10 Years after NNIS; the Use of Comparative Data as a Catalyst for Organizational Improvement

Abstract

BACKGROUND/OBJECTIVES: Surveillance data is used to drive interventions with the ultimate goal of improving both care processes and the outcomes associated with those processes. However, barriers surrounding such data are based on the acceptance of rates and outcomes as valid by care providers and administration. In 1995, our organization began to collect data using the National Nosocomial Infection Surveillance System [NNIS] methodologies. Our analysis objective was to determine the extent to which improvement in patient outcomes occurred during the subsequent 10 year period and the role in which the use of comparative data served as the catalyst for improvement projects. METHODS: We collected ongoing device associated infection data in 3 Intensive care units over a 10 year period. Infection rates and comparative data were provided to clinical staff, physicians, administration and the governing board on a regular basis. Based on the analysis, priorities were set and targeted interventions planned. RESULTS: Over the 10 year period, device associated infections decreased by statistically significant differences in all 3 intensive care units. The most significant decrease was in Central Venous Catheter Bloodstream Infections [CVC BSI]. Of note, the Medical Intensive Care CVC BSI rate decreased from 11.2 per 1,000 line days in 1995 to 4.3 in year one [ P = .001]. The 2005 rate was 0.75 per 1,000 line days with no CVC BSI'S for 6 consecutive months. Expansion of these efforts led to an organization-wide prevention program resulting in a 75% further reduction in CVC BSI'S over 3 years, attributable mortality which dropped from 4% to zero and cost-avoidance of over 1 million dollars. Rates of ventilator associated pneumonia were at the 75th to 90th percentile in 1995. The 2004 rate decreased to between the 10th and 25th percentile in 2 intensive care units, and to below the 10th percentile in the Cardiothoracic unit. Less dramatic decreases were identified in the rates of Urinary Tract Infections [UTI], where only the Cardiothoracic Intensive Care Unit rates were statistically significantly lower than the NNIS median. However, UTI rates in all units were significantly lower than baseline data. CONCLUSIONS: Although the reason for these decreases may be multifactoral, disseminating risk-adjusted, reliable infection rates to care providers is an essential part of a good surveillance program and may be the catalyst for organizational improvement projects. The NNIS program, now NHSN provides a framework for such activities.