Abstract
Advances in medicine have improved our understanding of sepsis, but it remains a major cause of morbidity and mortality. The detection of pathogens that cause sepsis remains a challenge for clinical microbiology laboratories.
Routine blood cultures are time-consuming and are negative in a large proportion of cases, leading to excessive use of broad-spectrum antimicrobials. Molecular testing direct from patient blood without the need for incubation has the potential to fill the gaps in our diagnostic armament and complement blood cultures to provide results in a timely manner. Currently available platforms show promise but have yet to definitively address gaps in sensitivity and specificity.
Significant strides have been made in the detection of pathogens directly from blood. A number of hurdles, however, remain before this technology can be adapted for routine use.
This review will enhance readers' understanding of the role of direct molecular testing on blood for the detection of sepsis pathogens. It will characterize the advantages and pitfalls of molecular testing and how it may be used in conjunction with traditional methods.
Sepsis has been widely acknowledged as a significant cause of morbidity and mortality around the world (1, 2). Despite improvements in diagnostic methods and aggressive therapeutic interventions, mortality related to sepsis remains high (3, 4). Patients who survive an episode of sepsis may face prolonged hospital stay and long term sequelae (5, 6). Neonates in particular are at risk of developmental delays and adverse outcomes (7, 8). The burden of sepsis on healthcare in the US is substantial, with the cost attributed directly to sepsis, reaching $15.4 billion in 2009 (9).
Delays in appropriate antibiotic therapy have shown to contribute toward increased mortality in sepsis (10, 11). Kumar et al. demonstrated that a 1-h delay in appropriate antimicrobial therapy translated to increases in mortality of 7%–10% (11). Others have shown that inappropriate initial therapy in patients with sepsis was associated with a 5-fold reduction in survival (12). The Surviving Sepsis campaign, which is an international collaborative effort by several professional societies to reduce sepsis mortality, is geared toward timely utilization of diagnostic testing and administration of antimicrobial therapy (4). In this regard, the diagnostic tools available to the clinicians for identification of the responsible pathogen do not facilitate timely intervention for a variety of reasons (4). Blood cultures (BC)2 require up to 5 days of incubation, with most bacteria detected after 12–48 h of incubation. Guidelines recommend a minimum of 2 sets of BC per septic episode with at least 40 mL of blood collected in adults with lower volume of blood in children and neonates (13, 14). BC sensitivity may be further optimized by up to 2 additional BC sets (13). The large volumes of blood are required for optimal sensitivity owing to the relatively low burden of bacteria in the blood (1–10 cfu/mL) during septic episodes (13, 15, 16). Despite the large sample volume, BC identifies a pathogen during sepsis in only 40%–60% of cases (1, 17). The failure of BC to yield a causative organism could be because of a number of reasons—one key cause could be prior antibiotic coverage, which could result in nonviable organisms. Phua et al. noted that respiratory tract infection was associated with culture-negative sepsis likely owing to prior antibiotic therapy in the community setting (1). Although guidelines recommend collection of BC before antibiotic administration, the frequency with which this happens is highly variable, with studies indicating that 28%–63% of patients have received antibiotics before BC collection (2, 4, 18). False-negative blood in the setting of sepsis cultures may also occur because of the presence of noncultivable or fastidious bacteria. In addition, sepsis may occur because of nonbacterial pathogens—such as viral or fungal organisms. De Prost et al. noted that some patients with culture negative sepsis had lower procalcitonin levels and positive detection of viruses in respiratory specimens (17). In addition, empiric coverage for sepsis does not target fungal pathogens unless the patients have appropriate risk factors (4, 19, 20).
Aside from false-negative BC, false-positive BC are also a cause for concern. BC may be contaminated because of errors during the specimen collection process. Typical contaminants include skin flora such as coagulase-negative staphylococci, α-hemolytic streptococci, corynebacterial, and Bacillus species. The impact of contaminated BC can be significant, leading to unnecessary antimicrobial treatment and diagnostic procedures and extended hospital stays, with average costs per episode of $8000–$25000 (21, 22). With average reported BC contamination rates of approximately 3%, this could mean millions of dollars of excess healthcare costs per institution and extended patient stays amounting to 1372–2200 extra hospital days per year (22, 23).
The Surviving Sepsis campaign encourages prompt initiation of empiric broad-spectrum antibiotic coverage within 1 h for both sepsis and septic shock (4). Empiric coverage would be continued until detection of the pathogen by BC followed by susceptibility testing. This would allow for narrowing down the therapy needed to target the responsible pathogen and thus avoid the adverse impact of prolonged use of broad-spectrum antimicrobials including the development of resistance and hospital-acquired infections (HAI) such as Clostridium difficile. Changes to empiric coverage, however, require a positive BC to detect the potential pathogen. BC may need up to 5 days for a negative result and suffer from limited sensitivity (11, 24). BC bottles are typically incubated in automated instruments that monitor for detection of growth in real time. Positive BC bottles are removed from the instrument and Gram-stained, and the report is typically communicated to the clinician as a critical value and it allows the clinician to narrow down treatment. Laboratories that are not staffed 24/7 may further delay the reporting of positive BC owing to inadequate staffing. Barenfanger et al. demonstrated significant increases in mortality (10.1% vs 19.2%) when positive BC Gram stains were delayed >1 h (25). Timely reporting of positive Gram stain results is essential but may not always be practical in small laboratories and community hospital settings.
More recently, laboratories have adopted probe and multiplex PCR-based technologies that allow for rapid identification of bacteria in positive BC bottles (Fig. 1). These panels can identify the most frequently isolated pathogens within 1–3 h (26). In addition, these panels can detect resistance markers such as mecA and KPC that would guide treatment choices a full day or two before routine susceptibility test results are available (26). Utilization of these panels has been shown to have a positive impact on patient management including reduction in length of intensive care unit (ICU) stay and 30-day mortality, although it still requires a positive BC (27).
Progression in blood culture diagnostics with the development of molecular technologies over time.
Molecular diagnostic testing has revolutionized the field of clinical microbiology for over 2 decades. The use of DNA- and RNA-detection techniques has served to nearly eliminate the role of traditional routine culture for viral pathogens. Cultures for cytomegalovirus routinely required 21 days of incubation for tissue culture, but with molecular methods, results are typically available within hours. Viral culture for influenza required 7 days of incubation in multiple cell lines but CLIA-waived point-of-care PCR-based devices are now available that detect influenza with increased sensitivity within 20 m (28). The adoption of these technologies for detection of bacterial pathogens has progressed at a slower pace in part due to the challenges of distinguishing normal bacterial flora from pathogens. The role of nucleic acid amplification test (NAAT) in advancing the detection of sepsis has been largely limited to identification of pathogens from BC bottles after amplification by culture.
NAAT-based testing holds significant promise because of the potential for high sensitivity and the ability to detect organisms that are noncultivable or nonviable owing to prior antibiotic treatment. One of the key limitations of BC is the low organism load (1–10 cfu/mL) in circulation during a septic episode. However, using culture-independent methods, Bacconi et al. determined that 103 or 104 GC (genome copies)/mL of bacteria was present in blood (16). This included DNA from nonviable cells that could not be detected by culture and could potentially serve to amplify the performance characteristics of NAAT testing. Some studies have suggested that patients with positive vs negative BC during sepsis do not demonstrate significant differences in mortality, although culture-positive patients may have longer ICU and hospital stay (1, 29, 30). O'Dwyer et al. demonstrated that patients with NAAT-positive results for bacterial DNA from blood had significantly higher mortality (42% vs 26%), whereas the same was not true for patients with positive BC (29). It is worth noting that these findings have not always been replicated in similar studies, and it remains to be seen if the presence of microbial DNA in the bloodstream is a predictor of mortality or rather a symptom of patient clinical status (31, 32).
Despite significant interest in the utilization of NAATs for direct detection of pathogens from the bloodstream of a patient with sepsis, attempts to adapt the technology to this purpose, although promising, have generally met with limited success and until recently there were no US Food and Drug Administration (FDA)-approved platforms for the direct detection of bacterial or fungal pathogens from the bloodstream. Many early studies failed to demonstrate comparable performance between NAATs and culture for detection of bloodstream pathogens (10, 33). NAAT-based testing was able to detect pathogens missed by culture but also lacked sensitivity in culture-positive cases. Although NAATs can be expected to have a lower limit of detection than BC, they may still struggle to match BC performance, in part because of the significant differences in the sample volume tested. These approaches may typically sample anywhere from 0.5 mL to 5 mL of whole blood, whereas a routine BC including a minimum of 2 correctly collected sets with 4 bottles could sample nearly 40 mL of whole blood. In addition, whole blood contains a significant amount of cellular material and other known PCR inhibitors such as iron, immunoglobulins, and lactoferrin (18). The presence of anticoagulants such as lactoferrin can also inhibit PCR reactions (18).
On the other hand, NAAT-based testing was also challenged with regards to specificity when compared to BC. Studies examining healthy donors have shown the presence of bacterial genetic material in the circulation of a significant proportion (31%) of donors tested (34). Transient bacteremia/fungemia may occur even in healthy individuals, and NAAT testing may detect the presence of pathogen DNA/RNA circulating even after successful treatment. In addition, contamination could occur from the environment during the specimen collection or testing process. Some studies have suggested the use of semiquantitative molecular assays to distinguish between true infection and contamination, but this may be challenging due to the fact that low organism loads are common during sepsis (18, 35).
NON-FDA-APPROVED PLATFORMS
A number of platforms are currently available for detection of pathogens directly from blood; however, many of these are not currently FDA-approved for use in the US, although some have regulatory approval in Europe (CE marked).
LightCycler SeptiFast
This system (Roche Molecular System) is an automated platform that can use 1.5 mL of whole blood to detect 19 bacterial and fungal pathogens using multiplexed PCR coupled with probe hybridization and DNA melt curve analysis. The identity of the pathogen is available in 3.5–5 h with additional detection of mecA gene associated with methicillin resistance in Staphylococcus species. The platform is capable of performing a semiquantitative analysis that may be useful to determine the significance of positive results (18).
MagicPlex Sepsis
The MagicPlex Sepsis instrument (Seegene) is capable of detecting >90 different bacteria and fungi, as well as methicillin and vancomycin resistance, directly from 1 mL of whole blood within 3 h using multiplex PCR. However, only a limited subset of the 90 bacteria/fungi are identified to the species level.
VYOO
The VYOO rapid pathogen identification system (Analytik Jena Gmbh) uses a combination of DNA amplification followed by electrophoresis-based analysis for the detection of 34 bacteria and 7 fungi using a sample volume of 5 mL of whole blood. In addition, the platform also detects 5 different resistance markers, providing information on methicillin resistance, vancomycin resistance, and the presence at least 2 classes of extended-spectrum β lactamases.
PLEX-ID
The PLEX-ID Pathogen Detector (Ibis Biosciences) has the most comprehensive approach of currently available platforms. It can detect and identify up to 800 different gram-positive and gram-negative bacteria, as well as fungi and 4 resistance markers: mecA, vanA, vanB, and blaKPC. These markers convey information with regard to methicillin, vancomycin, and carbapenem resistance. The platform targets the rRNA genes and other conserved regions of the bacterial and fungal genomes by using broad-range PCRs followed by amplicon analysis by ESI-MS (electrospray ionization mass spectrometry). Pathogen identification is facilitated by determination of the base composition of the amplified sequences rather than analysis of the sequence itself followed by comparison of base composition to a database. The platform uses 5 mL of whole blood to provide an answer in 6 h.
SepsiTest
The SepsiTest (Molzym Molecular Diagnostics) amplifies bacterial 16S rRNA and fungal 18S rRNA, which enables it to detect over 345 bacteria and 13 fungi (Sinha). This is followed by Sanger sequencing and BLAST analysis for identification. The test includes several steps, requires 1 mL of whole blood, and can take 8–12 h. The platform also does not include detection of any resistance markers. It can also be used for the detection of pathogens from other sterile fluids.
PERFORMANCE
These platforms all demonstrate varying performance characteristics in the detection of pathogens during sepsis. Differences in sample volume, processing steps, amplification, and detection methods, as well as number of targets, all contribute to assay performance. It would be reasonable to expect NAAT-based methods to outperform routine BC, but this has not been consistently demonstrated in published studies (18, 33). As discussed previously, this may be in part due to the differences in input volume for BC (40–80 mL) as compared to NAAT-based methods (1–5 mL) (18). The sensitivity of NAAT and the manual nature and multiple processing steps involved with many of these platforms can also increase the risk of contamination. Assessing the performance of these assays can also be challenging in the setting of NAAT+/BC− results because the lack of a gold standard leads to subjective interpretations of the significance of positive NAAT results based on clinical criteria.
The largest body of literature assessing the performance of NAAT-based testing in sepsis exists with the SeptiFast platform because of its longevity in the market in Europe. Although the positivity rate of SeptiFast is significantly higher than BC (25%–35% vs 12%–21%), a metaanalysis reported sensitivity and specificity of only 68% and 86%, respectively (33, 36). SeptiFast failed to detect the pathogen in 20%–30% of cases, and in 35% of cases, it provided a potential false-positive result (6, 10, 37). Similar performance characteristics were noted with most commercially available PCR platforms with the exception of the PLEX-ID, which demonstrated slightly better performance with 83% sensitivity and 94% specificity compared to BC (16).
The challenge of using PCR-based approaches to sepsis lies not just in performance characteristics but also in cost justification. Platforms such as SeptiFast cannot yet replace routine BC. The limited sensitivity means that negative results are not actionable and concerns about specificity mean that positive results have to be interpreted with caution (38).
Sequencing-based methods
Next-generation sequencing (NGS) offers the advantage of being able to take an unbiased approach toward the detection of bacterial, viral, or fungal pathogens via the analysis of cell-free DNA (cfDNA) in plasma (39–41). Increased levels of cfDNA have been reported in patients with inflammatory diseases, trauma, cancer, and surgery (42, 43), with some also suggesting that increased levels (44) of pathogen DNA potentially released from bacteria or phagocytes during sepsis in cfDNA may facilitate molecular detection (39). NGS offers the ability to quantitatively assess the level of bacterial cfDNA in plasma that may allow it to distinguish between pathogen and contaminant (40). This functionality is important because multiple studies using NGS have reported detection of commensal skin flora such as Propionibacterium acnes in plasma samples from healthy donors at varying levels (40, 41). NGS was able to complement BC by detecting a number of bacterial and viral pathogens that would have been otherwise missed, although there were also instances of BC+/NGS− results in patients with sepsis (40, 41). Long et al. demonstrated that the number of positive patients increased from 12.82% (12/78) by using routine BC to 30.77% (24/78) using NGS (41). In addition, 15 viruses were detected in 14 patients (41). NGS was also able to detect resistance genes such as mecA and VanA/VanB; however, the detection of such resistance markers would require high read coverage that may not always be possible depending on the platform (40).
A number of challenges remain with the use of NGS—the protocols are often complex, involving multiple steps with the concomitant risk of contamination increasing with each step. Trace contamination including bacterial DNA in DNA extraction kits and centrifuge tubes resulted in erroneous detection of bacterial DNA in clinical specimens (45–47). The use of a scoring system based on the number of reads could potentially play a role in interpreting positive results, but the lack of standardization across laboratories running different platforms and protocols may render this challenging (40). Traditional NGS can be labor-intensive, expensive, and time-consuming and therefore also limit the ability to obtain results in a timely manner. Newer platforms such as the MinION (Oxford Nanopore Technologies) have attempted to address these concerns with a portable DNA/RNA sequencer with relatively short library prep times, low cost, and the ability to potentially provide results in <4 h. It remains to be determined whether such platforms would provide adequate read coverage for the high sensitivity required for the detection of bloodstream pathogens and resistance markers (38). Commercial laboratories now offer NGS testing on plasma using proprietary methods that assess relative abundance of pathogen DNA and flag organisms that reach a threshold of significance (48). These assays demonstrate significantly higher positivity rates (60.1% vs 15.7%) when compared with routine BC (48). Further studies are needed to determine the significance of NGS+/BC− results, and the impact these results may have on patient management and outcomes.
FDA-APPROVED PLATFORMS
The first FDA-approved platform for the direct detection of a bacterial or fungal pathogen from bloodstream was the T2Candida (T2C) assay (T2 Biosystems). Invasive candidiasis continues to present a diagnostic challenge for laboratories, with BC being negative in 50% of proven and probable cases of invasive candidiasis (49, 50). For positive cultures, the median time to positivity is 2–3 days, and high mortality rates (40%) persist even when appropriate therapy is administered (51). T2C combines PCR amplification with magnetic resonance for the detection of the 5 Candida species most commonly implicated in invasive candidiasis. The assay uses approximately 4 mL of whole blood to provide a result in 3–5 h on a fully automated platform with the capability to detect >1 cfu/mL of Candida (52–54). Three possible positive results are provided—Candida albicans/Candida tropicalis, Candida glabrata/C. krusei, and Candida parapsilosis. The pairing of organisms in results is based on expected known resistance patterns and therapeutic interventions. During clinical trials, the assay demonstrated 91.1% sensitivity and 99.4% specificity, although only 6 patients with candidemia were detected among almost 1500 participants (52). A subsequent study in 152 patients with positive BC for Candida species reported sensitivity of 89% with 4/152 patients with T2C−/BC+ and 32/152 T2C+/BC− results (55). In this study, patients with candidemia but without prior antifungal therapy showed excellent correlation between positive T2C and BC results (32.5% vs 30%) as compared to patients on antifungal therapy for which T2C was positive in more than twice as many patients (50% vs 21%) with candidemia (55). These findings implied that T2C may perform better than BC in the detection of candidemia in patients on prior antifungal therapy.
An independent assessment of the impact of T2C on patients at a tertiary care center determined that appropriate antifungal therapy was initiated earlier after T2C implementation (22 h vs 39 h) (56). No significant difference was noted in mortality between 2 groups before and after T2C implementation, but the authors did observe a significant difference in the incidence of ocular candidiasis [16/53 (30%) before T2C vs 6/49 (12%) after T2C implementation] (56). Early administration (>24 h) of appropriate antifungal therapy has been shown to significantly reduce mortality (15.4%) vs delayed administration >72 h (41%) (19, 57). The authors speculated that early administration of appropriate antifungal therapy may have limited dissemination of Candida, resulting in reduction in ocular infection (56).
A retrospective analysis of 1273 patients tested with T2C over 2 years showed that T2C results were available significantly earlier than preliminary BC results (13 h vs 34 h) (58). T2C was positive in 11% (141/1273) patients with 26/112 (23%) of the positives being transplant patients vs 126/1161 (11%) being nontransplant patients. Only 28% (40/141) of the positives were T2C+/BC+. An additional 72% (102/141) were T2C+/BC−, with the remaining 28% (39/141) T2C−/BC+ (58). Candida was isolated by sterile site cultures from 10% of the T2C+/BC− patients but from none of the T2C−/BC+ patients. Sensitivity of T2C was 55% in transplant and 50% in the nontransplant population with specificity of 80% and 93%, respectively. Positive and negative predictive value for transplant and nontransplant groups were 23%/29% and 94%/97%, respectively (58).
More recently, the FDA approved T2Bacteria (T2B) for direct detection of 5 different organisms that are responsible for >50% of all bloodstream infections directly from whole blood—Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Escherichia coli. The system uses the same platform and technology as the T2C assay. The assay showed limits of detection for the target bacteria of 2–11 cfu/mL. In clinical trials comparing T2B against BC, the positivity rate was 5.7% (82/1427) for BC vs 13.3% (190/1427) for T2B (59). There were 35 concordant T2B+/BC+ samples. Among BC-positive patients, 28% (23/82) represented typical contaminants such as coagulase-negative staphylococci, diphtheroids, and Corynebacterium species that are not detected by the T2B panel. Of the remaining BC-positive patients, T2B identified 66% (39/59). However, there were 155/1427 (10.8%) of T2B+/BC− discordant results (59). According to analysis of clinical trial data, 44.5% (69/155) of the discordants had “strong evidence” of infection and an additional 23/155 had “other evidence of infection”; however, 41% (63/155) of T2B+/BC− results were classed as potential contaminants. After further review of antibiotic therapy received at the time of testing, it was determined that T2B could have facilitated more effective therapy earlier in 3.9% (56/1427) of patients tested (59). Mean time to result was 5.4 h for T2B and 38.5 h for BC (59). In an independent study by De Angelis et al., T2B showed 83.3% sensitivity and 97.6% specificity, although sensitivity improved to 89.5% when taking into consideration true infection criteria (e.g., same organism isolated from another site) (60).
CHALLENGES IN THE IMPLEMENTATION OF NAAT TESTING
The use of NAATs for direct detection of pathogens from blood raises a number of challenges. The prevalence of invasive candidiasis tends to be relatively low in most hospitalized patients. The majority of results from NAATs testing blood are likely to be negative, whether testing for Candida or bacteria. For a negative value to be an actionable result, i.e., for the physician to stop antimicrobial therapy, the assay has to have excellent negative predictive value. Clancy et al. determined that a PCR with performance characteristics of 85% sensitivity/97% specificity would still have a 6% chance of infection (61). In a high-risk patient, treatment might be continued despite a negative test result in the absence of any other explanation. For most of the molecular panels that target bacteria, the assays simply do not have adequate negative predictive value to justify discontinuation of treatment (31, 62–64). For assays such as T2B that have promising preliminary data, the limited number of target organisms restricts the value of a negative result and makes it unlikely that negative results will affect patient management. This would mean that a significant proportion (approximately 80%) of results offer no value to the caregiver. In an era where there in increased scrutiny of laboratory utilization and NAAT testing can potentially be very expensive, this can represent a significant challenge that needs to be addressed in future studies.
These concerns emphasize the need for a rational approach to utilization of these novel assays. In the ICU setting with an expected prevalence of 1% invasive candidiasis, T2C would be expected to have a positive predictive value of only 31% (61). One of the expected benefits of implementation of T2C is that the high negative predictive value would reduce antifungal usage (65); however, the high rate of NAAT+/BC− results seen across multiple platforms is a concern that could be particularly amplified if assays are overused in low-prevalence settings. This could in turn have the opposite effect of driving unnecessary antimicrobial usage rather than reducing it. Clancy et al. estimated that even in optimal scenarios, approximately 30% of positive T2C results may be false positives (61). These concerns were echoed by the findings of Hussain et al., in which the positive predictive value of T2C even among transplant patients was only 23% (58). This calls for the development of effective strategies and algorithms to target the use of these assays toward patients for which results will actually impact management.
A further concern is the impact of NAAT testing in bloodstream infections on the reporting of HAIs. Among other things, hospitals are required to report central line–associated bloodstream infections (CLABSI) to the Centers for Disease Control and Prevention National Healthcare Safety Network (4). Hospitals that rank poorly in the control of HAIs will be penalized by the Centers for Medicare and Medicaid services via the reduction of payments (66). Data generated thus far strongly suggest that hospitals performing NAAT testing directly on blood will have a significantly higher rate of positive results from blood as compared to institutions that only utilize routine BC. Data generated from both T2C and T2B suggest that a significant proportion of positives will represent false positives (50, 58, 59). In the T2B clinical trials, 41% of the positives were determined to be contaminants (59). In calculating the CLABSI rate, National Healthcare Safety Network includes both detection by culture and nonculture-based methods (66). As a result, hospitals that routinely use NAAT testing for bloodstream infections could potentially find themselves penalized for utilizing novel technologies.
Despite these concerns, the limitations of BC during sepsis are well known, and NAAT-based testing has the potential to address these gaps. Ideally, the platforms that perform this testing have the following characteristics:
Easy to use with limited manipulation required. This would allow it to be run across multiple shifts without the need for staff with specialized skills or training.
Results available in a timely manner to facilitate prompt antimicrobial therapy.
The ability to detect key resistance markers such as mecA.
Broad spectrum of targets to encompass as many of the key target organisms as possible and facilitate change in therapy.
Cost-effective testing that has the throughput required to process multiple specimens simultaneously if needed.
The costs of NAAT-based testing are likely to be significant, so laboratories that implement such assays should consider that they will need to develop algorithms to identify patients that would benefit the most. In addition, a decision support network that includes infectious disease physicians and pharmacists is essential for ensuring that these assays are effective. Studies have shown that efforts to expedite BC results using novel technology are often ineffective unless they are coupled with education of clinical staff and a network of support staff that ensure results are being used (67). The Surviving Sepsis campaign has incentivized hospitals to ensure timely administration of antimicrobial prophylaxis. In this setting, it important to determine the role of NAAT-based testing, as it is unlikely that deescalation of antibiotics alone will be adequate to cost-justify the use of NAAT testing. Further studies are also required to determine the impact of the use of NAAT-based testing in sepsis on the patient outcomes and mortality.
CONCLUSION
NAAT-based testing has the potential to not just expedite pathogen detection in patients with sepsis but also address the gaps caused by lack of sensitivity of BC. Additional studies are required to determine the significance of NAAT+/BC− results. Owing to the limitations of currently available assays, NAAT testing cannot yet serve as a surrogate for routine BC, but it can be a useful complementary tool for the detection of pathogens in sepsis.
2 Nonstandard abbreviations
- BC
blood culture
- HAI
hospital acquired infection
- ICU
intensive care unit
- NAAT
nucleic acid amplification test
- FDA
US Food and Drug Administration
- NGS
next-generation sequencing
- cf DNA
cell-free DNA
- T2C
T2Candida
- T2B
T2Bacteria
- CLABSI
central-line associated bloodstream infections.
Author Contributions: The author confirms they have contributed to the intellectual content of this paper and have met the following 4 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved.
L.P. Samuel, administrative support.
Authors' Disclosures or Potential Conflicts of Interest: The author declared no potential conflicts of interest.
References
