Sepsis with bacteremia is a leading cause of death among adults and children.¹ Blood culture (BC) is the gold standard in the diagnosis of sepsis but the turnaround time is long. Even automated, continuous-monitoring blood culture systems (CMBCS) followed by the use of standard laboratory agar media for subculturing positive BC bottles take 48-72 hours to identify microorganisms. Early identification and determination of antimicrobial susceptibilities of microorganisms growing in blood cultures is crucial as delay can lead to increased mortality, morbidity and cost.^2-4

The FilmArray blood culture identification (FA-BCID) Panel (BioFire Diagnostics, Salt Lake City, UT, USA) is a US FDA approved two stages, multiplexed, nested polymerase chain reaction (PCR)-based molecular platform that is carried out in a closed, disposable, single-use pouch. It has high analytical sensitivities of over 90% for rapid identification of bacteria and fungi from positive blood culture bottles with a turnaround time of one hour as stated by the manufacturer. FA-BCID is widely evaluated and studied in countries like USA, Sweden, South Africa and Japan.^5-8 The FA-BCID has been recently introduced in India and clinical performance data is lacking in India. This is the first ever study conducted in India to evaluate the clinical performance of FA-BCID panel in comparison to automated blood culture system in early identification, appropriate antimicrobial initiation and de-escalation.^6,9

Study design and setting

This was a single center, prospective study conducted in a 24-bed critical care unit (CCU) of a tertiary care hospital at Chennai, India between October 2016 and December 2016. All patients more than 18 years of age admitted in the CCU who had positive Gram stain from the flagged blood cultures in the clinical microbiology laboratory by automated blood culture system were included in the study. Patients were informed about the study and informed consent was obtained from stable patients. In clinically unstable patients, consent was obtained from legally acceptable representative. Simultaneously, patient data were collected which included demographic and clinical data.

BD BACTEC FX (BD, Franklin Lakes, NJ, USA) was used in comparison to assess the clinical performance of the FA-BCID assay. BD BACTEC FX is an automated, continuous-monitoring blood culture system which utilizes fluorescence technology for the detection of microorganisms. It depends on the amount of carbon dioxide released by the growing microorganisms. The instrument measures the level of fluorescence corresponding to the amount of CO₂released by the organisms. Positive cultures are immediately flagged by an indicator light in front of the system. Gram stain was performed from flagged blood culture bottle: when found positive it was subjected to identification by both the standard conventional method (agar culture followed by both VITEK 2 (bioMérieux, Marcy-l'Étoile, France) and disk diffusion) and the FA-BCID assay. Only one positive blood culture per patient was used for the FA-BCID panel.

FA-BCID is a closed diagnostic system based on nucleic acid extraction from clinical specimens, high-order nested multiplex PCR and post-PCR DNA melting curve analysis. It is done in a closed, disposable single pouch. The FA system performs a "nested multiplex” PCR (nmPCR) within the instrument, which provides increased sensitivity compared with conventional PCR. The test’s principle is described in the manufacturer’s approval documents.¹⁰ BCID results for a specimen were displayed only if the two internal pouch controls for the run were valid. The panel tests for 19 bacteria, 5 yeasts and 3 antibiotic resistance genes: mecA, vanA/vanB and the KPC gene (Table 1). Only positive blood culture bottles with positive Gram stain were analysed by FA-BCID.

Results of blood culture were analyzed by Vitek 2.0 for both identification and antimicrobial susceptibility testing (AST). Drug susceptibility was also assessed by disk diffusion (Kirby Bauer Method) which conformed to the standards of the Clinical and Laboratory Standards Institute (CLSI 2016).¹¹ Also blood culture results via FA-BCID were analyzed for sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for identification of bacteria. Statistical analysis was done using IBM Statistical Package for the Social Sciences (SPSS version 21, IBM Corp, Armonk, NY, USA). Sensitivity, specificity, PPV and NPV were determined from standard two-by-two performance tables.

The study was approved by the Institutional ethical committee on 19 August 2016, Main Apollo hospital, Greams road, Chennai.

A total of 36 positive blood culture bottles from 36 patients admitted in the critical care unit were analyzed; 72.2% of patients were males and 27.8% were females. The mean age of the study patients was 52.6 years (Table 1). The most common source of bacteremia was respiratory infection (25%), followed by liver and biliary tree (19.4%) (Table 1).

Table 1. FA-BCID panel constituents

A total of 36 positive blood culture bottles were analyzed by both FA-BCID and the conventional method. Of these, eight blood culture bottles showed no pathogen in FA-BCID.

Of the 40 different isolates that were retrieved during the study period, 32 (80%) organisms were accurately identified by FA-BCID (Table 2). FA-BCID identified 81.3% (26/32) of organisms to species level and a further 18.7% (6/32) to genus level (Table 3).

Table 2. Demographic and clinical data of the patients (n=36)

Of the 36 patients, 94.4% of patients had mono-microbial growth and 5.6% of patients had more than one organism, which was similar to the conventional method (Table 1). Of the total 32 isolates identified by FA-BCID, 50% (16/32) showed pure growth of Gram negative bacteria, 37.5% (12/32) showed pure growth of Gram positive bacteria, 6.25% (2/32) of Gram negative and Gram positive, and 6.25% (2/32) of Gram negative and Candida from the same culture bottle (Table 3).

Klebsiella pneumoniae was most commonly identified in FA-BCID among Gram negative bacteria (33.3%, 6/18), and Staphylococcus aureusamong Gram positive bacteria (38.5%, 5/13) (Table 4). Overall, the sensitivity and PPV were 100% for the identification of Gram positive bacteria and 69% and 100% for Gram negative bacteria, respectively (Table 3). Table 5 presents the pathogens that were identified in blood cultures, but that were not present in the FA-BCID panel.

Table 3. Comparative evaluation of the FA-BCID and the conventional method

The panel identified the mecA gene in 75% (6/8) of Staphylococcusisolates and these results were consistent with the Kirby Bauer method (Table 4). None of the Enterococcus isolates had the vanA/vanB gene and the results were consistent with antimicrobial susceptibility testing. Identification for the presence of methicillin resistance (mecA) and also identification for the absence of vancomycin resistance (van A/van B) was accurate.

Table 4. Identification of bacteria by FA-BCID

Table 5. FA-BCID negative report n=8 (not in the panel)

None of the Gram negative bacteria showed presence of the KPC gene but 7 Gram negative bacteria showed carbapenem resistance in antimicrobial susceptibility testing. Hence the resistance mechanism was not detected by FA-BCID, as the panel only includes KPC gene identification, so the drug susceptibility pattern could not be predicted.

The FA-BCID panel took a median time of 2 hours to obtain identification and appropriate susceptibility information of an organism from a positive blood culture, as opposed to a median time of 2 days for the conventional method. The overall performance of FA-BCID was compared with the conventional method in Table 3.

The antibiotic regimen was de-escalated or escalated in 17 patients (47.2%) based on FA-BCID results in a median time of 3 hours, whereas for 19 patients a change in the antibiotics was done after the result of the conventional method in a median time of 3 days.

Rapid blood culture identification methods are the need of the hour in this multidrug-resistant era, both from a management and an antimicrobial stewardship standpoint. However, these are just being introduced in India and are as yet not widely available. We hereby describe our experience with FA-BCID in critically ill patients with bacteremia. The panel as it is currently available detects common bacteria, yeasts and antimicrobial resistance genes encountered in the western scenario; we attempted to analyze its applicability in India.

In our study all the blood culture organisms included in the panel were accurately identified by FA-BCID, irrespective of mono or polymicrobial growth. FA-BCID accurately identified 80% (32/40) of all microorganisms from positive blood culture bottles that were included in the panel; 20% (8/40) of Gram negative bacteria which were not identified were not included in the panel.

Among the Gram positive cocci, S. aureus, coagulase negative staphylococciand Enterococcus spp. were the most frequently identified and all correlated completely with conventional identification. We did not have any vancomycin-resistant enterococcal isolates on conventional culture and therefore none of our isolates showed the presence of vanA/vanB resistance gene, similar to a study done by Altun et al.^7,12Sensitivity of the mecA gene detection for methicillin resistant staphylococci was 100%, which is similar to studies done by Buchan et al. and Wojewoda CM et al.^13,14 Our study favors the FA-BCID panel as an excellent test for identification and determination of resistance in Gram positive bacteremia. However, resistant Gram positive bacteremia is less of a problem in India compared to Gram negatives.¹⁵

Among the Gram negative bacteria, Klebsiella pneumoniae was the most common isolate followed by E. coli, Pseudomonas aeruginosa and Acinetobacter baumannii, all of which were successfully identified to species level as in other studies.^6,16 Although all Gram positive bacteria were identified by FA-BCID, 20% (8/40) of blood cultures identified as Gram negative bacteria by conventional method were not detected by FA-BCID as they were off panel microorganisms. In the present study off panel organisms are significantly important pathogen as they had clinical correlation. Salimnia et al. also detected off panel organisms isolated by conventional method and not by FA-BCID but the organisms were different compared to those in our study.⁵In a few studies common skin commensals were the common off panel organisms as compared to the present study.^6,12 Targets for important community acquired pathogens seen in India such as S. typhi and B. pseudomallei, and hospital acquired ones such as Burkholderia cepacia need addition to the FA-BCID panel in the future, to increase the clinical utility of the test.

The sensitivity for organisms identification included on the FA-BCID panel was 68% for Gram negative and 100% for Gram positive isolates, while specificity and positive predictive values for identification of both Gram positive and Gram negative isolates were 100%, almost similar to other studies.^12,16,17 In our study we considered the overall negative predictive value for all Gram negative bacteria including off panel organisms. Hence our NPV 76.4% is lower than the study done by Altun et al. and Salimnia et al.^5,6

Although E. coli and Klebsiella spp. were reliably identified, the absence of a probe for CTXM-1 or other ESBLs was a drawback, as identifying ESBL producers, which might require early carbapenem therapy, would be very helpful to clinicians. We suggest that the test be modified to include probes for ESBLs as well, so as to increase its validity and use in the Indian setting.

In our study, none of the seven CRE isolates identified on conventional culture had the KPC gene detected on FA-BCID. We therefore could not use the FA-BCID to detect carbapenem resistance and start appropriate therapy such as colistin early, which is an important limitation of the test in its current form in the Indian setting. Indian single center studies report an overall prevalence of carbapenem resistance in up to 12-15% ofEnterobacteriaceaeand 40-60% in A. baumanniiandP. aeruginosa.^18,19Carbapenemases produced by Gram negative bacteria seen in India are NDM-1 and OXA-48, rather than KPC, which is the predominant molecular mechanism of resistance in the west.^17,20 Hence, more information is hoped in the future for NDM-1, OXA, IMP and VIM probes in the FA-BCID for Gram negative bacteria.

One isolate of Candida albicans was the only yeast from patient blood culture and it was accurately identified by the FA-BCID. Findings of 100% yeast identification rate were also reported by Southern et al. and Altun et al., but the rate was different in the study by Blaschke et al., where only 75% sensitivity was reported for Candida albicans.^5,12,16

Altun et al. recorded an average turnaround time for FA-BCID as 75 min, which was shorter as compared to our study median time of 2 hours (120 minutes).¹² As the time to identify organisms in our study was short, it facilitated clinical decision making within a median time of 3 hours for escalating or de-escalating antibiotics for the patient. It will hopefully translate into earlier appropriate antibiotic therapy which in turn will reduce morbidity, mortality and cost.

Antimicrobial stewardship is an important component of strategies to ensure appropriate therapy and to reduce antimicrobial resistance.²¹ Banerjee et al. recommend that rapid blood culture diagnostics implemented together with antimicrobial stewardship can optimize antibiotic prescription for bloodstream infections.²²Antibiotics were de-escalated or escalated in 47.2% of our patients based on FA-BCID, despite the absence of probes for detection of important resistance mechanisms present in Gram negative bacilli in India; the median time taken was 3 hours as compared to the conventional method. Our findings support the use of FA-BCID as an antimicrobial stewardship tool, even more so with Gram positive bacteremia.

Our study had some limitations: we studied only 36 isolates and only a minority of patients with bacteremia during the study period, and we did not analyze the clinical outcomes.

FA-BCID is a significant advance in the early identification and escalation or de-escalation of treatment for bacteremia in critically ill patients, with a high sensitivity for Gram positive bacteria as compared to Gram negative bacteria. Our results suggest that the test be introduced more widely in Indian settings with new probes for some prevalent pathogens and resistance mechanisms in future modifications.

1. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign: international guidelines for management of severe sepsis and septic shock, 2012. Intensive Care Med 2013;39:165-228. [Crossref]

2. Wiest R, Krag A, Gerbes A. Spontaneous bacterial peritonitis: recent guidelines and beyond. Gut 2012;61:297-310. [Crossref]

3. van de Beek D, Brouwer MC, Thwaites GE, Tunkel AR. Advances in treatment of bacterial meningitis. Lancet 2012;380:1693-702. [Crossref]

4. Proulx N, Frechette D, Toye B, Chan J, Kravcik S. Delays in the administration of antibiotics are associated with mortality from adult acute bacterial meningitis. QJM 2005;98:291-8. [Crossref]

5. Salimnia H, Fairfax MR, Lephart PR, et al. Evaluation of the FilmArray blood culture identification panel: Results of a multicenter controlled trial. J Clin Microbiol 2016;54:687-98. [Crossref]

6. Altun O, Almuhayawi M, Ullberg M, Ozenci V. Clinical evaluation of the FilmArray Blood Culture Identification Panel in identification of bacteria and yeasts from positive blood culture bottles. J Clin Microbiol 2013;51:4130-6. [Crossref]

7. Fhooblall M, Nkwanyana F, Mlisana KP. Evaluation of the BioFire((R)) FilmArray((R)) Blood Culture Identification Panel on positive blood cultures in a regional hospital laboratory in KwaZulu-Natal. Afr J Lab Med 2016;5:411. [Crossref]

8.Saito K, Endo S, Katsumi M, et al. Evaluation of the FilmArray Blood Culture Identification Panel on detection of pathogenic microorganisms in positive blood cultures: the first clinical report in Japan. Jpn J Infect Dis 2018;71:145-7. [Crossref]

9.Rand KH, Delano JP. Direct identification of bacteria in positive blood cultures: comparison of two rapid methods, FilmArray and mass spectrometry. Diagn Microbiol Infect Dis 2014;79:293-7. [Crossref]

10. 510(k) Substantial equivalence determination decision summary assay and instrument combination template. Accessed on: 05 May 2019. Available at: https://www.accessdata.fda.gov/cdrh_docs/reviews/K130914.pdf

11. Clinical and Laboratory Standards Institute (CLSI). 2016. Performance standards or antimicrobial susceptibility testing Document M100, 26th edition. Wayne, PA: Clinical and Laboratory Standards Institute.

12. Altun O, Almuhayawi M, Ullberg M, Ozenci V. Rapid identification of microorganisms from sterile body fluids by use of FilmArray. J Clin Microbiol 2015;53:710-2. [Crossref]

13. Buchan BW, Ginocchio CC, Manii R, et al. Multiplex identification of gram-positive bacteria and resistance determinants directly from positive blood culture broths: evaluation of an automated microarray-based nucleic acid test. PLoS Med 2013;10:e1001478. [Crossref]

14. Wojewoda CM, Sercia L, Navas M, et al. Evaluation of the Verigene Gram-positive blood culture nucleic acid test for rapid detection of bacteria and resistance determinants. J Clin Microbiol 2013;51:2072-6.[Crossref]

15. Gopalakrishnan R, Sureshkumar D. Changing trends in antimicrobial susceptibility and hospital acquired infections over an 8 year period in a tertiary care hospital in relation to introduction of an infection control programme. J Assoc Physicians India 2010;58 Suppl:25-31.

16. Blaschke AJ, Heyrend C, Byington CL, et al. Rapid identification of pathogens from positive blood cultures by multiplex polymerase chain reaction using the FilmArray system. Diagn Microbiol Infect Dis 2012;74:349-55.[Crossref]

17. Molton JS, Tambyah PA, Ang BS, Ling ML, Fisher DA. The global spread of healthcare-associated multidrug-resistant bacteria: a perspective from Asia. Clin Infect Dis 2013;56:1310-8. [Crossref]

18. Nair PK, Vaz MS. Prevalence of carbapenem resistantEnterobacteriaceaefrom a tertiary care hospital in Mumbai, India.J Microbiol Infect Dis 2013;3:207-10. [Crossref]

19. Gupta E, Mohanty S, Sood S, Dhawan B, Das BK, Kapil A. Emerging resistance to carbapenems in a tertiary care hospital in North India.Indian J Med Res.2006;124:95-8.

20. Kumarasamy KK, Toleman MA, Walsh TR, et al. Emergence of a new antibiotic resistance mechanism in India, Pakistan, and the UK: a molecular, biological, and epidemiological study. Lancet Infect Dis 2010;10:597-602.[Crossref]

21. Bauer KA, Perez KK, Forrest GN, Goff DA. Review of rapid diagnostic tests used by antimicrobial stewardship programs. Clin Infect Dis 2014;59 Suppl 3:S134-45. [Crossref]

22. Banerjee R, Teng CB, Cunningham SA, et al. Randomized trial of rapid multiplex polymerase chain reaction-based blood culture identification and susceptibility testing. Clin Infect Dis 2015;61:1071-80.[Crossref]