Comparing the effects of non-homogenous mixing patterns on epidemiological outcomes in equine populations: A mathematical modelling study.

Disease transmission models often assume homogenous mixing. This assumption, however, has the potential to misrepresent the disease dynamics for populations in which contact patterns are non-random. A disease transmission model with an SEIR structure was used to compare the effect of weighted and unweighted empirical equine contact networks to weighted and unweighted theoretical networks generated using random mixing. Equine influenza was used as a case study. Incidence curves generated with the unweighted empirical networks were similar in epidemic duration (5-8 days) and peak incidence (30.8-46.4%). In contrast, the weighted empirical networks resulted in a more pronounced difference between the networks in terms of the epidemic duration (8-15 days) and the peak incidence (5-25%). The incidence curves for the empirical networks were bimodal, while the incidence curves for the theoretical networks were unimodal. The incorporation of vaccination and isolation in the model caused a decrease in the cumulative incidence for each network, however, this effect was only seen at high levels of vaccination and isolation for the complete network. This study highlights the importance of using empirical networks to describe contact patterns within populations that are unlikely to exhibit random mixing such as equine populations.

Comparison of the dynamic networks of four equine boarding and training facilities.

Contact networks can be analyzed to assess the potential for disease spread throughout the network. The lack of Canadian facility-level equine contact data makes the characterization of the equine contact structure difficult. Therefore, the purpose of this study was to use empirical contact data to characterize and compare equine network characteristics between equine facilities in Ontario. Contact pattern data from 4 equine facilities were collected using radio-frequency identification tags. The collected data were used to form 7 static contact networks (1 for each study day) for each facility. The assumption of homogenous mixing, where each individual in a population has an equal probability of coming in contact, was assessed for each network, since homogenous mixing is often used to describe mixing patterns in disease transmission models. At the facility level, neither the day-long static networks, nor a combined, week-long network were representative of homogenous mixing. The Jaccard Similarity Index indicated that 11-62% of the contacts were repeated throughout the study period. A network generated with survey-based data enabled the prediction of 8.7-79.6% of the contacts that were recorded with the RFID tags. With respect to the node centrality, the normalized node degree ranged from 0.0 to 0.96, with a mean of 0.31. The node strength ranged from 0 to 1 with a mean of 0.38. For both the node degree and node strength, a node's centrality score relative to the other nodes' centrality scores tended to be consistent throughout the study week. A significant (p < 0.05), weak positive correlation existed between the node degree and strength (0.41 < r < 0.54). The normalized betweenness centrality ranged from 0.00 to 1.00, with a mean of 0.11. Lastly, an exponential random graph model was used to quantify the relationship between the distance between the horses' stalls and edge formation. The distance parameter was not significant for all of the facilities. To conclude, the non-homogenous nature of the contact patterns, coupled with the large range of the centrality measures indicate the importance of using empirical data to understand processes such as disease spread potential within equine populations. Although the collection of a full set of data is optimal, the study results suggest an ability to infer contact networks using observational data in situations where little-to-no data exist. This study serves as a starting point for the characterization of equine contact networks in Ontario.

Validation of modified radio-frequency identification tag firmware, using an equine population case study.

BACKGROUND: Contact networks can be used to assess disease spread potential within a population. However, the data required to generate the networks can be challenging to collect. One method of collecting this type of data is by using radio-frequency identification (RFID) tags. The OpenBeacon RFID system generally consists of tags and readers. Communicating tags should be within 10m of the readers, which are powered by an external power source. The readers are challenging to implement in agricultural settings due to the lack of a power source and the large area needed to be covered. METHODS: OpenBeacon firmware was modified to use the tag's onboard flash memory for data storage. The tags were deployed within an equine facility for a 7-day period. Tags were attached to the horses' halters, worn by facility staff, and placed in strategic locations around the facility to monitor which participants had contact with the specified locations during the study period. When the tags came within 2m of each other, they recorded the contact event participant IDs, and start and end times. At the end of the study period, the data were downloaded to a computer and analyzed using network analysis methods. RESULTS: The resulting networks were plausible given the facility schedule as described in a survey completed by the facility manager. Furthermore, changes in the daily facility operations as described in the survey were reflected in the tag-collected data. In terms of the battery life, 88% of batteries maintained a charge for at least 6 days. Lastly, no consistent trends were evident in the horses' centrality metrics. DISCUSSION: This study demonstrates the utility of RFID tags for the collection of equine contact data. Future work should include the collection of contact data from multiple equine facilities to better characterize equine disease spread potential in Ontario.

Individual movements and contact patterns in a Canadian long-term care facility.

Contact networks of individuals in healthcare facilities are poorly understood, largely due to the lack of spatio-temporal movement data. A better understanding of such networks of interactions can help improve disease control strategies for nosocomial outbreaks. We sought to determine the spatio-temporal patterns of interactions between individuals using movement data collected in the largest veterans long-term care facility in Canada. We processed close-range contact data generated by the exchange of ultra-low-power radio signals, in a prescribed proximity, between wireless sensors worn by the participants over a two-week period. Statistical analyses of contact and movement data were conducted. We found a clear dichotomy in the contact network and movement patterns between residents and healthcare workers (HCWs) in this facility. Overall, residents tend to have significantly more distinct contacts with the mean of 17.3 (s.d. 3.6) contacts, versus 3.5 (s.d. 2.3) for HCWs (p-value < 10-12), for a longer duration of time (with mean contact duration of 8 minutes for resident-resident pair versus 4.6 minutes for HCW-resident pair) while being less mobile than HCWs. Analysis of movement data and clustering coefficient of the hourly aggregated network indicates that the contact network is loosely connected (mean clustering coefficient: 0.25, interquartile range 0-0.40), while being highly structured. Our findings bring quantitative insights regarding the contact network and movements in a long-term care facility, which are highly relevant to infer direct human-to-human and indirect (i.e., via the environment) disease transmission processes. This data-driven quantification is essential for validating disease dynamic models, as well as decision analytic methods to inform control strategies for nosocomial infections.

Assessing the benefits of early pandemic influenza vaccine availability: a case study for Ontario, Canada.

New vaccine production technologies can significantly shorten the timelines for availability of a strain-specific vaccine in the event of an influenza pandemic. We sought to evaluate the potential benefits of early vaccination in reducing the clinical attack rate (CAR), taking into account the timing and speed of vaccination roll-out. Various scenarios corresponding to the transmissibility of a pandemic strain and vaccine prioritization strategies were simulated using an agent-based model of disease spread in Ontario, the largest Canadian province. We found that the relative reduction of the CAR reached 60% (90%CI: 44-100%) in a best-case scenario, in which the pandemic strain was moderately transmissible, vaccination started 4 weeks before the first imported case, the vaccine administration rate was 4 times higher than its average for seasonal influenza, and the vaccine efficacy was up to 90%. But the relative reductions in the CAR decreased significantly when the vaccination campaign was delayed or the administration rate reduced. In urban settings with similar characteristics to our population study, early availability and high rates of vaccine administration has the potential to substantially reduce the number of influenza cases. Low rates of vaccine administration or uptake can potentially offset the benefits of early vaccination.

Asymptomatic Transmission and the Dynamics of Zika Infection.

Following the 2013-14 outbreak in French Polynesia, the Zika virus (ZIKV) epidemic spread widely to many countries where Aedes Aegypti as the main transmitting vector is endemic. The lack of a second wave of ZIKV infection in most affected regions may suggest that a sufficiently high level of herd immunity was reached during the first wave. We developed an agent-based transmission model to investigate the role of asymptomatic infection on the likelihood of observing a second wave, while accounting for its relative transmissibility. We found that, as the relative transmissibility of asymptomatic infection increases, a second wave is more likely to occur, despite an increase in the attack rate during the first wave. When the reproduction number varies between 1.9 and 2.8 based on estimates for Antioquia, Colombia, the attack rate varies between 4% and 26% for a low (below 10%) effectiveness of interventions in blunting the ZIKV transmission from symptomatic cases to mosquitoes. Moreover, the fraction of cases due to sexual transmission is estimated below 4% of the cumulative incidence. Our analyses underscore the need to quantify the transmissibility of asymptomatic infections, without which the overall attack rates and the level of herd immunity cannot be accurately estimated.

The Effect of Individual Movements and Interventions on the Spread of Influenza in Long-Term Care Facilities.

BACKGROUND: Nosocomial influenza poses a serious risk among residents of long-term care facilities (LTCFs). OBJECTIVE: We sought to evaluate the effect of resident and staff movements and contact patterns on the outcomes of various intervention strategies for influenza control in an LTCF. METHODS: We collected contact frequency data in Canada's largest veterans' LTCF by enroling residents and staff into a study that tracked their movements through wireless tags and signal receivers. We analyzed and fitted the data to an agent-based simulation model of influenza infection, and performed Monte-Carlo simulations to evaluate the benefit of antiviral prophylaxis and patient isolation added to standard (baseline) infection control practice (i.e., vaccination of residents and staff, plus antiviral treatment of residents with symptomatic infection). RESULTS: We calibrated the model to attack rates of 20%, 40%, and 60% for the baseline scenario. For data-driven movements, we found that the largest reduction in attack rates (12.5% to 27%; ANOVA P < 0.001) was achieved when the baseline strategy was combined with antiviral prophylaxis for all residents for the duration of the outbreak. Isolation of residents with symptomatic infection resulted in little or no effect on the attack rates (2.3% to 4.2%; ANOVA P > 0.2) among residents. In contrast, parameterizing the model with random movements yielded different results, suggesting that the highest benefit was achieved through patient isolation (69.6% to 79.6%; ANOVA P < 0.001) while the additional benefit of prophylaxis was negligible in reducing the cumulative number of infections. CONCLUSIONS: Our study revealed a highly structured contact and movement patterns within the LTCF. Accounting for this structure-instead of assuming randomness-in decision analytic methods can result in substantially different predictions.

Strategies for halting the rise of multidrug resistant TB epidemics: assessing the effect of early case detection and isolation.

Background: The increasing rates of multidrug resistant TB (MDR-TB) have posed the question of whether control programs under enhanced directly observed treatment, short-course (DOTS-Plus) are sufficient or implemented optimally. Despite enhanced efforts on early case detection and improved treatment regimens, direct transmission of MDR-TB remains a major hurdle for global TB control. Methods: We developed an agent-based simulation model of TB dynamics to evaluate the effect of transmission reduction measures on the incidence of MDR-TB. We implemented a 15-day isolation period following the start of treatment in active TB cases. The model was parameterized with the latest estimates derived from the published literature. Results: We found that if high rates (over 90%) of TB case identification are achieved within 4 weeks of developing active TB, then a 15-day patient isolation strategy with 50% effectiveness in interrupting disease transmission leads to 10% reduction in the incidence of MDR-TB over 10 years. If transmission is fully prevented, the rise of MDR-TB can be halted within 10 years, but the temporal reduction of MDR-TB incidence remains below 20% in this period. Conclusions: The impact of transmission reduction measures on the TB incidence depends critically on the rates and timelines of case identification. The high costs and adverse effects associated with MDR-TB treatment warrant increased efforts and investments on measures that can interrupt direct transmission through early case detection.

The Maillard reaction of bisoprolol fumarate with various reducing carbohydrates.

HPLC analysis of drug products containing bisoprolol fumarate and lactose revealed the presence of N-formylbisoprolol, which is a final product of the Maillard reaction. Formulations containing secondary amines and reducing carbohydrates are prone to the condensation of amine and carbonyl functional groups and formation of glycosylamines in pharmaceutically relevant conditions. Further rearrangement occurs in the presence of a nucleophile and leads to the formation of 1-deoxy-1-amino-2-ketose also known as the Amadori Rearrangement Product (ARP). The influence of water content, carbohydrate, and lubricant types on the reaction rate was tested. The reaction progress was monitored by HPLC and UV-Vis spectrophotometry. The structures of intermediates were confirmed by the LC/MS(2) analysis. N-formylbisoprolol - the final reaction product - was synthesised and characterised by LC/MS(2), H(1) and C(13) NMR.

Antiviral strategies for emerging influenza viruses in remote communities.

BACKGROUND: Due to the lack of timely access to resources for critical care, strategic use of antiviral drugs is crucial for mitigating the impact of novel influenza viruses with pandemic potential in remote and isolated communities. We sought to evaluate the effect of antiviral treatment and prophylaxis of close contacts in a Canadian remote northern community. METHODS: We used an agent-based, discrete-time simulation model for disease spread in a remote community, which was developed as an in-silico population using population census data. Relative and cumulative age-specific attack rates, and the total number of infections in simulated model scenarios were obtained. RESULTS: We found that early initiation of antiviral treatment is more critical for lowering attack rates in a remote setting with a low population-average age compared to an urban population. Our results show that a significant reduction in the relative, age-specific attack rates due to increasing treatment coverage does not necessarily translate to a significant reduction in the overall arrack rate. When treatment coverage varies from low to moderate, targeted prophylaxis has a very limited impact in reducing attack rates and should be offered at a low level (below 10%) to avoid excessive waste of drugs. CONCLUSIONS: In contrast to previous work, for conservative treatment coverages, our results do not provide any convincing evidence for the implementation of targeted prophylaxis. The findings suggest that public health strategies in remote communities should focus on the wider availability (higher coverage) and timely distribution of antiviral drugs for treatment of clinically ill individuals.

Active buildings: modelling physical activity and movement in office buildings. An observational study protocol.

INTRODUCTION: Health benefits of regular participation in physical activity are well documented but population levels are low. Office layout, and in particular the number and location of office building destinations (eg, print and meeting rooms), may influence both walking time and characteristics of sitting time. No research to date has focused on the role that the layout of the indoor office environment plays in facilitating or inhibiting step counts and characteristics of sitting time. The primary aim of this study was to investigate associations between office layout and physical activity, as well as sitting time using objective measures. METHODS AND ANALYSIS: Active buildings is a unique collaboration between public health, built environment and computer science researchers. The study involves objective monitoring complemented by a larger questionnaire arm. UK office buildings will be selected based on a variety of features, including office floor area and number of occupants. Questionnaires will include items on standard demographics, well-being, physical activity behaviour and putative socioecological correlates of workplace physical activity. Based on survey responses, approximately 30 participants will be recruited from each building into the objective monitoring arm. Participants will wear accelerometers (to monitor physical activity and sitting inside and outside the office) and a novel tracking device will be placed in the office (to record participant location) for five consecutive days. Data will be analysed using regression analyses, as well as novel agent-based modelling techniques. ETHICS AND DISSEMINATION: The results of this study will be disseminated through peer-reviewed publications and scientific presentations. Ethical approval was obtained through the University College London Research Ethics Committee (Reference number 4400/001).

Influenza H3N2 variant viruses with pandemic potential: preventing catastrophe in remote and isolated Canadian communities.

OBJECTIVE: To evaluate the impact of age-specific cross-reactive antibody protection levels on the outcomes of a pandemic outbreak of new variants of H3N2 influenza A viruses (H3N2v). METHODS: We calibrated a previously validated agent-based model of human-to-human transmission of influenza viruses to project the outcomes of various protection levels in a remote and isolated Canadian community, when demographics are drawn from the Statistics Canada census data. We then compared the outcomes with a scenario in which demographic variables were shifted to resemble an urban structure. This comparative evaluation was conducted using in-silico computer simulations, where the epidemiological data were drawn from relevant estimates in published literature. RESULTS: Simulations, using estimates of transmissibility for the 2009 H1N1 pandemic strain in the study population, show that the epidemic size is primarily affected by the cross-reactive protection levels of young children. A lower number of secondary infections at the early stages of an outbreak does not necessarily correspond to a lower epidemic size. CONCLUSIONS: Demographic variables could play a significant role in determining the outcomes of an outbreak. The findings strongly suggest that, when an H3N2v-specific vaccine becomes available, children below the age of 17 should be prioritized for vaccination. This prioritization is essential in population settings with a low average age, including aboriginal communities in northern latitudes.

Agent-based modeling of the spread of influenza-like illness in an emergency department: a simulation study.

The objective of this paper was to develop an agent-based modeling framework in order to simulate the spread of influenza virus infection on a layout based on a representative hospital emergency department in Winnipeg, Canada. In doing so, the study complements mathematical modeling techniques for disease spread, as well as modeling applications focused on the spread of antibiotic-resistant nosocomial infections in hospitals. Twenty different emergency department scenarios were simulated, with further simulation of four infection control strategies. The agent-based modeling approach represents systems modeling, in which the emergency department was modeled as a collection of agents (patients and healthcare workers) and their individual characteristics, behaviors, and interactions. The framework was coded in C++ using Qt4 libraries running under the Linux operating system. A simple ordinary least squares (OLS) regression was used to analyze the data, in which the percentage of patients that became infected in one day within the simulation was the dependent variable. The results suggest that within the given instance context, patient-oriented infection control policies (alternate treatment streams, masking symptomatic patients) tend to have a larger effect than policies that target healthcare workers. The agent-based modeling framework is a flexible tool that can be made to reflect any given environment; it is also a decision support tool for practitioners and policymakers to assess the relative impact of infection control strategies. The framework illuminates scenarios worthy of further investigation, as well as counterintuitive findings.

Improving Agent Based Models and Validation through Data Fusion.

This work is contextualized in research in modeling and simulation of infection spread within a community or population, with the objective to provide a public health and policy tool in assessing the dynamics of infection spread and the qualitative impacts of public health interventions. This work uses the integration of real data sources into an Agent Based Model (ABM) to simulate respiratory infection spread within a small municipality. Novelty is derived in that the data sources are not necessarily obvious within ABM infection spread models. The ABM is a spatial-temporal model inclusive of behavioral and interaction patterns between individual agents on a real topography. The agent behaviours (movements and interactions) are fed by census / demographic data, integrated with real data from a telecommunication service provider (cellular records) and person-person contact data obtained via a custom 3G Smartphone application that logs Bluetooth connectivity between devices. Each source provides data of varying type and granularity, thereby enhancing the robustness of the model. The work demonstrates opportunities in data mining and fusion that can be used by policy and decision makers. The data become real-world inputs into individual SIR disease spread models and variants, thereby building credible and non-intrusive models to qualitatively simulate and assess public health interventions at the population level.

Models of emergency departments for reducing patient waiting times.

In this paper, we apply both agent-based models and queuing models to investigate patient access and patient flow through emergency departments. The objective of this work is to gain insights into the comparative contributions and limitations of these complementary techniques, in their ability to contribute empirical input into healthcare policy and practice guidelines. The models were developed independently, with a view to compare their suitability to emergency department simulation. The current models implement relatively simple general scenarios, and rely on a combination of simulated and real data to simulate patient flow in a single emergency department or in multiple interacting emergency departments. In addition, several concepts from telecommunications engineering are translated into this modeling context. The framework of multiple-priority queue systems and the genetic programming paradigm of evolutionary machine learning are applied as a means of forecasting patient wait times and as a means of evolving healthcare policy, respectively. The models' utility lies in their ability to provide qualitative insights into the relative sensitivities and impacts of model input parameters, to illuminate scenarios worthy of more complex investigation, and to iteratively validate the models as they continue to be refined and extended. The paper discusses future efforts to refine, extend, and validate the models with more data and real data relative to physical (spatial-topographical) and social inputs (staffing, patient care models, etc.). Real data obtained through proximity location and tracking system technologies is one example discussed.