Monitoring COVID-19 in Belgian general practice: A tool for syndromic surveillance based on electronic health records

BACKGROUND

COVID-19 may initially manifest as flu-like symptoms. As such, general practitioners (GPs) will likely to play an important role in monitoring the pandemic through syndromic surveillance.

OBJECTIVES

To present a COVID-19 syndromic surveillance tool in Belgian general practices.

METHODS

We performed a nationwide observational prospective study in Belgian general practices. The surveillance tool extracted the daily entries of diagnostic codes for COVID-19 and associated conditions (suspected or confirmed COVID-19, acute respiratory infection and influenza-like illness) from electronic medical records. We calculated the 7-day rolling average for these diagnoses and compared them with data from two other Belgian population-based sources (laboratory-confirmed new COVID-19 cases and hospital admissions for COVID-19), using time series analysis. We also collected data from users and stakeholders about the syndromic surveillance tool and performed a thematic analysis.

RESULTS

4773 out of 11,935 practising GPs in Belgium participated in the study. The curve of contacts for suspected COVID-19 followed a similar trend compared with the curves of the official data sources: laboratory-confirmed COVID-19 cases and hospital admissions but with a 10-day delay for the latter. Data were quickly available and useful for decision making, but some technical and methodological components can be improved, such as a greater standardisation between EMR software developers.

CONCLUSION

The syndromic surveillance tool for COVID-19 in primary care provides rapidly available data useful in all phases of the COVID-19 pandemic to support data-driven decision-making. Potential enhancements were identified for a prospective surveillance tool.

Community-based symptom reporting among agro-pastoralists and their livestock in Chad in a One Health approach

One Health Syndromic Surveillance has a high potential for detecting early epidemiological events in remote and hard-to-reach populations. Chadian pastoralists living close to their animals and being socio-economically unprivileged have an increased risk for zoonosis exposure. Engaging communities in disease surveillance could also strengthen preparedness capacities for outbreaks in rural Chad. This study describes a retrospective cross-sectional survey that collected data on clinical symptoms reported in people and livestock in Chadian agro-pastoral communities. In January-February 2018, interviews were conducted in rural households living in nomadic camps or settled villages in the Yao and Danamadji health districts. The questionnaire covered demographic data and symptoms reported in humans and animals for the hot, wet, and cold seasons over the last 12 months. Incidence rates of human and animal symptoms were comparatively analyzed at the household level. Ninety-two households with a homogeneous socio-demographic distribution were included. We observed cough and diarrhea as the most frequent symptoms reported simultaneously in humans and animals. In all species, the incidence rate of cough was significantly higher during the cold season, and diarrhea tended to occur more frequently during the wet season. However, the incidence rate of cough and diarrhea in animals did not predict the incidence rate of these symptoms in humans. Overall, the variations in reported symptoms were consistent with known seasonal, regional, and sociological influences on endemic diseases. Our retrospective study demonstrated the feasibility of collecting relevant health data in humans and animals in remote regions with low access to health services by actively involving community members. This encourages establishing real-time community-based syndromic surveillance in areas such as rural Chad.

Early warning of potential epidemics: A pilot application of an early warning tool to data from the pulmonary clinic of the university hospital of Thessaly, Greece

BACKGROUND & METHODS

This paper describes a pilot application of the Epidemic Volatility Index (EVI) to data from the pulmonary clinic of the University Hospital of Thessaly, Greece, for monitoring respiratory infections, COVID-19, and flu cases. EVI, a simple and easily implemented early warning method based on the volatility of newly reported cases, exhibited consistent and stable performance in detecting new waves of epidemics. The study highlights the importance of implementing early warning tools to address the effects of epidemics, including containment of outbreaks, timely intervention strategies, and resource allocation within real-world clinical settings as part of a broader public health strategy.

RESULTS

The results presented in the figures demonstrate the association between successive early warnings and the onset of new waves, providing valuable insights for proactive decision-making. A web-based application enabling real-time monitoring and informed decision-making by healthcare professionals, public health officials, and policymakers was developed.

CONCLUSIONS

This study emphasizes the significant role of early warning methods in managing epidemics and safeguarding public health. Future research may explore extensions and combinations of multiple warning systems for optimal outbreak interventions and application of the methods in the context of personalized medicine.

Enhanced measures, including PCR-based screening and syndromic surveillance for nosocomial outbreaks of the COVID-19 Omicron variant, using descriptive epidemiology and whole-genome sequencing in a Japanese tertiary care hospital

INTRODUCTION

In this study, we aimed to analyze the effectiveness of enhanced preventive measures against nosocomial COVID-19 Omicron outbreaks based on those encountered.

METHODS

We introduced PCR-based screening and syndromic surveillance, in addition to standard and transmission-based precautions, during a COVID-19 outbreak in three wards of Kagoshima University Hospital, a Japanese tertiary care hospital, in February 2022, amid the Omicron variant endemic. Furthermore, we analyzed the descriptive epidemiology and whole-genome sequencing (WGS) of positive SARS-CoV-2 PCR samples from this outbreak.

RESULTS

PCR-based screening tests were conducted following the identification of three cases through syndromic surveillance. As a result, 30 individuals tested positive for SARS-CoV-2, including 13 inpatients, five attendant family members, and 12 healthcare workers across the three wards. Notably, no new infections were observed within eight days following the implementation of preventive measures. Among the SARS-CoV-2 genomes analyzed (n = 16; 53.3%), all strains were identified as belonged to BA.1.1 variant. Detailed analysis of descriptive and molecular epidemiology, incorporating single-nucleotide polymorphism analysis of WGS and clarification of transmission links, considering two potential entry routes to the hospital.

CONCLUSIONS

Introduction of additional preventive measures, including PCR-based screening and syndromic surveillance, in addition to WGS and descriptive epidemiology, is useful for the early intervention of nosocomial outbreaks and for revealing the transmission route of the COVID-19 Omicron variant.

Epidemiological surveillance of respiratory viral infections in SARS-CoV-2-negative samples during COVID-19 pandemic in Iran

BACKGROUND

To improve the patient care, public health surveillance, and infection control, it is crucial to identify the presence and frequency of the common respiratory infections in individuals with COVID-19 symptoms but tested negative for SARS-CoV-2. This study aimed to shed light on this during the COVID-19 pandemic in Iran.

METHODS

In this cross-sectional study, a total of 1,002 patients with acute respiratory infection who had negative SARS-CoV-2 test results and referred to Valfajr Health Center, the National Collaborating Laboratory of Influenza and COVID-19 National Reference Laboratory at Pasteur Institute of Iran were recruited between January 2020 and January 2022. Nasopharyngeal and oropharyngeal swab samples were collected to detect 17 common respiratory viruses via TaqMan one-step real-time multiplex PCR. Demographic and clinical data of the participants were obtained from their electronic medical records.

RESULTS

In total, 218 samples (21.8%) were tested positive for at least one respiratory virus infection. Most of the common investigated respiratory viruses belonged to the years 2020 and 2022. The number of investigated patients in 2021 was few, which highlights the impact of health measures following the COVID-19 pandemic in Iran. Influenza A was the most common virus (5.8%), while adenovirus had the lowest prevalence (0.1%). Although the rate of respiratory virus infection was higher in men (24%) compared to women (19.3%), this difference was not statistically significant (P = 0.069). The prevalence of respiratory viruses had an inverse association with increasing age, with the highest rate (55.6%) observed in the age group below 2 years and the lowest rate (12.7%) in those above 65 years.

CONCLUSION

Our findings underscore the significance of adopting a comprehensive approach to respiratory infections detection and management. These results can be employed for the development of syndromic surveillance systems and implementation of the effective infection control measures. Furthermore, the results contribute to better understanding of the dynamics of respiratory viruses, both during pandemic periods and in non-pandemic contexts.

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.

Understanding the leading indicators of hospital admissions from COVID-19 across successive waves in the UK

Following the end of universal testing in the UK, hospital admissions are a key measure of COVID-19 pandemic pressure. Understanding leading indicators of admissions at the National Health Service (NHS) Trust, regional and national geographies help health services plan for ongoing pressures. We explored the spatio-temporal relationships of leading indicators of hospitalisations across SARS-CoV-2 waves in England. This analysis includes an evaluation of internet search volumes from Google Trends, NHS triage calls and online queries, the NHS COVID-19 app, lateral flow devices (LFDs), and the ZOE app. Data sources were analysed for their feasibility as leading indicators using Granger causality, cross-correlation, and dynamic time warping at fine spatial scales. Google Trends and NHS triages consistently temporally led admissions in most locations, with lead times ranging from 5 to 20 days, whereas an inconsistent relationship was found for the ZOE app, NHS COVID-19 app, and LFD testing, which diminished with spatial resolution, showing cross-correlation of leads between -7 and 7 days. The results indicate that novel surveillance sources can be used effectively to understand the expected healthcare burden within hospital administrative areas though the temporal and spatial heterogeneity of these relationships is a key determinant of their operational public health utility.

Comparison of statistical methods for the early detection of disease outbreaks in small population settings

Objectives

This study examines the performance of 6 aberration detection algorithms for the early detection of disease outbreaks in small population settings using syndrome-based early warning surveillance data collected by the Pacific Syndromic Surveillance System (PSSS). Although previous studies have proposed statistical methods for detecting aberrations in larger datasets, there is limited knowledge about how these perform in the presence of small numbers of background cases.

Methods

To address this gap a simulation model was developed to test and compare the performance of the 6 algorithms in detecting outbreaks of different magnitudes, durations, and case distributions.

Results

The study found that while the Early Aberration Reporting System-C1 algorithm developed by Hutwagner et al. outperformed others, no single approach provided reliable monitoring across all outbreak types. Furthermore, aberration detection approaches could only detect very large and acute outbreaks with any reliability.

Conclusion

The findings of this study suggest that algorithm-based approaches to outbreak signal detection perform poorly when applied to settings with small numbers of background cases and should not be relied upon in these contexts. This highlights the need for alternative approaches for accurate and timely outbreak detection in small population settings, particularly those that are resource-constrained.

Establishing severe acute respiratory infection (SARI) surveillance in a sentinel hospital, Ireland, 2021 to 2022

BackgroundIn 2020, due to the COVID-19 pandemic, the European Centre for Disease Prevention and Control (ECDC) accelerated development of European-level severe acute respiratory infection (SARI) surveillance.AimWe aimed to establish SARI surveillance in one Irish hospital as part of a European network E-SARI-NET.MethodsWe used routine emergency department records to identify cases in one adult acute hospital. The SARI case definition was adapted from the ECDC clinical criteria for a possible COVID-19 case. Clinical data were collected using an online questionnaire. Cases were tested for SARS-CoV-2, influenza and respiratory syncytial virus (RSV), including whole genome sequencing (WGS) on SARS-CoV-2 RNA-positive samples and viral characterisation/sequencing on influenza RNA-positive samples. Descriptive analysis was conducted for SARI cases hospitalised between July 2021 and April 2022.ResultsOverall, we identified 437 SARI cases, the incidence ranged from two to 28 cases per week (0.7-9.2/100,000 hospital catchment population). Of 431 cases tested for SARS-CoV-2 RNA, 226 (52%) were positive. Of 349 (80%) cases tested for influenza and RSV RNA, 15 (4.3%) were positive for influenza and eight (2.3%) for RSV. Using WGS, we identified Delta- and Omicron-dominant periods. The resource-intensive nature of manual clinical data collection, specimen management and laboratory supply shortages for influenza and RSV testing were challenging.ConclusionWe successfully established SARI surveillance as part of E-SARI-NET. Expansion to additional sentinel sites is planned following formal evaluation of the existing system. SARI surveillance requires multidisciplinary collaboration, automated data collection where possible, and dedicated personnel resources, including for specimen management.

Changes in emergency department utilisation in Germany before and during different phases of the COVID-19 pandemic, using data from a national surveillance system up to June 2021

BACKGROUND

During the COVID-19 pandemic and associated public health and social measures, decreasing patient numbers have been described in various healthcare settings in Germany, including emergency care. This could be explained by changes in disease burden, e.g. due to contact restrictions, but could also be a result of changes in utilisation behaviour of the population. To better understand those dynamics, we analysed routine data from emergency departments to quantify changes in consultation numbers, age distribution, disease acuity and day and hour of the day during different phases of the COVID-19 pandemic.

METHODS

We used interrupted time series analyses to estimate relative changes for consultation numbers of 20 emergency departments spread throughout Germany. For the pandemic period (16-03-2020 - 13-06-2021) four different phases of the COVID-19 pandemic were defined as interruption points, the pre-pandemic period (06-03-2017 - 09-03-2020) was used as the reference.

RESULTS

The most pronounced decreases were visible in the first and second wave of the pandemic, with changes of - 30.0% (95%CI: - 32.2%; - 27.7%) and - 25.7% (95%CI: - 27.4%; - 23.9%) for overall consultations, respectively. The decrease was even stronger for the age group of 0-19 years, with - 39.4% in the first and - 35.0% in the second wave. Regarding acuity levels, consultations assessed as urgent, standard, and non-urgent showed the largest decrease, while the most severe cases showed the smallest decrease.

CONCLUSIONS

The number of emergency department consultations decreased rapidly during the COVID-19 pandemic, without extensive variation in the distribution of patient characteristics. Smallest changes were observed for the most severe consultations and older age groups, which is especially reassuring regarding concerns of possible long-term complications due to patients avoiding urgent emergency care during the pandemic.

Enhanced passive surveillance for early detection of African and classical swine fevers

African swine fever (ASF) and classical swine fever (CSF) are transboundary animal diseases (TADs) of pigs. Much effort and resources are regularly put into preventing these diseases' introduction in free areas. Passive surveillance activities bring the highest chances for the early detection of TAD incursions because they are routinely and widely conducted at farms, and because these activities focus on the time between introduction and when the first sample is sent for diagnostic testing. The authors proposed the implementation of an enhanced passive surveillance (EPS) protocol based on collecting data through participatory surveillance actions using an objective and adaptable scoring system to aid the early detection of ASF or CSF at the farm level. The protocol was applied in two commercial pig farms for ten weeks in the Dominican Republic, which is a CSF- and ASF-infected country. This study was a proof of concept, based on the EPS protocol to aid detection of substantial variations in the risk score triggering testing. One of the followed farms had score variation, which triggered testing of the animals, although the test results were negative. The study enables assessment of some of the weaknesses associated with passive surveillance and provides lessons applicable to the problem. Results demonstrate the potential for overcoming some issues preventing the broad application of EPS protocols and suggest that standardised approaches may contribute to the early detection of CSF and ASF introductions.

Syndromic surveillance system during mass gathering of Panchkroshi Yatra festival, Ujjain, Madhya Pradesh, India

Background

The health implications surrounding a mass gathering pose significant challenges to public health officials. The use of syndromic surveillance provides an ideal method for achieving the public health goals and objectives at such events. In the absence of published reports of systematic documentation of public health preparedness in mass gatherings in the local context, we describe the public health preparedness and demonstrate the operational feasibility of a tablet-based participatory syndromic surveillance among pilgrims during the annual ritual circumambulation- Panchkroshi Yatra.

Methods

A real-time surveillance system was established from 2017-2019 to capture all the health consultations done at the designated points (medical camps) in the Panchkroshi yatra area of the city Ujjain in Madhya Pradesh. We also surveyed a subset of pilgrims in 2017 to gauge satisfaction with the public health measures such as sanitation, water, safety, food, and cleanliness.

Results

In 2019, injuries were reported in the highest proportion (16.7%; 794/4744); most numbers of fever cases (10.6%; 598/5600) were reported in 2018, while 2017 saw the highest number of patient presentations of abdominal pain (7.73%; 498/6435).

Conclusion

Public health and safety measures were satisfactory except for the need for setting up urinals along the fixed route of the circumambulation. A systematic data collection of selected symptoms among yatris and their surveillance through tablet could be established during the panchkroshi yatra, which can complement the existing surveillance for detecting early warning signals. We recommend the implementation of such tablet-based surveillance during such mass gathering events.

Citywide wastewater SARS-CoV-2 levels strongly correlated with multiple disease surveillance indicators and outcomes over three COVID-19 waves

Public health surveillance systems for COVID-19 are multifaceted and include multiple indicators reflective of different aspects of the burden and spread of the disease in a community. With the emergence of wastewater disease surveillance as a powerful tool to track infection dynamics of SARS-CoV-2, there is a need to integrate and validate wastewater information with existing disease surveillance systems and demonstrate how it can be used as a routine surveillance tool. A first step toward integration is showing how it relates to other disease surveillance indicators and outcomes, such as case positivity rates, syndromic surveillance data, and hospital bed use rates. Here, we present an 86-week long surveillance study that covers three major COVID-19 surges. City-wide SARS-CoV-2 RNA viral loads in wastewater were measured across 39 wastewater treatment plants and compared to other disease metrics for the city of Houston, TX. We show that wastewater levels are strongly correlated with positivity rate, syndromic surveillance rates of COVID-19 visits, and COVID-19-related general bed use rates at hospitals. We show that the relative timing of wastewater relative to each indicator shifted across the pandemic, likely due to a multitude of factors including testing availability, health-seeking behavior, and changes in viral variants. Next, we show that individual WWTPs led city-wide changes in SARS-CoV-2 viral loads, indicating a distributed monitoring system could be used to enhance the early-warning capability of a wastewater monitoring system. Finally, we describe how the results were used in real-time to inform public health response and resource allocation.

Exposure to industrial hog and poultry operations and urinary tract infections in North Carolina, USA

An increasing share of urinary tract infections (UTIs) are caused by extraintestinal pathogenic Escherichia coli (ExPEC) lineages that have also been identified in poultry and hogs with high genetic similarity to human clinical isolates. We investigated industrial food animal production as a source of uropathogen transmission by examining relationships of hog and poultry density with emergency department (ED) visits for UTIs in North Carolina (NC). ED visits for UTI in 2016-2019 were identified by ICD-10 code from NC's ZIP code-level syndromic surveillance system and livestock counts were obtained from permit data and aerial imagery. We calculated separate hog and poultry spatial densities (animals/km²) by Census block with a 5 km buffer on the block perimeter and weighted by block population to estimate mean ZIP code densities. Associations between livestock density and UTI incidence were estimated using a reparameterized Besag-York-Mollié (BYM2) model with ZIP code population offsets to account for spatial autocorrelation. We excluded metropolitan and offshore ZIP codes and assessed effect measure modification by calendar year, ZIP code rurality, and patient sex, age, race/ethnicity, and health insurance status. In single-animal models, hog exposure was associated with increased UTI incidence (rate ratio [RR]: 1.21, 95 % CI: 1.07-1.37 in the highest hog-density tertile), but poultry exposure was associated with reduced UTI rates (RR: 0.86, 95 % CI: 0.81-0.91). However, the reference group for single-animal poultry models included ZIP codes with only hogs, which had some of the highest UTI rates; when compared with ZIP codes without any hogs or poultry, there was no association between poultry exposure and UTI incidence. Hog exposure was associated with increased UTI incidence in areas that also had medium to high poultry density, but not in areas with low poultry density, suggesting that intense hog production may contribute to increased UTI incidence in neighboring communities.

Expert consultation using the on-line Delphi method for the revision of syndromic groups compiled from emergency data (SOS Médecins and OSCOUR®) in France

Background

Consultation data from emergency general practitioners known as SOS Médecins and emergency departments (ED) from OSCOUR® network to the French syndromic surveillance system SurSaUD® (Surveillance sanitaire des urgences et décès). These data are aggregated and monitored on a daily basis through groupings of one or more medical symptoms or diagnoses (“syndromic groups” (SG)).

The objective of this study was to evaluate, revise and enrich the composition of SGs through a consensus of experts who contributed or have experience in syndromic surveillance.

Methods

Three rounds of a Delphi survey were organised, involving 15 volunteers from SOS Médecins and 64 ED physicians in the OSCOUR® network as well as 8 international epidemiologists. Thirty-four SOS Médecins and 40 OSCOUR® SGs covering major medical specialities were put to the experts, along with their diagnostic codes and their surveillance objectives. In each round, the experts could retain or reject the codes according to the surveillance objective. The panel could also put forward new diagnostic codes in the 1st round, included in subsequent rounds. Consensus was reached for a code if 80% of participants had chosen to keep it, or less than 20% to reject it.

Results

A total of 12 SOS Médecins doctors (80%), 30 ED doctors (47%) and 4 international experts (50%) participated in the three rounds. All of the SGs presented to the panel included 102 initial diagnostic codes and 73 additional codes for SOS Médecins, 272 initial diagnostic codes and 204 additional codes for OSCOUR®. At the end of the 3 rounds, 14 SOS Médecins (40%) and 11 OSCOUR® (28%) SGs achieved a consensus to maintain all of their diagnostic codes. Among these, indicators of winter seasonal surveillance (bronchiolitis and gastroenteritis) were included.

Conclusion

This study involved a panel of national experts with international representation and a good level of involvement throughout the survey. In the absence of a standard definition, the Delphi method has been shown to be useful in defining and validating syndromic surveillance indicators.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-022-14157-x.

Syndromic surveillance of respiratory infections during protracted conflict: experiences from northern Syria 2016-2021

OBJECTIVE

Northern Syria faces a large burden of influenza-like illness (ILI) and severe acute respiratory illness (SARI). This study aimed to investigate the trends of Early Warning and Response Network (EWARN) reported ILI and SARI in northern Syria between 2016 and 2021 and the potential impact of SARS-CoV-2.

METHODS

We extracted weekly EWARN data on ILI/ SARI and aggregated cases and consultations into 4-week intervals to calculate case positivity. We conducted a seasonal-trend decomposition to assess case trends in the presence of seasonal fluctuations.

RESULTS

It was observed that 4-week aggregates of ILI cases (n = 5,942,012), SARI cases (n = 114,939), ILI case positivity, and SARI case positivity exhibited seasonal fluctuations with peaks in the winter months. ILI and SARI cases in individuals aged ≥5 years surpassed those in individuals aged <5 years in late 2019. ILI cases clustered primarily in Aleppo and Idlib, whereas SARI cases clustered in Aleppo, Idlib, Deir Ezzor, and Hassakeh. SARI cases increased sharply in 2021, corresponding with a severe SARS-CoV-2 wave, compared with the steady increase in ILI cases over time.

CONCLUSION

Respiratory infections cause widespread morbidity and mortality throughout northern Syria, particularly with the emergence of SARS-CoV-2. Strengthened surveillance and access to testing and treatment are critical to manage outbreaks among conflict-affected populations.

Calculating incidence of Influenza-like and COVID-like symptoms from Flutracking participatory survey data

This article describes a new method for estimating weekly incidence (new onset) of symptoms consistent with Influenza and COVID-19, using data from the Flutracking survey. The method mitigates some of the known self-selection and symptom-reporting biases present in existing approaches to this type of participatory longitudinal survey data. The key novel steps in the analysis are: 1) Identifying new onset of symptoms for three different Symptom Groupings: COVID-like illness (CLI1+, CLI2+), and Influenza-like illness (ILI), for responses reported in the Flutracking survey. 2) Adjusting for symptom reporting bias by restricting the analysis to a sub-set of responses from those participants who have consistently responded for a number of weeks prior to the analysis week. 3) Weighting responses by age to adjust for self-selection bias in order to account for the under- and over-representation of different age groups amongst the survey participants. This uses the survey package [22] in R [30]. 4) Constructing 95% point-wise confidence bands for incidence estimates using weighted logistic regression from the survey package [21] in R [28]. In addition to describing these steps, the article demonstrates an application of this method to Flutracking data for the 12 months from 27th April 2020 until 25th April 2021.

The impact of SARS-CoV-2 on respiratory syndromic and sentinel surveillance in Israel, 2020: a new perspective on established systems

Background

The COVID-19 pandemic presented new challenges for the existing respiratory surveillance systems, and adaptations were implemented. Systematic assessment of the syndromic and sentinel surveillance platforms during the pandemic is essential for understanding the value of each platform in the context of an emerging pathogen with rapid global spread.

Aim

We aimed to evaluate systematically the performance of various respiratory syndromic surveillance platforms and the sentinel surveillance system in Israel from 1 January to 31 December 2020.

Methods

We compared the 2020 syndromic surveillance trends to those of the previous 3 years, using Poisson regression adjusted for overdispersion. To assess the performance of the sentinel clinic system as compared with the national SARS-CoV-2 repository, a cubic spline with 7 knots and 95% confidence intervals were applied to the sentinel network's weekly percentage of positive SARS-CoV-2 cases.

Results

Syndromic surveillance trends changed substantially during 2020, with a statistically significant reduction in the rates of visits to physicians and emergency departments to below previous years' levels. Morbidity patterns of the syndromic surveillance platforms were inconsistent with the progress of the pandemic, while the sentinel surveillance platform was found to reflect the national circulation of SARS-CoV-2 in the population.

Conclusion

Our findings reveal the robustness of the sentinel clinics platform for the surveillance of the main respiratory viruses during the pandemic and possibly beyond. The robustness of the sentinel clinics platform during 2020 supports its use in locations with insufficient resources for widespread testing of respiratory viruses.

Public health surveillance in the U.S. Department of Veterans Affairs: evaluation of the Praedico surveillance system

Background

Early threat detection and situational awareness are vital to achieving a comprehensive and accurate view of health-related events for federal, state, and local health agencies. Key to this are public health and syndromic surveillance systems that can analyze large data sets to discover patterns, trends, and correlations of public health significance. In 2020, Department of Veterans Affairs (VA) evaluated its public health surveillance system and identified areas for improvement.

Methods

Using the Centers for Disease Control and Prevention (CDC) Guidelines for Evaluating Public Health Surveillance Systems, we assessed the ability of the Praedico Surveillance System to perform public health surveillance for a variety of health issues and evaluated its performance compared to an enterprise data solution (VA Corporate Data Warehouse), legacy surveillance system (VA ESSENCE) and a national, collaborative syndromic surveillance platform (CDC NSSP BioSense).

Results

Review of system attributes found that the system was simple, flexible, and stable. Representativeness, timeliness, sensitivity, and Predictive Value Positive were acceptable but could be further improved. Data quality issues and acceptability present challenges that potentially affect the overall usefulness of the system.

Conclusions

Praedico is a customizable surveillance and data analytics platform built on big data technologies. Functionality is straightforward, with rapid query generation and runtimes. Data can be graphed, mapped, analyzed, and shared with key decision makers and stakeholders. Evaluation findings suggest that future development and system enhancements should focus on addressing Praedico data quality issues and improving user acceptability. Because Praedico is designed to handle big data queries and work with data from a variety of sources, it could be enlisted as a tool for interdepartmental and interagency collaboration and public health data sharing. We suggest that future system evaluations include measurements of value and effectiveness along with additional organizations and functional assessments.

Acute respiratory infection symptoms and COVID-19 testing behaviour: results based on South Australian health surveys

Background

Effective syndromic surveillance alongside COVID-19 testing behaviours in the population including in higher risk and hard to reach subgroups is vital to detect re-emergence of COVID-19 transmission in the community. The aim of this paper was to identify the prevalence of acute respiratory infection symptoms and coronavirus testing behaviour among South Australians using data from a population based survey.

Methods

We used cross-sectional data from the 2020 state-wide population level health survey on 6857 respondents aged 18 years and above. Descriptive statistics were used to explore the risk factors and multivariable logistic regression models were used to assess the factors associated with the acute respiratory infection symptoms and coronavirus testing behaviour after adjusting for gender, age, household size, household income, Aboriginal and/or Torres Strait Islander status, SEIFA, Country of birth, number of chronic diseases, wellbeing, psychological distress, and mental health.

Results

We found that 19.3% of respondents reported having symptoms of acute respiratory infection and the most commonly reported symptoms were a runny nose (11.2%), coughing (9.9%) and sore throat (6.2%). Fever and cough were reported by 0.8% of participants. Of the symptomatic respondents, 32.6% reported seeking health advice from a nurse, doctor or healthcare provider. Around 18% (n = 130) of symptomatic respondents had sought testing and a further 4.3% (n = 31) reported they intended to get tested. The regression results suggest that older age, larger household size, a higher number of chronic disease, mental health condition, poor wellbeing, and psychological distress were associated with higher odds of ARI symptoms. Higher household income was associated with lower odds of being tested or intending to be tested for coronavirus after adjusting for other explanatory variables.

Conclusions

There were relatively high rates of self-reported acute respiratory infection during a period of very low COVID-19 prevalence and low rate of coronavirus testing among symptomatic respondents. Ongoing monitoring of testing uptake, including in higher-risk groups, and possible interventions to improve testing uptake is key to early detection of disease.

Evidence of co-circulation of multiple arboviruses transmitted by Aedes species based on laboratory syndromic surveillance at a health unit in a slum of the Federal District, Brazil

Background

Vector-borne diseases, especially arboviruses transmitted by Aedes sp. mosquitos, should be a health policy priority in Brazil. Despite this urgency, there are significant limitations in the traditional surveillance system, mainly in vulnerable areas. This study aimed to investigate the circulation of dengue (DENV), Zika (ZIKV), and chikungunya viruses (CHIKV) by laboratory syndromic surveillance (LSS) in a slum area of the Federal District of Brazil, comparing the results with traditional surveillance data.

Methods

LSS for acute febrile and/or exanthematous symptoms was developed at a health unit of Cidade Estrutural, in order to identify the circulation of arboviruses transmitted by Aedes sp. mosquitos. Between June 2019 and March 2020, 131 valid participants were identified and sera tested by reverse transcription polymerase chain reaction (RT-PCR) for DENV (by serotype), ZIKV, and CHIKV acute infection and by immunoglobulin M enzyme-inked immunosorbent assay (ELISA-IgM) for DENV and CHIKV 15–21 days after symptom onset, when the participant reported no respiratory signs (cough and/or coryza). The results obtained were compared with traditional surveillance data for the study area and period.

Results

At least three DENV-1 (2.3%), four DENV-2 (3%), and one CHIKV (0.7%) cases were confirmed in the laboratory, showing evidence of hyperendemicity even though LSS had not reached the historic peak dengue fever months in the Federal District (April–May). When the results obtained here were compared with traditional surveillance, a significant discrepancy was observed, including underreporting of CHIKV infection.

Conclusions

In addition to the risks posed to the study population, the area investigated with its respective socio-environmental profile may be a potential site for spread of the virus, given the cosmopolitan presence of Aedes sp. and human mobility in the Federal District. It is also suggested that traditional epidemiological surveillance may be reporting acute viral infections other than DENV as dengue fever, while underreporting other arboviruses transmitted by Aedes sp. mosquitos in the Federal District.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13071-021-05110-9.

Describing the indirect impact of COVID-19 on healthcare utilisation using syndromic surveillance systems

Background

Since the end of January 2020, the coronavirus (COVID-19) pandemic has been responsible for a global health crisis. In England a number of non-pharmaceutical interventions have been introduced throughout the pandemic, including guidelines on healthcare attendance (for example, promoting remote consultations), increased handwashing and social distancing. These interventions are likely to have impacted the incidence of non–COVID-19 conditions as well as healthcare seeking behaviour. Syndromic Surveillance Systems offer the ability to monitor trends in healthcare usage over time.

Methods

This study describes the indirect impact of COVID-19 on healthcare utilisation using a range of syndromic indicators including eye conditions, mumps, fractures, herpes zoster and cardiac conditions. Data from the syndromic surveillance systems monitored by Public Health England were used to describe the number of contacts with NHS 111, general practitioner (GP) In Hours (GPIH) and Out-of-Hours (GPOOH), Ambulance and Emergency Department (ED) services over comparable periods before and during the pandemic.

Results

The peak pandemic period in 2020 (weeks 13–20), compared to the same period in 2019, displayed on average a 12% increase in NHS 111 calls, an 11% decrease in GPOOH consultations, and a 49% decrease in ED attendances. In the GP In Hours system, conjunctivitis consultations decreased by 64% and mumps consultations by 31%. There was a 49% reduction in attendance at EDs for fractures, and there was no longer any weekend increase in ED fracture attendances, with similar attendance patterns observed across each day of the week. There was a decrease in the number of ED attendances with diagnoses of myocardial ischaemia.

Conclusion

The COVID-19 pandemic drastically impacted healthcare utilisation for non-COVID-19 conditions, due to a combination of a probable decrease in incidence of certain conditions and changes in healthcare seeking behaviour. Syndromic surveillance has a valuable role in describing and understanding these trends.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-021-12117-5.

The impact of the national stay-at-home order on emergency department visits for suspected opioid overdose during the first wave of the COVID-19 pandemic

BACKGROUND

Although national syndromic surveillance data reported declines in emergency department (ED) visits after the declaration of the national stay-at-home order for COVID-19, little is known whether these declines were observed for suspected opioid overdose.

METHODS

This interrupted time series study used syndromic surveillance data from four states participating in the HEALing Communities Study: Kentucky, Massachusetts, New York, and Ohio. All ED encounters for suspected opioid overdose (n = 48,301) occurring during the first 31 weeks of 2020 were included. We examined the impact of the national public health emergency for COVID-19 (declared on March 14, 2020) on trends in ED encounters for suspected opioid overdose.

RESULTS

Three of four states (Massachusetts, New York and Ohio) experienced a statistically significant immediate decline in the rate of ED encounters for suspected opioid overdose (per 100,000) after the nationwide public health emergency declaration (MA: -0.99; 95 % CI: -1.75, -0.24; NY: -0.10; 95 % CI, -0.20, 0.0; OH: -0.33, 95 % CI: -0.58, -0.07). After this date, Ohio and Kentucky experienced a sustained rate of increase for a 13-week period. New York experienced a decrease in the rate of ED encounters for a 10-week period, after which the rate began to increase. In Massachusetts after a significant immediate decline in the rate of ED encounters, there was no significant difference in the rate of change for a 6-week period, followed by an immediate increase in the ED rate to higher than pre-COVID levels.

CONCLUSIONS

The heterogeneity in the trends in ED encounters between the four sites show that the national stay-at-home order had a differential impact on opioid overdose ED presentation in each state.

[The different periods of COVID-19 in Germany: a descriptive analysis from January 2020 to February 2021]

The first case of coronavirus SARS-CoV‑2 infection in Germany was diagnosed on 27 January 2020. To describe the pandemic course in 2020, we regarded four epidemiologically different periods and used data on COVID-19 cases from the mandatory reporting system as well as hospitalized COVID-19 cases with severe acute respiratory infection from the syndromic hospital surveillance.Period 0 covers weeks 5 to 9 of 2020, where mainly sporadic cases of younger age were observed and few regional outbreaks emerged. In total, 167 cases with mostly mild outcomes were reported. Subsequently, the first COVID-19 wave occurred in period 1 (weeks 10 to 20 of 2020) with a total of 175,013 cases throughout Germany. Increasingly, outbreaks in hospitals and nursing homes were registered. Moreover, elderly cases and severe outcomes were observed more frequently. Period 2 (weeks 21 to 39 of 2020) was an interim period with more mild cases, where many cases were younger and often travel-associated. Additionally, larger trans-regional outbreaks in business settings were reported. Among the 111,790 cases, severe outcomes were less frequent than in period 1. In period 3 (week 40 of 2020 to week 8 of 2021), the second COVID-19 wave started and peaked at the end of 2020. With 2,158,013 reported cases and considerably more severe outcomes in all age groups, the second wave was substantially stronger than the first wave.Irrespective of the different periods, more elderly persons and more men were affected by severe outcomes.

School-based surveillance of acute infectious disease in children: a systematic review

BACKGROUND

Syndromic surveillance systems are an essential component of public health surveillance and can provide timely detection of infectious disease cases and outbreaks. Whilst surveillance systems are generally embedded within healthcare, there is increasing interest in novel data sources for monitoring trends in illness, such as over-the-counter purchases, internet-based health searches and worker absenteeism. This systematic review considers the utility of school attendance registers in the surveillance of infectious disease outbreaks and occurrences amongst children.

METHODS

We searched eight databases using key words related to school absence, infectious disease and syndromic surveillance. Studies were limited to those published after 1st January 1995. Studies based in nursery schools or higher education settings were excluded. Article screening was undertaken by two independent reviewers using agreed eligibility criteria. Data extraction was performed using a standardised data extraction form. Outcomes included estimates of absenteeism, correlation with existing surveillance systems and associated lead or lag times.

RESULTS

Fifteen studies met the inclusion criteria, all of which were concerned with the surveillance of influenza. The specificity of absence data varied between all-cause absence, illness absence and syndrome-specific absence. Systems differed in terms of the frequency of data submissions from schools and the level of aggregation of the data. Baseline rates of illness absence varied between 2.3-3.7%, with peak absences ranging between 4.1-9.8%. Syndrome-specific absenteeism had the strongest correlation with other surveillance systems (r = 0.92), with illness absenteeism generating mixed results and all-cause absenteeism performing the least well. A similar pattern of results emerged in terms of lead and lag times, with influenza-like illness (ILI)-specific absence providing a 1-2 week lead time, compared to lag times reported for all-cause absence data and inconsistent results for illness absence data.

CONCLUSION

Syndrome-specific school absences have potential utility in the syndromic surveillance of influenza, demonstrating good correlation with healthcare surveillance data and a lead time of 1-2 weeks ahead of existing surveillance measures. Further research should consider the utility of school attendance registers for conditions other than influenza, to broaden our understanding of the potential application of this data for infectious disease surveillance in children.

SYSTEMATIC REVIEW REGISTRATION

PROSPERO 2019 CRD42019119737.

Development and evaluation of a new method to combine clinical impression survey data with existing laboratory data for veterinary syndromic surveillance with the Canada West Swine Health Intelligence Network (CWSHIN)

The Canada West Swine Health Intelligence Network (CWSHIN) is a surveillance system imbedded in an intelligence network. It has been conducting syndromic surveillance in the four western provinces of Canada since 2012. The quarterly activities include repeated clinical impression surveys, compilation of laboratory data, discussion of trends with an expert group (practitioners, laboratory diagnosticians) and swine health reports for producers and swine practitioners. However, due to the lack of standardized population identifiers across data sources usual methods of combining data could not be applied and the collated data were not being fully utilized and analysed. Therefore in 2019, CWSHIN underwent a substantial review resulting in the "Next Generation CWSHIN". The objectives of this study were to develop and evaluate a new data merging method to combine CWSHIN's clinical impression survey and laboratory data; and to provide examples of analyses and modeling based on these data. The data for analysis were restricted to repeated clinical impression surveys (2019-2020) from veterinary practitioners and laboratory diagnostic data (2016-2020). Merging surveillance data from existing sources can be challenging. Therefore, as an alternative to merge data using a hierarchy of population identifiers, we developed a Disease Map to link surveillance data from all our data-sources. The resulting Data Repository allowed monitoring of temporal trends of syndromes, clinical diseases, and laboratory identified organisms, but it cannot provide estimates of disease occurrence. Two main reasons were the lack of denominators and using existing data on routine diagnostic tests. Therefore, discussion in the expert group (veterinary practitioners, laboratory diagnosticians, swine health experts) was critical to the system's success. Based on repeated clinical impression surveys a stochastic scenario tree model for freedom from Foot and Mouth Disease (CWSHIN Blister model) was also developed. In conclusion, the method to link existing data systems from multiple divergent sources by means of a Disease Map improved CWSHIN's veterinary syndromic surveillance. Together the Data Repository and Disease map provided flexibility to monitor temporal trends, define populations and diseases, and allow analysis. However, it is critical that the surveillance is coupled with a good intelligence network that can help interpret the results and disseminate knowledge to veterinarians and producers.

The effectiveness of syndromic surveillance for the early detection of waterborne outbreaks: a systematic review

Background

Waterborne outbreaks are still a risk in high-income countries, and their early detection is crucial to limit their societal consequences. Although syndromic surveillance is widely used for the purpose of detecting outbreaks days earlier than traditional surveillance systems, evidence of the effectiveness of such systems is lacking. Thus, our objective was to conduct a systematic review of the effectiveness of syndromic surveillance to detect waterborne outbreaks.

Method

We searched the Cochrane Library, Medline/PubMed, EMBASE, Scopus, and Web of Science for relevant published articles using a combination of the keywords ‘drinking water’, ‘surveillance’, and ‘waterborne disease’ for the period of 1990 to 2018. The references lists of the identified articles for full-text record assessment were screened, and searches in Google Scholar using the same key words were conducted. We assessed the risk of bias in the included articles using the ROBINS-I tool and PRECEPT for the cumulative body of evidence.

Results

From the 1959 articles identified, we reviewed 52 articles, of which 18 met the eligibility criteria. Twelve were descriptive/analytical studies, whereas six were simulation studies. There is no clear evidence for syndromic surveillance in terms of the ability to detect waterborne outbreaks (low sensitivity and high specificity). However, one simulation study implied that multiple sources of signals combined with spatial information may increase the timeliness in detecting a waterborne outbreak and reduce false alarms.

Conclusion

This review demonstrates that there is no conclusive evidence on the effectiveness of syndromic surveillance for the detection of waterborne outbreaks, thus suggesting the need to focus on primary prevention measures to reduce the risk of waterborne outbreaks. Future studies should investigate methods for combining health and environmental data with an assessment of needed financial and human resources for implementing such surveillance systems. In addition, a more critical thematic narrative synthesis on the most promising sources of data, and an assessment of the basis for arguments that joint analysis of different data or dimensions of data (e.g. spatial and temporal) might perform better, should be carried out.

Trial registration

PROSPERO: International prospective register of systematic reviews. 2019. CRD42019122332.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-021-06387-y.

Development and validation of an automated emergency department-based syndromic surveillance system to enhance public health surveillance in Yukon: a lower-resourced and remote setting

BACKGROUND

Automated Emergency Department syndromic surveillance systems (ED-SyS) are useful tools in routine surveillance activities and during mass gathering events to rapidly detect public health threats. To improve the existing surveillance infrastructure in a lower-resourced rural/remote setting and enhance monitoring during an upcoming mass gathering event, an automated low-cost and low-resources ED-SyS was developed and validated in Yukon, Canada.

METHODS

Syndromes of interest were identified in consultation with the local public health authorities. For each syndrome, case definitions were developed using published resources and expert elicitation. Natural language processing algorithms were then written using Stata LP 15.1 (Texas, USA) to detect syndromic cases from three different fields (e.g., triage notes; chief complaint; discharge diagnosis), comprising of free-text and standardized codes. Validation was conducted using data from 19,082 visits between October 1, 2018 to April 30, 2019. The National Ambulatory Care Reporting System (NACRS) records were used as a reference for the inclusion of International Classification of Disease, 10th edition (ICD-10) diagnosis codes. The automatic identification of cases was then manually validated by two raters and results were used to calculate positive predicted values for each syndrome and identify improvements to the detection algorithms.

RESULTS

A daily secure file transfer of Yukon's Meditech ED-Tracker system data and an aberration detection plan was set up. A total of six syndromes were originally identified for the syndromic surveillance system (e.g., Gastrointestinal, Influenza-like-Illness, Mumps, Neurological Infections, Rash, Respiratory), with an additional syndrome added to assist in detecting potential cases of COVID-19. The positive predictive value for the automated detection of each syndrome ranged from 48.8-89.5% to 62.5-94.1% after implementing improvements identified during validation. As expected, no records were flagged for COVID-19 from our validation dataset.

CONCLUSIONS

The development and validation of automated ED-SyS in lower-resourced settings can be achieved without sophisticated platforms, intensive resources, time or costs. Validation is an important step for measuring the accuracy of syndromic surveillance, and ensuring it performs adequately in a local context. The use of three different fields and integration of both free-text and structured fields improved case detection.

Development and assessment of a hospital admissions-based syndromic surveillance system for COVID-19 in Ontario, Canada: ACES Pandemic Tracker

BACKGROUND

The COVID-19 pandemic has continued to pose a major global public health risk. The importance of public health surveillance systems to monitor the spread and impact of COVID-19 has been well demonstrated. The purpose of this study was to describe the development and effectiveness of a real-time public health syndromic surveillance system (ACES Pandemic Tracker) as an early warning system and to provide situational awareness in response to the COVID-19 pandemic in Ontario, Canada.

METHODS

We used hospital admissions data from the Acute Care Enhanced Surveillance (ACES) system to collect data on pre-defined groupings of symptoms (syndromes of interest; SOI) that may be related to COVID-19 from 131 hospitals across Ontario. To evaluate which SOI for suspected COVID-19 admissions were best correlated with laboratory confirmed admissions, laboratory confirmed COVID-19 hospital admissions data were collected from the Ontario Ministry of Health. Correlations and time-series lag analysis between suspected and confirmed COVID-19 hospital admissions were calculated. Data used for analyses covered the period between March 1, 2020 and September 21, 2020.

RESULTS

Between March 1, 2020 and September 21, 2020, ACES Pandemic Tracker identified 22,075 suspected COVID-19 hospital admissions (150 per 100,000 population) in Ontario. After correlation analysis, we found laboratory-confirmed hospital admissions for COVID-19 were strongly and significantly correlated with suspected COVID-19 hospital admissions when SOI were included (Spearman's rho = 0.617) and suspected COVID-19 admissions when SOI were excluded (Spearman's rho = 0.867). Weak to moderate significant correlations were found among individual SOI. Laboratory confirmed COVID-19 hospital admissions lagged in reporting by 3 days compared with suspected COVID-19 admissions when SOI were excluded.

CONCLUSIONS

Our results demonstrate the utility of a hospital admissions syndromic surveillance system to monitor and identify potential surges in severe COVID-19 infection within the community in a timely manner and provide situational awareness to inform preventive and preparatory health interventions.

Differences and similarities between emergency department syndromic surveillance and hospital discharge data for nonfatal drug overdose

PURPOSE

Emergency department syndromic surveillance and hospital discharge data have been used to detect and monitor nonfatal drug overdose, yet few studies have assessed the differences and similarities between these two data sources.

METHODS

The Centers for Disease Control and Prevention Drug Overdose Surveillance and Epidemiology system data from 14 states were used to compare these two sources at estimating monthly overdose burden and trends from January 2018 through December 2019 for nonfatal all drug, opioid-, heroin-, and stimulant-involved overdoses.

RESULTS

Compared to discharge data, syndromic data captured 13.3% more overall emergency department visits, 67.8% more all drug overdose visits, 15.6% more opioid-involved overdose visits, and 78.8% more stimulant-involved overdose visits. Discharge data captured 18.9% more heroin-involved overdoses. Significant trends were identified for all drug (Average Monthly Percentage Change [AMPC]=1.1, 95% CI=0.4,1.8) and stimulant-involved overdoses (AMPC=2.4, 95% CI=1.2,3.7) in syndromic data; opioid-involved overdoses increased in both discharge and syndromic data (AMPC_Discharge=0.9, 95% CI=0.2,1.7; AMPC_Syndromic=1.9, CI=1.1,2.8).

CONCLUSIONS

Results demonstrate that discharge data may be better for reporting counts, yet syndromic data are preferable to detect changes quickly and to alert practitioners and public health officials to local overdose clusters. These data sources do serve complementary purposes when examining overdose trends.

Utilizing the electronic health records to create a syndromic staff surveillance system during the COVID-19 outbreak

Objectives

Since December 2019, COVID-19 has caused a worldwide pandemic and Singapore has seen escalating cases with community spread. Aggressive contact tracing and identification of suspects has helped to identify local community clusters, surveillance being the key to early intervention. Healthcare workers (HCWs) have contracted COVID-19 infection both at the workplace and community. We aimed to create a prototype staff surveillance system for the detection of acute respiratory infection (ARI) clusters amongst our HCWs and describe its effectiveness.

Methods

A prototypical surveillance system was built on existing electronic health record infrastructure.

Results

Over a 10-week period, we investigated 10 ARI clusters amongst 7 departments. One of the ARI clusters was later determined to be related to COVID-19 infection. We demonstrate the feasibility of syndromic surveillance to detect ARI clusters during the COVID-19 outbreak.

Conclusion

The use of syndromic surveillance to detect ARI clusters amongst HCWs in the COVID-19 pandemic may enable early case detection and prevent onward transmission. It could be an important tool in infection prevention within healthcare institutions.

Syndromic surveillance using monthly aggregate health systems information data: methods with application to COVID-19 in Liberia

BACKGROUND

Early detection of SARS-CoV-2 circulation is imperative to inform local public health response. However, it has been hindered by limited access to SARS-CoV-2 diagnostic tests and testing infrastructure. In regions with limited testing capacity, routinely collected health data might be leveraged to identify geographical locales experiencing higher than expected rates of COVID-19-associated symptoms for more specific testing activities.

METHODS

We developed syndromic surveillance tools to analyse aggregated health facility data on COVID-19-related indicators in seven low- and middle-income countries (LMICs), including Liberia. We used time series models to estimate the expected monthly counts and 95% prediction intervals based on 4 years of previous data. Here, we detail and provide resources for our data preparation procedures, modelling approach and data visualisation tools with application to Liberia.

RESULTS

To demonstrate the utility of these methods, we present syndromic surveillance results for acute respiratory infections (ARI) at health facilities in Liberia during the initial months of the COVID-19 pandemic (January through August 2020). For each month, we estimated the deviation between the expected and observed number of ARI cases for 325 health facilities and 15 counties to identify potential areas of SARS-CoV-2 circulation.

CONCLUSIONS

Syndromic surveillance can be used to monitor health facility catchment areas for spikes in specific symptoms which may indicate SARS-CoV-2 circulation. The developed methods coupled with the existing infrastructure for routine health data systems can be leveraged to monitor a variety of indicators and other infectious diseases with epidemic potential.

Emergency department admissions for myocardial infarction and stroke in France during the first wave of the COVID-19 pandemic: National temporal trends and regional disparities

BACKGROUND

The coronavirus disease 2019 (COVID-19) pandemic and the national lockdown have led to significant changes in the use of emergency care by the French population.

AIMS

To describe the national and regional temporal trends in emergency department (ED) admissions for myocardial infarction (MI) and stroke, before, during and after the first national lockdown.

METHODS

The weekly numbers of ED admissions for MI and stroke were collected from the OSCOUR® network, which covers 93.3% of all ED admissions in France. National and regional incidence rate ratios from 02 February until 31 May (2020 versus 2017-2019) were estimated using Poisson regression for MI and stroke, before, during and after lockdown.

RESULTS

A decrease in ED admissions was observed for MI (-20% for ST-segment elevation MI and-25% for non-ST-segment elevation MI) and stroke (-18% for ischaemic and-22% for haemorrhagic) during the lockdown. The decrease became significant earlier for stroke than for MI. No compensatory increase in ED admissions was observed at the end of the lockdown for these diseases. Important regional disparities in ED admissions were observed, without correlation with the regional levels of COVID-19 cases. The impact of lockdown on ED admissions was particularly significant in six regions (Ile-de France, Occitanie, Provence-Alpes-Côte d'Azur, Nouvelle Aquitaine, Hauts-de-France and Bretagne).

CONCLUSIONS

The decrease in ED admissions for MI and stroke observed during the lockdown was probably caused by fear of COVID-19 and augmented by the lockdown, and was heterogeneous across the French territory. ED admissions were slow to return to the usual levels from previous years, without a compensatory increase. These results underline the need to reinforce messages directed at the population to encourage them to seek care without delay in case of cardiovascular symptoms.

Successful management of COVID-19 outbreak in a long-term care facility in Jeddah, Saudi Arabia: Epidemiology, challenges for prevention and adaptive management strategies

The global transmission of SARS-COV-2 constitutes a highly challenging situation for long-term care facilities, especially with the lack of standardized and approved procedures. Residents in these facilities are at high risk for contamination due to proximity, and to morbidity and mortality given their advanced age and critical baseline health conditions. This paper exposes the experience and outcomes of a COVID-19 outbreak in a long-term facility in Jeddah, Saudi Arabia, which occurred after admission of a new resident despite high baseline level of alertness including systematic isolation and screening of all newly admitted residents. We highlight the challenges for case detection and application of protective measures, and describe the adaptive management strategies implemented to contain the outbreak.

[Monitoring of COVID-19 by extending existing surveillance for acute respiratory infections]

Im Rahmen der nationalen Influenzapandemieplanung wurden in Deutschland neben dem Meldewesen gemäß Infektionsschutzgesetz (IfSG) weitere Überwachungssysteme etabliert. Ziel dieser Systeme sind die Beschreibung, Analyse und Bewertung der Situation bei akuten respiratorischen Erkrankungen (ARE), die Identifikation der hauptsächlich zirkulierenden Atemwegserreger und die Beschreibung des zeitlichen Verlaufs. Seit Beginn der COVID-19-Pandemie wurden die Systeme erweitert, um auch Infektionen mit SARS-CoV‑2 erfassen zu können.

In diesem Beitrag werden drei verschiedene Surveillance-Systeme für ARE vorgestellt: GrippeWeb, die Arbeitsgemeinschaft Influenza mit dem SEED^ARE-Modul (Sentinel zur elektronischen Erfassung von Diagnosecodes) und das Krankenhaus-Sentinel ICOSARI (ICD-10-code-basierte Krankenhaus-Surveillance schwerer akuter respiratorischer Infektionen). Mit diesen Systemen können ARE auf Bevölkerungsebene, im ambulanten und im stationären Bereich überwacht werden. Zusammen mit dem Monitoring der Mortalität liefern sie wichtige Hinweise zur Häufigkeit verschieden schwerer Krankheitsverläufe in der Bevölkerung. Um die Systeme für SARS-CoV‑2 zu erweitern, waren nur wenige Anpassungen notwendig.

Da die Falldefinitionen für ARE nicht geändert wurden, können in den beschriebenen Systemen historische Zeitreihen zum Vergleich herangezogen werden. Alle Systeme sind so aufgebaut, dass stabile und etablierte Bezugsgrößen für die Berechnung von wöchentlichen Anteilen und Raten zur Verfügung stehen. Dies ist eine wichtige Ergänzung zum Meldewesen gemäß IfSG, welches stark von Testkapazitäten und -strategien sowie veränderten Falldefinitionen abhängt. Die Surveillance-Systeme haben sich in der COVID-19-Pandemie auch im internationalen Vergleich als praktikabel und effizient erwiesen.

Predicting epidemics using search engine data: a comparative study on measles in the largest countries of Europe

Background

In recent years new forms of syndromic surveillance that use data from the Internet have been proposed. These have been developed to assist the early prediction of epidemics in various cases and diseases. It has been found that these systems are accurate in monitoring and predicting outbreaks before these are observed in population and, therefore, they can be used as a complement to other methods. In this research, our aim is to examine a highly infectious disease, measles, as there is no extensive literature on forecasting measles using Internet data,

Methods

This research has been conducted with official data on measles for 5 years (2013–2018) from the competent authority of the European Union (European Center of Disease and Prevention - ECDC) and data obtained from Google Trends by using scripts coded in Python. We compared regression models forecasting the development of measles in the five countries.

Results

Results show that measles can be estimated and predicted through Google Trends in terms of time, volume and the overall spread. The combined results reveal a strong relationship of measles cases with the predicted cases (correlation coefficient R= 0.779 in two-tailed significance p< 0.01). The mean standard error was relatively low 45.2 (12.19%) for the combined results. However, major differences and deviations were observed for countries with a relatively low impact of measles, such as the United Kingdom and Spain. For these countries, alternative models were tested in an attempt to improve the results.

Conclusions

The estimation of measles cases from Google Trends produces acceptable results and can help predict outbreaks in a robust and sound manner, at least 2 months in advance. Python scripts can be used individually or within the framework of an integrated Internet surveillance system for tracking epidemics as the one addressed here.

Using farmer observations for animal health syndromic surveillance: Participation and performance of an online enhanced passive surveillance system

The challenge of animal health surveillance is to provide the information necessary to appropriately inform disease prevention and control activities within the constraints of available resources. Syndromic surveillance of farmers' disease observations can improve animal health data capture from extensive livestock farming systems, especially where data are not otherwise being systematically collected or when data on confirmed aetiological diagnoses are unavailable at the disease level. As it is rarely feasible to recruit a truly random sample of farmers to provide observational reports, directing farmer sampling to align with the surveillance objectives is a reasonable and practical approach. As long as potential bias is recognised and managed, farmers who will report reliably can be desirable participants in a surveillance system. Thus, one early objective of a surveillance program should be to identify characteristics associated with reporting behaviour. Knowledge of the demographic and managerial characteristics of good reporters can inform efforts to recruit additional farms into the system or aid understanding of potential bias of system reports. We describe the operation of a farmer syndromic surveillance system in Victoria, Australia, over its first two years from 2014 to 2016. Survival analysis and classification and regression tree analysis were used to identify farm level factors associated with 'reliable' participation (low non-response rates in longitudinal reporting). Response rate and timeliness were not associated with whether farmers had disease to report, or with different months of the year. Farmers keeping only sheep were the most reliable and timely respondents. Farmers < 43 years of age had lower response rates than older farmers. Farmers with veterinary qualifications and those working full-time on-farm provided less timely reports than other educational backgrounds and farmers who worked part-time on-farm. These analyses provide a starting point to guide recruitment of participants for surveillance of farmers' observations using syndromic surveillance, and provide examples of strengths and weaknesses of syndromic surveillance systems for extensively-managed livestock. Once farm characteristics associated with reliable participation are known, they can be incorporated into surveillance system design in accordance with the objectives of the system.

An aberration detection-based approach for sentinel syndromic surveillance of COVID-19 and other novel influenza-like illnesses

Coronavirus Disease 2019 has emerged as a significant global concern, triggering harsh public health restrictions in a successful bid to curb its exponential growth. As discussion shifts towards relaxation of these restrictions, there is significant concern of second-wave resurgence. The key to managing these outbreaks is early detection and intervention, and yet there is a significant lag time associated with usage of laboratory confirmed cases for surveillance purposes. To address this, syndromic surveillance can be considered to provide a timelier alternative for first-line screening. Existing syndromic surveillance solutions are however typically focused around a known disease and have limited capability to distinguish between outbreaks of individual diseases sharing similar syndromes. This poses a challenge for surveillance of COVID-19 as its active periods tend to overlap temporally with other influenza-like illnesses. In this study we explore performing sentinel syndromic surveillance for COVID-19 and other influenza-like illnesses using a deep learning-based approach. Our methods are based on aberration detection utilizing autoencoders that leverages symptom prevalence distributions to distinguish outbreaks of two ongoing diseases that share similar syndromes, even if they occur concurrently. We first demonstrate that this approach works for detection of outbreaks of influenza, which has known temporal boundaries. We then demonstrate that the autoencoder can be trained to not alert on known and well-managed influenza-like illnesses such as the common cold and influenza. Finally, we applied our approach to 2019-2020 data in the context of a COVID-19 syndromic surveillance task to demonstrate how implementation of such a system could have provided early warning of an outbreak of a novel influenza-like illness that did not match the symptom prevalence profile of influenza and other known influenza-like illnesses.

Emergency department syndromic surveillance systems: a systematic review

BACKGROUND

Syndromic surveillance provides public health intelligence to aid in early warning and monitoring of public health impacts (e.g. seasonal influenza), or reassurance when an impact has not occurred. Using information collected during routine patient care, syndromic surveillance can be based on signs/symptoms/preliminary diagnoses. This approach makes syndromic surveillance much timelier than surveillance requiring laboratory confirmed diagnoses. The provision of healthcare services and patient access to them varies globally. However, emergency departments (EDs) exist worldwide, providing unscheduled urgent care to people in acute need. This provision of care makes ED syndromic surveillance (EDSyS) a potentially valuable tool for public health surveillance internationally. The objective of this study was to identify and describe the key characteristics of EDSyS systems that have been established and used globally.

METHODS

We systematically reviewed studies published in peer review journals and presented at International Society of Infectious Disease Surveillance conferences (up to and including 2017) to identify EDSyS systems which have been created and used for public health purposes. Search criteria developed to identify "emergency department" and "syndromic surveillance" were applied to NICE healthcare, Global Health and Scopus databases.

RESULTS

In total, 559 studies were identified as eligible for inclusion in the review, comprising 136 journal articles and 423 conference abstracts/papers. From these studies we identified 115 EDSyS systems in 15 different countries/territories across North America, Europe, Asia and Australasia. Systems ranged from local surveillance based on a single ED, to comprehensive national systems. National EDSyS systems were identified in 8 countries/territories: 2 reported inclusion of ≥85% of ED visits nationally (France and Taiwan).

CONCLUSIONS

EDSyS provides a valuable tool for the identification and monitoring of trends in severe illness. Technological advances, particularly in the emergency care patient record, have enabled the evolution of EDSyS over time. EDSyS reporting has become closer to 'real-time', with automated, secure electronic extraction and analysis possible on a daily, or more frequent basis. The dissemination of methods employed and evidence of successful application to public health practice should be encouraged to support learning from best practice, enabling future improvement, harmonisation and collaboration between systems in future.

PROSPERO NUMBER

CRD42017069150 .

Investigating novel approaches to tick-borne encephalitis surveillance in Sweden, 2010-2017

Tick-borne encephalitis (TBE) is a vaccine-preventable, high-priority disease in Sweden, with increasing incidence. However, surveillance is limited to case reports. We investigated relationships between reported TBE incidence and syndromic surveillance data to determine if these novel data sources could provide earlier indications of disease activity. We retrospectively compared national, weekly (2010-2017) reported TBE incidence to the percentage of TBE-related a) searches on the main Swedish healthcare information website and b) calls to its telehealth service using Spearman's ρ to determine the most strongly correlated lags. We conducted a sub-analysis (2012-2017) of TBE-related Google Trends queries and compared the number of TBE-related media stories to each novel surveillance dataset. Healthcare website searches for "tbe" and "vaccine" combined, "tbe", "tick", and "tick bite" led case data by 12, 8, 7, and 6 weeks, respectively (ρ = 0.87-0.89); telehealth calls led by 4 weeks (ρ = 0.92; all p < 0.001). Correlations and lags for Google Trends and healthcare website searches were fairly similar to each other. In comparison, correlation between the different syndromic surveillance datasets and the number of media stories was lower (ρ = 0.25-0.56). We observed volume discrepancies between TBE incidence and the novel surveillance datasets during some years, particularly for web searches. Syndromic surveillance data were strongly correlated with and preceded case data by 4-12 weeks. Syndromic data may provide advanced awareness and earlier indications of TBE activity, which can improve timing and specificity of public health communications. The use of these data as supplements to notifiable disease data for national planning and preparedness in real-time should be investigated.

Increase of scabies infestations, Norway, 2006 to 2018

Between October and December 2018, several clinicians in Norway reported an increase in scabies diagnoses. We compared data from the Norwegian Syndromic Surveillance System on medical consultations for mite infestations with scabies treatment sales data to investigate this reported increase. From 2013 to 2018, consultations and sales of scabies treatments had almost increased by threefold, particularly affecting young adults 15–29 years. We recommend to increase awareness among clinicians to ensure timely diagnosis and treatment.

A scoping review of the use of Twitter for public health research

Public health practitioners and researchers have used traditional medical databases to study and understand public health for a long time. Recently, social media data, particularly Twitter, has seen some use for public health purposes. Every large technological development in history has had an impact on the behaviour of society. The advent of the internet and social media is no different. Social media creates public streams of communication, and scientists are starting to understand that such data can provide some level of access into the people's opinions and situations. As such, this paper aims to review and synthesize the literature on Twitter applications for public health, highlighting current research and products in practice. A scoping review methodology was employed and four leading health, computer science and cross-disciplinary databases were searched. A total of 755 articles were retreived, 92 of which met the criteria for review. From the reviewed literature, six domains for the application of Twitter to public health were identified: (i) Surveillance; (ii) Event Detection; (iii) Pharmacovigilance; (iv) Forecasting; (v) Disease Tracking; and (vi) Geographic Identification. From our review, we were able to obtain a clear picture of the use of Twitter for public health. We gained insights into interesting observations such as how the popularity of different domains changed with time, the diseases and conditions studied and the different approaches to understanding each disease, which algorithms and techniques were popular with each domain, and more.

Opportunities and Challenges for Developing Syndromic Surveillance Systems for the Detection of Social Epidemics

This commentary explores the potential and challenges of developing syndromic surveillance systems with the ability to more rapidly detect epidemics of addiction, poverty, housing instability, food insecurity, social isolation and other social determinants of health (SDoH). Epidemiologists tracking SDoH heavily rely on expensive government surveys released annually, delaying for months if not years the timely detection of social epidemics, defined as sudden, rapid or unexpected changes in social determinants of population health. Conversely, infectious disease syndromic surveillance is an effective early warning tool for epidemic diseases using various types of non-traditional epidemiological data from emergency room chief complaints to search query data. Based on such experience, novel social syndromic surveillance systems for early detection of social epidemics with health implications are not only possible but necessary. Challenges to their widespread implementation include incorporating disparate proprietary data sources and database integration. Significantly more resources are critically needed to address these barriers to allow for accessing, integrating and rapidly analyzing appropriate data streams to make syndromic surveillance for social determinants of health widely available to public health professionals.

Enhancing the monitoring of fallen stock at different hierarchical administrative levels: an illustration on dairy cattle from regions with distinct husbandry, demographical and climate traits

Background

The automated collection of non-specific data from livestock, combined with techniques for data mining and time series analyses, facilitates the development of animal health syndromic surveillance (AHSyS). An example of AHSyS approach relates to the monitoring of bovine fallen stock. In order to enhance part of the machinery of a complete syndromic surveillance system, the present work developed a novel approach for modelling in near real time multiple mortality patterns at different hierarchical administrative levels. To illustrate its functionality, this system was applied to mortality data in dairy cattle collected across two Spanish regions with distinct demographical, husbandry, and climate conditions.

Results

The process analyzed the patterns of weekly counts of fallen dairy cattle at different hierarchical administrative levels across two regions between Jan-2006 and Dec-2013 and predicted their respective expected counts between Jan-2014 and Jun- 2015. By comparing predicted to observed data, those counts of fallen dairy cattle that exceeded the upper limits of a conventional 95% predicted interval were identified as mortality peaks. This work proposes a dynamic system that combines hierarchical time series and autoregressive integrated moving average models (ARIMA). These ARIMA models also include trend and seasonality for describing profiles of weekly mortality and detecting aberrations at the region, province, and county levels (spatial aggregations). Software that fitted the model parameters was built using the R statistical packages.

Conclusions

The work builds a novel tool to monitor fallen stock data for different geographical aggregations and can serve as a means of generating early warning signals of a health problem. This approach can be adapted to other types of animal health data that share similar hierarchical structures.

Relationship between the number of pediatric patients with rotavirus and telephone triage for associated symptoms

BACKGROUND

Earlier syndromic surveillance may be effective in preventing the spread of infectious disease. However, there has been no research on syndromic surveillance for rotavirus. The study aimed to assess the relationship between the incidence of rotavirus infections and the number of telephone triages for associated symptoms in pediatric patients under 4 years old in Osaka prefecture, Japan.

METHODS

This was a retrospective observational study for which the study period was the 3 years between January 2015 and December 2017. We analyzed data on children under 4 years old who were triaged by telephone triage nurses using software. The primary endpoint was the number of rotavirus patients under 4 years triaged old per week. Using a linear regression model, we calculated the R square value of the regression model to assess the relationship between the number of patients with rotavirus and the number of telephone triages made for associated symptoms. Covariates in the linear regression model were the week number indicating seasonality and the weekly number of telephone triages related to rotavirus symptoms such as stomachache and vomiting.

RESULTS

During the study period, there were 102,336 patients with rotavirus, and the number of people triaged by telephone was 123,720. The highest correlation coefficient was 0.921 in the regression model with the number of telephone triages for "stomachache + nausea/vomiting" and "stomachache + diarrhea + nausea/vomiting".

CONCLUSION

The number of telephone triage symptoms was positively related to the incidence of pediatric patients with rotavirus in a large metropolitan area of Japan.

Syndromic surveillance using ambulance transfer data in Tokyo, Japan

Bioterrorism attacks become more probable when important high-profile international or political events are held, such as G7 summit meetings or mass gathering events including Olympic and Paralympic games and FIFA World Cup tournaments. Outbreaks of infectious disease and widespread incidents of food poisoning are also public health concerns at such times. In Japan, the Tokyo Metropolitan Government operates Ambulance Transfer Syndromic Surveillance (ATSS), which can help monitor such incidents. The present study presents and assesses the ATSS framework. During the study period of October 2017 through November 2018, we monitored 33 areas for symptoms of 9 categories: vomiting/nausea, dizziness, palpitation, unconsciousness, breathing disorder, fever, spasm/paralysis, collapse/weakness, and bloody emesis/nasal hemorrhage. Among all symptoms, we found 9929 low-level aberrations, 2537 medium-level aberrations, and 577 high-level aberrations, with respective frequencies of 9.2%, 2.3%, and 0.5%. Of those, Tokyo Metropolitan Institute of Public Health reported the information to Tokyo Metropolitan Government 28 times during the period. Of the 28 identified clusters, Tokyo Metropolitan Government judged the necessity for investigating 7. All of those were investigated at hospitals by the jurisdictional public health center. Because ATSS covers almost the entire Tokyo metropolitan area, with about 13.8 million residents, it is definitely the largest syndromic surveillance in the world.

Multivariate syndromic surveillance for cattle diseases: Epidemic simulation and algorithm performance evaluation

Multivariate Syndromic Surveillance (SyS) systems that simultaneously assess and combine information from different data sources are especially useful for strengthening surveillance systems for early detection of infectious disease epidemics. Despite the strong motivation for implementing multivariate SyS and there being numerous methods reported, the number of operational multivariate SyS systems in veterinary medicine is still very small. One possible reason is that assessing the performance of such surveillance systems remains challenging because field epidemic data are often unavailable. The objective of this study is to demonstrate a practical multivariate event detection method (directionally sensitive multivariate control charts) that can be easily applied in livestock disease SyS, using syndrome time series data from the Swiss cattle population as an example. We present a standardized method for simulating multivariate epidemics of different diseases using four diseases as examples: Bovine Virus Diarrhea (BVD), Infectious Bovine Rhinotracheitis (IBR), Bluetongue virus (BTV) and Schmallenberg virus (SV). Two directional multivariate control chart algorithms, Multivariate Exponentially Weighted Moving Average (MEWMA) and Multivariate Cumulative Sum (MCUSUM) were compared. The two algorithms were evaluated using 12 syndrome time series extracted from two Swiss national databases. The two algorithms were able to detect all simulated epidemics around 4.5 months after the start of the epidemic, with a specificity of 95%. However, the results varied depending on the algorithm and the disease. The MEWMA algorithm always detected epidemics earlier than the MCUSUM, and epidemics of IBR and SV were detected earlier than epidemics of BVD and BTV. Our results show that the two directional multivariate control charts are promising methods for combining information from multiple time series for early detection of subtle changes in time series from a population without producing an unreasonable amount of false alarms. The approach that we used for simulating multivariate epidemics is relatively easy to implement and could be used in other situations where real epidemic data are unavailable. We believe that our study results can support the implementation and assessment of multivariate SyS systems in animal health.

Novel use of syndromic surveillance to monitor the impact of synthetic cannabinoid control measures on morbidity

Background

Using data from syndromic surveillance, the New York City Department of Health and Mental Hygiene (DOHMH) identified an increase in the number of emergency department (ED) visits related to synthetic cannabinoids. Syndromic surveillance data were used to target community-level interventions and assess the real-time impact of control measures in reducing synthetic cannabinoid (“K2”)-related morbidity.

Methods

From April 2015 through September 2015, DOHMH implemented 3 separate interventions to reduce K2-related morbidity by limiting the availability of K2 products. Difference-in-difference analyses compared pre- and post-intervention differences in cannabinoid-related ED visit rates between neighborhoods and controls for Interventions A and B. City-wide count data were used to compare K2-related ED visits before and after Intervention C.

Results

Syndromic data showed a reduction in K2-related ED visits following the 3 interventions. Respective decreases in rates of synthetic cannabinoid-related ED visits of 33 and 38% were detected at the neighborhood-level due to Interventions A and B, respectively. A decrease of 29% was calculated at the city level following Intervention C.

Conclusions

In addition to identifying emerging public health concerns, syndromic data can provide valuable real-time evidence on the effectiveness of public health interventions.

Electronic supplementary material

The online version of this article (10.1186/s40621-019-0210-2) contains supplementary material, which is available to authorized users.

Automatic classification of free-text medical causes from death certificates for reactive mortality surveillance in France

BACKGROUND

Mortality surveillance is of fundamental importance to public health surveillance. The real-time recording of death certificates, thanks to Electronic Death Registration System (EDRS), provides valuable data for reactive mortality surveillance based on medical causes of death in free-text format. Reactive mortality surveillance is based on the monitoring of mortality syndromic groups (MSGs). An MSG is a cluster of medical causes of death (pathologies, syndromes or symptoms) that meets the objectives of early detection and impact assessment of public health events. The aim of this study is to implement and measure the performance of a rule-based method and two supervised models for automatic free-text cause of death classification from death certificates in order to implement them for routine surveillance.

METHOD

A rule-based method was implemented using four processing steps: standardization rules, splitting causes of death using delimiters, spelling corrections and dictionary projection. A supervised machine learning method using a linear Support Vector Machine (SVM) classifier was also implemented. Two models were produced using different features (SVM1 based solely on surface features and SVM2 combining surface features and MSGs classified by the rule-based method as feature vectors). The evaluation was conducted using an annotated subset of electronic death certificates received between 2012 and 2016. Classification performance was evaluated on seven MSGs (Influenza, Low respiratory diseases, Asphyxia/abnormal respiration, Acute respiratory disease, Sepsis, Chronic digestive diseases, and Chronic endocrine diseases).

RESULTS

The rule-based method and the SVM2 model displayed a high performance with F-measures over 0.94 for all MSGs. Precision and recall were slightly higher for the rule-based method and the SVM2 model. An error-analysis shows that errors were not specific to an MSG.

CONCLUSION

The high performance of the rule-based method and SVM2 model will allow us to set-up a reactive mortality surveillance system based on free-text death certificates. This surveillance will be an added-value for public health decision making.

Machine learning to refine decision making within a syndromic surveillance service

Background

Worldwide, syndromic surveillance is increasingly used for improved and timely situational awareness and early identification of public health threats. Syndromic data streams are fed into detection algorithms, which produce statistical alarms highlighting potential activity of public health importance. All alarms must be assessed to confirm whether they are of public health importance. In England, approximately 100 alarms are generated daily and, although their analysis is formalised through a risk assessment process, the process requires notable time, training, and maintenance of an expertise base to determine which alarms are of public health importance. The process is made more complicated by the observation that only 0.1% of statistical alarms are deemed to be of public health importance. Therefore, the aims of this study were to evaluate machine learning as a tool for computer-assisted human decision-making when assessing statistical alarms.

Methods

A record of the risk assessment process was obtained from Public Health England for all 67,505 statistical alarms between August 2013 and October 2015. This record contained information on the characteristics of the alarm (e.g. size, location). We used three Bayesian classifiers- naïve Bayes, tree-augmented naïve Bayes and Multinets - to examine the risk assessment record in England with respect to the final ‘Decision’ outcome made by an epidemiologist of ‘Alert’, ‘Monitor’ or ‘No-action’. Two further classifications based upon tree-augmented naïve Bayes and Multinets were implemented to account for the predominance of ‘No-action’ outcomes.

Results

The attributes of each individual risk assessment were linked to the final decision made by an epidemiologist, providing confidence in the current process. The naïve Bayesian classifier performed best, correctly classifying 51.5% of ‘Alert’ outcomes. If the ‘Alert’ and ‘Monitor’ actions are combined then performance increases to 82.6% correctly classified. We demonstrate how a decision support system based upon a naïve Bayes classifier could be operationalised within an operational syndromic surveillance system.

Conclusions

Within syndromic surveillance systems, machine learning techniques have the potential to make risk assessment following statistical alarms more automated, robust, and rigorous. However, our results also highlight the importance of specialist human input to the process.