Incidence of ICU-Bacteremia Is Decreasing

The results of a new analysis have revealed that incidence of intensive care unit (ICU) bacteremia is decreasing over time. The study was presented at the 47th Critical Care Congress, which took place February 25 to 28, 2018, in San Antonio, Texas.

According to the Centers for Disease Control and Prevention (CDC), between 2008 and 2014, there was a 50% decrease in central line-associated bloodstream infections (CLABSIs) in the United States. Furthermore, long-term acute care hospitals saw a 9% decrease in CLABSIs.

In 2008, the Centers for Medicare & Medicaid Services began to tie reimbursement with incidence of CLABSIs and other hospital-acquired conditions as an incentive to reduce their rates. Trends can be seen between incidence of CLABSIs and blood culture ordering practices and so investigators from the National Institutes of Health, Harvard Medical School, and Commonwealth Informatics, Inc. set up a study to better understand these relationships.

Using Cerner Healthfacts database data from 2009 through 2013, the investigators identified “ICU stays spanning ≥3 days in adult (≥ 20 years) patients admitted to US hospitals,” according to the study. The investigators used ICD-9 codes listed in the database to identify those patients who were given a central venous catheter (CVC). Microbiology data was used to identify those patients who had blood culture (BC) orders as well as whether or not a bloodstream infection (BSI) was present.

“CVC stays (CVC placed ± 1 day of ICU admission) were 1:1 matched to non-CVC stays (no CVC during hospitalization) on institution, age, admit year and ICU type, ICU admission Sequential Organ Failure Assessment (SOFA) score and ICU length of stay (LOS),” the study authors wrote.

A case of CLASBI was identified as a CVC case having 1 or more positive non-contaminant BC between day 3 of the patient’s ICU stay and discharge. Additionally, according to the authors, “adjusting for BC ordering trends over time, the adjusted Annual Percent Change (aAPC; [95%CI]) in CLABSI vs non-CVC BSI cases over the 5-year period was calculated using Posisson regression separately for all hospitals and those reporting data for all 5 years.”

A total of 9,022 matched-encounter pairs were identified across 63 hospitals over 5 years out of a total of 10,599 overall CVC cases and 35,790 non-CVC cases. The matched-encounter patients had a median age of 65 (54,76], a SOFA score of 4 [2,7], and LOS in ICU of 6 days [4,9]. Septicemia, acute respiratory failure, and pneumonia were the most frequent diagnoses among both matched-CVC and non-CVC patients.

The frequency of BC ordering “decreased comparably in CVC (APC = -4.1% [-5.9, -2.3]) and non-CVC cases (APC = -4.4% [-6.3, -2.5]); P = .81,” according to the study authors. “After adjusting for the changing trend in BC ordering, CLABSI incidence decreased over time (aAPC = -5.8% [-8.0, -3.6]), albeit at a rate that was comparable to non-CVC BSI (aAPC = -5.8% [-8.1, -3.4]); P = .97. Similar trends in CLABSI (aAPC = -5.5% [-8.0, -3.0]) and non-CVC BSI (aAPC = -4.7% [-8.1, -1.3]; P = .77) were observed at 17 continuously reporting hospitals as well.”

Although the reasoning why fewer cultures are being drawn in ICU patients is not known, the authors found a decrease across the board, in both CVC and non-CVC patients. Likewise, comparable decreases were seen in CLABSIs and non-CVC BSIs.