To examine the impact of cleaning and directional airflow on environmental contamination with Aspergillus species in hospital rooms filtered with high-efficiency particulate air (HEPA) filters that house patients with hematologic malignancy.

Detailed environmental assessment.

A 475-bed tertiary cancer center in the southern United States.

From April to October 2004, 1,258 surface samples and 627 bioaerosol samples were obtained from 74 HEPA-filtered rooms (in addition, 88 outdoor bioaerosol samples were obtained). Samples were collected from rooms cleaned within 1 hour after patient discharge and from rooms before cleaning. Positive and negative airflows were evaluated using air-current tubes at entrances to patient rooms.

Of 1,258 surface samples, 3.3% were positive for Aspergillus species. Univariate analysis showed no relationship between cleaning status and occurrence of Aspergillus species. Of 627 bioaerosol samples, 7.3% were positive for Aspergillus species. Multiple logistic analysis revealed independently significant associations with detection of Aspergillus species. Cleaned rooms positive for Aspergillus species had a higher geometric mean density of colonies than that of rooms sampled before cleaning (18.9 vs 5.5 colony-forming units [cfu] per cubic meter; P = .0047). Rooms with positive airflow had a detection rate for bioaerosol samples equivalent to that of rooms with negative airflow (7.3% vs 7.8%; P = .8). There was no significant difference in the density of Aspergillus species between rooms with negative airflow and rooms with positive airflow (12.5 vs 8.4 cfu/m3; P = .33).

Concentration of bioaerosol contamination with Aspergillus species was increased in rooms sampled 1 hour after cleaning compared with rooms sampled before cleaning, suggesting a possible correlation between reentrained bioaerosols (ie, those suspended by activity in the room) after cleaning and the risk of nosocomial invasive aspergillosis.

To determine the impact of stool surveillance cultures of critically ill patients on controlling vancomycin-resistant enterococci (VRE) outbreak bacteremia.

Stool surveillance cultures were performed on patients who had hematologic malignancy or were critically ill at the time of hospital admission to identify those colonized with VRE. Hence, contact isolation was initiated.

A tertiary-care cancer center with a high prevalence of VRE.

All patients with hematologic malignancy who were admitted to the hospital as well as all of those admitted to the intensive care unit were eligible.

Active stool surveillance cultures performed between 1997 and 2001 decreased the incidence density of VRE bacteremias eightfold while vancomycin use remained constant. In fiscal year (FY) 1997 and FY 1998, there were five and three VRE outbreak bacteremias, respectively. The outbreak clones were responsible for infection in 69% of those patients with VRE bacteremia. However, the stool surveillance program resulted in the complete control of VRE bacteremia by FY 1999 until the end of the study.

Despite the steady use of vancomycin, the active surveillance program among high-risk patients with hematologic malignancy and those who were critically ill resulted in the complete control of VRE outbreak bacteremia at our institution.