Data Use

Syndromic surveillance originally was originally developed to detect outbreaks around 2001. Public Health practitioners have improved the utilization of these data by applying them to larger population trends in illness to inform surveillance

Common uses of the data

• Influenza-like illness
• Drug-related morbidity (opioid and heroin)
• Weather –related illness (heat and cold)
• Case detection (e.g., rabies, varicella)
• Gastro-intestinal illness
• Mass gathering events (e.g., Eclipse, marathon)
• Environmental health (e.g., CO poisoning, mercury)

Specific customized searched for unique events can be developed, which may involve searches of the Triage Note or national data.  IDPH can assist in developing these queries.
 

Important considerations for understanding and interpreting the data

• Best practice for looking at the data is by trends in the percent of visits over time.
• The data represent visits to IL hospitals only; if an IL resident seeks care at an out of state facility, or a location that does not report (such as a physician office), it will not be reflected in the data
• Counts alone are NOT exact case counts for a population. Categorized visits are not based on specific case definitions or laboratory results, as reportable communicable diseases are. Visits are classified into illness or injury categories based on their chief complaint or the diagnoses coded for the visit.
  • A content of a chief complaint can be a detailed summary in the person’s own words or it may be from a drop down menu or pick list of symptoms. A chief complaint may be too limited or non-specific, and may not be categorized into a specific illness (e.g., nosebleed or loss of conscience may be the only complaint, but it could be due to an overdose, or a flu visit may only say fever and be insufficient to categorize the visit).
  • Alternatively, sometimes a visit will be misclassified and included in a category, but is not truly do to that illness, such as an STI acronym may be grouped into a sexually transmitted infection when it was due to a soft tissue infection.
  • A query that includes a search for text terms and diagnosis codes generally captures more visits than a chief complaint query only.
• Time series comparisons of counts or trends may increase or decrease artificially due to changes in data quality.
  • If a facility fails to report during a specific time period, the overall missing data will show up as a decrease in visit counts due to a specific condition. A user should be sure the underlying total visits submitted are consistent over time. 
  • Comparisons to historic trends over several years should take into consideration when a facility began reporting. Before 2016, new sites began sending more visits at various times. The addition of a new facility could lead to an increase that is not due to more illness. Percent trends are best practice to avoid this problem.
  • Illness or injury queries that include a diagnosis may show a drop in trends for recent days, as diagnosis may not be received until long after the visits occurred.
• Spatial comparisons across regions of IL should be done with caution and with understanding of the similarity of utilization patterns and access to hospitals or alternative care sites in a region as well as verification that data is consistent in quality and completeness between geographic regions.
  • Note that there may be differences between Patient Location and Facility Location counts. This could be due to missing patient ZIP code, so that the Patient Location data source will assign this visit to ‘other’ region.