Lead/Presenter: Haley Appaneal,
COIN - Providence
All Authors: Appaneal HJ (Providence VAMC, Providence, RI, United States, University of Rhode Island, Kingston, RI, United States), Caffrey AR (Providence VAMC, Providence, RI, United States, University of Rhode Island, Kingston, RI, United States), Hughes MA (Providence VAMC, Providence, RI, United States) Lopes VV (Providence VAMC, Providence, RI, United States) Jump RLP (Cleveland VAMC, Cleveland, OH, United States) LaPlante KL (Providence VAMC, Providence, RI, United States, University of Rhode Island, Kingston, RI, United States) Dosa DM (Providence VAMC, Providence, RI, United States)
Objectives:
Microbiological cultures are critical in the diagnosis of infection, identification of pathogenic organisms, and tailoring antibiotic use. However, unnecessary collection of cultures, particularly from the urine, may lead to overuse of antibiotics. There have been no national studies to evaluate trends in the collection of cultures in acute and long-term care settings. Here we describe changes in the collection of cultures nationally across Veterans Affairs medical centers (VAMCs) and Community Living Centers (CLCs).
Methods:
All positive and negative cultures collected from 2010-2017 among Veterans admitted to VAMCs or CLCs were included. Cultures were categorized by specimen source (urine, blood, skin and soft tissue, or lung). Joinpoint Software was used for regression analyses of trends over time and to estimate annual average percent changes with 95% confidence intervals (CI).
Results:
A total of 5,089,640 cultures from 158 VAMCS and 342,850 cultures from 146 CLCs were identified. The number of cultures collected for all culture types in VAMCs and CLCs decreased significantly. The number of cultures collected per admission decreased significantly by 5.5% annually among VAMCs (95% CI -7.0 - -4.0%) and by 8.4% annually among CLCs (95% CI -10.1 - -6.6%). The proportion of positive cultures decreased 1.6% annually among VAMCs (95%CI -2.3 - -0.9%) and remained stable among CLCs (-0.4% annually, 95% CI, -1.1 to 0.4%). The most common culture source among VAMCs was blood (36.2%), followed by urine (31.8%), and among CLCs was urine (56.9%), followed by blood (16.0%). Urine cultures decreased by 4.5% annually among VAMCs (95% CI -5.4 - -3.6%) and 7.0% annually among CLCs (95% CI -7.6 - -6.4%).
Implications:
Our study demonstrates a significant reduction in the number of cultures collected over time. Positive cultures decreased significantly in VAMCs, possibly indicating fewer culture-positive infections.In both VAMCs and CLCs, decreases in cultures taken may represent an important reduction in the collection of unnecessary cultures nationally driven by increased awareness about over-testing and over-treatment of presumed infection, particularly urinary tract infections.
Impacts:
Further work is warranted to investigate if the reductions in cultures collected has led to less inappropriate antibiotic prescribing and improved Veteran care.