Mandl Kenneth

Bordetella Pertussis outbreaks cause morbidity in all age groups, but the infection is most dangerous for young infants. Pertussis is difficult to diagnose, especially in its early stages, and definitive test results are not available for several days. Because of temporal and geographic variability of pertussis outbreaks, delay in diagnostic test results and ramifications of incorrect management decisions at the point of care, pertussis represents a prototypical disease where realtime public health surveillance data might inform, guide and improve medical decision making. Previously, we showed that diagnostic accuracy for meningitis can be improved when information about recent, local disease incidence is accounted for. Here, we quantify the contribution of epidemiologic context to a clinical prediction model for pertussis using a state public health data stream.

Objective

To explore the integration of epidemiological context – current population-level disease incidence data – into a clinical prediction model for pertussis.

Objective: To investigate seasonal patterns of gastrointestinal (GI) illness among children and adults.

With the widespread deployment of near real time population health monitoring, there is increasing focus on spatial cluster detection for identifying disease outbreaks. These spatial epidemiologic methods rely on knowledge of patient location to detect unusual clusters. In hospital administrative data, patient location is collected as home address but use of this precise location raises privacy concerns. Regional locations, such as center points of zip codes, have been deployed in many existing systems. However, this practice could distort the geographic properties of the raw data and affect subsequent spatial analyses. The impact of location error due to centroid assignment on the statistical analyses underlying these systems requires study.

Objective

To investigate the impact of address precision (exact latitude and longitude versus the center points of zip codes) on spatial cluster detection.

The performance of even the most advanced syndromic surveillance systems can be undermined if the monitored data is delayed before it arrives into the system.  In such cases, an outbreak may be detected only after it is too late for appropriate public health response. Surveillance systems can experience delays in data availability for a number of reasons: The process of transmitting data from data sources to the surveillance system can involve delays, especially in large systems where data is first aggregated across a national network of data sources before being transmitted to the surveillance system. Delays can also arise in the course of care, where, for example, a diagnosis is not available for a few days after the healthcare encounter.  It is important to minimize delays in data availability in order to maintain timeliness of detection [1].  When this is not possible, it is desirable to compensate for these data delays to minimize their effects.

Objective

This paper describes an approach to improving the detection timeliness of real-time health surveillance systems by modeling and correcting for delays in data availability.

While traditional means of surveillance by governments, multi-national agencies, and institutional networks assist in reporting and confirming infectious disease outbreaks, these formal sources of information are limited by their geographic coverage and timeliness of information flow. In contrast, rapid global reach of electronic communication has resulted in the advent of informal sources of information on outbreaks. Informal resources include discussion sites, online news media, individual and organization reports and even individual search records. The earliest descriptions of the severe acute respiratory syndrome outbreak in Guangdon Province, south China came from informal reports. However, system development to date has been geared toward knowledge management and strategies for interpreting these data are underdeveloped. There is a need to move from simple knowledge reorganization to an analytic approach for disseminating timely yet specific signals.

Objective

Internet-based resources such as discussion sites and online news sources have become invaluable sources for a new wave of surveillance systems. The WHO relies on these informal sources for about 65% of their outbreak investigations. Despite widespread use of unstructured information there has been little, if any, data evaluation.

This paper describes the syndromic networks paradigm and its application to various surveillance settings.

To measure the impact, within administrative geographic areas, of household income on rates of visits due influenza and RSV among children using real-time syndromic surveillance.

To evaluate the robustness of a spatial anonymization algorithm for syndromic surveillance data against a triangulation vulnerability attack. `BACKGROUND We have published an anonymization algorithm that takes precise point locations for patients and moves them a randomized distance according to a 2D Gaussian distribution that is inversely adjusted by the underlying population density. Before such algorithms can be integrated into live systems, assurances are needed so that patients cannot be reidentified through systematic vulnerabilities. Here we investigate the ease with which a spatial anonymization algorithm can be compromised by triangulating the original points with multiple repeated data requests. Obfuscative and cryptographic algorithms may be susceptible to weakening when it is possible for an adversary to produce output from the algorithm according to adversary-provided input. Under this threat model, an adversary could use a syndromic surveillance system to request anonymized patient data from a RHIO or other health network several different times. If the anonymized results are produced each time they are requested, triangulation of original addresses may be possible or the anonymity afforded by the algorithm may be reduced.

This paper describes a modular approach to surveillance system design, ensuring flexibility, scalability and fault tolerance.

Methods for locating spatial clusters of diseases are typically variations of the circular scan statistic method. They restrict the number of potential clusters by considering all circular, rectangular, or elliptical regions, and then apply a likelihood ratio test to evaluate the statistical significance of each potential cluster. Because disease outbreaks may have variable shapes, there has been recent interest in developing methods to detect irregularly-shaped clusters. Starting with a neighborhood graph of the administrative regions in the study area, certain sub-graphs are evaluated. These include all connected subgraphs within a circular window and sub-graphs of the minimum spanning tree of a weighted neighborhood graph formed by deleting one edge. These methods restrict the maximum cluster size or identify large clusters having greater likelihood ratios than true clusters in the data, suggesting a limitation of using the likelihood ratio to detect arbitrarily-shaped clusters.

Objective

A method for detecting spatial clusters of diseases of any shape based on the Euclidean minimum spanning tree is described and compared to the circular scan statistic.