Disease Surveillance

Modern surveillance systems use statistical process control (SPC) charts such as Cumulative Sum and Exponentially Weighted Moving Average charts for monitoring daily counts of such quantities as ICD-9 codes from ED visits, sales of medications, and doctors’ office visits. The working assumption is that such pre-clinical data contain an early signature of disease outbreaks, manifested as an increase in the count levels. However, the direct application of SPC charts to the raw counts leads to unreliable performance. A popular statistical solution is to precondition the data before applying the charts by modeling or removing explainable patterns from the data and then monitoring the residuals. Although the general idea is common practice, the specifics of how to identify the existing explainable components and how to account for them are domain-specific. Therefore, we seek to present a set of modeling and data-driven tools that are useful for syndromic data.

Objective

SPC charts are widely used in disease surveillance. The charts are very effective when monitored data meet the requirements of temporal independence, stationarity, and normality. However, when these assumptions are violated, the SPC charts will either fail to detect special cause variations or will alert frequently even in the absence of anomalies. Currently collected biosurveillance data contain predictable factors such as day-of-week effects, seasonal effects, holidays, autocorrelation, and global trends that cause the data to violate these assumptions. This work (1) describes a set of tools for identifying such explainable patterns and (2) examines several data preconditioning methods that account for these factors, yielding data better suited for monitoring by traditional SPC charts.

Modern biosurveillance relies on multiple sources of both prediagnostic and diagnostic data, updated daily, to discover disease outbreaks. Intrinsic to this effort are two assumptions: (1) the data being analyzed contain early indicators of a disease outbreak and (2) the outbreaks to be detected are not known a priori. However, in addition to outbreak indicators, syndromic data streams include such factors as day-of-week effects, seasonal effects, autocorrelation, and global trends. These explainable factors obscure unexplained outbreak events, and their presence in the data violates standard control-chart assumptions. Monitoring tools such as Shewhart, cumulative sum, and exponentially weighted moving average control charts will alert based largely on these explainable factors instead of on outbreaks. The goal of this paper is 2-fold: first, to describe a set of tools for identifying explainable patterns such as temporal dependence and, second, to survey and examine several data preconditioning methods that significantly reduce these explainable factors, yielding data better suited for monitoring using the popular control charts.

This paper describes lessons learned from a regional tabletop exercise (TTX) of the National Capital Region (NCR) Syndromic Surveillance Network, from the perspective of the Maryland Department of Health and Mental Hygiene (DHMH).

Real-time disease surveillance is critical for early detection of the covert release of a biological threat agent (BTA). Numerous software applications have been developed to detect emerging disease clusters resulting from either naturally occurring phenomena or from occult acts of bioterrorism. However, these do not focus adequately on the diagnosis of BTA infection in proportion to the potential risk to public health.

GUARDIAN is a real-time, scalable, extensible, automated, knowledge-based BTA detection and diagnosis system.  GUARDIAN conducts real-time analysis of multiple pre-diagnostic parameters from records already being collected within an emergency department (ED).  The goal of this system is to assist clinicians in detecting potential BTAs as quickly and effectively as possible in order to better respond to and mitigate the effects of a large-scale outbreak.  

GUARDIAN improves the diagnostic process by moving away from simple trend anomaly detection and towards the development of a BTA-specific infectious disease expert system [1].  Through the capture and automated application of specific clinical expertise, GUARDIAN provides the focus and accuracy necessary for effective BTA infection diagnosis.  The continuity of this process improves the efficiency by which diagnoses of BTA infections can be made.

 

Many heuristics were developed recently to find arbitrarily shaped clusters (see  review  [1]). The most popular statistic is the spatial scan  [2]. Nevertheless, even if all cluster solutions could be known, the problem  of selecting the best cluster is ill posed. This happens because other measures, such as geometric regularity  [3-5] or topology  [6] must be taken intoconsideration. Most cluster finding  methods does not address  this last problem. A genetic multi-objective algorithm was developed elsewhere to identify irregularlyshaped clusters [5]. That method conducts a search aiming to maximize two objectives, namely the scan  statistic and the regularity of shape (using the compactness concept).The solution presented is a Pareto-set, consisting of all the clusters found which are not simultaneously worse in both objectives. The significance evaluation is conducted in parallel for all the  clusters  in  the  Pareto-set  through a  Monte Carlo simulation, determining the best cluster solution.

Objective

Irregularly shaped clusters occur naturally in disease surveillance, but they are not well defined. The number of possible clusters increases exponentially with the number of regions in a map. This concurs to reduce the power of detection, motivating the utilization of some kind of penalty function to avoid excessive freedom of shape. We introduce a weak link based correction which penalizes inconsistent clusters, without forbidding the presence of the geographically interesting irregularly shaped ones.

Electronic  Health  Record  (EHR)  data  offers  the  researcher a potentially rich source of data for tracking disease  syndromes. Procedures  performed  on  the  patient, medications prescribed (not necessarily filled by  the  patient),  and  reason  for  visit  are  just  some  characteristics of the patient encounter that are available  through  an  EHR  that  can  be  used  to  define  surveillance  syndromes.    Since  procedures  have  not  been used frequently in defining syndromes, encounter  level  procedures  data,  extracted  from  the  EHR  of  a   large   local   primary   care   practice   with   about   200,000 patient encounters per year was used to identify  procedures  associated  with  an  established  respiratory syndrome.

Objective

To investigate the utility of different sources of patient encounter information, particularly in the primary care setting, that can be used to characterize surveillance syndromes, such as respiratory or flu.

 Internet-based technologies have been used to assist in disease surveillance and reporting.  The Public Health Agency of Canada operates the Global Public Health Information Network, credited with early notification of many outbreaks (including SARS) through automated multilingual analysis of internet media sources such as news wires and web pages(www.phac-aspc.gc.ca/media/nr-rp/2004/2004_gphinrmispbk_e.html). An innovative web-based forum (www.RUsick2.msu.edu) collects foodborne illness reports from visitors to a web site and has been used to identify foodborne outbreaks in Michigan (1).   Health-related topics are among the most popular Internet searches. Many individuals experiencing symptoms of illness conduct Internet searches prior to seeking medical attention.  An early site-based study found limited utility to monitoring of Internet queries (2), but recent developments merit re-examination of the potential of internet searches for public health surveillance purposes.

Objective:

To evaluate whether trends in internet searches might provide useful data for public health surveillance.

We present a new method for multivariate outbreak detection, the ìnonparametric scan statisticî (NPSS). NPSS enables fast and accurate detection of emerging space-time clusters using multiple disparate data streams, including nontraditional data sources where standard parametric model assumptions are incorrect.

Multiple surveillance activities have been conducted in Great Britain (GB) with the objective of estimating the occurrence of scrapie, a fatal neurological infectious disease of small ruminants: statutory reporting of clinical cases, annual surveys on sections of the population and occasional anonymous postal surveys. None of the surveillance sources is either unbiased or comprehensive and if the progress of control schemes is to be closely monitored, better estimates of disease occurrence are required. With this objective, the Department for Food, Environment and Rural Affairs (Defra) funded a project to: i)provide estimates of the frequency of scrapie that integrate currently available surveillance data; and ii)inform the most effective surveillance strategies that will result in sensitive systems for the detection of changes in disease prevalence in time. To make this review as comprehensive as possible it should also: i)consider clinical disease and infection at both individual animal and holding level; ii) subject to data availability, extend all analyses to the recently detected atypical form of scrapie and iii) in a context of scarce and competitive resources, approach the problem efficiently. The approaches used within this project, outlined below, describe the efficient use and integration of all existing sources to evaluate the surveillance effort. Three surveillance attributes were of particular interest in the evaluation process: sensitivity, representativeness and cost.