Development and validation of a standardized blood culture contamination definition and metric dashboard for a large health care system

• Published in: American journal of clinical pathology Vol. 160; no. 3; pp. 255 - 260
• Main Authors: Hancock, Jill A, Campbell, Sheldon, Jones, Makoto M, Wang-Rodriguez, Jessica, Klutts, J Stacey
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 01.09.2023
• Subjects: Automation; Blood; Blood culture; Codes; Contamination; Data collection; Data entry; Data warehouses; Etiology; Health care; Health care policy; Information systems; Medical examination; Medical laboratories; Medicine; Organisms; Pathogens; Pathology; Patient safety; Quality control; Quality improvement; Queries; Structured Query Language-SQL; Topography; SQL data queries; Power BI data dashboards; standardized health care data; blood culture contamination; quality improvement data
• ISSN: 0002-9173; 1943-7722
• DOI: 10.1093/ajcp/aqad044

Abstract Objectives Blood culture contamination is a major problem in health care, with significant impacts on both patient safety and cost. Initiatives to reduce blood culture contamination require a reliable, consistent metric to track the success of interventions. The objective of our project was to establish a standardized definition of blood culture contamination suitable for use in a Veterans Health Administration (VHA) national data query, then to validate this definition and query. A secondary objective was to construct a national VHA data dashboard to display the data from this query that could be used in VHA quality improvement projects aimed at reducing blood culture contamination. Methods A VHA microbiology expert work group was formed to generate a standardized definition and oversee the validation studies. The standardized definition was used to generate data for calendar year 2021 using a Structured Query Language data query. Twelve VHA hospital microbiology laboratories compared the data from the query against their own locally derived contamination data and recorded those data in a data collection worksheet that all sites used. Data were collated and presented to the work group. Results More than 50,000 blood culture accessions were in the validation data set, with more than 1,200 contamination events. The overall blood culture contamination rate for the 12 facilities participating was 2.56% with local definitions and data and 2.43% with the standardized definitions and data query. The main differences noted between the 2 data sets were deemed to be issues in local definitions. The query and definition were then converted into a national data dashboard that all VHA facilities can now access. Conclusions A standardized definition for blood culture contamination and a national data query were validated for enterprise-wide VHA use. To our knowledge, this represents the first reported standardized, validated, and automated approach for calculating and tracking blood culture contamination. This tool will be key in quality initiatives aimed at reducing contamination events in VHA.