Use of computer algorithms to conduct surveillance

Use of computer algorithms to conduct surveillance 

A few surveillance systems have been developed that employ computer algorithms to screen electronic data sources for disease cases and apply automated statistical methods to assess data trends and changes in case activity. For example, a component of the Infectious Diseases Surveillance and Information System (ISIS) in the Netherlands runs automated algorithms on electronically-transmitted laboratory data to identify selected cases of public health interest (e.g., new positive Neisseria gonorrhea test results). Automated time-series analyses process these and other surveillance data to detect variations from expected rates; statistically significant changes automatically generate and distribute alerts (see Chapter 22). Syndromic surveillance systems use automated data extraction and analyses methods to detect aberrations from expected levels of various syndromes

Although these systems exhibit the powerful capacity of technologies to automatically process enormous quantities of data, humans must still verify, investigate, and prioritize these reports. Research is needed to refine these automated data processing systems and capitalize on their strengths.