2 Clinical need and practice

The problem addressed

2.1 In current practice, people who are clinically unwell and who have a suspected bloodstream infection have empirically prescribed broad‑spectrum antibiotics, that is, antibiotics that are prescribed based on clinical presentation, until the identity of the pathogen causing the infection is known. Broad‑spectrum antibiotics and, if appropriate, antifungals, are used because they are effective against a wide range of bacterial and fungal pathogens and are likely to achieve a therapeutic response. But, although clinically effective, broad‑spectrum antibiotic use is associated with people developing superinfection and with antimicrobial resistance. Rapidly identifying the bacterial and fungal pathogen may allow earlier targeted treatment and shorten the length of use of broad‑spectrum antibiotics and antifungals, which may help antimicrobial stewardship by conserving the effectiveness of existing antimicrobials.

2.2 Three molecular tests, the LightCycler SeptiFast Test MGRADE, SepsiTest and IRIDICA BAC BSI assay, were identified during scoping as relevant to the assessment (see section 3 for additional details). These tests are designed to rapidly detect and identify bacterial and fungal DNA that may be in the bloodstream in people who are suspected of having sepsis. These tests are intended to be used with clinical assessment and established microbiology techniques that provide information on which antimicrobials are likely to be effective against the identified pathogen. The tests are designed to be run on whole blood samples and without the prior incubation or the pre‑culture steps that are needed for tests used in current standard practice. The absence of these steps means that pathogens may be identified earlier. It is possible that blood culture would still be needed to give definitive antimicrobial‑susceptibility data, if this is not provided by the rapid diagnostic test. The rapid detection and identification of bacterial and fungal DNA may be particularly beneficial in people who are suspected of having a severe infection and who need quick medical intervention.

2.3 The purpose of this assessment is to evaluate the clinical and cost effectiveness of using the LightCycler SeptiFast Test MGRADE, SepsiTest and IRIDICA BAC BSI assay for rapidly identifying bloodstream bacteria and fungi in the NHS.

The condition

Sepsis and bloodstream infection

2.4 Sepsis is a life‑threatening condition characterised by the body's inflammatory response to an infection. According to the Surviving Sepsis Campaign's International guidelines for the management of severe sepsis and septic shock, sepsis is diagnosed if there is evidence of systemic inflammation, in addition to a documented or presumed infection in the body. Systemic illness often happens if bacteria invade normally sterile parts of the body. One example of this is when bacteria or fungi invade the bloodstream (bloodstream infection); a process that often causes an inflammatory immune response.

2.5 Bacterial infections are the most common cause of sepsis and bloodstream infection, but they can also be caused by fungal infections, and less commonly by viral infections. The most common sites of infection associated with sepsis are the lungs, urinary tract, abdomen and pelvis. Other sources of infection leading to sepsis include skin infections (such as cellulitis), post‑surgical infections and infections of the nervous system (such as meningitis or encephalitis).

2.6 People who have recently been admitted to hospital are at risk of getting hospital‑acquired infections that can lead to sepsis and bloodstream infection. The increased use of invasive procedures, such as catheterisation and life support measures, as well as immunosuppressive therapy and antibiotic therapy may have resulted in more healthcare‑associated bloodstream infections. Community‑acquired bloodstream infections may also occur in people who have not had recent contact with healthcare services. The pathogens infecting these people may differ from those associated with hospital‑acquired bloodstream infection.

2.7 The bacteria most commonly associated with bloodstream infection in adults include gram‑negative species such as Escherichia coli, Klebsiella and Pseudomonas, and gram‑positive species such as Staphylococcus aureus, non-pyogenic streptococci, Enterococcus and Streptococcus pneumoniae. The types of pathogens causing bloodstream infection can differ in children compared with those causing infection in adults, and can include Neisseria meningitidis. Polymicrobial infection and anaerobic bacteraemia are also thought to occur less often in children.

The diagnostic and care pathways

Diagnosing sepsis and bloodstream infection

2.8 Diagnostic criteria for sepsis are listed in the Surviving Sepsis Campaign's International guidelines for the management of severe sepsis and septic shock. In summary, regular observations of all vital signs should be taken and recorded, kidney and liver function tests should be done, and inflammatory biomarkers and serum lactate should be measured. These guidelines state that a diagnosis of sepsis should be based on infection, documented or suspected, with hyperthermia or hypothermia, tachycardia and at least 1 indication of altered organ function.

2.9 The guidelines also make the following specific recommendations relating to detecting localised and bloodstream infection:

  • At least 2 samples for blood culture should be collected (aerobic and anaerobic) before antimicrobial therapy is started if such cultures do not cause significant delay (greater than 45 minutes) in the start of antimicrobial administration. At least 1 sample should be drawn percutaneously and 1 drawn through each vascular access device, unless the device was recently (less than 48 hours) inserted. The blood cultures can be drawn at the same time if they are taken from different sites. Cultures from other sites that may be the source of infection, such as urine, cerebrospinal fluid, wounds, respiratory secretions or other bodily fluids, should be collected before starting antimicrobial therapy, if doing so does not cause significant delay in the start of antimicrobial treatment.

  • Imaging studies such as CT or X‑ray should be done to confirm a potential source of infection.

  • Assays to diagnose systemic fungal infection should be used if available and invasive candidiasis is suspected.

Blood cultures

2.10 Public Health England's standards for the investigation of blood cultures are available. A blood culture set for diagnosing bloodstream infection is defined as 1 aerobic and 1 anaerobic bottle. For adults it is recommended that 20–30 ml of blood is cultured per set, and that 2 consecutive blood culture sets from 2 separate venepuncture sites should be collected during any 24‑hour period for each septic episode. The first set should be taken before starting antimicrobial treatment because the presence of antibiotics or antifungals may inhibit the growth of pathogens in blood culture. Blood culture sample collection differs for infants and neonates, for whom a single aerobic bottle or low‑volume blood culture bottle may be requested. The criterion for calculating total blood‑culture volume in neonates and children is based on weight rather than age and relates to total patient blood volume. It has been suggested that the volume of blood drawn should be no more than 1% of the patient's total blood volume. In infants and children, the level of bacteraemia is usually higher than in adults and so the sensitivity of detection is not thought to be substantially reduced by a lower blood‑to‑medium ratio.

2.11 Blood culture bottles should be incubated within 4 hours of the blood sample being taken. Many laboratories now use automated culture systems that alert laboratory staff once growth has been detected.

2.12 When a blood culture has been detected as positive it is recommended that:

  • Gram staining and rapid antigen testing should be done within 2 hours.

  • Direct or automated isolate identification should be done within 24 hours (extending to 48 hours if traditional microbiology techniques such as morphological identification are used). Rapid species identification may be done after blood culture using techniques such as MALDI‑TOF mass spectrometry.

  • Identification should be followed by sensitivity testing to determine the antimicrobials that the identified pathogen is susceptible to. If direct or automated sensitivity testing is used, a report should be made within 24 hours, extended to 48 hours if traditional techniques, such as the disc diffusion method, are used.

  • A preliminary positive report is made within 2 hours of identification and sensitivity testing, and a final positive report should be made within 5 days of the sample arriving in the laboratory.

2.13 If a blood culture is negative, it is recommended that a preliminary negative report is provided within 48 hours of the sample arriving in the laboratory and a final negative report should be issued within 5 days unless extended culture is being done, such as if fungi or unusual, fastidious or slow growing organisms are suspected.

Treating sepsis and bloodstream infection

2.14 Sepsis treatment varies based on the initial infection, the organs affected and the extent of tissue damage. The management of severe sepsis and septic shock is described in the Surviving Sepsis Campaign's International guidelines for the management of severe sepsis and septic shock.

2.15 The guidelines recommend that effective intravenous antimicrobials should be given within the first hour of recognising severe sepsis and septic shock. Initial empirical antimicrobial therapy should include 1 or more drugs that have activity against all likely pathogens (bacterial, fungal or viral) and that penetrate in adequate concentrations into the tissues thought to be the source of sepsis. Frequently used broad‑spectrum antibiotics for more serious infections include cephalosporins and aminoglycosides.

2.16 The guidelines recommend that the choice of empirical antimicrobial therapy be based on:

  • the patient's history, including drug intolerances

  • recent antibiotic treatments (previous 3 months)

  • underlying disease

  • the clinical syndrome

  • susceptibility patterns of pathogens in the community and hospital

  • previous microbiology reports identifying pathogens that have previously colonised or infected the patient.

2.17 Clinicians prescribing antimicrobial therapy should take into account the Department of Health's guidance on antimicrobial stewardship, which is based on the 'start smart then focus' strategy. The guidance recommends that when empirical antimicrobials are prescribed, the clinical diagnosis should be reviewed after 48 to 72 hours to allow an antimicrobial prescribing decision to be made. This decision should take into account available microbiology results to determine if therapy can be stopped or changed; that is, the de‑escalation, substitution or addition of antimicrobial agents to the treatment plan.

2.18 Narrowing the spectrum of antimicrobial coverage and shortening the duration of therapy may reduce the risk of a person developing a superinfection, and reduce treatment‑related adverse events. Adverse events associated with using broad‑spectrum antimicrobials may include diarrhoea, nausea, vomiting, hearing loss, damage to the kidneys and an increased risk of superinfection with Clostridium difficile. Narrowing the spectrum of antimicrobial coverage may also be associated with an increase in treatment efficacy in some scenarios.

2.19 Reducing the spectrum of antimicrobial coverage and duration of antibiotic therapy may also contribute to antimicrobial stewardship and protect the effectiveness of existing antibiotics. Surveillance data for England for the period 2010 to 2013 suggest that rates of methicillin‑resistant Staphylococcus aureus (MRSA) have fallen while the incidence of bloodstream infections caused by resistant gram-negative Enterobacteriaceae bacteria, such as Klebsiella and Escherichia coli, has increased (English surveillance programme for antimicrobial utilisation and resistance, 2014). Of particular concern in some regions of England is the increasing resistance to carbapenem antibiotics, which are often used as a last resort for treating severe infections when other antibiotics have not brought the infection under control.

  • National Institute for Health and Care Excellence (NICE)