Benefit of systematic selection of pairs of cases matched by surgical specialty for surveillance of bacterial transmission in operating rooms

Background Bacterial transmission within and between successive surgical cases occurs in operating rooms (ORs), often includes anesthesia equipment as a reservoir, and can be monitored by collecting samples and identifying bacteria by genetic testing. We evaluated how to choose cases for active surveillance to quantify the effectiveness of interventions in 2 groups of ORs (eg, rooms with germicidal lighting vs those without). Methods Data were from a 7 OR single-specialty gastrointestinal endoscopy suite and from a typical 8 OR multispecialty surgical suite. Results At the multispecialty hospital, 40.3% (SE 1.2%) of the total number of cases could be used for surveillance (ie, followed by another case of the same specialty and matched with a corresponding pair of cases from the other OR group). Random selection obtained fewer matched pairs than deliberate selection: mean ratio of random/deliberate = 0.64 (0.01) for the single-specialty and 0.51 (0.02) for the multispecialty suite (P Conclusions The efficiency of sampling to obtain pairs of successive surgical cases of the same specialty is impaired markedly by randomly selecting pairs of cases (or using convenience sampling) as compared to choosing pairs deliberately. This is important because the number of cases that can be suitably used for surveillance of bacterial transmission will typically be less than one-half the total case number.