Airborne bacteria in hospitals have been implicated in nosocomial infections. This investigation studied the characteristics of airborne bacteria and the effect of patient visitation on the bacteria profile in intensive care units (ICUs). Air at a medical ICU and surgical ICU was sampled for one year. Airborne bacteria before and during visits to patients in ICUs were collected using a Six-Stage Viable Andersen Cascade Impactor to analyze the concentration and size distribution of airborne bacteria and the percentage thereof that were antibioticresistant. During patient visitation in the ICUs in this study, the number of visitors was 20-80. Airborne bacteria concentration during visiting hours (total averaging 168.5 CFU/m(3)) was three to four times than before visiting hours (p = 0.043). With increasing the visitors, most of the airborne human-associated bacteria (HAB) concentrations during visitations were higher than before visitations in each season. The two-way ANOVA of HAB concentration before and during visitation (p = 0.028) of combining MICU and SICU in various season (p = 0.007) all revealed statistical agreement. The proportion of particles, from 1.1 to 4.7 mu m, during the visits was almost 1-2.4 times that before the visits in most sampling periods (p = 0.028). In addition, the opportunistic pathogens such as Micrococcus spp., Staphylococcus spp. and Acinetobacter spp. were found in the air during visiting times. Small proportions of some environmental strains with a high antibiotic-resistance percentage (42- 78%), including Brevundimonas spp., Elizabethkingia spp. and others, were detected during patient visitation. Patient visitation activities affect the bacterial profile in air in ICUs. During the visitation, visitors might bring or generate bacteria into ICUs. Limiting the number of patient visitors to ICUs, wearing respirators and gowns or increasing ventilation rate during and after patient visitation is required to maintain indoor air quality and probably decrease the risk of patient infection.
