Bacteremia Detection in Second or Subsequent Blood Cultures Among Hospitalized Patients in a Tertiary Care Hospital

Few data exist on the yield and utility of additional blood culture (BC) sets during extravascular infections. The low yield and false-positivity rate owing to contaminants during sample obtainment or handling¹ support optimizing patient management and resource use. We assessed the clinical relevance of collecting an additional BC set while the first set is still incubating with results pending.

This retrospective cohort study analyzed all BCs performed at a tertiary care hospital from January 1, 2019, through January 1, 2020, among adult patients with at least 2 BC sets collected during hospitalization, with the first collected in the emergency department within 24 hours after admission. Hematopoietic stem cell transplant recipients were excluded. The ethics committee of Geneva approved this study and waived informed consent owing to the retrospective design. We followed the STROBE reporting guideline.

We defined the first BC set as the first 2 bottles collected. Additional bottles were collected at least 24 hours later. Microbiological time to positivity (mTTP) was the time from start to end of incubation for a sample with a positive BC result (eFigure in the Supplement). The primary outcome was the likelihood of a positive BC result or detection of bacteremia with additional sets while the first was incubating after 24 hours, assessed using conditional probabilities for independent events (eMethods in the Supplement). Logistic regression and the Wald test assessed factors associated with bacteremia (secondary outcome). Kruskal-Wallis and χ² tests compared continuous and categorical variables, respectively. Significance was set at 2-sided α = .05. Analyses were performed with R, version 4.1.0.

Among 23 088 BC bottles (2863 unique patients, 3214 care episodes) (Figure), the positivity rate was 8.34% (95% CI, 7.90%-8.70%), with 0.29% (95% CI, 0.28%-0.30%) of positive results owed to contaminants (Table). mTTP was 24 hours or less in 76.8% (95% CI, 74.8%-78.6%) of positive BC results, with mostly gram-negative bacilli and anaerobic microorganisms (Figure). Of 446 BC sets with mTTP longer than 24 hours, 317 (71.1%) were first sets (contaminants, 14.5%; 95% CI, 11.4%-17.9%) and 129 (28.9%) additional sets. In the latter, by care episode, the most common microorganisms were Staphylococcus aureus (34.8%; 95% CI, 22.7%-49.2%) and Staphylococcus epidermidis (8.7%; 95% CI, 3.4%-20.3%) and diagnoses were endovascular (58.7%; 95% CI, 44.3%-71.7%) and osteoarticular (19.6%; 95% CI, 10.7%-33.2%) infections. The probability of detecting bacteremia with additional BC sets was 4.1% (95% CI, 3.9%-4.4%) and 2.6% (95% CI, 2.4%-2.8%) when excluding contaminants and BC sets collected for endovascular infection. Female sex (odds ratio [OR], 0.34; 95% CI, 0.18-0.63; P < .001), hourly mTTP increment (OR, 0.98; 95% CI, 0.97-0.99; P = .005), sampling from a catheter (OR, 0.22; 95% CI, 0.10-0.47; P < .001), and growth of gram-positive microorganisms (OR, 0.04; 95% CI, 0.02-0.13; P < .001) were associated with lower odds for bacteremia.

The probability of detecting bacteremia with additional BC sets while the first was incubating after 24 hours was 4.1% (2.6% when excluding endovascular infections). The contamination rate was consistent with our institution’s rate for bloodstream infections. This quality-of-care indicator reflects adherence to the measures implemented in our institution for BC sample collection. Similar to other studies,^2-4 most BCs (76.8%) had an mTTP of 24 hours or less. Preanalytical time was significantly higher for positive BC results with an mTTP greater than 24 hours.

A limitation was the retrospective design, which did not allow assessment of associations with outcomes. The review and extraction of data from medical records cannot exclude report and misclassification bias.

Campaigns such as Choosing Wisely and Smarter Medicine⁵ advocate avoiding unnecessary tests for patients, and medical resource preservation was highlighted by the COVID-19 pandemic, which challenged hospital and laboratory capacities.⁶ These findings support development of novel guidelines for BC sample collection to improve diagnostic resource use and patient management and to reduce costs and support rapid transportation and incubation of BCs to improve bacteremia diagnosis.

Accepted for Publication: March 02, 2022.

Published: April 20, 2022. doi:10.1001/jamanetworkopen.2022.8065

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Zanella MC et al. JAMA Network Open.

Corresponding Author: Marie-Céline Zanella, MD, Bacteriology Laboratory, Department of Diagnostics, Geneva University Hospitals, 4 Rue Gabrielle-Perret-Gentil, 1211 Geneva 14, Switzerland (marie-celine.zanella@hcuge.ch).

Author Contributions: Drs Zanella and de Lorenzi-Tognon contributed equally to this work. Drs Zanella and de Lorenzi-Tognon had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Zanella, de Lorenzi-Tognon, Schrenzel.

Acquisition, analysis, or interpretation of data: Zanella, de Lorenzi-Tognon, Fischer, Vernaz.

Drafting of the manuscript: Zanella, de Lorenzi-Tognon.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: de Lorenzi-Tognon.

Administrative, technical, or material support: Zanella, de Lorenzi-Tognon, Fischer, Vernaz.

Supervision: Schrenzel.

Conflict of Interest Disclosures: None reported.

Additional Contributions: Antoine Poncet, MS, University Hospitals of Geneva, contributed to the statistical methods, and Rosemary Sudan, Geneva University Hospitals, provided editorial assistance. Both received compensation.