274 Correlation of Hand Hygiene Compliance and Nosocomial Infections in a University Hospital Adult ICUs

Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
John Shepard, BA/MBA , Johns Hopkins University, Baltimore, MD
Aaron Milestone, MD , Johns Hopkins University, Baltimore, MD
Elaine Fadden , Johns Hopkins University, Baltimore, MD
Polly Ristaino, MPH , Johns Hopkins University, Baltimore, MD
Lisa Maragakis, MD , Johns Hopkins University, Baltimore, MD
Trish Perl, MD, MSc , The Johns Hopkins Hospital, Baltimore, MD


Background: HH is considered the single most important intervention to prevent infection.  While data exist showing HH can decrease infections in specific settings, there are limited data on the impact a HH program can have on HAIs.
Objective: To correlate HH compliance with hospital acquired transmissions or infections for Methicillin-resistant Staphylococcus aureus (MRSA), vancomycin resistant enterococcu (VRE), Clostridium difficile, and central line associated blood-stream infections (CLA-BSI) in adult ICUs.
Methods: Trained, unknown observers gathered data on compliance with the HH policy on all 6 adult ICUs for 2 years. ICUs were separated into 2 groups: (Group A) those with a >30% median increase HH compliance and (Group B) those units with < 30% median increase in HH compliance over 2 years. HA CLABSI, C-Diff, MRSA, & VRE data was gathered electronically and validated manually using NHSN guidelines. Patients with no previous history of C-Diff, MRSA, or VRE that are admitted >48 hours and then colonized/infected was said to have had HAI. For C-Diff, a patient had to have a positive Cytotoxin or PCR test and the unit where the positive culture was collected was where the infection was counted. CLABSI infections were reviewed by trained epidemiologists to verify HAI. The total # of infections were combined into a composite number reflecting all four infections and then converted to a rate by multiplying the total # of infections by 1,000 and then dividing by the total # of patient days.   Results: Group A had a statistically significantly lower median infection rate than the Group B, P-Value <.00001 using a Mann-Whitney test. The units in Group A had a median hand hygiene compliance of 60%, a 40% increase in the median hand hygiene compliance from year 1 to year 2, a median infection rate of 6.57 over 2 years, and a 1.15/ 1,000 patients days decrease in infection rate, a 16% decrease, from year 1 to year 2. The units in Group B had a median hand hygiene compliance of 54%, a 26% increase in the median hand hygiene compliance from year 1 to year 2, a median infection rate of 11.50 over 2 years, and a 2.18/ 1,000 patients days decrease in infection rate, a 17% decrease, from year 1 to year 2.

Conclusions: The correlation between HH and HAIs will be important to examine.  For all 6 adult ICU's there was an inverse linear correlation between HH compliance and the infection rate. The Persons Co-Efficient for the data was P=-0.26. If the HH data is shifted to correlate the infection rate with the hand hygiene compliance of two months prior the then P=-.33 We found that HH increased from year 1 to year 2 and the HAI rate decreased in an approximate ratio of 2:1. The incremental benefits of increased HH and their impact on HAI rates are important to understand within healthcare systems and to promote patient safety.