Standardized definition of contamination and evidence-based target necessary for high-quality blood culture contamination rate audit

Delayed Treatment of Bloodstream Infection at Admission is Associated With Initial Low Early Warning Score and Increased Mortality

OBJECTIVES

To identify factors associated with antibiotic treatment delay in patients admitted with bloodstream infections (BSIs).

DESIGN

Retrospective cohort study.

SETTING

North Zealand Hospital, Denmark.

PATIENTS

Adult patients with positive blood cultures obtained within the first 48 hours of admission between January 1, 2015, and December 31, 2015 (n = 926).

MEASUREMENTS AND MAIN RESULTS

First recorded Early Warning Score (EWS), patient characteristics, time to antibiotic treatment, and survival at day 60 after admission were obtained from electronic health records and medicine module. Presence of contaminants and the match between the antibiotic treatment and susceptibility of the cultured microorganism were included in the analysis. Data were stratified according to EWS quartiles. Overall, time from admission to prescription of antibiotic treatment was 3.7 (3.4-4.0) hours, whereas time from admission to antibiotic treatment was 5.7 (5.4-6.1) hours. A gap between prescription and administration of antibiotic treatment was present across all EWS quartiles. Importantly, 23.4% of patients admitted with BSI presented with an initial EWS 0-1. Within this group of patients, time to antibiotic treatment was markedly higher among nonsurvivors at day 60 compared with survivors. Furthermore, time to antibiotic treatment later than 6 hours was associated with increased mortality at day 60. Among patients with an initial EWS of 0-1, 51.3% of survivors received antibiotic treatment within 6 hours, whereas only 19.0% of nonsurvivors received antibiotic treatment within 6 hours.

CONCLUSIONS

Among patients with initial low EWS, delay in antibiotic treatment of BSIs was associated with increased mortality at day 60. Lag from prescription to administration may contribute to delayed antibiotic treatment. A more frequent reevaluation of patients with infections with a low initial EWS and reduction of time from prescription to administration may reduce the time to antibiotic treatment, thus potentially improving survival.

Diagnostic stewardship for blood cultures in the emergency department: A multicenter validation and prospective evaluation of a machine learning prediction tool

Background

Overuse of blood cultures (BCs) in emergency departments (EDs) leads to low yields and high numbers of contaminated cultures, accompanied by increased diagnostics, antibiotic usage, prolonged hospitalization, and mortality. We aimed to simplify and validate a recently developed machine learning model to help safely withhold BC testing in low-risk patients.

Methods

We extracted data from the electronic health records (EHR) for 44.123 unique ED visits with BC sampling in the Amsterdam UMC (locations VUMC and AMC; the Netherlands), Zaans Medical Center (ZMC; the Netherlands), and Beth Israel Deaconess Medical Center (BIDMC; United States) in periods between 2011 and 2021. We trained a machine learning model on the VUMC data to predict blood culture outcomes and validated it in the AMC, ZMC, and BIDMC with subsequent real-time prospective evaluation in the VUMC.

Findings

The model had an Area Under the Receiver Operating Characteristics curve (AUROC) of 0.81 (95%-CI = 0.78–0.83) in the VUMC test set. The most important predictors were temperature, creatinine, and C-reactive protein. The AUROCs in the validation cohorts were 0.80 (AMC; 0.78–0.82), 0.76 (ZMC; 0.74–0.78), and 0.75 (BIDMC; 0.74–0.76). During real-time prospective evaluation in the EHR of the VUMC, it reached an AUROC of 0.76 (0.71–0.81) among 590 patients with BC draws in the ED. The prospective evaluation showed that the model can be used to safely withhold blood culture analyses in at least 30% of patients in the ED.

Interpretation

We developed a machine learning model to predict blood culture outcomes in the ED, which retained its performance during external validation and real-time prospective evaluation. Our model can identify patients at low risk of having a positive blood culture. Using the model in practice can significantly reduce the number of blood culture analyses and thus avoid the hidden costs of false-positive culture results.

Funding

This research project was funded by the Amsterdam Public Health – Quality of Care program and the Dutch “Doen of Laten” project (project number: 839205002).

Using machine learning to predict blood culture outcomes in the emergency department: a single-centre, retrospective, observational study

OBJECTIVES

To develop predictive models for blood culture (BC) outcomes in an emergency department (ED) setting.

DESIGN

Retrospective observational study.

SETTING

ED of a large teaching hospital in the Netherlands between 1 September 2018 and 24 June 2020.

PARTICIPANTS

Adult patients from whom BCs were collected in the ED. Data of demographic information, vital signs, administered medications in the ED and laboratory and radiology results were extracted from the electronic health record, if available at the end of the ED visits.

MAIN OUTCOME MEASURES

The primary outcome was the performance of two models (logistic regression and gradient boosted trees) to predict bacteraemia in ED patients, defined as at least one true positive BC collected at the ED.

RESULTS

In 4885 out of 51 399 ED visits (9.5%), BCs were collected. In 598/4885 (12.2%) visits, at least one of the BCs was true positive. Both a gradient boosted tree model and a logistic regression model showed good performance in predicting BC results with area under curve of the receiver operating characteristics of 0.77 (95% CI 0.73 to 0.82) and 0.78 (95% CI 0.73 to 0.82) in the test sets, respectively. In the gradient boosted tree model, the optimal threshold would predict 69% of BCs in the test set to be negative, with a negative predictive value of over 94%.

CONCLUSIONS

Both models can accurately identify patients with low risk of bacteraemia at the ED in this single-centre setting and may be useful to reduce unnecessary BCs and associated healthcare costs. Further studies are necessary for validation and to investigate the potential clinical benefits and possible risks after implementation.

Pathogens distribution and antimicrobial resistance in bloodstream infections in twenty-five neonatal intensive care units in China, 2017-2019

BACKGROUND

Overcrowding, abuse of antibiotics and increasing antimicrobial resistance negatively affect neonatal survival rates in developing countries. We aimed to define pathogens and their antimicrobial resistance (AMR) of early-onset sepsis (EOS), hospital-acquired late-onset sepsis (HALOS) and community-acquired late-onset sepsis (CALOS) in 25 neonatal intensive care units (NICUs) in China.

STUDY DESIGN

This retrospective descriptive study included pathogens and their AMR from all neonates with bloodstream infections (BSIs) admitted to 25 tertiary hospitals in China from January 1, 2017, and December 31, 2019. We defined EOS as the occurrence of BSI at or before 72 h of life and late-onset sepsis (LOS) if BSI occurred after 72 h of life. LOS were classified as CALOS if occurrence of BSI was ≤ 48 h after admission, and HALOS, if occurrence was > 48 h after admission.

RESULTS

We identified 1092 pathogens of BSIs in 1088 infants from 25 NICUs. Thirty-two percent of all pathogens were responsible for EOS, 64.3% HALOS, and 3.7% CALOS. Gram-negative (GN) bacteria accounted for a majority of pathogens in EOS (56.7%) and HALOS (62.2%). The most frequent pathogens causing EOS were Escherichia coli (27.2%) and group B streptococcus (GBS; 14.6%) whereas in CALOS they were GBS (46.3%) and Staphylococcus aureus (41.5%). Klebsiella pneumoniae (27.9%), Escherichia coli (15.7%) and Fungi (12.8%) were the top three isolates in HALOS. Third-generation cephalosporin resistance rates in GN bacteria ranged from 9.7 to 55.6% in EOS and 26% to 63.3% in HALOS. Carbapenem resistance rates in GN bacteria ranged from 2.7 to 31.3% in HALOS and only six isolates in EOS were carbapenem resistant. High rates of multidrug resistance were observed in Klebsiella pneumoniae (60.7%) in HALOS and in Escherichia coli (44.4%) in EOS. All gram-positive bacteria were susceptible to vancomycin except for three Enterococcus faecalis in HALOS. All-cause mortality was higher among neonates with EOS than HALOS (7.4% VS 4.4%, [OR] 0.577, 95% CI 0.337-0.989; P = 0.045).

CONCLUSIONS

Escherichia coli, Klebsiella pneumoniae and GBS were the leading pathogens in EOS, HALOS and CALOS, respectively. The high proportion of pathogens and high degree of antimicrobial resistance in HALOS underscore understanding of the pathogenesis and emphasise the need to devise effective interventions in developing countries.

Factors associated with the risk of positive blood culture in neonatal foals presented to a referral center (2000-2014)

Background

Bloodstream infections (BSI) are common in sick foals and increase foal morbidity and mortality when they occur. Recognition of risk factors for BSI could be an important means to limit their occurrence, but studies on this topic are limited.

Objectives

Historical as well as maternal and foal physical examination findings will predict risk of BSI in neonatal foals.

Animals

Foals <14 days of age admitted to a referral equine hospital for care.

Methods

Retrospective case‐control study with univariate and multivariable logistic regression analysis.

Results

Four hundred twenty‐nine (143 cases and 286 controls) foals <14 days of age were studied. Risk of a foal having a BSI was increased in foals with umbilical disease (adjusted odds ratio [OR], 11.01; P = .02), hypoglycemia (adjusted OR, 13.51; P = .03), and the combined presence of umbilical disease and low hematocrit (adjusted OR, >999.99; P = .04). Factors not found to be risk factors for development of BSI included prematurity, hypothermia, abdominal disease, diarrhea, failure of passive transfer, and maternal uterine infection.

Conclusions and Clinical Importance

Several historical and physical examination findings increase the risk of foals being blood culture positive at presentation to the hospital. This knowledge may aid early identification of blood culture status, thus aiding in treatment decisions.

Blood culture contamination in hospitalized pediatric patients: a single institution experience

PURPOSE

Blood culture is the most important tool for detecting bacteremia in children with fever. However, blood culture contamination rates range from 0.6% to 6.0% in adults; rates for young children have been considered higher than these, although data are limited, especially in Korea. This study determined the contamination rate and risk factors in pediatric patients visiting the emergency room (ER) or being admitted to the ward.

METHODS

We conducted a retrospective chart review of blood cultures obtained from children who visited Yonsei Severance Hospital, Korea between 2006 and 2010. Positive blood cultures were labeled as true bacteremia or contamination according to Centers for Disease Control and Prevention/National Healthcare Safety Network definitions for laboratory-confirmed bloodstream infection, after exclusion of cultures drawn from preexisting central lines only.

RESULTS

Among 40,542 blood cultures, 610 were positive, of which 479 were contaminations and 131 were true bacteremia (overall contamination rate, 1.18%). The contamination rate in the ER was significantly higher than in the ward (1.32% vs. 0.66%, P<0.001). The rate was higher in younger children (2.07%, 0.94%, and 0.61% in children aged <1 year, 1-6 years, and >6 years, respectively).

CONCLUSION

Overall, contamination rates were higher in younger children than in older children, given the difficulty of performing blood sampling in younger children. The contamination rates from the ER were higher than those from the ward, not accounted for only by overcrowding and lack of experience among personnel collecting samples. Further study to investigate other factors affecting contamination should be required.