Eliminating CLABSI, A National Patient Safety Imperative: Final Report Companion Guide

Data Collection

This report uses data stored in the Care Counts database created and maintained by the Michigan Health & Hospital Association (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 were submitted to this database. Some facilities submitted infection rate data directly into this database, while others contributed data through the Centers for Disease Control and Prevention's (CDC's) National Healthcare Safety Network (NHSN). All participants used CDC definitions to count central line days and to determine the number of CLABSIs observed.

All analyses are based on data extracted from Care Counts on October 15, 2012. The baseline period consisted of 12 consecutive data collection months and post-baseline data collection periods consisted of 3 consecutive months, referred to as quarters. One should note that quarters are in reference to project quarters, not calendar quarters. Since cohorts entered the project at different times, not all cohorts have completed data entry for all six post-baseline project quarters. The October 2012 data extract included data submitted for all periods through August 2012. The project quarters by cohort included in the October 2012 data extract, and used for this report, are detailed in Table 1. For cohort 6, only two of the three months of its sixth project quarter were available.

Table 1. Project data by cohort

Calculations

The project calculated central line days as the sum of all reported central line days during the reporting period, while the number of CLABSIs equaled the sum of all CLABSIs reported during the period. The unit level CLABSI rate equals the ratio of CLABSIs to central line days multiplied by 1,000. Relative reduction is calculated as baseline rate for the sample, minus the quarter of interest rate for the sample, divided by the baseline rate. All analyses were conducted using SAS version 9.3 (SAS Institute, Cary, North Carolina).