Development (and Validation) of a Novel Electronic Surveillance System for Monitoring of Bloodstream Infections
Objective: Electronic surveillance systems (ESS ) that utilize existing information in databases are more efficient than conventional infection surveillance methods. We developed an ESS for bloodstream infections (BSI ) and assessed its agreement with traditional manual chart review.
Methods: An ESS was developed by linking data from regional laboratory and hospital administrative databases. Definitions for excluding contaminants and duplicative episodes were developed and applied. Infections were classified as nosocomial (NI), healthcare-associated community onset (HCA), or community acquired (CA). A random sample was then compared with independent chart review.
Results: Among the 306 patients selected for comparative review, ESS identified 323 episodes of BSI of which 107 (33%) were HCA, 108 (33%) were CA, 107 (33%) were NI and 1 failed to be classified. In comparison, chart review identified 310 episodes of which 116 (37%) were HCA, 95 (31%) were CA, and 99 (32%) were classified as NI. True episodes of BSI were concordant by ESS and chart review in 302 cases. Of the 21 discordant cases further identified by ESS, 17 (81%) were classified as contaminants by chart review. Of the 8 discordant cases further identified by chart review, most were due to classification as repeat or poly-microbial episodes of disease. There was an overall 85% agreement between ESS and chart review (kappa=0.78, standard error 0.04) for classification as to location of acquisition.
Conclusion: Our novel ESS identifies and classifies BSI with a high degree of accuracy. This system requires validation in other cohorts and settings.
