Syndromic Surveillance

As of October 1, 2015, all HIPAA covered entities transition from the use of International Classification of Diseases version 9 (ICD-9-CM) to version 10 (ICD-10-CM/PCS). Many Public Health surveillance entities receive, interpret, analyze, and report ICD-9 encoded data, which will all be significantly impacted by the transition. Public health agencies will need to modify existing database structures, extraction rules, and messaging guides, as well as revise established syndromic surveillance definitions and underlying analytic and business rules to accommodate this transition. Implementation challenges include resource, funding, and time constraints for code translation and syndrome classification, and developing statistical methodologies to accommodate changes to coding practices.

To address these challenges, the International Society for Disease Surveillance (ISDS), in consultation with the Centers for Disease Control and Prevention (CDC) and the Council of State and Territorial Epidemiologists (CSTE), has conducted a project to develop consensus-driven syndrome definitions based on ICD- 10-CM codes. The goal was to have the newly created ICD-9-CM to-ICD-10-CM mappings and corresponding syndromic definitions fully reviewed and vetted by the syndromic surveillance community, which relies on these codes for routine surveillance, as well as for research purposes. The mappings may be leveraged by other federal, state, and local public health entities to better prepare and improve the surveillance, analytics, and reporting activities impacted by the ICD-10-CM transition.

Objective

To describe the process undertaken to translate syndromic surveillance syndromes and sub-syndromes consisting of ICD-9 CM diagnostic codes to syndromes and sub-syndromes consisting of ICD-10-CM codes, and how these translations can be used to improve syndromic surveillance practice.

In 2012, an estimated 2.5 million people presented to the ED for a MVC injury in the U.S. National injury surveillance is commonly captured using E-codes. However, use of E-codes alone to capture MVC-related ED visits may result in a different picture of MVC injuries compared to using text searches of triage or chief compliant notes.

Objective

Identify and describe how the case definition used to identify MVC patients can impact results when conducting MVC surveillance using ED data. We compare MVC patients identified using external cause of injury codes (E-codes), text searches of triage notes and chief complaint, or both criteria together.

Dengue fever is a dynamic infectious disease, allowing the patient to rapidly move from one stage to another during its course. Proper management of patients depends on early recognition of warning signs, continuous monitoring and re-staging cases and prompt fluid replacement. The telemedicine and Electronic Patient Records (EPR) belong to a series of advances of new features such as decision-making support systems including efforts on health monitoring, in view of the EPR as a support tool to allow the association of welfare activities as a database for the management of epidemiological information and monitoring. In addition, telemonitoring systems can be used for the monitoring of patients with chronic diseases in their homes which leads to cost savings in hospitalization and ensures appropriate care and the proper development of these patients. The continuous remote monitoring of these patients decreases the amount of hospital visits for monitoring procedures, also facilitating successful treatment, as in the fever dengue cases.

Objective

Report successful experience in fighting dengue fever in the Hospital and Emergency Services in São Bernardo do Campo, joining the flowchart included, telephone monitoring and Electronic Patient Records.

While results from syndromic surveillance systems are commonly presented in the literature, few systems appear to have been thoroughly evaluated to examine which events can and cannot be detected, the time to detection and the efficacy of different syndromic surveillance data streams. Such an evaluation framework is presented.

Objective

To devise a methodology for evaluating the effectiveness of syndromic surveillance systems

Flu Near You allows individuals to volunteer to be a sentinel node of the syndromic surveillance (SyS) network. The platform has the potential to provide insight into the spread of influenza-like illness (ILI). CDC’s ILINet is the gold standard for tracking ILI at the national level, but does not track into the local level. Local health departments (LHD) frequently express a need for granular data specific to their jurisdictions. FNY attempts to meet this need by collecting and sharing data at the zip code level. Knowing how well FNY data correlates to ILINet data will give local health departments an important tool to communicate the arrival of influenza to their jurisdiction. However, there is significant skepticism at the quality of FNY data as compared to validated datasets.

Objective

Our objective is to provide evidence for the data quality of Flu Near You (FNY) by evaluating the national and Houston datasets against CDC ILI data.

The Risk Identification Unit (RIU) of the US Dept. of Agriculture’s Center for Epidemiology and Animal Health (CEAH) conducts weekly surveillance of national livestock health data and routine coordination with agricultural stakeholders. In an initiative to increase the monitored species, health issues, and data sources, CEAH epidemiologists are building a surveillance system based on weekly counts of laboratory test orders along with Colorado State Univ. laboratorians and statistical analysts from the Johns Hopkins Univ. Applied Physics Lab. Initial efforts used 12 years of equine test records from 3 state labs covering most Colorado horse testing. Trial syndrome groups were formed based on RIU experience and published articles. Data analysis, stakeholder input, and discovery of laboratory workflow details were needed to modify these groups and filter test records to eliminate alerting bias. Customized statistical monitoring methods were sought based on specialized lab information characteristics and on likely presentation and health significance of syndrome-associated diseases.

The mission of the Maricopa County Department of Public Health (MCDPH; Arizona) is to protect and promote the health and well-being of its residents and visitors. Surveillance efforts allow epidemiologists to quantify and characterize public health threats, but traditional methods take time. In an effort to enhance situational awareness, the Office of Epidemiology dedicated resources to begin developing a robust syndromic surveillance program. This abstract outlines steps for enhancing syndromic surveillance at a local public health department.

Objective

To demonstrate how a local public health department used the Centers for Disease Control and Prevention (CDC) Framework for Program Evaluation and a logic model to enhance its syndromic surveillance program.

The EpiCenter syndromic surveillance platform currently uses Java libraries for time series analysis. Expanding the data quality capabilities of EpiCenter requires new analysis methods. While the Java ecosystem has a number of resources for general software engineering, it has lagged behind on numerical tools. As a result, including additional analytics requires implementing the methods de novo.

The R language and ecosystem has emerged as one of the leading platforms for statistical analysis. A wide range of standard time series analysis methods are available in either the base system or contributed packages, and new techniques are regularly implemented in R. Previous attempts to integrate R with EpiCenter were hampered by the limitations of available R/Java interfaces, which were not actively developed for a long time.

An alternative bridge is via the PostgreSQL database used by EpiCenter on the backend. An R extension for PostgreSQL exists, which can expose the entire R ecosystem to EpiCenter with minimal development effort.

Objective To demonstrate the broader analytical capabilities available by making the R language available to EpiCenter reporting