreak transmission tree, which visualizes transmission between individuals as a network with nodes representing individuals and branches representing transmission from person to person. We compiled a database of 383 published, standardized transmission trees consisting of 16 directly-transmitted diseases ranging in size from 2 to 286 cases. For each tree and disease we calculated several key statistics, such as outbreak size, average number of secondary infections, the dispersion parameter, and the number of superspreaders. We demonstrated the potential utility of the database through short analyses addressing questions about superspreader epidemiology for a variety of diseases, including COVID-19. First, we compared the frequency and contribution of superspreaders to onward transmission across diseases. COVID-19 outbreaks had significantly fewer superspreaders than outbreaks of SARS and MERS and a dispersion parameter between that of SARS and MERS. Across diseases the presence of more superspreaders was associated with greater outbreak size. Second, we further examined how early spread impacts tree size. Generally, trees sparked by a superspreader had larger outbreak sizes than those trees not sparked by a superspreader, and this trend was significant for COVID-19 trees. Third, we investigated patterns in how superspreaders are infected. Across trees with more than one superspreader, we found support for the theory that superspreaders generate other superspreaders, even when controlling for number of secondary infections. In sum, our findings put the role of superspreading to COVID-19 transmission in perspective with that of SARS and MERS and suggest an avenue for further research on the generation of superspreaders. These data have been made openly available to encourage reuse and further scientific inquiry.Author SummaryPublic health investigations often aim to identify who infected whom, or the transmission tree, during outbreaks of infectious diseases. These investigations tend to be resource intensive but valuable as they contain epidemiological information, including the average number of infections caused by each individual and the variation in this number. To date, there remains no standardized format nor comprehensive database of infectious disease transmission trees. To fill this gap, we standardized and compiled more than 350 published transmission trees for 16 directly-transmitted diseases into a database that is publicly available. In this paper, we give an overview of the database construction process, as well as a demonstration of the types of questions that the database can be used to answer related to superspreader epidemiology. For example, we show that COVID-19 outbreaks have fewer superspreaders than outbreaks of SARS and MERS. We also find support for the theory that superspreaders generate other superspreaders. In the future, this database can be used to answer other outstanding questions in the field of epidemiology.