Background:

Healthcare-associated infections represent a public health problem, and horizontal transmission has led to an increase in morbidity and mortality as well as higher health care costs. Active surveillance is expensive and carries high risk of failing to detect outbreaks. Virtual surveillance (mathematical models) allows a systematic search for alerts to outbreaks. The objective of this study is to evaluate the cost-effectiveness of the SaTScan-Whonet tool for the early detection of outbreaks of bacterial infection, compared with traditional surveillance, in an institution of high complexity in Colombia.

Methodology:

In a university hospital of high complexity a retrospective study was performed, identifying a bacterial outbreak that was characterised clinically and by molecular biology techniques. Databases of automated systems of identification and microbiological susceptibility were extracted. Retrospective analyses were performed using SaTScan-Whonet and daily simulations during the first semester of 2011 in a prospective manner. The date for the alert to the detection of the outbreak for both active and virtual surveillance was also identified.

Results:

A total of 4,584 microorganisms were isolated both inside and outside the ICU bet-ween 2010 and 2011 (2,288 and 2,296, respectively). An outbreak of Enterococcus faecium was identified by active surveillance on March 28, 2011. Using molecular biology techniques, the outbreak was characterised, showing the presence of the vanA gene, which confers resistance to glycopeptides. An alert to an Enterococcus faecium outbreak was retrospectively identified between March 14 and May 10, 2011 with a recurrence interval of 609,384. The first alert to outbreak for this bacterium was identified in a prospective simulated analysis on April 13, 2011 with a recurrence interval of 3,897 (P=.0002655).

Conclusion:

The use of such a tool prospectively is not superior to active surveillance in regard to timely detection of bacterial outbreaks. Retrospective analyses have high diagnostic ability and could be very helpful in systems of surveillance and control of regulatory entities.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )