Participation in a Central Line Associated Blood Stream Infection [CLABSI] Prevention Collaborative and CLABSI Rates- Tennessee, 2008

Background: In 2006, the Tennessee [TN] legislature passed legislation that requires acute care hospitals with an average daily census of 25+ to report CLABSIs from critical care units [ICUs] and surgical site infection [SSI] for coronary artery  bypass graft surgery to the National Healthcare Safety Network [NHSN]. Reporting started 1/2008.  In late 2007, the Tennessee Hospital Association launched the TN Center for Patient Safety [TCPS] to implement several infection prevention collaboratives [PC], including a CLABSI -PC modeled after the Michigan Keystone Project that uses the Comprehensive Unit-based Safety Program [CUSP].

Objective: To determine if participation in the TCPS CLABSI-PC was associated with lower CLABSI rates.

Methods: We calculated a metric called 'Standardized Infection Ratio' [SIR]. SIR is an indirect standardization method and is a summary measure used to compare the CLABSI event among one or more groups of patients to that of a standard population (National NHSN data from 2006 through 2007). Computationally, SIR is the ratio of 'observed' CLABSI to 'statistically expected' CLABSI derived from the standard population, and accounts for differences in risk of CLABSI (i.e., number of central line days and different ICU types) among the two comparison groups (participation or not in the TCPS CLABSI-PC). We analyzed 2008 NHSN data, and included adult and pediatric ICUs, but not neonatal ICUs.

Results: 75 facilities reported CLABSI data; 49 (65%) participated in the TCPS CLABSI-PC. Overall SIR among participating facilities was not statistically different from the standard population (SIR=1.1, 95% confidence interval [CI]: 1.0-1.2).  The SIR remained insignificant even after stratification by ICU location type, a proxy measure of risk factor at the facility level. In contrast, overall SIR among non-participants was significantly higher than that of the standard population (SIR=1.4, 95%CI: 1.2-1.7), particularly in the two types of medical-surgical ICUs (SIR= 2.8; 95% CI 1.7-4.2 for medical surgical ICUs in major teaching hospitals; and SIR= 1.6; 95% CI= 1.2-2.1 for medical surgical ICUs in non-major teaching hospitals).

Conclusions: The SIR among facilities participating in the TCPS CLABSI-PC was lower than the SIR among facilities that did not participate.  This association may not be causal; facilities committed to reducing CLABSI may have been more inclined to participate in the PC. A remaining challenge is identifying methods to reduce CLABSI rates among non-participants of a PC.