An algorithm to build synthetic temporal contact networks based on close-proximity interactions data.

Small populations (e.g., hospitals, schools or workplaces) are characterised by high contact heterogeneity and stochasticity affecting pathogen transmission dynamics. Empirical individual contact data provide unprecedented information to characterize such heterogeneity and are increasingly available, but are usually collected over a limited period, and can suffer from observation bias. We propose an algorithm to stochastically reconstruct realistic temporal networks from individual contact data in healthcare settings (HCS) and test this approach using real data previously collected in a long-term care facility (LTCF). Our algorithm generates full networks from recorded close-proximity interactions, using hourly inter-individual contact rates and information on individuals' wards, the categories of staff involved in contacts, and the frequency of recurring contacts. It also provides data augmentation by reconstructing contacts for days when some individuals are present in the HCS without having contacts recorded in the empirical data. Recording bias is formalized through an observation model, to allow direct comparison between the augmented and observed networks. We validate our algorithm using data collected during the i-Bird study, and compare the empirical and reconstructed networks. The algorithm was substantially more accurate to reproduce network characteristics than random graphs. The reconstructed networks reproduced well the assortativity by ward (first-third quartiles observed: 0.54-0.64; synthetic: 0.52-0.64) and the hourly staff and patient contact patterns. Importantly, the observed temporal correlation was also well reproduced (0.39-0.50 vs 0.37-0.44), indicating that our algorithm could recreate a realistic temporal structure. The algorithm consistently recreated unobserved contacts to generate full reconstructed networks for the LTCF. To conclude, we propose an approach to generate realistic temporal contact networks and reconstruct unobserved contacts from summary statistics computed using individual-level interaction networks. This could be applied and extended to generate contact networks to other HCS using limited empirical data, to subsequently inform individual-based epidemic models.

Using contact network dynamics to implement efficient interventions against pathogen spread in hospital settings: A modelling study.

BACKGROUND: Long-term care facilities (LTCFs) are hotspots for pathogen transmission. Infection control interventions are essential, but the high density and heterogeneity of interindividual contacts within LTCF may hinder their efficacy. Here, we explore how the patient-staff contact structure may inform effective intervention implementation. METHODS AND FINDINGS: Using an individual-based model (IBM), we reproduced methicillin-resistant Staphylococcus aureus colonisation transmission dynamics over a detailed contact network recorded within a French LTCF of 327 patients and 263 staff over 3 months. Simulated baseline cumulative colonisation incidence was 21 patients (prediction interval: 11, 31) and 35 staff (prediction interval: 19, 54). We examined the potential impact of 3 types of interventions against transmission (reallocation reducing the number of unique contacts per staff, reinforced contact precautions, and hypothetical vaccination protecting against acquisition), targeted towards specific populations. All 3 interventions were effective when applied to all nurses or healthcare assistants (median reduction in MRSA colonisation incidence up to 35%), but the benefit did not exceed 8% when targeting any other single staff category. We identified "supercontactor" individuals with most contacts ("frequency-based," overrepresented among nurses, porters, and rehabilitation staff) or with the longest cumulative time spent in contact ("duration-based," overrepresented among healthcare assistants and patients in elderly care or persistent vegetative state (PVS)). Targeting supercontactors enhanced interventions against pathogen spread in the LTCF. With contact precautions, targeting frequency-based staff supercontactors led to the highest incidence reduction (20%, 95% CI: 19, 21). Vaccinating a mix of frequency- and duration-based staff supercontactors led to a higher reduction (23%, 95% CI: 22, 24) than all other approaches. Although based on data from a single LTCF, when varying epidemiological parameters to extend to other pathogens, our results suggest that targeting supercontactors is always the most effective strategy, indicating this approach could be applied to prevent transmission of other nosocomial pathogens. CONCLUSIONS: By characterising the contact structure in hospital settings and identifying the categories of staff and patients more likely to be supercontactors, with either more or longer contacts than others, interventions against nosocomial spread could be more effective. We find that the most efficient implementation strategy depends on the intervention (reallocation, contact precautions, vaccination) and target population (staff, patients, supercontactors). Importantly, both staff and patients may be supercontactors, highlighting the importance of including patients in measures to prevent pathogen transmission in LTCF.

Modelling the synergistic effect of bacteriophage and antibiotics on bacteria: Killers and drivers of resistance evolution.

Bacteriophage (phage) are bacterial predators that can also spread antimicrobial resistance (AMR) genes between bacteria by generalised transduction. Phage are often present alongside antibiotics in the environment, yet evidence of their joint killing effect on bacteria is conflicted, and the dynamics of transduction in such systems are unknown. Here, we combine in vitro data and mathematical modelling to identify conditions where phage and antibiotics act in synergy to remove bacteria or drive AMR evolution. We adapt a published model of phage-bacteria dynamics, including transduction, to add the pharmacodynamics of erythromycin and tetracycline, parameterised from new in vitro data. We simulate a system where two strains of Staphylococcus aureus are present at stationary phase, each carrying either an erythromycin or tetracycline resistance gene, and where multidrug-resistant bacteria can be generated by transduction only. We determine rates of bacterial clearance and multidrug-resistant bacteria appearance, when either or both antibiotics and phage are present at varying timings and concentrations. Although phage and antibiotics act in synergy to kill bacteria, by reducing bacterial growth antibiotics reduce phage production. A low concentration of phage introduced shortly after antibiotics fails to replicate and exert a strong killing pressure on bacteria, instead generating multidrug-resistant bacteria by transduction which are then selected for by the antibiotics. Multidrug-resistant bacteria numbers were highest when antibiotics and phage were introduced simultaneously. The interaction between phage and antibiotics leads to a trade-off between a slower clearing rate of bacteria (if antibiotics are added before phage), and a higher risk of multidrug-resistance evolution (if phage are added before antibiotics), exacerbated by low concentrations of phage or antibiotics. Our results form hypotheses to guide future experimental and clinical work on the impact of phage on AMR evolution, notably for studies of phage therapy which should investigate varying timings and concentrations of phage and antibiotics.

Longitudinal social contact data analysis: insights from 2 years of data collection in Belgium during the COVID-19 pandemic.

BACKGROUND: During the COVID-19 pandemic, the CoMix study, a longitudinal behavioral survey, was designed to monitor social contacts and public awareness in multiple countries, including Belgium. As a longitudinal survey, it is vulnerable to participants' "survey fatigue", which may impact inferences. METHODS: A negative binomial generalized additive model for location, scale, and shape (NBI GAMLSS) was adopted to estimate the number of contacts reported between age groups and to deal with under-reporting due to fatigue within the study. The dropout process was analyzed with first-order auto-regressive logistic regression to identify factors that influence dropout. Using the so-called next generation principle, we calculated the effect of under-reporting due to fatigue on estimating the reproduction number. RESULTS: Fewer contacts were reported as people participated longer in the survey, which suggests under-reporting due to survey fatigue. Participant dropout is significantly affected by household size and age categories, but not significantly affected by the number of contacts reported in any of the two latest waves. This indicates covariate-dependent missing completely at random (MCAR) in the dropout pattern, when missing at random (MAR) is the alternative. However, we cannot rule out more complex mechanisms such as missing not at random (MNAR). Moreover, under-reporting due to fatigue is found to be consistent over time and implies a 15-30% reduction in both the number of contacts and the reproduction number ([Formula: see text]) ratio between correcting and not correcting for under-reporting. Lastly, we found that correcting for fatigue did not change the pattern of relative incidence between age groups also when considering age-specific heterogeneity in susceptibility and infectivity. CONCLUSIONS: CoMix data highlights the variability of contact patterns across age groups and time, revealing the mechanisms governing the spread/transmission of COVID-19/airborne diseases in the population. Although such longitudinal contact surveys are prone to the under-reporting due to participant fatigue and drop-out, we showed that these factors can be identified and corrected using NBI GAMLSS. This information can be used to improve the design of similar, future surveys.

Importance of patient bed pathways and length of stay differences in predicting COVID-19 hospital bed occupancy in England.

BACKGROUND: Predicting bed occupancy for hospitalised patients with COVID-19 requires understanding of length of stay (LoS) in particular bed types. LoS can vary depending on the patient's "bed pathway" - the sequence of transfers of individual patients between bed types during a hospital stay. In this study, we characterise these pathways, and their impact on predicted hospital bed occupancy. METHODS: We obtained data from University College Hospital (UCH) and the ISARIC4C COVID-19 Clinical Information Network (CO-CIN) on hospitalised patients with COVID-19 who required care in general ward or critical care (CC) beds to determine possible bed pathways and LoS. We developed a discrete-time model to examine the implications of using either bed pathways or only average LoS by bed type to forecast bed occupancy. We compared model-predicted bed occupancy to publicly available bed occupancy data on COVID-19 in England between March and August 2020. RESULTS: In both the UCH and CO-CIN datasets, 82% of hospitalised patients with COVID-19 only received care in general ward beds. We identified four other bed pathways, present in both datasets: "Ward, CC, Ward", "Ward, CC", "CC" and "CC, Ward". Mean LoS varied by bed type, pathway, and dataset, between 1.78 and 13.53 days. For UCH, we found that using bed pathways improved the accuracy of bed occupancy predictions, while only using an average LoS for each bed type underestimated true bed occupancy. However, using the CO-CIN LoS dataset we were not able to replicate past data on bed occupancy in England, suggesting regional LoS heterogeneities. CONCLUSIONS: We identified five bed pathways, with substantial variation in LoS by bed type, pathway, and geography. This might be caused by local differences in patient characteristics, clinical care strategies, or resource availability, and suggests that national LoS averages may not be appropriate for local forecasts of bed occupancy for COVID-19. TRIAL REGISTRATION: The ISARIC WHO CCP-UK study ISRCTN66726260 was retrospectively registered on 21/04/2020 and designated an Urgent Public Health Research Study by NIHR.

Investigating the feasibility and potential of combining industry AMR monitoring systems: a comparison with WHO GLASS.

Background: Efforts to estimate the global burden of antimicrobial resistance (AMR) have highlighted gaps in existing surveillance systems. Data gathered from hospital networks globally by pharmaceutical industries to monitor antibiotic efficacy in different bacteria represent an underused source of information to complete our knowledge of AMR burden.. We analysed available industry monitoring systems to assess to which extent combining them could help fill the gaps in our current understanding of AMR levels and trends. Methods: We analysed six industry monitoring systems (ATLAS, GEARS, SIDERO-WT, KEYSTONE, DREAM, and SOAR) obtained from the Vivli platform and reviewed their respective isolates collection and analysis protocols. Using the R software, we designed a pipeline to harmonise and combine these into a single dataset. We assessed the reliability of resistance estimates from these sources by comparing the combined dataset to the publicly available subset of WHO GLASS for shared bacteria-antibiotic-country-year combinations. Results: Combined, the industry monitoring systems cover 18 years (4 years for GLASS), 85 countries (71), 412 bacterial species (8), and 75 antibiotics (25). Although all industry systems followed a similar centralised testing approach, the patient selection protocol and associated sampling period were unclear. Over all reported years and countries, E.coli, K. pneumoniae and S. aureus resistance rates were in >65% of cases within 0.1 of the corresponding estimate in GLASS. We did not identify systemic bias towards resistance in industry systems compared to GLASS. Conclusions: High agreement values for available comparisons with GLASS suggest that data for other bacteria-antibiotic-country-year combinations only present in industry systems could complement GLASS; however, for this purpose patient and isolate selection criteria must first be clarified to understand the representativeness of industry systems. This additional source of information on resistance levels could help clinicians and stakeholders prioritize testing and select appropriate antibiotics in settings with limited surveillance data.

Antimicrobial resistance (AMR) is a growing problem worldwide, but we don't always have enough information to fully understand its extent and how it's changing over time. In this study, we looked at data collected by pharmaceutical companies from hospitals around the world to see how well antibiotics are working against different bacteria. We wanted to see if combining these data sources could help us fill in gaps in global AMR surveillance. We reviewed the methods of six different systems that collect this data and developed an approach to combine them. Then, we compared this combined data to publicly available GLASS data from the WHO to check if it was reliable. We found that the data from the pharmaceutical companies covered more years, countries, bacterial species, and antibiotics than GLASS. Even though the way the data was collected by the companies wasn't always clear, we saw that the resistance estimates were similar to those from GLASS for some common bacteria like E.coli, K. pneumoniae, and S. aureus. Overall, combining data from these different sources could improve our understanding of AMR worldwide, especially in places where surveillance is currently limited, and for Priority Pathogens not covered by GLASS.

Growth-Dependent Predation and Generalized Transduction of Antimicrobial Resistance by Bacteriophage.

Bacteriophage (phage) are both predators and evolutionary drivers for bacteria, notably contributing to the spread of antimicrobial resistance (AMR) genes by generalized transduction. Our current understanding of this complex relationship is limited. We used an interdisciplinary approach to quantify how these interacting dynamics can lead to the evolution of multidrug-resistant bacteria. We cocultured two strains of methicillin-resistant Staphylococcus aureus, each harboring a different antibiotic resistance gene, with generalized transducing phage. After a growth phase of 8 h, bacteria and phage surprisingly coexisted at a stable equilibrium in our culture, the level of which was dependent on the starting concentration of phage. We detected double-resistant bacteria as early as 7 h, indicating that transduction of AMR genes had occurred. We developed multiple mathematical models of the bacteria and phage relationship and found that phage-bacteria dynamics were best captured by a model in which phage burst size decreases as the bacteria population reaches stationary phase and where phage predation is frequency-dependent. We estimated that one in every 108 new phage generated was a transducing phage carrying an AMR gene and that double-resistant bacteria were always predominantly generated by transduction rather than by growth. Our results suggest a shift in how we understand and model phage-bacteria dynamics. Although rates of generalized transduction could be interpreted as too rare to be significant, they are sufficient in our system to consistently lead to the evolution of multidrug-resistant bacteria. Currently, the potential of phage to contribute to the growing burden of AMR is likely underestimated. IMPORTANCE Bacteriophage (phage), viruses that can infect and kill bacteria, are being investigated through phage therapy as a potential solution to the threat of antimicrobial resistance (AMR). In reality, however, phage are also natural drivers of bacterial evolution by transduction when they accidentally carry nonphage DNA between bacteria. Using laboratory work and mathematical models, we show that transduction leads to evolution of multidrug-resistant bacteria in less than 8 h and that phage production decreases when bacterial growth decreases, allowing bacteria and phage to coexist at stable equilibria. The joint dynamics of phage predation and transduction lead to complex interactions with bacteria, which must be clarified to prevent phage from contributing to the spread of AMR.

Analysis of temporal trends in potential COVID-19 cases reported through NHS Pathways England.

The National Health Service (NHS) Pathways triage system collates data on enquiries to 111 and 999 services in England. Since the 18th of March 2020, these data have been made publically available for potential COVID-19 symptoms self-reported by members of the public. Trends in such reports over time are likely to reflect behaviour of the ongoing epidemic within the wider community, potentially capturing valuable information across a broader severity profile of cases than hospital admission data. We present a fully reproducible analysis of temporal trends in NHS Pathways reports until 14th May 2020, nationally and regionally, and demonstrate that rates of growth/decline and effective reproduction number estimated from these data may be useful in monitoring transmission. This is a particularly pressing issue as lockdown restrictions begin to be lifted and evidence of disease resurgence must be constantly reassessed. We further assess the correlation between NHS Pathways reports and a publicly available NHS dataset of COVID-19-associated deaths in England, finding that enquiries to 111/999 were strongly associated with daily deaths reported 16 days later. Our results highlight the potential of NHS Pathways as the basis of an early warning system. However, this dataset relies on self-reported symptoms, which are at risk of being severely biased. Further detailed work is therefore necessary to investigate potential behavioural issues which might otherwise explain our conclusions.

Comparing the Analysis and Results of a Modified Social Accounting Matrix Framework with Conventional Methods of Reporting Indirect Non-Medical Costs.

BACKGROUND: Assessing the societal perspective in economic evaluations of new interventions requires estimates of indirect non-medical costs caused by the disease. Different methods exist for measuring the labor input function as a surrogate for these costs. They rarely specify the effect of health on labor and who gains and who loses money. Social accounting matrix (SAM) is an established framework that evaluates public policies with multiple perspectives that could help. OBJECTIVES: We evaluated the use of a modified SAM to assess money flows between different economic agents resulting in economic transactions following policy changes of medical interventions. METHODS: We compared conventional methods of measuring indirect non-medical costs related to rotavirus vaccination in the Netherlands with a modified SAM framework. To compare the outcome of each method, we calculated returns on investment (ROI) as the net amount of money per euro invested in the vaccine. One-way and probabilistic sensitivity analyses were carried out for each method, focusing on critical variables with the largest impact on indirect cost estimates. RESULTS: The ROI was higher for the modified SAM (1.33) than for the conventional methods assessing income calculations (range - 0.178 to 1.22). Probabilistic sensitivity analyses showed wide distributions in the ROI estimates, with variation in the variable impact on the indirect cost results per method selected. CONCLUSIONS: In contrast to conventional methods, the SAM approach provides detailed and comprehensive assessments of the impact of new interventions on the indirect non-medical costs and the financial interactions between agents, disclosing useful information for different stakeholders.

Antimicrobial resistance and COVID-19: Intersections and implications.

Before the coronavirus 2019 (COVID-19) pandemic began, antimicrobial resistance (AMR) was among the top priorities for global public health. Already a complex challenge, AMR now needs to be addressed in a changing healthcare landscape. Here, we analyse how changes due to COVID-19 in terms of antimicrobial usage, infection prevention, and health systems affect the emergence, transmission, and burden of AMR. Increased hand hygiene, decreased international travel, and decreased elective hospital procedures may reduce AMR pathogen selection and spread in the short term. However, the opposite effects may be seen if antibiotics are more widely used as standard healthcare pathways break down. Over 6 months into the COVID-19 pandemic, the dynamics of AMR remain uncertain. We call for the AMR community to keep a global perspective while designing finely tuned surveillance and research to continue to improve our preparedness and response to these intersecting public health challenges.

Real-time monitoring of COVID-19 dynamics using automated trend fitting and anomaly detection.

As several countries gradually release social distancing measures, rapid detection of new localized COVID-19 hotspots and subsequent intervention will be key to avoiding large-scale resurgence of transmission. We introduce ASMODEE (automatic selection of models and outlier detection for epidemics), a new tool for detecting sudden changes in COVID-19 incidence. Our approach relies on automatically selecting the best (fitting or predicting) model from a range of user-defined time series models, excluding the most recent data points, to characterize the main trend in an incidence. We then derive prediction intervals and classify data points outside this interval as outliers, which provides an objective criterion for identifying departures from previous trends. We also provide a method for selecting the optimal breakpoints, used to define how many recent data points are to be excluded from the trend fitting procedure. The analysis of simulated COVID-19 outbreaks suggests ASMODEE compares favourably with a state-of-art outbreak-detection algorithm while being simpler and more flexible. As such, our method could be of wider use for infectious disease surveillance. We illustrate ASMODEE using publicly available data of National Health Service (NHS) Pathways reporting potential COVID-19 cases in England at a fine spatial scale, showing that the method would have enabled the early detection of the flare-ups in Leicester and Blackburn with Darwen, two to three weeks before their respective lockdown. ASMODEE is implemented in the free R package trendbreaker. This article is part of the theme issue 'Modelling that shaped the early COVID-19 pandemic response in the UK'.

What settings have been linked to SARS-CoV-2 transmission clusters?

Background: Concern about the health impact of novel coronavirus SARS-CoV-2 has resulted in widespread enforced reductions in people's movement ("lockdowns"). However, there are increasing concerns about the severe economic and wider societal consequences of these measures. Some countries have begun to lift some of the rules on physical distancing in a stepwise manner, with differences in what these "exit strategies" entail and their timeframes. The aim of this work was to inform such exit strategies by exploring the types of indoor and outdoor settings where transmission of SARS-CoV-2 has been reported to occur and result in clusters of cases. Identifying potential settings that result in transmission clusters allows these to be kept under close surveillance and/or to remain closed as part of strategies that aim to avoid a resurgence in transmission following the lifting of lockdown measures. Methods: We performed a systematic review of available literature and media reports to find settings reported in peer reviewed articles and media with these characteristics. These sources are curated and made available in an editable online database. Results: We found many examples of SARS-CoV-2 clusters linked to a wide range of mostly indoor settings. Few reports came from schools, many from households, and an increasing number were reported in hospitals and elderly care settings across Europe. Conclusions: We identified possible places that are linked to clusters of COVID-19 cases and could be closely monitored and/or remain closed in the first instance following the progressive removal of lockdown restrictions. However, in part due to the limits in surveillance capacities in many settings, the gathering of information such as cluster sizes and attack rates is limited in several ways: inherent recall bias, biased media reporting and missing data.

Implication of backward contact tracing in the presence of overdispersed transmission in COVID-19 outbreaks.

Introduction: Contact tracing has the potential to control outbreaks without the need for stringent physical distancing policies, e.g. civil lockdowns. Unlike forward contact tracing, backward contact tracing identifies the source of newly detected cases. This approach is particularly valuable when there is high individual-level variation in the number of secondary transmissions (overdispersion). Methods: By using a simple branching process model, we explored the potential of combining backward contact tracing with more conventional forward contact tracing for control of COVID-19. We estimated the typical size of clusters that can be reached by backward tracing and simulated the incremental effectiveness of combining backward tracing with conventional forward tracing. Results: Across ranges of parameter values consistent with dynamics of SARS-CoV-2, backward tracing is expected to identify a primary case generating 3-10 times more infections than a randomly chosen case, typically increasing the proportion of subsequent cases averted by a factor of 2-3. The estimated number of cases averted by backward tracing became greater with a higher degree of overdispersion. Conclusion: Backward contact tracing can be an effective tool for outbreak control, especially in the presence of overdispersion as is observed with SARS-CoV-2.

Mathematical modelling to study the horizontal transfer of antimicrobial resistance genes in bacteria: current state of the field and recommendations.

Antimicrobial resistance (AMR) is one of the greatest public health challenges we are currently facing. To develop effective interventions against this, it is essential to understand the processes behind the spread of AMR. These are partly dependent on the dynamics of horizontal transfer of resistance genes between bacteria, which can occur by conjugation (direct contact), transformation (uptake from the environment) or transduction (mediated by bacteriophages). Mathematical modelling is a powerful tool to investigate the dynamics of AMR; however, the extent of its use to study the horizontal transfer of AMR genes is currently unclear. In this systematic review, we searched for mathematical modelling studies that focused on horizontal transfer of AMR genes. We compared their aims and methods using a list of predetermined criteria and used our results to assess the current state of this research field. Of the 43 studies we identified, most focused on the transfer of single genes by conjugation in Escherichia coli in culture and its impact on the bacterial evolutionary dynamics. Our findings highlight the existence of an important research gap in the dynamics of transformation and transduction and the overall public health implications of horizontal transfer of AMR genes. To further develop this field and improve our ability to control AMR, it is essential that we clarify the structural complexity required to study the dynamics of horizontal gene transfer, which will require cooperation between microbiologists and modellers.

Feasibility of informing syndrome-level empiric antibiotic recommendations using publicly available antibiotic resistance datasets.

Background: Antibiotics are most often prescribed empirically, meaning that they are used to treat infection syndromes prior to identification of the causative bacteria and their susceptibility to antibiotics. The effectiveness of antibiotic therapies is now compromised by the emergence and spread of antibiotic-resistant bacteria. Guidelines on empiric antibiotic therapy are a key component of effective clinical care for infection syndromes, as treatment needs to be informed by knowledge of likely aetiology and bacterial resistance patterns. Methods: We used open-access antimicrobial resistance (AMR) surveillance datasets, including the newly available ATLAS dataset from Pfizer, to derive a composite index of antibiotic resistance for common infection syndromes. Results: We developed a framework that integrated data on antibiotic prescribing guidelines, aetiology of infections, access to and cost of antibiotics, with antibiotic susceptibilities from global AMR surveillance datasets to create an empirical prescribing index. The results are presented in an interactive web app to allow users to visualise underlying resistance rates to first-line empiric antibiotics for their infection syndromes and countries of interest. Conclusions: We found that whilst an index for empiric antibiotic therapy based on resistance data can technically be created, the ATLAS dataset in its current form can only inform on a limited number of infection syndromes. Other open-access AMR surveillance datasets (ECDC Surveillance Atlas, CDDEP ResistanceMap and WHO GLASS datasets) are largely limited to bacteraemia-derived specimens and cannot directly inform treatment of other infection syndromes. With improving data availability on international rates of AMR and better understanding of infection aetiology, our approach may prove useful for informing empiric prescribing decisions in settings with limited local AMR surveillance data. Syndrome-level resistance could be a more clinically relevant measure of resistance to inform on the appropriateness of empiric antibiotic therapies at the country-level.