Skip to content


Critical Care

Volume 12 Supplement 5

Sepsis 2008

Open Access

Automation of Septifast®for molecular diagnosis of infection in septic patients

  • Benito Regueiro1,
  • Eduardo Varela1,
  • Lucia Martinez Lamas1,
  • Antonio Santos2 and
  • Julian Alvarez3
Critical Care200812(Suppl 5):P12

Published: 18 November 2008


Blood CultureSystemic Inflammatory Response SyndromeSimultaneous DetectionClinical Risk FactorHigh Clinical Risk


Time is a critical issue in confirming infection in systemic inflammatory response syndrome patients with suspected sepsis. Early diagnosis, followed by prompt implementation of an appropriate treatment, improves the prognosis of these patients. The SeptiFast® test is a new multiplex PCR test for simultaneous detection of sepsis-relevant microorganisms. It technically consists of three well-defined phases: extraction, detection on the LightCycler 2.0 machine and data analysis. The first one is a manual procedure based on filtration that takes 3.5 to 4 hours. The whole procedure including detection and data analysis takes approximately 6 hours. The aim of our study was to automate the current protocol using MagNa pur® compact and to evaluate the potential advantages of reducing the time to perform the method.

Materials and study design

Inclusion criteria were age ≥18 years and suspected sepsis. The Ethics Committees from our institution approved the study, and patient or relatives' consent was obtained for blood sampling for PCR. Patient population (ICU): 100 Septifast® determinations and correlative blood cultures and microbiological cultures were performed in 64 patients with systemic inflammatory response syndrome and high clinical risk factors for bloodstream infection.


We used a new multiplex PCR-based test, LightCycler® SeptiFast® Test MGRADE (Roche Diagnostics, Penzberg, Germany), including software for simultaneous detection of sepsis-relevant microorganisms directly from blood, according to the manufacturer's recommendations. We followed the standard manufacturer's recommended procedure for extraction and compared it with the procedure using the MagNa pure® compact (Roche Applied Science, Penzberg, Germany) extraction column based on magnetic nanoparticles. Samples for blood culture analysis were drawn from a fresh venipuncture site according to the common guidelines (DGHM). Blood cultures were analyzed using a semi-automated blood system (BacT/ALERT®; bioMerieux, Marcy-Etoile, France). VITEK II (bioMerieux) was the system used to phenotypically identify pathogens growing from blood cultures and other microbiological tests (stool, urine, respiratory samples, catheter, etc.).


The results were both retrospective and observational. The quality of PCR results versus other laboratory findings and clinical evaluation, using sensitivity and specificity analysis and kappa evaluation to compare the results obtained, was determined. One hundred determinations were studied comparing both methods for extraction. Table 1 presents comparative data for both methods. For sensitivity and specificity analysis, the blood culture data define the disease condition (infection).
Table 1

Comparative values of Septifast® (manual extraction) versus Septifast® (MagNa Pure® compact extraction)


Blood culture vs. Septifast®

(manual extraction)

Blood culture vs. Septifast®

(MagNa Pure® compact extraction)


50 (27 to 72)

80 (56 to 94)


98.75 (93 to 99)

95 (87 to 98)

Positive predictive value

90.9 (58 to 99)

80 (56 to 94)

Negative predictive value

88.76 (80 to 94)

95 (87 to 98)




Likelihood ratio positive test

40 (5.43 to 294.5)

15.99 (6 to 43.63)

Likelihood ratio negative test

0.506 (0.326 to 0.78)

0.21 (0.087 to 0.5)

Proportion agreement (strength of agreement) [1]

0.89 (moderate)

0.92 (substantial)

Bias index



Prevalence index



Kappa (8)



Extraction volume

1,000 μl

400 μl

Time (extraction)

4 to 3.5 hours

25 to 30 minutes

95% confidence interval calculated with binomial expansion.


MagNa pure® extraction versus the standard procedure for extraction used in the SeptiFast® test reduces the total test time from 6 to 3 or 3.5 hours. Use of this alternative protocol does not affect the sensitivity or specificity of the method.

Authors’ Affiliations

Microbiologia, Hospital Clinico Universitario, Complejo Hospitalario Universitario de Santiago, Universidad de Santiago, Spain
Unidad Cuidados Intensivos, Hospital de Conxo, Complejo Hospitalario Universitario de Santiago, Universidad de Santiago, Spain
Anestesia y Reanimacion, Hospital Clinico Universitario, Complejo Hospitalario Universitario de Santiago, Universidad de Santiago, Spain


  1. Landis RJ, Koch GG: The measurement of observer agreement for categorical data. Biometrics. 1977, 33: 159-174. 10.2307/2529310.View ArticlePubMedGoogle Scholar


© Regueiro et al; licensee BioMed Central Ltd. 2008

This article is published under license to BioMed Central Ltd.