Eliminating CLABSI: A National Patient Safety Imperative
This report uses data stored in the On the CUSP: Stop BSI National Database created and maintained by the MHA Keystone Center in Lansing, Michigan. Each month, the number of central line days and the number of CLABSIs observed in the participating hospital units are submitted into the national database. Some hospitals submit infection rate data directly into this database, and others contribute data through the Centers for Disease Control and Prevention's (CDC) National Healthcare Safety Network (NHSN). All hospitals use CDC definitions to count central line days and determine the number of CLABSIs observed in participating units.
All analyses except the list of participating hospitals in Appendix A are based on data drawn from the National Database on October 7, 2010. Data from participating hospitals in Tennessee were added to the National Database in November 2010, so results from these hospitals will be included in the next project update. Because we want to recognize as many of the hospitals that have chosen to participate as possible, the hospital list in Appendix A reflects all participants confirmed as of January 2011.
For our assessments of project impact, data were limited to teams representing adult ICUs that began participating in the project in 2009. These teams were drawn from eight States that began the project in May 2009 (cohort 1) and an additional 14 States that began the project in September 2009 (cohort 2). We limited our results to adult ICUs because of challenges in counting central line days outside the ICU setting. As such, CLABSI data for ICUs are more valid than CLABSI rates outside the ICU. Results are currently limited to the first two cohorts because they have at least four quarters of data that can be used to assess CLABSI rate changes over time.
Data from the On the CUSP: Stop BSI National Database were linked to information in the 2008 AHA Annual Survey of hospitals based on hospital name, location, and AHA hospital identifier (when available and accurate). All hospitals that could be linked were classified by size using the "total beds staffed" variable from the AHA Annual Survey. Hospitals were also classified based on whether they were critical access hospitals, rural referral hospitals, members of a health care system, located in one of the 100 largest cities in the United States, or members of the Council of Teaching Hospitals (COTH). Because hospitals in North Carolina and Ohio use blinded hospital identifiers, hospital characteristics were not available for those hospitals or for an additional nine hospitals that could not be matched to data in the AHA Annual Survey.
CLABSI rates for the two cohorts were calculated for a baseline time period of up to 12 months and for each subsequent time period. At the time of the production of this report each cohort had at least four full quarters of data. The time periods are shown below:
|Cohort 1||Cohort 2|
|Baseline||May 1, 2008 to Apr. 30, 2009||Sept. 1, 2008 to Aug. 31, 2009|
|Time period 1||May 1, 2009 to July 31, 2009||Sept. 1, 2009 to Nov. 30, 2009|
|Time period 2||Aug. 1, 2009 to Oct. 31, 2009||Dec. 1, 2009 to Feb. 28, 2010|
|Time period 3||Nov. 1, 2009 to Jan. 31, 2010||Mar. 1, 2010 to May 31, 2010|
|Time period 4||Feb. 1, 2010 to July 31, 2010||June 1, 2010 to Sept. 31, 2010|
Central line days were calculated as the sum of all reported central line days during the reporting period, while the number of CLABSIs equaled the sum total of all CLABSIs reported during the period. The CLABSI rate equals the ratio of CLABSIs to central line days multiplied by 1,000 (to create a rate per 1,000 central line days). If a team reported zero central line days, the CLABSI rate for that period was set to "missing." However, the team still was classified as having reported CLABSI data for that period.