Infectious Disease Prevention and Control Using an Integrated Health Big Data System in China | BMC Infectious Diseases
The advent of big data platforms and advances in machine learning algorithms have enabled researchers to learn from vast amounts of information, which has the potential to create methods for monitoring and controlling infectious diseases. more and more efficient. . The term bigdata refers to complex datasets that are too large and sophisticated to be processed by traditional analytics [2, 3] and public health agencies increasingly rely on big data to improve screening for infectious diseases. There is a wide variety of data structures relevant to public health issues, including electronic health records (EHRs), national and local clinical databases, social media, and public health agency reports. such as the Centers for Disease Control and Prevention (CDC) . The ease of large-scale storage, manipulation and analysis of these wide-ranging types of data has the potential to empower healthcare organizations and public health officials to prevent the spread of infectious diseases and to respond and manage outbreaks in a timely manner. .
While researchers have recognized the promise of big data in improving infectious disease surveillance and control, the effectiveness of big data in preparing public health officials at state and national levels to respond to the epidemic of infectious diseases largely depends on the existence of successful partnerships between different stakeholders, such as government health agencies, infectious disease control departments, and private and public hospitals and clinics. A concerted effort to effectively store, transfer, manipulate and analyze data for infectious disease control requires the support of IT and organizational infrastructures and the streamlining of processes between different stakeholders.
The overall objective of this study is to demonstrate how building an integrated health and big data infrastructure and platform at the district level can significantly improve the ability of public health officials to leverage knowledge big data for the prevention and control of infectious diseases. To achieve this goal, we compared the effectiveness of a new tool with an integrated big data platform located in Yinzhou, a major district of Ningbo city in Zhejiang province in China. , with a traditional method of infectious disease surveillance – which relies on laboratory and health facility reports (e.g. doctors’ diagnosis after face-to-face consultation) – in identifying cases for control and prevention of infectious diseases. For this study, we focused on three types of infectious disease categories: (a) acute infectious diseases (i.e. dengue), (b) chronic infectious diseases (i.e. say pulmonary tuberculosis (TB)) and (c) immunization gaps. adoption among migrant children.
Challenges of traditional infectious disease surveillance methodologies
Prior to the introduction of big data, public health officials traditionally dealt with infectious disease surveillance and control in three ways. The first was the use of doctors to report suspected cases. These reports would then be compiled and sent to a centralized infectious disease agency (e.g., the CDC) to coordinate control and prevention measures. Second, public health officials have relied on predictive models drawn from sources such as weather and vector surveillance data. [7, 8]or in recent years, social media sources . Third, public health officials would rely on vaccination records to examine vaccination uptake. This would allow government and public health agencies to target specific populations that are still unvaccinated and therefore most at risk for outbreaks.
However, these traditional approaches have several limitations. First, there is a high probability of human error in the diagnosis of infectious diseases, even if the diagnosis was made by highly qualified doctors. For example, research has shown that a significant portion of TB patients received non-TB respiratory diagnoses in hospitals . In low-resource health care settings, doctors can sometimes misdiagnose dengue patients and treat them as having a common upper respiratory infection. The inaccuracy of traditional diagnostic methods would be magnified if hospitals and clinics were overwhelmed by a sudden increase in cases.
Second, although predictive modeling using various data sources can potentially help public health organizations prevent the spread of infectious diseases, these models have several limitations that impede their effectiveness. There are modeling issues associated with data deluge and hubris causing statistical models to over or under estimate . On the other hand, in rural areas that do not have access to high-speed Internet, there may be problems related to the lack of quality data. where data from underserved communities is not represented, severely compromising the quality of forecasting models.
Third, although identifying immunization gaps using local immunization records is a powerful step in prevention, the effectiveness of this method is diluted if there is an influx of migrants, who could be carriers of infectious diseases if they are not vaccinated. In the Yinzhou context, immunization records do not capture immunization rates for migrant children who are not listed as children in a local urban area. hukou— a household registration system — in China, making it difficult to track infectious diseases from other geographic regions. As such, primary care providers at local health facilities are tasked with screening migrant children by going door to door to identify those who may not have completed all basic vaccinations. which leads to huge financial and labor costs for medical institutions.
Overcoming the Challenges – Implementing an Infectious Disease Big Data Platform in Yinzhou
To circumvent the weaknesses of traditional infectious disease prevention and control methods, Yinzhou recently set up a synchronized district-wide health big data platform to coordinate data collection from healthcare providers. premises, such as clinics and hospitals, in the private and public sectors. The big data platform was developed by CDC-mandated Wonders Information Co Ltd to coordinate data sharing between a network of clinics and hospitals for CDC public health surveillance purposes. . The big data system consolidates different sources of data such as (a) daily clinical data produced by local health providers, (b) EHRs, (c) data from annual medical examinations, (d) data related to public health such as vaccination records, chronic and infectious diseases. disease notification data and mortality surveillance data, and (e) other types of data such as student register data (see Fig. 1). Individuals’ data has been linked and connected using their unique identification number and new data entries (eg visits to medical facilities) are uploaded into the big data system daily. The system performs daily scans to detect potential cases of infectious diseases among the population and notifies relevant public health agencies and authorities for further action.
For this study, we illustrate the accuracy and performance of the integrated big data system in identifying (a) dengue fever cases, (b) pulmonary tuberculosis among university students and (c) migrant children without vaccination records. comprehensive, compared to traditional monitoring methodologies. . While the integrated big data system contains information on all infectious diseases, we specifically chose to focus only on dengue fever and pulmonary tuberculosis, as these are the top public health priorities in Zhejiang province. Located on the east coast of China, Zhejiang province has a humid subtropical climate, with heavy rainfall in spring and summer, with an abundance of Aedes albopictus . Additionally, the high human population density, as well as the metropolitan nature of the city with a large influx of travelers, the province is susceptible to individuals who import dengue fever from dengue-endemic regions, which have been the cause of some dengue epidemics in recent years . In addition, in the Chinese context, pulmonary tuberculosis is a cause for concern because outbreaks often occur in educational institutions where there are overcrowded dormitories or close proximity and contact between students in classrooms. – in 2018, 48,289 students were reported to have had pulmonary tuberculosis, which was an incidence of 17.97/100,000 .