Reassessing Google Flu Trends data for detection of seasonal and pandemic influenza: a comparative epidemiological study at three geographic scales. PLoS Comput Biol ; 9 : e Salathe M , Khandelwal S. Assessing vaccination sentiments with online social media: implications for infectious disease dynamics and control. PLoS Comput Biol ; 7 : e A case study of the New York City — influenza season with daily geocoded Twitter data from temporal and spatiotemporal perspectives. J Medical Internet Res ; 16 : e Wikipedia usage estimates prevalence of influenza-like illness in the United States in near real-time.
PLoS Comput Biol ; 10 : e Forecasting the — influenza season using Wikipedia. Incidence and risk factors for influenza-like-illness in the UK: online surveillance using Flusurvey. BMC Infect Dis ; 14 : Ten-year performance of Influenzanet: ILI time series, risks, vaccine effects, and care-seeking behaviour.
Epidemics ; 13 : 28 — Flu near you: crowdsourced symptom reporting spanning 2 influenza seasons. Am J Public Health ; : — Harvard School of Public Health. HealthMap , GPHIN , A smartphone dongle for diagnosis of infectious diseases at the point of care. Sci Transl Med ; 7 : re1. UK National Statistics. Weekly National Flu Reports. Severe respiratory disease concurrent with the circulation of H1N1 influenza. Euro Surveill ; 14 : pii PLoS Curr ; 6.
Oprah book club , PLoS One ; 6 : e Big data. The parable of Google Flu: traps in big data analysis. Science ; : — 5. The signature features of influenza pandemics--implications for policy. N Engl J Med ; : — 8. Evaluation of Internet-based dengue query data: Google Dengue Trends. Digital epidemiology reveals global childhood disease seasonality and the effects of immunization.
Spatial transmission of pandemic influenza in the US. Forecast and control of epidemics in a globalized world. Mitigation strategies for pandemic influenza in the United States. Surveillance for the pandemic influenza A H1N1 virus and seasonal influenza viruses - New Zealand, Real-time influenza forecasts during the — season.
Nat Commun ; 4 : Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Search Menu. Article Navigation.
Close mobile search navigation Article Navigation. Volume Article Contents Abstract. Lone Simonsen , Lone Simonsen. Oxford Academic. Julia R. Don Olson. Select Format Select format. Permissions Icon Permissions. Abstract While big data have proven immensely useful in fields such as marketing and earth sciences, public health is still relying on more traditional surveillance systems and awaiting the fruits of a big data revolution. Disease s. Institution s. Open in new tab.
In the future, the many types of data sources described above will continue to be developed and used: traditional approaches laboratory-based and sentinel surveillance, as well as death certificates , electronic health data medical insurance claims from outpatient visits , and online nontraditional sources social media, search queries, and massive participatory surveys.
The relative merits of different types of source can be thought of in terms of 1 the specificity of the disease signal and 2 the volume and granularity of coverage achieved. A diagram of the relative placing of these for some of the discussed sources is given in Figure 1. Figure 1. Open in new tab Download slide. In contrast, with population-based surveillance, every appropriate health facility reports on the predefined diseases with the goal of identifying all cases in a specific geographic area.
Population-based surveillance can either represent the whole country national or a defined subnational population area. Since the population is defined, these surveillance sites can produce rates of disease for example, incidence and mortality rates , which allows for comparison of rates of disease between other population-based surveillance sites.
Population-based surveillance is more costly than sentinel site surveillance, but produces more generalizable data on incidence of disease. Aggregate surveillance data can exist in a variety of forms, but the main feature is that it lacks detailed information on specific cases. Aggregate data typically include the number of cases for example, number of suspect and confirmed neonatal tetanus cases, or by age group for a specific region and time period.
This information can monitor the number of cases but lacks the individual-level data required for specific analyses. An example of this is the Integrated Disease Surveillance and Response IDSR system which asks clinicians to report the number of cases of specific diseases. Case-based surveillance refers to surveillance systems that collect information about each case at the individual level. This type of surveillance system has a case investigation form where information can be gathered from the patient or their family members, their medical records, and their laboratory records.
At a minimum, more detailed information on person who is infected , place where they live, where they might have been infected , and time when they became ill is collected. A line list from this investigation form is created and reported up their normal reporting channels. In some scenarios, a case-based surveillance system might transition to aggregate as the number of cases becomes large as it overwhelms the system, like what happened during the H1N1 outbreak.
In contrast, an aggregate surveillance system might become case-based temporarily in an outbreak to understand more of the epidemiology of the disease. Certain diseases, such as polio and measles, are recommended to be case-based. Measles surveillance has seen a movement away from aggregate and toward case-based surveillance WHO, Initially, when the United Nations UN development goals were established in , measles was endemic in many countries, and mortality reduction was the primary goal.
Given this, aggregate data were the most feasible approach and were conducted in most countries. By , all six WHO regions have measles elimination goals. As measles has moved away from control and toward elimination, case-based surveillance is needed to ensure every case is reported and investigated. When disease was endemic, case-based surveillance would quickly be overwhelmed given the time and resources, but as countries have fewer and fewer cases, it is relatively easier to conduct an investigation on every single case.
WHO recommends the type of data to be collected in an investigation. One key advantage of case-based surveillance is that it allows one to analyze which age cohorts are being infected and their individual vaccination status to help to target vaccination efforts and close existing immunity gaps. Surveillance networks identify and enroll cases that meet a specific case definition. Case definitions have three essential components: person, place, and time.
Case definitions vary in sensitivity and specificity. Sensitive case definitions are more inclusive, are less likely to miss cases, but will include patients that do not have the disease. Specific case definitions have stricter criteria and exclude more patients that do not have the disease but can also miss patients with milder or atypical disease presentations. Both sensitive and specific case definitions can be used in infectious disease surveillance depending on the goals of surveillance.
For example, sensitive case definitions may be preferred if it is important not to miss cases. In general, case definitions should be as sensitive and specific as possible. However, since a highly sensitive and specific case definition is not always possible, it is important that the case definition is at least applied systematically and consistently over the surveillance period.
Syndromic surveillance involves monitoring cases that meet a clinical case definition for the disease under surveillance, typically without laboratory confirmation Henning, This allows for rapid identification of a cluster of cases that might warrant further investigation. As field investigations are ongoing, laboratory testing can be performed on some or all of the cases identified by syndromic surveillance to determine the etiology.
A well-established global WHO-coordinated measles laboratory network provides support to monitoring measles cases and provide genotype information globally. Syndromic surveillance case definitions can be used in emergency or outbreak situations as an alert system to identify suspect cases that meet a broad case definition to then be further investigated. During the Ebola outbreak in —15, airport security was increased to identify people with a fever and a history of travel to an Ebola-affected country in order to stop the disease from traveling between countries.
In contrast, some surveillance case definitions are based on confirmed cases in a laboratory where the etiologic agent can be identified through a variety of laboratory tests e. As an example, virologic influenza surveillance networks use laboratory-confirmed influenza to determine the circulating strains to provide information for vaccine composition.
A critical objective of laboratory-based surveillance is to monitor for emerging drug resistance in pathogens or shifts in serotype distribution. Cases meeting a suspect case definition a sensitive case definition may undergo laboratory testing leading to a more specific case definition. For example, the case definition for suspect meningitis as part of the WHO Invasive Bacterial Vaccine-Preventable Disease network is very sensitive: a hospitalized patient at a surveillance hospital with sudden onset of fever and at least one meningeal sign during the surveillance period.
After being enrolled into surveillance, additional clinical and laboratory information can reclassify a patient as having probable bacterial meningitis namely having abnormal white cell count, protein or glucose levels in cerebrospinal fluid.
This definition has a greater specificity but lower sensitivity. The most specific meningitis definition is confirmed meningitis by polymerase chain reaction assay or other laboratory test. This definition may lose some sensitivity because confirmatory tests can have false negatives, especially in areas with high antibiotic usage.
Zoonotic diseases cause disease in humans and can be challenging to control since both animals and humans can be hosts.
Many zoonotic diseases of public health importance are covered in other articles of this encyclopedia, including West Nile Virus, avian influenza, Ebola and other hemorrhagic fevers , Lyme disease, SARS, Nipah virus, and rabies. Historically, zoonotic and human disease surveillance existed separately, but there is a push to harmonize these systems to improve surveillance for diseases affecting both populations.
Illness in one species might be a harbinger of illness in humans, and an integrated comprehensive surveillance system can help identify potential disease transmission that might be ongoing. For example, surveillance for Borrelia burgdorferi , the causative agent of Lyme disease, in the tick population can help public health authorities determine proper interventions to decrease the transmission from ticks to humans. One Health emphasizes the link of human health to the surrounding environment and animals.
One of the mission statements of One Health is to improve the lives of all species by harmonizing both animal and human disease surveillance and control efforts. Serosurveillance involves the use of blood specimens to determine the burden of disease or immunity gaps in a population. Serosurveillance is frequently done as a periodic survey for multiple diseases of interest simultaneously.
However, serosurveillance cannot provide information in a timely manner; thus an outbreak might have occurred that is discovered by serosurveillance, but it might be potentially too late for an intervention to decrease disease transmission. Serosurveillance is sometimes the only type of surveillance conducted for an infectious disease.
For example, hepatitis B is frequently asymptomatic in children, making evaluating the impact of vaccination efforts extremely challenging WHO, The standard has become to perform serosurveillance among cohorts of vaccinated children to identify the burden of disease and determine the impact of vaccination efforts.
In some countries, national health surveys, such as the National Health and Nutrition Examination Survey NHANES and Malaria Indicator Surveys, are conducted periodically and include a serologic component, allowing one to monitor trends in diseases and immunity over time. Adverse events following immunization AEFI surveillance is a critical component of ensuring vaccine safety in the populations where the vaccines are being used.
Surveillance often begins at the health facility level, where health workers are trained to recognize adverse events from immunizations, and is reported to national regulatory agencies and WHO. This surveillance is critical for investigating problems that could occur with bad lots of vaccines and mishandling of vaccines in the cold chain improper storage which can contribute to the public perception of the vaccine program.
Technology is increasing the availability of data on health that can be used for infectious disease surveillance, including sources that go beyond that of traditional passive or active surveillance systems. New sources of data include mobile data, electronic health records, and social media. It is to the work of the Eastern Bureau that we now turn. While the early work of the League of Nations Health Organisation was directed towards the emergency situation in Eastern Europe — a situation that had begun to wane by — the Epidemiological Intelligence Service continued to collect and publish data that provided a world view of epidemic diseases of international importance.
Under the initial directorship of Gilbert E. Brooke Figure 2. The logic behind this fourfold division was that maritime communications within each group were mainly self-contained, with connections between the groups being chiefly the concern of larger ports.
Source : based on League of Nations Health Section , unnumbered figure, p. As the upper boxes in the diagram show, epidemiological information reached the Bureau through three main channels cf. Figure 3. A immediate notification by telegram.
On the basis of the epidemiological information received, the Bureau determined which countries were in direct communication with the infected ports and sent a summary of the situation by telegram to their administrations;. B weekly notification by telegram. All countries were required to submit a weekly telegram, to reach the Bureau by no later than Wednesday midday , summarising the epidemiological situation in the important ports and other territories in the seven-day period up to the preceding Saturday midnight ;.
C letter by first available post. The weekly telegram in B was to be confirmed by a letter, to be sent to the Bureau by the first available post, that contained supplementary information on the epidemiological status of the ports and other territories.
The information so received was to be used by the administrations to inform decisions on how to prevent entry of a given disease by sea and, in later years, by air. As noted in Section 2.
The Convention came into force in and, from thereon, the Bureau p. Routine telegraphic communications, however, were a significant financial liability. Through these facilities, wireless broadcasting of the Weekly Bulletin began in early April The first general return was despatched by the Bureau to Saigon on Thursday April 2nd, in clear [uncoded], and was broadcasted by their powerful wireless installation and has continued to transmit the weekly message in code, ever since, each Friday at G.
Most unfortunately the atmospherics of the East would appear to be indifferently good, and the messages can only be picked up by Japan, British North Borneo, and Bandong.
The matter is being considered by the Bureau with a view to possible extension of the wireless messages. Were a large number of administrations of the Arena able to get the broadcast, much expense would be saved by the Bureau. Throughout , improvements in the wireless reception of the Weekly Bulletin were pursued, with increasing numbers of stations being able to pick up the Saigon and Malabar broadcasts.
By , the Director of the Eastern Bureau could note that:. The development of our wireless service has been satisfactory. Its network has become denser and has greatly extended through the rebroadcast by new stations of our weekly wireless bulletin. We consider it a sound policy to use powerful stations with wide range of action for the broadcast of our bulletin in code, and to leave to stations of less range the broadcast of the bulletin in clear intended for shipping at sea.
A total of 30 health administrations in the Eastern Arena were able to pick up the weekly wireless transmission in and, through the s, a fully functional broadcast system for the Weekly Bulletin became established Figure 2.
As hostilities in Europe and Asia intensified, however, there was a decline in the regularity and quality of information being remitted to the Bureau, while wireless broadcasts became impossible. Source : redrawn from Eastern Bureau of the League of Nations Health Organisation , unnumbered figure, between [link] — [link]. In preparation for the establishment of the permanent WHO, p. Article Each Member [State] shall communicate promptly to the Organization important laws, regulations, official reports and statistics pertaining to health which have been published by the State concerned ;.
Article Each Member shall provide statistical and epidemiological reports in a manner to be determined by the Health Assembly ; and. Article Each Member shall transmit upon the request of the Board such additional information pertaining to health as may be practicable. This regional structure has continued through to the present day Figure 2.
Epidemiological Intelligence and Surveillance, s — s. The inheritance of the newly-created WHO included responsibility for international epidemic control in terms of: i quarantine and the International Sanitary Conventions; and ii epidemic intelligence and epidemiological services.
Further details are provided in Section 3. The Regulations continued to require national health administrations to notify the WHO by telegram or, in some instances, airmail of the appearance of quarantinable diseases in their territories.
Information received under the Regulations was collated by the four WHO quarantine and p. The radio bulletins, in turn, were confirmed via weekly publications. The Weekly Epidemiological Record formed the main WHO publication for this purpose, containing notifications of quarantinable diseases, information on international sanitary legislation and notes on the incidence of non-quarantinable diseases whenever their prevalence became of international importance.
Source : Goodman , Figure 36, [link]. Important developments included the establishment in of a WHO study group to advise on the collection of immunological information through immunological and haematological surveys. WHO serum reference banks were established and antibody studies of a range of diseases including poliomyelitis, measles, rubella, influenza, arbovirus infections, rickettsial infections, pertussis, typhoid and diphtheria, among other diseases were undertaken in the s in Africa, the Americas, Asia and Europe.
Alongside diseases for which surveillance activities were already in place including cholera, influenza, malaria, smallpox and tuberculosis , surveillance of a range of other diseases was initiated, including: dengue haemorrhagic fever started in , salmonellosis started in and, in cooperation with the Food and Agriculture Organization FAO and the World Organisation for Animal Health OIE , rabies.
The new approach had grown out of improved epidemio logical methods including data processing and analysis, laboratory and field studies and was defined as. The exercise of continuous scrutiny of and watchfulness over the distribution and spread of infections and factors related thereto, of sufficient accuracy and completeness to be pertinent to effective control.
World Health Organization, a , [link]. With this development, the Unit of Quarantine was merged with the Unit of Epidemiological Surveillance in and, by , the programme was referred to as Epidemiological Surveillance. This strategy was adopted as a means of moving away from the concept that disease prevention and control could only be achieved through the application of quarantine measures.
As a consequence of this development, the existing International Sanitary Regulations were revised and adopted as the International Health Regulations in Figure 2. Information collected by p. Additionally, the Epidemiological Surveillance programme placed emphasis on the strengthening of national surveillance; training of epidemiologists in Member States in surveillance methods was provided, along with the availability of technicians and allied health personnel to observe and report on outbreaks.
Technical manuals were prepared for the surveillance of four diseases under the international health regulations cholera, plague, smallpox and yellow fever and five other diseases influenza, louse-borne typhus fever, louse-borne relapsing fever, malaria and paralytic poliomyelitis , while a general guide on disease surveillance was prepared in This same period witnessed great strides in disease prevention and control through the eradication of smallpox see Section 5.
With the advent of computer technology, a computer was installed at the WHO HQ in the mids and computerisation of statistical work was undertaken. Computers, in turn, permitted information processing and the development of epidemiological models in relation to such diseases as typhoid, cholera, tetanus and diphtheria Uemura, What was the practical reality on the ground of all this international disease surveillance activity in the six decades from the formation of the League of Nations Health Organisation?
Which communicable diseases were recorded by which countries? To address these questions, Cliff, Smallman-Raynor, Haggett, et al.
While details of the data matrix are given by Cliff and colleagues, we note here that it is based on two categories of data:. For the period —58, Annual Epidemiological Report and its successor publication, Annual p.
Between and , the matrices of notifiable diseases were not published, but equivalent matrices of reported diseases by country can be constructed from the raw returns of recorded non-zero mortality and morbidity published in later issues of Annual Epidemiological and Vital Statistics and its successor publication, World Health Statistics Annual.
Source : League of Nations Health Organisation , [link] — [link] , detail. For 29 sample years —37, , —55, , , , , , and ; see Cliff, et al. The same interval was associated with a steady and progressive increase in the number of WHO Member States for which surveillance activities for the sample disease categories were documented, from 85 44 percent of WHO Member States at the start of the observation period to a peak of 83 percent of WHO Member States in the early s.
The time series of diseases in Figure 2. It is important to recognise that not all countries undertook surveillance for all the diseases, and Member States varied in terms of the number of disease categories under surveillance in any given year. To capture this variability, the heavy line trace in Figure 2. As the graph shows, the first four decades of the observation period were associated with a steady increase in the median number of disease categories, rising from 15 to 30 , but with a sharp reduction thereafter.
As judged by the inter-quartile range, many WHO Member States approximated this general surveillance pattern. For the remaining 70, Figure 2. As a rule, this latter development reflected the extension of ICD recording classes, graphed in Figure 2. Shigellosis A03 ; amoebiasis A06 ; tularemia A21 ; rat-bite fevers A25 ; rhinoscleroma A48 ; other spirochaetal infections A69 ; psittacosis A70 ; foot-and-mouth disease B08 ; coccidioidomycosis B38 ; other fluke infections B66 ; other helminthiases B Other zoonotic bacterial diseases, not elsewhere classified A28 ; infection due to other mycobacteria A31 ; Boutonneuse fever A77 ; tick-borne viral encephalitis A84 ; acute hepatitis B B16 ; other acute viral hepatitis B17 ; chronic viral hepatitis B18 ; other and unspecified infectious diseases B One important geographical question that can be asked of the data set relates to the regional specificity of disease surveillance activities.
More particularly, is it possible to identify distinctive regional concentrations of disease categories for which surveillance was undertaken and, if so, did these regional concentrations change over time? Disease 1. Source : Cliff , Smallman-Raynor, Haggett, et al.
The table identifies a series of distinctive marker diseases for several of the world regions. In Africa, for example, African trypanosomiasis s—80s , schistosomiasis s—60s and relapsing fever s—70s feature in the list in three or more decadal periods. Only in relatively rare instances, however, does a single disease category remain important throughout the entire observation period.
A further feature of Table 2. During the s and s, Cliff, Smallman-Raynor, Haggett, et al. From the s, the geographical focus of monitoring activities for newly-defined disease categories evolved to include Africa, Eastern Mediterranean and South-East Asia. This new programme placed emphasis on a target-oriented approach to information, with priority given to only the most essential information for the improvement of health systems Uemura, This approach, built around vaccine-controllable and potentially globally-eradicable diseases, had been presaged by the Smallpox Eradication Programme in the s and s and was followed by the Expanded Programme on Immunization from and the Global Polio Eradication Initiative from Here, we examine aspects of programme-oriented surveillance in relation to the global smallpox and poliomyelitis eradication campaigns.
Overviews of the eradication campaigns are provided in Sections 5. As described in Section 5. The eradication campaign started with mass vaccination, but rapidly recognised the importance of selective control. As Fenner, et al. The primary objective of the smallpox programme is the eradication of this disease. WHO Handbook, cited in Fenner, et al. Fully satisfactory networks of notification took 1—2 years to develop. Initially, reliance was placed on surveillance of those attending health units.
From September , however, the nature of both surveillance and containment began to change. At this stage, surveillance-containment operations and mass vaccination had halted transmission of smallpox in all but five countries Bangladesh, Ethiopia, India, Nepal and Pakistan. In the summer of , WHO devised an intensified system of case detection and later containment, applied first in India and, subsequently in Pakistan late , Bangladesh early , Ethiopia late and Somalia mid This intensified system rested with the more complete and prompt detection of outbreaks and involved supplementing the existing notification system by engaging health staff from other programmes in national village-by-village and house-to-house searches.
Echoing aspects of the local detective work undertaken by searchers employed in the compilation of the London Bills centuries earlier Section 2.
When a suspected smallpox case was found, the search worker notified the supervisor or nearest health unit so that containment staff could move in. Searches were accompanied by an intensive publicity campaign, and rewards were offered to the populace for information on cases. Special search programmes were implemented for areas that were difficult to access Fenner, et al.
Active search operations for smallpox in India are described by Basu, et al. The first village-to-village search for smallpox cases was undertaken in West Bengal in September and, in subsequent months, in the highly endemic states of Bihar, Madhya Pradesh and Uttar Pradesh.
The search operations were later extended beyond the four endemic states to include 11 low incidence states in November—December and 16 smallpox-free states from December Search operations continued until the last all-India search in October—November Figure 2.
Searches involved the questioning of households and people in prominent places e. All told, these operations involved p. The active search operations yielded evidence of 20, outbreaks and 77, smallpox cases in the entire search period to November ; the last detected smallpox case was discharged from hospital in July and India was declared free of smallpox by an International Commission in April Basu, et al.
Source : Basu, et al. Source : redrawn from Basu, et al. Methods of syndromic surveillance are frequently adopted where early detection of a disease event is considered a high priority. Syndromic surveillance focuses on the monitoring of health indicators that are available prior to the confirmation of diagnosis or laboratory confirmation of infection.
The information is in near-real time and is often available sooner than a laboratory test can be completed. Acute flaccid paralysis is the principal clinical syndrome of paralytic poliomyelitis and is characterised by paralysis, muscle flaccidity and sudden onset. Although AFP is a feature of all cases of paralytic poliomyelitis, the syndrome has a number of aetiologies. These range from other neurotropic viruses to trauma and chemical exposure. Surveillance for AFP is considered the gold standard for detecting cases of poliomyelitis and involves four steps: i finding and reporting children with AFP; ii transporting stool samples for analysis; iii isolating and identifying poliovirus in the laboratory; and iv mapping the virus to determine the origin of the virus strain.
This serves in the detection of typical and atypical cases of poliomyelitis due to both wild and vaccine-derived strains of poliovirus. Laboratory-based investigation of samples is coordinated through the Global Polio Laboratory Network, a global and interdependent network of laboratories that was formalised in —93 Figure 2.
AFP surveillance provides a basis for assessment of the quality of disease surveillance for certification purposes Table 2. Molecular epidemiologic methods have enhanced the precision and reliability of laboratory-based poliomyelitis surveillance, allowing wild viruses to be classified into genetic families from which inferences on the geographical source of isolates can be drawn Cochi, et al.
Principal performance indicators for acute flaccid paralysis AFP surveillance. The distribution of reporting sites should be representative of the geography and demography of the country. Adequate stool specimens are defined as: two specimens of sufficient quantity for laboratory analysis, collected at least 24 hours apart, within 14 days after the onset of paralysis, and arriving in the laboratory by reverse cold chain and with proper documentation.
Source : abstracted from World Health Organization b, [link] — [link]. The surveillance cycle is not complete until disease information is relayed to those who have responsibilities for the various public health and medical actions. Surveillance assessment reports should serve to inform and motivate. Infectious disease surveillance endeavors to:.
0コメント