Rapid identification and subsequent contextualization of an outbreak of methicillin-resistant Staphylococcus aureus in a neonatal intensive care unit using nanopore sequencing.

Outbreaks of methicillin-resistant Staphylococcus aureus (MRSA) are well described in the neonatal intensive care unit (NICU) setting. Genomics has revolutionized the investigation of such outbreaks; however, to date, this has largely been completed retrospectively and has typically relied on short-read platforms. In 2022, our laboratory established a prospective genomic surveillance system using Oxford Nanopore Technologies sequencing for rapid outbreak detection. Herein, using this system, we describe the detection and control of an outbreak of sequence-type (ST)97 MRSA in our NICU. The outbreak was identified 13 days after the first MRSA-positive culture and at a point where there were only two known cases. Ward screening rapidly defined the extent of the outbreak, with six other infants found to be colonized. There was minimal transmission once the outbreak had been detected and appropriate infection control measures had been instituted; only two further ST97 cases were detected, along with three unrelated non-ST97 MRSA cases. To contextualize the outbreak, core-genome single-nucleotide variants were identified for phylogenetic analysis after de novo assembly of nanopore data. Comparisons with global (n=45) and national surveillance (n=35) ST97 genomes revealed the stepwise evolution of methicillin resistance within this ST97 subset. A distinct cluster comprising nine of the ten ST97-IVa genomes from the NICU was identified, with strains from 2020 to 2022 national surveillance serving as outgroups to this cluster. One ST97-IVa genome presumed to be part of the outbreak formed an outgroup and was retrospectively excluded. A second phylogeny was created using Illumina sequencing, which considerably reduced the branch lengths of the NICU isolates on the phylogenetic tree. However, the overall tree topology and conclusions were unchanged, with the exception of the NICU outbreak cluster, where differences in branch lengths were observed. This analysis demonstrated the ability of a nanopore-only prospective genomic surveillance system to rapidly identify and contextualize an outbreak of MRSA in a NICU.

Prospective Spatiotemporal Cluster Detection Using SaTScan: Tutorial for Designing and Fine-Tuning a System to Detect Reportable Communicable Disease Outbreaks.

Staff at public health departments have few training materials to learn how to design and fine-tune systems to quickly detect acute, localized, community-acquired outbreaks of infectious diseases. Since 2014, the Bureau of Communicable Disease at the New York City Department of Health and Mental Hygiene has analyzed reportable communicable diseases daily using SaTScan. SaTScan is a free software that analyzes data using scan statistics, which can detect increasing disease activity without a priori specification of temporal period, geographic location, or size. The Bureau of Communicable Disease's systems have quickly detected outbreaks of salmonellosis, legionellosis, shigellosis, and COVID-19. This tutorial details system design considerations, including geographic and temporal data aggregation, study period length, inclusion criteria, whether to account for population size, network location file setup to account for natural boundaries, probability model (eg, space-time permutation), day-of-week effects, minimum and maximum spatial and temporal cluster sizes, secondary cluster reporting criteria, signaling criteria, and distinguishing new clusters versus ongoing clusters with additional events. We illustrate how to support health equity by minimizing analytic exclusions of patients with reportable diseases (eg, persons experiencing homelessness who are unsheltered) and accounting for purely spatial patterns, such as adjusting nonparametrically for areas with lower access to care and testing for reportable diseases. We describe how to fine-tune the system when the detected clusters are too large to be of interest or when signals of clusters are delayed, missed, too numerous, or false. We demonstrate low-code techniques for automating analyses and interpreting results through built-in features on the user interface (eg, patient line lists, temporal graphs, and dynamic maps), which became newly available with the July 2022 release of SaTScan version 10.1. This tutorial is the first comprehensive resource for health department staff to design and maintain a reportable communicable disease outbreak detection system using SaTScan to catalyze field investigations as well as develop intuition for interpreting results and fine-tuning the system. While our practical experience is limited to monitoring certain reportable diseases in a dense, urban area, we believe that most recommendations are generalizable to other jurisdictions in the United States and internationally. Additional analytic technical support for detecting outbreaks would benefit state, tribal, local, and territorial public health departments and the populations they serve.

Methods for Genomic Epidemiology of Bacterial Pathogens: Example Salmonella.

Genomics has revolutionized how we characterize and monitor infectious diseases for public health. The surveillance and characterization of Salmonella has improved drastically within the past decade. In this chapter, we discuss the prerequisites for good bacterial genomics studies and make note of advantages and disadvantages of this research approach. We discuss methods for outbreak detection and the evolutionary and epidemiological characterization of Salmonella spp. We provide an outline for determining the sequence type and serotype of isolates, building a core genome phylogenetic tree, and detecting antimicrobial resistance genes, virulence factors, and mobile genetic elements. These methods can be used to study other pathogenic bacterial species.

Enhancing Public Health Surveillance: Outbreak Detection Algorithms Deployed for Syndromic Surveillance During Arbaeenia Mass Gatherings in Iraq.

BACKGROUND: Large gatherings often involve extended and intimate contact among individuals, creating environments conducive to the spread of infectious diseases. Despite this, there is limited research utilizing outbreak detection algorithms to analyze real syndrome data from such events. This study sought to address this gap by examining the implementation and efficacy of outbreak detection algorithms for syndromic surveillance during mass gatherings in Iraq. METHODS: For the study, 10 data collectors conducted field data collection over 10 days from August 25, 2023, to September 3, 2023. Data were gathered from 10 healthcare clinics situated along Ya Hussein Road, a major route from Najaf to Karbala in Iraq. Various outbreak detection algorithms, such as moving average, cumulative sum, and exponentially weighted moving average, were applied to analyze the reported syndromes. RESULTS: During the 10 days from August 25, 2023, to September 3, 2023, 12202 pilgrims visited 10 health clinics along a route in Iraq. Most pilgrims were between 20 and 59 years old (77.4%, n=9444), with more than half being foreigners (58.1%, n=7092). Among the pilgrims, 40.5% (n=4938) exhibited syndromes, with influenza-like illness (ILI) being the most common (48.8%, n=2411). Other prevalent syndromes included food poisoning (21.2%, n=1048), heatstroke (17.7%, n=875), febrile rash (9.0%, n=446), and gastroenteritis (3.2%, n=158). The cumulative sum (CUSUM) algorithm was more effective than exponentially weighted moving average (EWMA) and moving average (MA) algorithms for detecting small shifts. CONCLUSION: Effective public health surveillance systems are crucial during mass gatherings to swiftly identify and address emerging health risks. Utilizing advanced algorithms and real-time data analysis can empower authorities to improve their readiness and response capacity, thereby ensuring the protection of public health during these gatherings.

Setting clinically relevant thresholds for the notification of canine disease outbreaks to veterinary practitioners: an exploratory qualitative interview study.

Introduction: The Small Animal Veterinary Surveillance Network (SAVSNET) has developed mathematical models to analyse the veterinary practice and diagnostic laboratory data to detect genuine outbreaks of canine disease in the United Kingdom. There are, however, no validated methods available to establish the clinical relevance of these genuine statistical outbreaks before their formal investigation is conducted. This study aimed to gain an actionable understanding of a veterinary practitioner's preferences regarding which outbreak scenarios have a substantial impact on veterinary practice for six priority canine diseases in the United Kingdom. Methodology: An intensity sampling approach was followed to recruit veterinary practitioners according to their years of experience and the size of their practice. In-depth semi-structured and structured interviews were conducted to describe an outbreak notification and outbreak response thresholds for six canine endemic diseases, exotic diseases, and syndromes. These thresholds reflected participants' preferred balance between the levels of excess case incidence and predictive certainty of the detection system. Interviews were transcribed, and a thematic analysis was performed using NVivo 12. Results: Seven interviews were completed. The findings indicate higher preferred levels of predictive certainty for endemic diseases than for exotic diseases, ranging from 95 to 99% and 80 to 90%, respectively. The levels of excess case incidence were considered clinically relevant at values representing an increase of two to four times in the normal case incidence expectancy for endemic agents, such as parvovirus, and where they indicated a single case in the practice's catchment area for exotic diseases such as leishmaniosis and babesiosis. Conclusion: This study's innovative methodology uses veterinary practitioners' opinions to inform the selection of a notification threshold value in real-world applications of stochastic canine outbreak detection models. The clinically relevant thresholds derived from participants' needs will be used by SAVSNET to inform its outbreak detection system and to improve its response to canine disease outbreaks in the United Kingdom.

Genomic surveillance of STEC/EHEC infections in Germany 2020 to 2022 permits insight into virulence gene profiles and novel O-antigen gene clusters.

Shiga toxin-producing E. coli (STEC), including the subgroup of enterohemorrhagic E. coli (EHEC), are important bacterial pathogens which cause diarrhea and the severe clinical manifestation hemolytic uremic syndrome (HUS). Genomic surveillance of STEC/EHEC is a state-of-the-art tool to identify infection clusters and to extract markers of circulating clinical strains, such as their virulence and resistance profile for risk assessment and implementation of infection prevention measures. The aim of the study was characterization of the clinical STEC population in Germany for establishment of a reference data set. To that end, from 2020 to 2022 1257 STEC isolates, including 39 of known HUS association, were analyzed and lead to a classification of 30.4 % into 129 infection clusters. Major serogroups in all clinical STEC analyzed were O26, O146, O91, O157, O103, and O145; and in HUS-associated strains were O26, O145, O157, O111, and O80. stx1 was less frequently and stx2 or a combination of stx, eaeA and ehxA were more frequently found in HUS-associated strains. Predominant stx gene subtypes in all STEC strains were stx1a (24 %) and stx2a (21 %) and in HUS-associated strains were mainly stx2a (69 %) and the combination of stx1a and stx2a (12.8 %). Furthermore, two novel O-antigen gene clusters (RKI6 and RKI7) and strains of serovars O45:H2 and O80:H2 showing multidrug resistance were detected. In conclusion, the implemented surveillance tools now allow to comprehensively define the population of clinical STEC strains including those associated with the severe disease manifestation HUS reaching a new surveillance level in Germany.

Genome-Based Characterization of Listeria monocytogenes, Costa Rica.

Genomic data on the foodborne pathogen Listeria monocytogenes from Central America are scarce. We analyzed 92 isolates collected during 2009-2019 from different regions in Costa Rica, compared those to publicly available genomes, and identified unrecognized outbreaks. Our findings suggest mandatory reporting of listeriosis in Costa Rica would improve pathogen surveillance.

Evaluation of the application of sequence data to the identification of outbreaks of disease using anomaly detection methods.

Anomaly detection methods have a great potential to assist the detection of diseases in animal production systems. We used sequence data of Porcine Reproductive and Respiratory Syndrome (PRRS) to define the emergence of new strains at the farm level. We evaluated the performance of 24 anomaly detection methods based on machine learning, regression, time series techniques and control charts to identify outbreaks in time series of new strains and compared the best methods using different time series: PCR positives, PCR requests and laboratory requests. We introduced synthetic outbreaks of different size and calculated the probability of detection of outbreaks (POD), sensitivity (Se), probability of detection of outbreaks in the first week of appearance (POD1w) and background alarm rate (BAR). The use of time series of new strains from sequence data outperformed the other types of data but POD, Se, POD1w were only high when outbreaks were large. The methods based on Long Short-Term Memory (LSTM) and Bayesian approaches presented the best performance. Using anomaly detection methods with sequence data may help to identify the emergency of cases in multiple farms, but more work is required to improve the detection with time series of high variability. Our results suggest a promising application of sequence data for early detection of diseases at a production system level. This may provide a simple way to extract additional value from routine laboratory analysis. Next steps should include validation of this approach in different settings and with different diseases.

Epidemic intelligence activities among national public and animal health agencies: a European cross-sectional study.

Epidemic Intelligence (EI) encompasses all activities related to early identification, verification, analysis, assessment, and investigation of health threats. It integrates an indicator-based (IBS) component using systematically collected surveillance data, and an event-based component (EBS), using non-official, non-verified, non-structured data from multiple sources. We described current EI practices in Europe by conducting a survey of national Public Health (PH) and Animal Health (AH) agencies. We included generic questions on the structure, mandate and scope of the institute, on the existence and coordination of EI activities, followed by a section where respondents provided a description of EI activities for three diseases out of seven disease models. Out of 81 gatekeeper agencies from 41 countries contacted, 34 agencies (42%) from 26 (63%) different countries responded, out of which, 32 conducted EI activities. Less than half (15/32; 47%) had teams dedicated to EI activities and 56% (18/34) had Standard Operating Procedures (SOPs) in place. On a national level, a combination of IBS and EBS was the most common data source. Most respondents monitored the epidemiological situation in bordering countries, the rest of Europe and the world. EI systems were heterogeneous across countries and diseases. National IBS activities strongly relied on mandatory laboratory-based surveillance systems. The collection, analysis and interpretation of IBS information was performed manually for most disease models. Depending on the disease, some respondents did not have any EBS activity. Most respondents conducted signal assessment manually through expert review. Cross-sectoral collaboration was heterogeneous. More than half of the responding institutes collaborated on various levels (data sharing, communication, etc.) with neighbouring countries and/or international structures, across most disease models. Our findings emphasise a notable engagement in EI activities across PH and AH institutes of Europe, but opportunities exist for better integration, standardisation, and automatization of these efforts. A strong reliance on traditional IBS and laboratory-based surveillance systems, emphasises the key role of in-country laboratories networks. EI activities may benefit particularly from investments in cross-border collaboration, the development of methods that can automatise signal assessment in both IBS and EBS data, as well as further investments in the collection of EBS data beyond scientific literature and mainstream media.

ClustFinder: A tool for threshold-delineated clustering of microbial isolates by pairwise genomic distance.

This paper presents ClustFinder, a command line tool designed to automate clustering of genomes based on genomic distance. This tool will aid researchers and public health professionals in the identification of epidemiological clusters. Here, we demonstrate the usage of ClustFinder with example datasets. ClustFinder is available at github.com/Denes-Lab/ClustFinder.

Automatic case cluster detection using hospital electronic health record data.

Case detection through contact tracing is a key intervention during an infectious disease outbreak. However, contact tracing is an intensive process where a given contact tracer must locate not only confirmed cases but also identify and interview known contacts. Often these data are manually recorded. During emerging outbreaks, the number of contacts could expand rapidly and beyond this, when focused on individual transmission chains, larger patterns may not be identified. Understanding if particular cases can be clustered and linked to a common source can help to prioritize contact tracing effects and understand underlying risk factors for large spreading events. Electronic health records systems are used by the vast majority of private healthcare systems across the USA, providing a potential way to automatically detect outbreaks and connect cases through already collected data. In this analysis, we propose an algorithm to identify case clusters within a community during an infectious disease outbreak using Bayesian probabilistic case linking and explore how this approach could supplement outbreak responses; especially when human contact tracing resources are limited.

DiD IT?: a differences-in-differences investigation tool to quantify the impact of local incidents on public health using real-time syndromic surveillance health data.

Syndromic surveillance was originally developed to provide early warning compared to laboratory surveillance, but it is increasing used for real-time situational awareness. When a potential threat to public health is identified, a rapid assessment of its impact is required for public health management. When threats are localised, analysis is more complex as local trends need to be separated from national trends and differences compared to unaffected areas may be due to confounding factors such as deprivation or age distributions. Accounting for confounding factors usually requires an in-depth study, which takes time. Therefore, a tool is required which can provide a rapid estimate of local incidents using syndromic surveillance data.Here, we present 'DiD IT?', a new investigation tool designed to measure the impact of local threats to public health. 'DiD IT?' uses a difference-in-differences statistical approach to account for temporal and spatial confounding and provide a direct estimate of impact due to incidents. Temporal confounding differences are estimated by comparing unaffected locations during and outside of exposure periods. Whilst spatial confounding differences are estimated by comparing unaffected and exposed locations outside of the exposure period. Any remaining differences can be considered to be the direct effect of the local incident.We illustrate the potential utility of the tool through four examples of localised health protection incidents in England. The examples cover a range of data sources including general practitioner (GP) consultations, emergency department (ED) attendances and a telehealth call and online health symptom checker; and different types of incidents including, infectious disease outbreak, mass-gathering, extreme weather and an industrial fire. The examples use the UK Health Security Agency's ongoing real-time syndromic surveillance systems to show how results can be obtained in near real-time.The tool identified 700 additional online difficulty breathing assessments associated with a severe thunderstorm, 53 additional GP consultations during a mumps outbreak, 2-3 telehealth line calls following an industrial fire and that there was no significant increase in ED attendances during the G7 summit in 2021.DiD IT? can provide estimates for the direct impact of localised events in real-time as part of a syndromic surveillance system. Thus, it has the potential for enhancing surveillance and can be used to evaluate the effectiveness of extending national surveillance to a more granular local surveillance.

Detecting early signals of COVID-19 outbreaks in 2020 in small areas by monitoring healthcare utilisation databases: first lessons learned from the Italian Alert_CoV project.

BackgroundDuring the COVID-19 pandemic, large-scale diagnostic testing and contact tracing have proven insufficient to promptly monitor the spread of infections.AimTo develop and retrospectively evaluate a system identifying aberrations in the use of selected healthcare services to timely detect COVID-19 outbreaks in small areas.MethodsData were retrieved from the healthcare utilisation (HCU) databases of the Lombardy Region, Italy. We identified eight services suggesting a respiratory infection (syndromic proxies). Count time series reporting the weekly occurrence of each proxy from 2015 to 2020 were generated considering small administrative areas (i.e. census units of Cremona and Mantua provinces). The ability to uncover aberrations during 2020 was tested for two algorithms: the improved Farrington algorithm and the generalised likelihood ratio-based procedure for negative binomial counts. To evaluate these algorithms' performance in detecting outbreaks earlier than the standard surveillance, confirmed outbreaks, defined according to the weekly number of confirmed COVID-19 cases, were used as reference. Performances were assessed separately for the first and second semester of the year. Proxies positively impacting performance were identified.ResultsWe estimated that 70% of outbreaks could be detected early using the proposed approach, with a corresponding false positive rate of ca 20%. Performance did not substantially differ either between algorithms or semesters. The best proxies included emergency calls for respiratory or infectious disease causes and emergency room visits.ConclusionImplementing HCU-based monitoring systems in small areas deserves further investigations as it could facilitate the containment of COVID-19 and other unknown infectious diseases in the future.

Efficient algorithms for real-time syndromic surveillance.

OBJECTIVE: Outbreaks of influenza-like diseases often cause spikes in the demand for hospital beds. Early detection of these outbreaks can enable improved management of hospital resources. The objective of this study was to test whether surveillance algorithms designed to be responsive to a wide range of anomalous decreases in the time between emergency department (ED) presentations with influenza-like illnesses provide efficient early detection of these outbreaks. METHODS: Our study used data on ED presentations to major public hospitals in Queensland, Australia across 2017-2020. We developed surveillance algorithms for each hospital that flag potential outbreaks when the average time between successive ED presentations with influenza-like illnesses becomes anomalously small. We designed one set of algorithms to be responsive to a wide range of anomalous decreases in the time between presentations. These algorithms concurrently monitor three exponentially weighted moving averages (EWMAs) of the time between presentations and flag an outbreak when at least one EWMA falls below its control limit. We designed another set of algorithms to be highly responsive to narrower ranges of anomalous decreases in the time between presentations. These algorithms monitor one EWMA of the time between presentations and flag an outbreak when the EWMA falls below its control limit. Our algorithms use dynamic control limits to reflect that the average time between presentations depends on the time of year, time of day, and day of the week. RESULTS: We compared the performance of the algorithms in detecting the start of two epidemic events at the hospital-level: the 2019 seasonal influenza outbreak and the early-2020 COVID-19 outbreak. The algorithm that concurrently monitors three EWMAs provided significantly earlier detection of these outbreaks than the algorithms that monitor one EWMA. CONCLUSION: Surveillance algorithms designed to be responsive to a wide range of anomalous decreases in the time between ED presentations are highly efficient at detecting outbreaks of influenza-like diseases at the hospital level.

Insights from Review and Content Analysis of Current COVID-19 Mobile Apps and Recommendations for Future Pandemics.

A number of mobile health apps related to coronavirus infectious disease 2019 (COVID-19) have been developed, but research into app content analytics for effective surveillance and management is still in its preliminary stages. The present study aimed to identify the purpose and functions of the currently available COVID-19 apps using content analysis. The secondary aim was to propose directions for the future development of apps that aid infectious disease surveillance and control with a focus on enhancing the app content and quality. Prior to conducting an app search in the App Store and the Google Play Store, we reviewed previous studies on COVID-19 apps found in Google Scholar and PubMed to examine the main purposes of the apps. Using the five selected keywords based on the review, we searched the two app stores to retrieve eligible COVID-19 apps including those already addressed in the reviewed literature. We conducted descriptive and content analyses of the selected apps. We classified the purpose types of the COVID-19 apps into the following five categories: Information provision, tracking, monitoring, mental health management, and engagement. We identified 890 apps from the review articles and the app stores: 47 apps met the selection criteria and were included in the content analysis. Among the selected apps, iOS apps outnumbered Android apps, 27 apps were government-developed, and most of the apps were created in the United States. The most common function for the iOS apps (63.6%) and Android apps (62.5%) was to provide COVID-19-related knowledge. The most common function among the tracking apps was to notify users of contact with infected people by the iOS apps (40.9%) and Android apps (37.5%). About 29.5% of the iOS apps and 25.0% of the Android apps were used to record symptoms and self-diagnose. Significantly fewer apps targeted mental health management and engagement. Six iOS apps (6/44, 13.6%) and four Android apps (4/24, 16.7%) provided behavioral guidelines about the pandemic. Two iOS apps (2/44, 4.5%) and two Android apps (2/24, 8.3%) featured communication functions. The present content analysis revealed that most of the apps provided unilateral information and contact tracing or location tracking. Several apps malfunctioned. Future research and development of COVID-19 apps or apps for other emerging infectious diseases should address the quality and functional improvements, which should begin with continuous monitoring and actions to mitigate any technical errors.

Guiding the design of SARS-CoV-2 genomic surveillance by estimating the resolution of outbreak detection.

Genomic surveillance of SARS-CoV-2 has been essential to inform public health response to outbreaks. The high incidence of infection has resulted in a smaller proportion of cases undergoing whole genome sequencing due to finite resources. We present a framework for estimating the impact of reduced depths of genomic surveillance on the resolution of outbreaks, based on a clustering approach using pairwise genetic and temporal distances. We apply the framework to simulated outbreak data to show that outbreaks are detected less frequently when fewer cases are subjected to whole genome sequencing. The impact of sequencing fewer cases depends on the size of the outbreaks, and on the genetic and temporal similarity of the index cases of the outbreaks. We also apply the framework to an outbreak of the SARS-CoV-2 Delta variant in New South Wales, Australia. We find that the detection of clusters in the outbreak would have been delayed if fewer cases had been sequenced. Existing recommendations for genomic surveillance estimate the minimum number of cases to sequence in order to detect and monitor new virus variants, assuming representative sampling of cases. Our method instead measures the resolution of clustering, which is important for genomic epidemiology, and accommodates sampling biases.

Delayed outbreak detection: a wake-up call to evaluate a surveillance system.

During May, 83 of the 120 districts in Uganda had reported malaria cases above the upper limit of the normal channel. Across all districts, cases had exceeded malaria normal channel upper limits for an average of six months. Yet no alarms had been raised! Starting in 2000, Uganda adopted the World Health Organization (WHO) Integrated Disease Surveillance and Response (IDSR) strategy for disease reporting, including for malaria. Even early on, however, it was unclear how effectively IDSR and DHIS2 were being used in Uganda. Outbreaks were consistently detected late, but the underlying cause of the late detection was unclear. Suspecting there might be gaps in the surveillance system that were not immediately obvious, the Uganda FETP was asked to evaluate the malaria surveillance system in Uganda. This case study teaches trainees in Field Epidemiology and Laboratory Training Programs, public health students, public health workers who may participate in evaluation of public health surveillance systems, and others who are interested in this topic on reasons, steps, and attributes and uses the surveillance evaluation approach to identify gaps and facilitates discussion of practical solutions for improving a public health surveillance system.

Early detection of COVID-19 outbreaks using textual analysis of electronic medical records.

PURPOSE: Our objective was to develop a tool promoting early detection of COVID-19 cases by focusing epidemiological investigations and PCR examinations during a period of limited testing capabilities. METHODS: We developed an algorithm for analyzing medical records recorded by healthcare providers in the Israeli Defense Forces. The algorithm utilized textual analysis to detect patients presenting with suspicious symptoms and was tested among 92 randomly selected units. Detection of a potential cluster of patients in a unit prompted a focused epidemiological investigation aided by data provided by the algorithm. RESULTS: During a month of follow up, the algorithm has flagged 17 of the units for investigation. The subsequent epidemiological investigations led to the testing of 78 persons and the detection of eight cases in four clusters that were previously gone unnoticed. The resulting positive test rate of 10.25% was five time higher than the IDF average at the time of the study. No cases of COVID-19 in the examined units were missed by the algorithm. CONCLUSIONS: This study depicts the successful development and large scale deployment of a textual analysis based algorithm for early detection of COVID-19 cases, demonstrating the potential of natural language processing of medical text as a tool for promoting public health.

Automated detection of hospital outbreaks of multi-drug resistant pathogens in one Italian region.

BACKGROUND: Automated tools for antimicrobial resistance surveillance are critical for improving detection of drug-resistant organisms and informing prevention and control interventions. In this study, the WHONET-SaTScan software was used at a multihospital level in Tuscany, Italy, to identify case clusters consistent with hospital outbreaks caused by drug-resistant pathogens. METHODS: Antimicrobial resistance surveillance data from all Tuscany hospitals between January 2018 and December 2020 were analyzed using WHONET. The SaTScan package was used to detect case clusters applying a simulated prospective approach and the space-time permutation algorithm. Clusters were identified using resistance profiles and two distinct spatial variables: single medical services ('service') or groups of related services ('metaservice'). RESULTS: Data from eight bacterial pathogens were provided from 49 hospitals for 312,779 isolates from 158,809 patients. Single service-based analysis detected 693 hospital clusters, while metaservice-based analysis identified 635. There was no evidence for a difference between the two methods in terms of cluster length, cluster size, recurrence intervals, number of alerts, distribution across years or hospitals. Among clusters involving multiple services identified by both analyses, metaservice-detected clusters were usually larger and more statistically significant. CONCLUSIONS: WHONET-SaTScan proved to be a valuable multi-facility cluster detection tool that can be implemented for real-time surveillance.