749 Risk Factors for Co-Colonization with Multidrug-Resistant Gram-Negative Bacteria in a Long-Term Care Facility

Sunday, March 21, 2010: 11:00 AM
International North (Hyatt Regency Atlanta)
Graham M. Snyder, MD , Beth Israel Deaconess Medical Center, Boston, MA
Erin O'Fallon, MD , Hebrew Senior Life Institute for Aging Research, Boston, MA
Erika M. C. D'Agata, MD, MPH , Beth Israel Deaconess Medical Center, Boston, MA
Background: The rate of colonization with multidrug-resistant bacteria (MDRGN) is rapidly increasing. Patients co-colonized with multiple different MDRGN species may be the “super-spreaders” of MDRGN. Characterizing this subgroup of patients would have important infection control implications. Objective: To identify variables associated with MDRGN co-colonization among long-term care facility (LTCF) residents. Methods: A cross-sectional study was performed to compare risk factors among residents colonized with ≥2 different MDRGN species to residents colonized with a single MDRGN species. MDRGN were defined as isolates resistant to ≥3 of the following drug classes: 3rd or 4th generation cephalosporins, beta-lactam/beta-lactamase inhibitor combinations, carbapenems, fluoroquinolones, and aminoglycosides. Rectal swabs, demographics, length of LTCF stay, hospitalization and antibiotic use in the previous 12 months, and validated scales assessing functional, health and cognitive status were collected at baseline. The global deterioration score (GDS), a validated assessment of dementia with ordinal values ranging from 1 to 7, was used and dichotomized at ≥5, representing patients with severe dementia who can no longer handle activities of daily life (ADL) without assistance. The correlation between variables among patients colonized with 1 MDRGN versus ≥2 MDRGN species was evaluated using Fisher's exact test. Independent risk factors were identified by stepwise logistic regression. Results: Of 212 enrolled patients, 53 (25%) patients were colonized with ≥1 MDRGN. Among these, 11 (20.8%) were colonized with ≥2 different MDRGN species. Ten and 1 residents were colonized with 2 and 3 different MDRGN species, respectively. Among the 42 singly-colonized patients, 6 (14.3%) were hospitalized and 32 (76.2%) received antibiotics in the preceding 12 months, while among 11 co-colonized patients, 1 (9.1%) was hospitalized and 9 (81.8%) received antibiotics. Length of LTCF stay for singly-colonized patients was 3.6 years (range, 0.04-14.5 years), and for co-colonized patients 3.9 years (range, 0.3-9.3 years). Differences in recent hospitalization, antibiotic use, and duration of LTCF stay were not significant between groups. Among the 11 patients co-colonized with MDRGN, all 11 had a GDS score ≥ 5, while only 30 (71%) of 42 singly-colonized patients had a GDS score ≥ 5 (p=0.04). Conclusions: Patients with severe dementia who require assistance with ADL are at high risk of MDRGN co-colonization and may be the “superspreaders” of MDRGN transmission in LTCF.