Precisión diagnóstica de extracción única de sangre en el diagnóstico de bacteriemia, comparada con la técnica estándar de extracción múltiple en pacientes con sepsis, Bogotá Colombia Thesis

short description

  • Postgraduate thesis

Thesis author

  • Camilo Andres, Rodriguez Garcia
  • Juan Camilo, Motta Valencia

external tutor

  • Arango Duque, Alvaro Ignacio


  • Introduction Sepsis is an uncontrolled complete-body response to an infection and bacteremia is the viable bacteria in the bloodstream. Mortality for those conditions is around 50% of the cases. Collecting blood cultures for an accurate identification of bacteria, allow physicians to early onset of a specific antibiotic treatment that can reduce mortality. The standard technique for collecting blood cultures requires three venopunctures (10 ml of blood each) with 20 to 30 minutes intervals between each blood-draw. Multiple venopunctures result uncomfortable for patients, requires longer times for laboratory personnel to completed samples and increase contamination probabilities. It has been suggested that an unique blood-draw of 30 ml may have similar diagnostic performance than the standard technique. Materials and methods A diagnostic test study was conducted at Fundación Cardioinfantil, Bogotá, Colombia, between April 2019 and May 2020. The Institutional Ethics Committee approved the study. A sample size of 118 patients with sepsis was calculated. All patients underwent three blood-draws (in 2, 10ml were collected for the standard technique, and in another venipuncture, 40 ml were collected (10 ml for the standard technique and 30ml for the unique blood culture technique). The blood-draw in which the 40 ml were collected the distribution of the 40ml for both methods (10ml and 30ml, respectively) were randomly decided. Five blood cultures were not considered for analysis because contamination was reported (3 obtained using the standard technique, one collected using the unique blood culture technique, and 1 in both methods). Sensitivity, specificity, predictive values (PPV and NPV), likelihood ratios (LR+ and LR-), and the Area Under the ROC Curve (AUC) were estimated. Results Most patients were men (60.2%), and blood samples were mainly collected at the emergency department (49.2%). Bacteremia prevalence in the study population was 36.0% (95%CI: 27.0%;45.9%). The following diagnostic performance measurements were obtained for the unique blood culture technique: Sensitivity: 97.6% (87.1%;99.9%); Specificity: 95.8% (88.3%;99.1%); PPV: 93% (80.90%;98.5); NPV: 98.60% (92.3%;100), and AUC: 0.967 (0.934;1) Discussion Unique blood culture collection (30ml) had a sensitivity of 97% for the diagnosis of bacteremia in patients with sepsis when compared to the standard blood cultures technique (3 venipunctures, 10 ml each). The unique blood-drawn showed high sensitivity and specificity and in addition, this alternative reduces patients’ pain and discomfort, reduces overall time for laboratory personnel to collect samples, in consequence, physicians get culture reports earlier and they start with an accurate antibiotic’s administration.

publication date

  • November 23, 2020 10:34 PM


  • Bacteremia
  • Bacteria in the bloodstream
  • Blood contamination
  • Blood culture
  • Comparison of diagnostic techniques
  • Diagnostic test study
  • Single extraction

Document Id

  • b54ae120-1631-4237-803f-e87e84664aff