An outbreak of Serratia marcescens infections occurred in a university tertiary-care hospital. Alcohol-free chlorhexidine solutions were contaminated with S marcescens. The majority of patient and chlorhexidine strains had similar pulsed field-gel electrophoresis banding patterns. Chlorhexidine was recalled, and the rate of S marcescens isolation returned to baseline. Chlorhexidine without alcohol should not be used as an antiseptic

A survey among hospital personnel regarding influenza vaccination (response rates 44% and 80% for vaccinated and unvaccinated subjects, respectively) identified fear of side effects as the main reason for refusing vaccination, whereas a convenient vaccination schedule fostered participation.

Even with a good surveillance program, nosocomial infections may be not recognized because of several reasons: absence of symptoms or prolonged incubation period (eg, viral bloodborne infections, tuberculosis); problems with the microbiological diagnosis, because adequate specimens may be difficult to obtain or special methods should be used (eg, fungal infections, virus, new agents); shorter hospital stays (eg, surgical-site infections); difficulty in distinguishing between nosocomial and community-acquired infections (eg, influenza); and failure to detect clinically relevant colonization (eg, multiresistant microorganisms). Because of the important potential consequences of occult nosocomial infections, specific surveillance programs should be designed to address these problems.

The most common route of entry of pathogens into the lung in patients managed in the intensive-care unit is aspiration of contaminated oropharyngeal secretions or gastric contents. In intubated patients, the risk of this type of infection is particularly high. Knowledge of specific risk factors for specific microorganisms, along with the origin of acquisition (primary endogenous, secondary endogenous, or exogenous), will permit a more rational and effective method of prevention. Attributable mortality is highly dependent on the institution of a correct initial antibiotic choice, as well as the interaction between the virulence of the pathogen responsible and host defenses. However, survival in these patients is determined primarily by the degree of severity at the time of pneumonia diagnosis and the response to initial therapy. As a consequence, the number of preventable deaths is likely to be much smaller than the total. Therapy requires both supportive and specific measures. When diagnostic information becomes available, it permits the rescue of some patients with inadequate therapy or simplifies the spectrum of the empirical therapy. Initial antibiotic choice should be based on expected etiologic pathogens, while knowledge of local microbial epidemiology and susceptibility patterns is crucial.