Search

Purpose: Contact tracing is a crucial tool in infection prevention and control (IPC), which aims to identify outbreaks and prevent onward transmission. What constitutes a contact is typically based on strict binary criteria (i.e., being at a location at the same time). Missing data, indirect contacts and background sources can however substantially alter contact-tracing investigations. Here, we present StEP, a Spatial-temporal Epidemiological Proximity model that accounts for imperfect data by introducing a network-based notion of contact based on spatial-temporal proximity derived from background flows of patient movement. Methods & Materials: We showcase StEP by analysing outbreaks of multidrug-resistant bacteria and COVID-19 within a large hospital Trust in London (UK).StEP utilises spatial-temporal patient trajectories and the background hospital movement flows to recover enhanced contact networks. Firstly, we study a well-characterised outbreak of carbapenemase-producing Enterobacteriaceae (CPE) involving 116 hospitalised patients where genetic sequencing is used to learn model parameters. Secondly, our trained model is deployed in an unsupervised manner on three unseen outbreaks involving 867 patients of related CPE-types. Thirdly, we test application to an altogether novel pathogen by analysing a hospital outbreak of COVID-19 among 90 hospital patients, and demonstrate the power of StEP when characterising newly emerging diseases, even when there is a lack of sequencing data. Results: In addition to recovering core contact structures, StEP identifies missing contacts that link seemingly unconnected infection clusters, revealing a larger extent of transmission than conventional methods. Via genomic analyses we confirm that the additional contacts detected through StEP lead to improved alignment to the plasmid phylogeny (the major outbreak driving force). Hence the StEP contact network is most aligned to the transmission structure. Conclusion: By considering spatial-temporal information in a continuous manner, StEP tackles several challenges associated with traditional contact-tracing. StEP allows both direct and indirect contacts as possible routes of disease transmission and is tuneable to a pathogen's epidemiological characteristics. Such flexible use of heterogeneous data in uncertain situations can significantly enhance IPC.

Purpose: Predicting healthcare-acquired infections (HAIs) has the potential to revolutionise the prevention and control of transmissible infections. Existing prediction models for HAIs, however, fail to capture fully the contact-driven nature of infectious diseases. Here, we investigate the epidemiological predictivity of patient contact patterns through a forecasting model for hospital-onset COVID-19 infection (HOCI). Methods & Materials: Our cohort comprises all patient admissions at a large London NHS Trust between 1/04/2020 and 1/04/2021. For patients, we consider (i) their hospital pathway, (ii) patient contacts, and (iii) date of COVID-19 infection. We consider rolling 14-day windows and forecast patient infection over the subsequent 7 days. Over each window, we construct a patient contact network and compute network features that capture contact centrality. We then combine network features, hospital environmental variables and patient clinical data to predict subsequent infections. Results: A total of 51,157 patient admissions/episodes were observed during the study. Across all models, we find that contact-network features showed the highest performance (0.91 AUC-ROC). A reduced model with the six most predictive variables was almost as predictive and contained five features from patient contact (including direct contact with and network proximity to infectious cases) and only one environmental variable (length of stay). Conclusion: Our results reveal that the number of direct contacts and network proximity to infectious patient(s) are highly predictive of HOCI. Such contact-based risk factors are easily extracted from routinely collected electronic health records providing a highly accessible route to improve personalised disease prognostics in future models.

Platelet transfusions are given to patients in hospital who have a low blood platelet count (thrombocytopenia) either because of major bleeding (following trauma or surgery) or because the bone marrow production of platelets is impaired often due to chemotherapy, infiltration with malignant cells, fibrosis or genetic disorders. We are currently entirely reliant on blood donors as a source of platelets in transfusion medicine. However, the demand for platelets continues to rise, driven by an aging population, advances in medical procedures and ever more aggressive cancer therapies, while the supply of blood donors continues to remain static. In recent years, several groups have made major advances toward the generation of platelets in vitro for human transfusion. Recent successes include results in both generating mature human megakaryocytes as well as in developing bioreactors for extracting platelets from these megakaryocytes. Platelets made in vitro could address several issues inherent to platelets derived from blood donors – the ability to scale up/down more flexibly according to demand and therefore less precarious supply line, reduction of the risk of exposure to infectious agents and finally the possibility of engineering stem cells to reduce immunogenicity. Here we define the quality control tools and suggest measures for implementation across the field for in vitro platelet genesis, to aid collaboration between laboratories and to aid production of the burdens of proof that will eventually be required by regulators for efficacy and biosafety. We will do this firstly, by addressing the quality control of the nucleated cells used to make the platelets with a particular emphasis to safety issues and secondly, we will look at how platelet function measurement are addressed particularly in the context of platelets derived in vitro.