Target condition being diagnosed

Sepsis, which consists of systemic inflammatory response to an infection, is increasing in incidence (Bone 2009; Martin 2012). There was more than a double‐fold rise in the hospitalization rate, from 11.6 to 24.0 per 10,000 population between 2000 and 2008, for people primarily diagnosed with sepsis. This increase may be due to increasing antimicrobial resistance, greater use of invasive medical procedures and immunosuppressive drugs, and the increasing elderly population (Martin 2012). The sepsis spectrum: sepsis, severe sepsis and septic shock, is a leading cause of mortality in critically ill people (Angus 2001). The number of deaths from severe sepsis may be equivalent to, or surpass, those from cancer, stroke or acute myocardial infarction (Angus 2001). The likelihood of in‐hospital mortality is at least eight times higher in people with sepsis than in people with other medical conditions (Hall 2011). The average length of stay in hospital for people who are hospitalized for sepsis is 75% longer than for those hospitalized for other diagnoses (Hall 2011).

Early signs of sepsis are quite variable and non‐specific, making the routine diagnosis of sepsis challenging. Delay in diagnosing sepsis further worsens the outcome in people with sepsis. The challenge lies in the immediate and accurate distinction of sepsis from the non‐infectious systemic inflammatory response syndrome (SIRS) which may occur in critically ill people. In adults, SIRS consists of a core temperature over 38^oC or lower than 36^oC, a respiratory rate over 20 breaths per minute or PCO₂ (partial pressure of carbon dioxide in the blood) below 32 mmHg, a pulse rate over 90 beats per minute and a leucocyte count less than 4000/mm³ or more than 12,000/mm³ or more than 10% band cells (Bone 2009). In children, the cut‐off values for the four criteria of temperature, pulse rate/heart rate, respiratory rate and leucocyte count vary for different age groups (Goldstein 2005). Additionally, the confirmation of bloodstream infections (which may arise from bacterial, fungal, viral or parasitic origins) by culture, is positive in only about 30% to 50% of sepsis cases (Murray 2012). In 1991, a consensus panel of the American College of Chest Physicians (ACCP) and the Society of Critical Care Medicine (SCCM) provided a practical framework for the definition of sepsis (Bone 2009). The panel defined sepsis as the presence of two or more SIRS criteria, with documented or suspected infection. However, in 2003, an international consensus panel of the SCCM, the European Society of Intensive Care Medicine (ESICM), the ACCP, the American Thoracic Society (ATS) and the Surgical Infection Society (SIS), having noted that the SIRS criteria published in 1992 were unduly sensitive and non‐specific, provided an expanded list of variables for the diagnosis of sepsis, severe sepsis and septic shock (Levy 2003). Despite the revised definition for sepsis, the ACCP/SCCM definition which entails SIRS criteria may be preferred by most clinicians, because of its concise nature.

Reports from a recent study among people with severe sepsis also question the sensitivity of the SIRS criteria for sepsis diagnosis. The use of the SIRS criteria for the diagnosis of sepsis would miss out one in eight people with infection and organ failure. These people with SIRS‐negative severe sepsis have similar epidemiologic trends to people who have the typical SIRS criteria (Kaukonen 2015).

Early in 2016, another set of updated criteria for sepsis was proposed by a consensus task force of the ESCIM and the SCCM defining sepsis as a life‐threatening organ dysfunction resulting from a dysregulated host response to inflammation. The new definition recommended discarding the term severe sepsis and focused on organ dysfunction as a distinguishing point between an infection and sepsis. It proposed a tool ‐ quick SOFA (qSOFA)‐ Sequential (sepsis‐related) Organ Failure Assessment. The qSOFA is based on the presence of two of three warning signs for the rapid identification of organ dysfunction and people at increased risk of mortality or prolonged stay in the intensive care unit. The three warning signs include an altered mental status, a systolic blood pressure of 100 mmHg or less and a respiratory rate of 22 breaths per minute or more (Singer 2016).

These continuous changes in the definition of sepsis reflect the challenges experienced in the diagnosis of sepsis. In order to aid the rapid distinction of sepsis from SIRS, the use of various biochemical tests has been proposed and some are in clinical use for this purpose (Dellinger 2013).

Index test(s)

We will evaluate the diagnostic performance of three different biochemical blood tests: Procalcitonin (PCT), C‐reactive protein (CRP), and presepsin which are used as biomarkers for sepsis. Because we anticipate many primary studies evaluating these biomarkers, we will present the results of these tests as three separate reviews. Presenting these test results as three separate reviews will also enable us to interpret and discuss the findings with sufficient detail. In light of the need relevant for clinical practice, we shall conduct and present these reviews in the following order: PCT, CRP and presepsin.

Procalcitonin (PCT) is a precursor of the hormone calcitonin produced by the parafollicular cells of the thyroid and the neuroendocrine cells of the lung and the intestine. Due to its being released in response to sepsis, the PCT assay is presently in use as a diagnostic tool for sepsis, in the USA, Europe, Australia, Asia and in some parts of Africa (Lloyd 2012; Schneider 2007).

The PCT test quantifies PCT in serum or plasma and results are usually available within one hour. Assays can be performed at point of care or in the routine laboratory. Commonly‐observed PCT values in healthy individuals who are aged three days old and above, are less than 0.05 ng/mL. While PCT levels 0.5 ng/mL to 2.0 ng/mL may indicate sepsis, levels above 2.0 ng/mL but less than 10 ng/mL indicate a high risk for progression to organ dysfunction (Meisner 2014; Nargis 2014). In the absence of sepsis, a rise in the levels of PCT has been reported in people with Addisonian crises (Schumm 2010), people undergoing conditioning treatment for stem cell transplantation using anti‐thymocyte globulin (Brodska 2009) and in people undergoing transplants administered with pan T‐cell antibody (Sabat 2001).

CRP is a plasma protein synthesized by the hepatocytes. It rises in response to inflammation (cell injury) and various pathogens (infection) because of its cell‐membrane‐binding capability which occurs through its attachment to the phosphocholine in exposed cell membranes during cell injury, and the phosphocholine in the polysaccharides of pathogens present in infections (Volanakis 2001). In humans, plasma CRP levels are typically below 5 mg/L but may rise exponentially, within a few hours, in response to an acute inflammatory stimulus (Black 2004). The typical cut off for CRP test is 10 mg/L and most assays have a lower limit of sensitivity of 5 mg/L. However, plasma CRP levels may not rise to 10 mg/L until after 24 hours, consequently, two values more than 10 mg/L taken 24 hours apart contend the diagnosis of sepsis (Zecca 2009). The CRP test results are available within minutes for the point‐of‐care testing (POCT) assays or within an hour for the laboratory‐based assays. The POCT assays require very minute volumes of blood and may be semi‐quantitative, such that values less than 10 mg/L give a negative result, ruling out sepsis. The laboratory assays are quantitative hence, they are ideal for the serial monitoring of patients (Vallance 1991; Zecca 2009).

Presepsin is also known as soluble CD14 subtype (sCD14‐ST). It is a glycoprotein‐fragment derived from monocytes and macrophages and produced in association with infections. Though its diagnostic accuracy has not been as extensively studied as that of CRP and PCT, a few reports have found that it may have a better prognostic value in sepsis than PCT (Masson 2014; Ulla 2013). Presepsin assays are available as POCT or routine laboratory tests and are based on immunochemical methods (Okamura 2011; Shirakawa 2011). Results are available as early as 20 minutes after sample collection. Reported levels of presepsin in plasma of healthy individuals, people with SIRS, local infection or sepsis are several folds higher than the assay limit of quantitation (Masson 2014; Ulla 2013).

Clinical pathway

When a person has signs of systemic inflammation and a suspected infection, blood samples and relevant body fluid samples are taken for culture to confirm the presence of an infection. The results become available in 24 hours to 48 hours after sample collection. In some cases, the source of infection may be obvious, such as some respiratory tract infections in which there may be signs of a pneumonia on a chest radiograph; urinary tract infection where the person may present with symptoms and signs like dysuria and urinary frequency respectively. In other cases, the presence of predisposing factors in an individual raises the suspicion for sepsis. These factors include the presence of in‐dwelling catheters or medical devices; elderly people; immunosuppression such as seen in severe burns, transplant recipients, people receiving chemotherapy, people with poorly‐controlled diabetes mellitus; cellulitis; recent surgery or invasive procedure, perforated viscus and syndromes associated with high risk of infection such as ascending cholangitis (Wacker 2013).

Sepsis is classified as severe when there is cardiovascular dysfunction or acute respiratory distress syndrome (ARDS) or two or more other acute organ dysfunctions manifested as acute kidney injury, or acute liver failure, thrombocytopaenia or coagulopathy. Hypoperfusion from cardiovascular dysfunction may manifest as oliguria, lactic acidosis or an acute mental status alteration (Dellinger 2008).

Septic shock is severe sepsis with arterial hypotension, in spite of adequate fluid resuscitation of 20 mL/kg crystalloid, in the presence of perfusion abnormalities. Hypotension occurs when the mean arterial pressure is less than 70 mmHg or the systolic blood pressure (SBP) is less than 90 mmHg or falls by more than 40 mmHg from the baseline SBP. People receiving inotropics may not be hypotensive (Bone 2009).

Current management guidelines recommend immediate institution of therapy once sepsis is suspected. Some of the therapeutic measures include the administration of antimicrobials within one hour of recognition of septic shock, the quantitative resuscitation of patients within six hours of recognition of sepsis, and the collection of samples for culture before administration of antibiotics (Dellinger 2013).

Rationale

In recent years, various biomarkers such as CRP, PCT and presepsin, have been studied and are proposed as helpful diagnostic tools for sepsis (Harbarth 2001; Rothenburger 1999; Ulla 2013). The early distinction of sepsis from non‐infectious conditions with similar clinical signs, is required for instituting a prompt and appropriate intervention. This ensures a favourable patient outcome and prevents unnecessary antibiotic usage, which is a driving force for antibiotic resistance. The results of CRP, PCT or presepsin as biomarkers for sepsis are usually obtained within a shorter time compared to the traditional laboratory test for sepsis, the blood culture or culture of other relevant body fluids. Assaying these biomarkers may be expensive and there is a need to objectively review the diagnostic performance of these biomarkers in order to determine what role each may play in the prompt and accurate distinction of sepsis from SIRS. There is currently no Cochrane systematic review that has evaluated the test accuracy of these biomarkers for diagnosis of sepsis in adults and children.