Kola A, Schwab F, Bärwolff S, Eckmanns T, Weist K, Dinger E, Klare I, Witte W, Rüden H, Gastmeier P (2010): Is there an association between nosocomial infection rates and bacterial cross transmissions?
Crit. Care Med. 38 (1): 46-50. Epub 2009 Sep 16.

Objective: Surveillance data of nosocomial infection rates are increasingly used for public reporting and interhospital comparisons. Approximately 15% of nosocomial infections on intensive care units are the result of patient-to-patient transmissions of the causative organisms. These exogenous infections could be prevented by adherence to basic infection control measures. The association between bacterial cross transmissions and nosocomial infection rates was analyzed.

Design: Prospective cohort study during 24 months.

Setting: Eleven intensive care units from two university hospitals.

Patients: All inpatients.

Interventions: None.

Measurements and Main Results: Primary isolates of six indicator organisms (Acinetobacter baumannii, Enterococcus faecalis and E. faecium, Klebsiella pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus) cultured from clinical samples or methicillin-resistant S. aureus surveillance testing of all inpatients were genotyped. Indistinguishable isolates in >=2 patients defined potential episodes of transmissions. Surveillance of nosocomial infection rates was performed according to the German nosocomial infection surveillance system, Krankenhaus Infektions Surveillance System (KISS). Transmission events and nosocomial infection rates were pooled by intensive care unit to calculate Spearman's rank-correlation test. During 100,781 patient days, 100,829 microbiological specimens from 24,362 patients were sampled (average investigation density: 1.0 sample per patient and day) and 3419 primary indicator organisms were cultured. Altogether, 462 transmissions (incidence density of 4.6 transmissions per 1000 patient days; range, 1.4-8.4 days) and 1216 nosocomial infections (incidence density of 12.1 per 1000 patient days; range, 6.2-16.6 days) were discerned. Correlation analysis was unable to reveal any association between the incidence of cross transmissions and nosocomial infections, duration of hospitalization, or device use.

Conclusions: Differences in nosocomial infection rates between study intensive care units are not explained solely by cross transmissions. Other factors, like the severity of the patient's underlying diseases, the patient's endogenous flora, or invasive procedures, likely have a dominant effect on the magnitude of nosocomial infection rates.