Issue 21 - Reviewing The Process of Syndromic Surveillance

____________________________________________________________________________

Reviewing The Process of Syndromic Surveillance
_____________________________________________________________________________

Here are 11 important points made in this article about syndromic surveillance…

1) Syndromic surveillance is based on the assumption that “ill persons may exhibit behavioral patterns, symptoms, signs, or laboratory findings that can be tracked through a variety of data sources”. This assumption holds regardless of whether the illness is caused by a covert bioterrorist attack or a natural illness.

2) Syndromic surveillance “relies on detection of clinical case features that are discernible before confirmed diagnosis is made”. In many cases, this is done by text mining emergency room or other data for symptoms and then developing syndrome scores or probabilities from these symptom data. Most syndromic surveillance systems capture “age, gender, free-text chief complaint, ICD9 coded discharge diagnosis (if available), and some form of spatial location (most often zip code).”

3) “Contemporary syndromic surveillance relies on data which is continuously acquired through protocols or automated routines”. That is, manual data entry is not a part of a modern syndromic surveillance system. “Practicality has dictated use of data already collected for other purposes,” because developing new data collection processes is costly and past experience has indicated that compliance is poor.

4) To begin a regional syndromic surveillance system, you need a provider (hospitals, clinics) and a public health authority. And “sufficient regional coverage may be achieved with data from a few large data providers”. However, “The geographic, demographic, and temporal coverage must be sufficient to support anomaly detection”.

5) The best data sources are “electronically stored, allow robust syndromic grouping and are available in a timely fashion”. ICD9 codes, for example, appeared promising early on, but have not been practical in most settings because there is a delay of days or weeks before a record is coded.

6) “In the case of a single hospital system reporting surveillance data to public health authorities, the HIPAA privacy regulations permit the unencumbered transmission of such information if it meets the criteria for public health activity.” However, data transmission should be encrypted or sent over a private network.

7) There are a number of different analysis methods that can be used. The CuSUM control chart, which relies on “cumulative differences between observed and expected data in a time window when compared to a threshold,” is a common statistical approach.

8) Three important issues in evaluating syndromic surveillance systems are: “sensitivity & specificity (data quality) and timeliness… “The potential data source should be judged by the combination of data quality and timeliness, as well as knowledge of the cost of false alarms versus the cost of delays in triggering true alarms…”

9) Since syndromic surveillance is based on non-diagnostic data, it simply provides signals that an event may be underway that deserves attention. In most cases no outbreak has occurred, but each signal needs to be investigated by an individual trained for this purpose. “In New York City, results of syndromic analyses are examined every day…365 days a year.”

10) When a public health investigation is needed, the ability to re-identify and examine individual cases is important. Syndromic surveillance data which provide only counts for each syndrome category (no patient level records) makes followup difficult or impossible.

11) “Dual use” systems, which combine bioterrorism syndrome tracking with the ability to automate data collection and provide new avenues to data, are the best. “If a surveillance system is designed to only detect bioterrorism or very rare outbreaks, its use and funding allocation will diminish over time if there are no events”. However, systems that provide dual use are more likely to be “maintained, improved and used. Furthermore, they are more likely to be up and running should a bioterrorist attack occur”.

1. Mandl, Kenneth, et al. Implementing Syndromic Surveillance: A Practical Guide Informed by the Early Experience. Journal of American Medical Informatics Association. 2004; 11:141-150

Back to the top