Babies and children have an increased risk of getting an infection with a bacteria in the bloodstream (sepsis). It is often difficult for the doctor to determine whether a child has an infection of the bloodstream, because the symptoms are often unclear and can also occur in children who are not sick. To determine whether there is an infection, a little blood is currently taken for a blood test (the blood culture) to investigate whether there is a bacteria in the blood. However, it often takes at least 36 hours before the results of this blood culture are available. That is why antibiotics are usually started immediately to treat the possible infection.

However, it often turns out that the blood culture is negative after 36 hours, which means that no bacteria have been found in the blood. Usually the antibiotics are then stopped because it turns out that there was no infection at all. There is currently no good test that can predict whether (newborn) children have an infection or not. That is why too many children are currently wrongly receiving antibiotics. These antibiotics can damage the healthy bacteria in the intestines. There are many billions of 'beneficial bacteria' in the intestine. These play an important role in the digestion of food and protect against external infections. Antibiotics aim to kill bacteria that cause inflammation or infection. Unfortunately, antibiotics also kill some of these beneficial bacteria. In addition, unnecessary use of antibiotics contributes to antibiotic resistance. The aim of this research is to investigate whether Molecular Culture, a PCR based test that can identify bacterial pathogens in bodily fluids within 4 hours, has greater accuracy than traditional culturing techniques for bacteria in blood. If proven, this could lead to faster identification or exclusion of sepsis in children.

Newborns and children often receive antibiotics for a suspicion of sepsis. Sepsis has high morbidity and mortality in newborns and children. Up to 50% of children in the Netherlands are prescribed at least one course of antibiotics in the first 4 years of life. Postnatally, antibiotics are often prescribed for presumed bacterial infections on neonatal and pediatric wards, but in approximately 30% of these patients bacterial infection is not proven.

Rapid diagnosis of sepsis in newborns and children is problematic because clinical signs start subtle and are non-specific. The gold standard for diagnosis of bacterial sepsis is a conventional blood culture. Unfortunately, bacterial culture is time-consuming (time to result up to 36-72 hours) and lacks sensitivity in this population for sepsis. For this reason, children and infants with risk factors for infection or clinical signs and symptoms of infection are treated with antibiotics empirically at initial sepsis suspicion, awaiting results of the conventional blood culture. Currently more than 85% of very preterm born infants (gestational age <30 weeks) receive antibiotics for the risk of early-onset sepsis (EOS) in the Netherlands and approximately two-third are evaluated at least once for late-onset sepsis (LOS) . Also older children are often prescribed antibiotics empirically for presumed sepsis, awaiting results of the blood culture.

There is increasing evidence that, apart from antibiotic resistance, the use of antibiotics in the neonatal period and during childhood alters the microbiome with an increased risk of immediate and long-term adverse effects, such as increased risk for asthma, obesity, allergies and inflammatory bowel diseases (IBD). To avoid unnecessary treatment of non-infected children, an early, quick, sensitive and specific laboratory test would be helpful to guide clinicians to decide when to discontinue antibiotics as soon as possible.

Another technique to detect neonatal sepsis quickly and in an earlier stage compared to conventional blood culture is Molecular Culture (MC). MC is a rapid molecular based culturing technique that is able to identify bacteria within 4 hours after blood sampling. In short, MC is a DNA-based profiling technique, differentiating between bacterial species based on species-specific differences in 16S-23S rDNA interspacer (IS) region nucleotide length by using phylum-specific fluorescently labelled polymerase chain reaction (PCR) primers. A standard MC procedure consists of two separate PCRs. In the first PCR two different primers are added, one primer fluorescently labeling members of the phyla Firmicutes, Actinobacter, Fusobacteria and Verrucomicrobia (FAFV), whereas the second primer labels members of the Bacteroidetes phylum. In the second PCR, a third labeled primer is added targeting members of the phylum Proteobacteria. Subsequently, these PCR products can be amplified and DNA fragments can be separated based on their nucleotide length. Eventually, a typical MC microbial profile will be created, consisting of a set of color-labeled peaks. Each peak representing an individual bacterial operational taxonomic unit (OTU) depending on the nucleotide length of the IS fragment, length of these peaks demonstrating the concentration of this particular OTU, whereas peak colors provides information about the present phyla (FAFV, Bacteroidetes or Proteobacteria).

In bacterial sepsis, a bacterium has reached the otherwise sterile bloodstream causing sepsis as a result of dysregulation of the host immune response and/or a reaction of bacterial endotoxins. Both blood culture and MC can detect bacteria and give a positive result (in case a bacterium has been cultured) or a negative result (in case no bacteria was detected). In case the tests are positive, both techniques also show which bacteria was found, which could be used to change antibiotic regimen to target that specific cultured bacterium. However, the process of MC can be finished within 4 hours, which is much shorter than the incubation period of the gold standard blood culture which is 36-72 hours. When there is no sepsis (and thus no bacteria in the bloodstream), the MC will turn out negative also within 4 hours and thus may guide clinicians to stop antibiotics in uninfected children much faster compared to the conventional blood culture.

Blood cultures are still gold standard, but are generally assumed to have a low sensitivity for the diagnosis of sepsis in newborns and children and are time consuming. Cases of sepsis may be missed by cultures and a more sensitive diagnostic test such as molecular tests as the MC may be useful. Advances in microbial technology have led to the development of rapid molecular methods such as MC, that may be more sensitive than culture. Multiple novel molecular techniques, such as quantitative PCR, broad range conventional PCR and multiplex PCR, have been studied to detect neonatal sepsis. However, these techniques are directed at specific species which makes it impossible to detect all bacterial species. Thus, anything that is not explicitly searched for will be missed. In contrast, MC has the ability to detect every bacterial species that can cause bacterial neonatal sepsis.

The MC technique is validated to detect these pathogenic bacteria in bodily fluids. A series of papers validating all aspects of the MC technique have been published last years. Furthermore, MC is being used in hospitals on the intensive care unit to detect bacteria in otherwise sterile specimens like blood, but also on samples obtained from abscesses. A proof of principle study on 39 neonates, suspicious for EOS, of whom additional blood samples were taken from both umbilical cords as well as from peripheral phlebotomy, showed that MC was able to detect a pathogenic bacterial strain that was highly likely the causative organism for sepsis in one infant that was clinically ill. Conventional culture did not yield any results for this patient. 2 additional MC samples showed strains that were likely to be contaminants in infants that were clinically well-appearing, were conventional culture remained negative. No other discrepancies were seen. This study also showed that MC on blood samples that were spiked with prevalent bacterial strains for neonatal sepsis, showed a very high agreement with quantitative PCR as a control diagnostic.Larger studies are needed to corroborate diagnostic accuracy given the very low incidence of culture proven sepsis.

Results from another study using the MC technique to detect bacteria in human body fluid are promising. In this study 66 samples were collected and tested by conventional culture and MC. In 100% of samples with a positive culture, the MC was also positive. In five samples, the conventional culture turned out to be negative, whereas the MC was positive. The case histories of these five patients were obtained and suggested that the MC findings were highly clinically relevant, and thus may have higher sensitivity compared to conventional blood culture.

In summary, the quick detection makes MC potentially better equipped to guide clinical decision making for management of sepsis in newborns and children. However, the suitability of MC in this specific population has not been investigated adequately. Therefore, a high quality study should be performed to determine the diagnostic accuracy of the MC for sepsis in neonates and children and whether this technique may replace the conventional blood culture.

Inclusion Criteria:

• Undergoing collection of blood for a conventional blood culture as part of standard care OR
• Having undergone sepsis evaluation collection of blood for a conventional blood culture as part of standard care in the past 72 hours

Exclusion Criteria:

• Apart from an age criterion, there are no strict exclusion criteria. However, for the analysis of the secondary outcome (I.e. the testing of diagnostic accuracy of both MC as well as conventional culture for clinical sepsis), we plan to exclude all children who ultimately have a clear alternative cause for clinical illness that does not directly result from bacteraemia or bacterial sepsis. This will remain true in the case of conventional culture positivity, either when considered a contaminant as well as when considered a contributing factor in the presence of any of the causes of clinical illness mentioned below. A potential subject who meets any of the following criteria will be excluded from participation in this study These causes include, but are not limited to:
• In case of the potential inclusion of a neonate suspicious for EOS, confirmed congenital infection with TORCHES (toxoplasmosis, rubella, cytomegalovirus, syphilis and herpes) will lead to exclusion particularly for the neonatal population
• Auto inflammatory disease
• Hemophagocytic syndrome
• SIRS (Systemic Inflammatory Response Syndrome following a severe viral infection

All minors aged <18 years