Abstract
Background
Various molecular methods are used to rapidly detect gastrointestinal pathogens, highlighting the importance of understanding the performance of the associated kits in detail. We comprehensively assessed the performance of Kogene PowerChek multiplex real-time PCR kits (PowerChek Bacterial/Viral Kits) with the FilmArray GI Panel.
Methods
Residual stool specimens (N=246), initially tested utilizing the FilmArray GI Panel (May 2023–Jan 2024), were reanalyzed using PowerChek Bacterial/Viral Kits. Discrepancies were resolved by performing additional molecular assays and reviewing culture results when available. True positives (TPs)/true negatives were defined by concordant results in at least two assays. We determined cycle threshold (Ct)/crossing point (Cp) distributions between the TP and false positive (FP) groups and analyzed melting curves for the FilmArray GI Panel FPs.
Results
The positive-percent agreement (PPA) of the PowerChek Bacterial/Viral Kits was 50–100%, with lower values for Salmonella spp., rotavirus, and astrovirus, whereas the FilmArray GI Panel showed 100% PPA across all targets. Both platforms demonstrated >99% negative-percent agreement, except for enteropathogenic Escherichia coli (EPEC) and adenovirus (PowerChek Bacterial/Viral Kits) or EPEC, enteroaggregative E. coli, norovirus, and Salmonella spp. (FilmArray GI Panel). The FPs showed higher Ct/Cp values with both kits, and these values were significantly higher for adenovirus (PowerChek Viral Kit), EPEC, and norovirus (FilmArray GI Panel). Melting curve analysis of four norovirus FPs (FilmArray GI Panel) revealed atypical patterns in three cases.
Conclusions
The FilmArray GI Panel demonstrated higher sensitivity than the PowerChek Kits. For norovirus, melting curve analysis will help avoid FPs.
Keywords: Crossing point, False positive, FilmArray GI Panel, Melting curve, Norovirus
INTRODUCTION
Acute gastroenteritis is a common cause of morbidity and mortality in infants and children [1]. Timely and accurate detection of gastrointestinal (GI) pathogens is essential for initiating appropriate therapeutic interventions and implementing effective infection control [2]. Although conventional stool culture is traditionally used to identify bacterial enteropathogens in clinical microbiology laboratories, this approach has limitations, including prolonged turnaround times and the need for various selective and enrichment media [3]. Recently, multiplex molecular assays were developed with remarkable clinical and operational advantages, including rapid turnaround times and ease of use [4].
The Food and Drug Administration-cleared BioFire FilmArray GI Panel (FilmArray GI Panel; bioMérieux, Marcy-l’Étoile, France) was designed as a cartridge-based system. The panel integrates automated specimen processing, nucleic acid extraction, and a nested multiplex PCR method that includes melting curve analysis to directly detect 22 gastroenteritis-related agents in stool specimens [5].
However, 20–36% false positive (FP) results were reported for norovirus with the FilmArray GI Panel [6–8]. Previously, we observed a low number of FP cases with the FilmArray GI Panel for Salmonella species [9], although the specificity was not 100% based on comparisons with results obtained with the BD MAX Enteric Panel (BD Diagnostics, Baltimore, MD, USA) and the Allplex GI-Bacteria(l) Assay (Seegene Inc., Seoul, Korea). Considering the clinical importance of accurately identifying bacterial infection (in that FP results lead to the unnecessary use of antibiotics), both high sensitivity and high specificity are important.
In our laboratory, we use both Kogene PowerChek multiplex real-time PCR kits (PowerChek Kits; Kogene Biotech, Seoul, Korea) and the FilmArray GI Panel to analyze specimens from patients with diarrhea. The former kits offer the advantage of detecting toxin-producing organisms, such as Staphylococcus aureus, Bacillus cereus, and Clostridium perfringens, and the latter provides the benefit of rapid execution and procedural simplicity.
We assessed the performance of both types of molecular tests, focusing on pathogens common to both kits. Discrepancy analysis was performed using alternative molecular assays, including the BD MAX Panel and in-house conventional PCR for diarrheagenic Escherichia coli. To the best of our knowledge, this is the first study to comprehensively assess both crossing point (Cp) values and melting curve patterns associated with true-positive (TP) and FP results obtained using the FilmArray GI Panel, offering a more detailed understanding of discordant molecular findings than previous data. While research has been conducted to evaluate the general diagnostic performance of multiplex molecular assays [6–9], few investigators have examined the analytical characteristics underlying FP results. Furthermore, this study is the first to evaluate the clinical performance of the PowerChek Bacterial/Viral Kits using residual stool specimens.
MATERIALS AND METHODS
Clinical specimens
Stool specimens from patients with diarrhea, which were subjected to FilmArray GI Panel testing per the clinician’s request, were selectively collected from both Uijeongbu St. Mary’s Hospital and Seoul St. Mary’s Hospital between May 2023 and January 2024. After conducting routine tests, the remaining stool specimens were stored at −70°C until evaluation via a comparator assay (PowerChek Kits) and for discrepancy analysis at Seoul St. Mary’s Hospital. This prospective study was approved by the Institutional Review Board (IRB) of the Catholic University of Korea (IRB approval number XC23DIDI0068). The IRB waived the requirement for informed consent, given that the specimens were collected for diagnostic purposes and the outcomes did not influence patient care.
PowerChek Kits
Viral DNA/RNA was extracted from residual stool specimens using a PowerPrep Viral DNA/RNA Extraction Kit (Kogene Biotech), per the manufacturer’s instructions. The PowerChek 19 Pathogen Multiplex Real-Time PCR Kit (PowerChek Bacterial Kit, Kogene Biotech) was used to detect 19 bacterial pathogens. The PowerChek Adeno/Astro/Rotavirus and PowerChek Norovirus GI/GII Real-Time PCR Kits (PowerChek Viral Kits, Kogene Biotech) were used to detect viral GI pathogens. All PowerChek Bacterial/Viral multiplex real-time PCR kits (PowerChek Bacterial/Viral Kits) were used according to the manufacturer’s instructions with an ABI 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) (Table 1).
Table 1. Pathogens and genes detected with PowerChek Bacterial/Viral Kits, FilmArray GI Panel, and discrepancy analysis.
| Target pathogen | PowerChek | FilmArray GI Panel | Discrepancy analysis |
|---|---|---|---|
| Bacteria | |||
| Campylobacter spp. | C. jejuni/C. coli | C. jejuni/C. coli/C. upsaliensis | BD MAX Bacterial |
| Salmonella spp. | Salmonella spp. | Salmonella spp. | BD MAX Bacterial |
| Vibrio spp. | V. parahaemolyticus/V. vulnificus/V. cholerae | V. parahaemolyticus/V. vulnificus/V. cholerae | BD MAX Bacterial |
| Yersinia enterocolitica | Y. enterocolitica | Y. enterocolitica | BD MAX Bacterial |
| Shigella spp./EIEC | Yes (ipaH) | Yes (ipaH) | Conventional E. coli PCR (ipaH) |
| ETEC | Yes (lt, st) | Yes (lt, st) | Conventional E. coli PCR (lt) |
| EHEC/STEC | Yes (stx1, stx2) | Yes (stx1, stx2) | Conventional E. coli PCR (stx1, stx2) |
| EPEC | Yes (eaeA, bfpA) | Yes (eaeA, bfpA) | Conventional E. coli PCR (eaeA, escV) |
| EAEC | Yes (aggR) | Yes (aggR, aatA) | Conventional E. coli PCR (aggR) |
| Virus | |||
| Rotavirus | Yes | Yes | BD MAX Viral |
| Norovirus | Yes | Yes | BD MAX Viral |
| Adenovirus | Yes | Yes | BD MAX Viral |
| Astrovirus | Yes | Yes | BD MAX Viral |
Abbreviations: PowerChek, Kogene PowerChek multiplex real-time PCR kit; FilmArray GI Panel, BioFire FilmArray Gastrointestinal Panel; BD MAX Bacterial, BD MAX Enteric Bacterial Panel; EIEC, enteroinvasive E. coli; ETEC, enterotoxigenic E. coli; EHEC, enterohemorrhagic E. coli; STEC, Shiga toxin-producing E. coli; EPEC, enteropathogenic E. coli; EAEC, enteroaggregative E. coli; BD MAX Viral, BD MAX Enteric Viral Panel.
FilmArray GI Panel
Utilizing nested PCR and melting curve analysis, the FilmArray GI Panel operates as an integrated automated system, facilitating nucleic acid extraction, amplification, and detection of 22 GI pathogens in a single platform, including 13 bacteria, five viruses, and four parasites. Stool specimens were suspended in Cary–Blair medium and 200 µL aliquots of each specimen were used for FilmArray GI Panel testing, per the manufacturer’s instructions. Amplification-cycle values were derived from the fluorescence data generated after the first PCR stage. Using a recently launched application (BioFire FIREWORKS), users can visualize amplification curves and Cp values for positive qualitative results, which are conceptually similar to cycle threshold (Ct) or quantification cycle values in real-time PCR systems. However, BioFire panels utilize nested PCR, with the second stage of PCR generally completing in <30 cycles; thus, the cycle values are not directly comparable to those of standard PCR assays [10]. Melting curves were reviewed in cases where discrepancies were found.
Discrepancy analysis with a BD MAX Panel and conventional E. coli PCR
We compared the results obtained with the PowerChek Bacterial/Viral Kits and FilmArray GI Panel for pathogens targeted by both platforms. These pathogens included nine bacteria (Campylobacter spp., Salmonella spp., Vibrio spp., Yersinia enterocolitica, Shigella spp./enteroinvasive E. coli [EIEC], enterotoxigenic E. coli [ETEC], enterohemorrhagic E. coli/Shiga toxin-producing E. coli [STEC], enteropathogenic E. coli [EPEC], and enteroaggregative E. coli [EAEC]) and four viruses (rotavirus, norovirus, adenovirus, and astrovirus).
Discrepancies between the PowerChek Bacterial/Viral Kits and FilmArray GI Panel were analyzed further using a BD MAX Enteric Bacterial Panel and a BD MAX Enteric Viral Panel (BD MAX Panel) per the manufacturer’s instructions (Table 1). An in-house conventional PCR test was performed for diarrheagenic E. coli [11]. This PCR detects aggR in EAEC, escV, and eaeA in EPEC; lt in ETEC; stx1/stx2 in STEC; and ipaH in EIEC. ‘Consensus’ TP and true negative (TN) results were defined as positive/negative results from at least two molecular assays that included the target organisms. These estimates were defined in terms of the positive-percent agreement (PPA) or negative-percentage agreement (NPA). The performance characteristics of the assays were calculated using the following formulas: PPA=TP/(TP+false negative [FN]) and NPA=TN/(TN+FP). The terms PPA and NPA are used instead of sensitivity and specificity when the comparator analysis involves a non-gold-standard assay, such as PCR.
However, clinicians occasionally request both the FilmArray GI Panel testing and bacterial culture; therefore, we analyzed the culture results for discrepant cases when the results were available. The bacteria were cultured as described by Jo, et al. [9]. Furthermore, for cases showing discrepancies with diarrheagenic E. coli, medical records were reviewed for demographic characteristics, clinical symptoms, and treatment history.
Statistical analysis
The PPA, NPA, and 95% confidence intervals were calculated using GraphPad QuickCalcs. Significant differences in the distribution of Ct/Cp values between the TP and FP groups for both kits were assessed using the Mann–Whitney U-test, using SPSS (v28.0; IBM Corporation, Armonk, NY, USA). In all cases, P<0.05 was considered to reflect a statistically significant difference.
RESULTS
We selected 247 stool specimens for this comparison study. One specimen was excluded because a DNA specimen was unavailable for discrepancy analysis. Positive results were obtained for 105 of 246 stool specimens (42.7%) using the PowerChek Bacterial/Viral Kits, with multiple pathogens detected in 20 specimens (8.1%). The FilmArray GI Panel detected one or more GI pathogens in 132 specimens (53.7%), with multiple pathogens identified in 18 specimens (7.3%). As shown in Table 2, the PPA of the PowerChek Bacterial/Viral Kits ranged from 50% to 100%, whereas the PPA of the FilmArray GI Panel was 100% across all targets. With the PowerChek Bacterial/Viral Kits, three pathogens displayed low PPA values (<90%), namely Salmonella spp. (14/16, 87.5%), rotavirus (3/6, 50.0%), and astrovirus (22/25, 88.0%). For both kits, the NPA of most targets was >99%. However, the PowerChek Bacterial/Viral Kits exhibited slightly lower NPAs for two pathogens: 97.8% for EPEC (five FPs) and 97.0% for adenovirus (seven FPs). The NPA values of the FilmArray GI Panel were <99% for four pathogens, with EPEC showing the lowest NPA (95.7% with 10 FPs), followed by EAEC (97.1% with seven FPs), norovirus (97.8% with five FPs), and Salmonella spp. (98.7% with three FPs) (Table 2).
Table 2. Performance summary and characteristics of the PowerChek Bacterial/Viral Kits and FilmArray GI Panel versus those of comparator assays (BD MAX Bacterial/Viral Panel and conventional Escherichia coli PCR).
| Pathogen | PowerChek | FilmArray GI Panel | |||||
|---|---|---|---|---|---|---|---|
| PPA (95% CI) | NPA (95% CI) | TP/FP/TN/FN | PPA (95% CI) | NPA (95% CI) | TP/FP/TN/FN | ||
| Bacteria | |||||||
| Campylobacter spp. | 95.8 (78.9–99.9) | 100 (98.4–100) | 23/0/222/1 | 100 (85.8–100) | 99.1 (96.8–99.9) | 24/2/220/0 | |
| Salmonella spp. | 87.5 (61.7–98.5) | 100 (98.4–100) | 14/0/230/2 | 100 (79.4–100) | 98.7 (96.2–99.7) | 16/3/227/0 | |
| Vibrio spp. | NA | 99.6 (97.8–100) | 0/1/245/0 | NA | 100 (98.5–100) | 0/0/246/0 | |
| Yersinia enterocolitica | 100 (15.8–100) | 100 (98.5–100) | 2/0/244/0 | 100 (15.8–100) | 100 (98.5–100) | 2/0/244/0 | |
| Shigella spp./EIEC | NA | 100 (98.5–100) | 0/0/246/0 | NA | 99.6 (97.8–100) | 0/1/245/0 | |
| ETEC | 100 (2.5–100) | 100 (98.5–100) | 1/0/245/0 | 100 (2.5–100) | 99.6 (97.8–100) | 1/1/244/0 | |
| EHEC/STEC | 100 (15.8–100) | 100 (98.5–100) | 2/0/244/0 | 100 (15.8–100) | 99.2 (97.1–99.9) | 2/2/242/0 | |
| EPEC | 93.3 (68.1–99.8) | 97.8 (95.0–99.3) | 14/5/226/1 | 100 (78.2–100) | 95.7 (92.2–97.9) | 15/10/221/0 | |
| EAEC | 100 (54.1–100) | 99.6 (97.7–100) | 6/1/239/0 | 100 (54.1–100) | 97.1 (94.1–98.8) | 6/7/233/0 | |
| Virus | |||||||
| Rotavirus | 50.0 (11.8–88.2) | 100 (98.5–100) | 3/0/240/3 | 100 (54.1–100) | 99.6 (97.7–100) | 6/1/239/0 | |
| Norovirus | 92.9 (66.1–99.8) | 100 (98.4–100) | 13/0/232/1 | 100 (76.8–100) | 97.8 (95.0–99.3) | 14/5/227/0 | |
| Adenovirus | 100 (69.2–100) | 97.0 (94.0–98.8) | 10/7/229/0 | 100 (69.2–100) | 100 (98.5–100) | 10/0/236/0 | |
| Astrovirus | 88.0 (68.8–97.5) | 100 (98.3–100) | 22/0/221/3 | 100 (86.3–100) | 99.5 (97.5–100) | 25/1/220/0 | |
Abbreviations: PowerChek Bacterial Kit, Kogene PowerChek 19 Pathogen Multiplex Real-Time PCR Kit; PowerChek Viral Kits; PowerChek Adeno/Astro/Rotavirus and PowerChek Norovirus GI/GII Real-Time PCR Kits; FilmArray GI Panel, BioFire FilmArray Gastrointestinal Panel; BD MAX Bacterial, BD MAX Enteric Bacterial Panel; PPA, positive-percent agreement; CI, confidence interval; NPA, negative-percent agreement; TP, true positive; FP, false positive; TN, true negative; FN, false negative; NA, not available; EIEC, enteroinvasive E. coli; ETEC, enterotoxigenic E. coli; EHEC, enterohemorrhagic E. coli; STEC, Shiga toxin-producing E. coli; EPEC, enteropathogenic E. coli; EAEC, enteroaggregative E. coli.
Discrepancies in bacterial detection were observed with 30 specimens. Among these, eight cases with discrepancies for Salmonella spp., Campylobacter spp., and Vibrio spp. were tested further using the BD MAX Panel, and their culture results were also reviewed (Table 3). Follow-up discrepancy analysis with the BD MAX Panel indicated that the PowerChek Bacterial Kit yielded two FN results for Salmonella spp. and one FN result for Campylobacter spp. Bacterial stool cultures for Salmonella spp. were positive in both cases, although bacterial culture was not performed originally for the specimen with an FN for Campylobacter spp. The FilmArray GI Panel yielded three FP results for Salmonella spp. and two FP results for Campylobacter spp. Among these, bacterial cultures were positive for Salmonella spp. in one case and for Campylobacter spp. in another. In one case, the FilmArray GI Panel detected Campylobacter spp., and the PowerChek Bacterial Kit identified Vibrio parahaemolyticus with a Ct value of 32.28. The BD MAX Panel yielded a negative result; however, the bacterial culture was positive for Campylobacter spp. (Table 3).
Table 3. Discrepancy analysis results, including conventional culture results for Salmonella spp., Campylobacter spp., and Vibrio spp.
| No. case | PowerChek (Ct value)* | FilmArray GI Panel (Cp value)* | BD MAX Bacterial | Conventional culture |
|---|---|---|---|---|
| 15 | Negative (Salmonella spp.)† | Salmonella spp. | Salmonella spp. | Salmonella spp. |
| 56 | Negative (Salmonella spp.)† | Salmonella spp. | Salmonella spp. | Salmonella spp. |
| 84 | Negative | Salmonella spp.‡ (27.3) | Negative | Salmonella spp. |
| 67 | Negative | Campylobacter spp.‡ (ND) | Negative | No growth |
| 99 | Negative | Salmonella spp.‡ (16.5) | Negative | Not done |
| 114 | Negative | Salmonella spp.‡ (NA) | Negative | No growth |
| 116 | Negative (Campylobacter spp.)† | Campylobacter spp. | Campylobacter spp. | Not done |
| 121 | V. parahaemolyticus (32.28)‡ | Campylobacter spp.‡ (16.2) | Negative | Campylobacter spp. |
*Ct or Cp values were provided only for false positive cases.
†Specimens showing false negative results are indicated.
‡Specimens showing false positive results are indicated.
Abbreviations: PowerChek, Kogene PowerChek Multiplex Real-Time PCR Kit; FilmArray GI Panel, BioFire FilmArray Gastrointestinal Panel; BD MAX Bacterial, BD MAX Enteric Bacterial Panel; Ct: cycle threshold; Cp, crossing point; ND, Cp value not detected with the FilmArray GI Panel; NA, not available.
Conventional PCR was performed to further analyze 22 cases with discrepancies between the PowerChek Bacterial Kit and the FilmArray GI Panel, in terms of diarrheagenic E. coli detection. Discrepancy analysis indicated that the PowerChek Bacterial Kit yielded five FP results (four for EPEC and one for EPEC/EAEC) and one FN result for EPEC. The FilmArray GI Panel yielded 16 FP results, including eight for EPEC, five for EAEC, one for EPEC/EAEC, one for EPEC/EAEC/EIEC, and one for ETEC. All patients exhibited clinical symptoms of GI infection. Among the 11 patients for whom the FilmArray GI Panel showed FP results for EPEC or EPEC/EAEC as the sole bacterial pathogen, seven received antibiotic treatment (Table 4).
Table 4. Discrepancy analysis results for clinical findings for diarrheagenic Escherichia coli .
| Case No. | PowerChek (Ct value)* |
FilmArray GI Panel (Cp value)* |
Conventional diarrheagenic E. coli PCR | Age (yrs) | Clinical symptoms (abdominal pain, fever, loose stools, nausea, and/or vomiting) | Treatment history |
|---|---|---|---|---|---|---|
| 20 | Negative | EAEC† (27.1) | Negative | 4 | Yes | Cefpodoxime (3 days) |
| 25 | Campylobacter jejuni | Campylobacter spp., EAEC† (8.9) | Negative | 58 | Yes | Ciprofloxacin (7 days) |
| 40 | Astrovirus |
Astrovirus EAEC† (24.3) |
Negative | 1 | Yes | Amoxicillin/clavulanate (3 days) |
| 85 | Negative | EAEC† (26.9) | Negative | 7 | Yes | Cefditoren (6 days) |
| 87 | Negative | EAEC† (25.6) | Negative | 53 | Yes |
Piperacillin/tazobactam (6 days) |
| 29 | EHEC, EPEC† (37.66) | STEC | Negative | 13 | Yes | No |
| 37 | Campylobacter jejuni, EPEC† (31.76) | Campylobacter spp. | Negative | 11 | Yes | Cefotaxime (2 days) |
| 45 | Negative | EPEC†, EAEC†, Shigella spp./EIEC† (26.4, 24.9, 22.1) | Negative | 8 | Yes |
Cefotaxime (2 days) Cefpodoxime (4 days) |
| 52 | Negative | EPEC† (26.6) | Negative | 2 | Yes | Cefpodoxime (8 days) |
| 53 | Astrovirus |
Astrovirus EPEC† (18.4) |
Negative | 9 | Yes | No |
| 70 | Negative | EPEC† (25.9) | Negative | 83 | Yes | No |
| 76 | Negative | EPEC† (19.4) | Negative | 44 | Yes | Ciprofloxacin (10 days) |
| 77 | Negative | EPEC† (16.3) | Negative | 13 | Yes | No |
| 86 | Negative | EPEC† (16.1) | Negative | 70 | Yes | No |
| 89 | Negative | EPEC† (11.1) | Negative | 7 | Yes |
Cefotaxime (2 days) Cefditoren (3 days) |
| 55 | Negative | EPEC | EPEC | 67 | Yes | No |
| 63 | EAEC† (25.47), EPEC† (28.7) | STEC† (NA) | Negative | 87 | Yes |
Piperacillin/tazobactam (4 days) Ertapenem (2 days) |
| 73 | Negative | EAEC†, EPEC† (14.0, 16.1) | Negative | 64 | Yes | No |
| 78 | Campylobacter jejuni, EAEC | Campylobacter spp., EAEC, EPEC† (16.1) | Negative | 12 | Yes | Clarithromycin (4 days) |
| 79 | EHEC, EPEC† (31.46) | STEC | Negative | 3 | Yes |
Azithromycin (4 days) Meropenem (3 days) |
| 115 | Negative | ETEC† (21.6) | Negative | 69 | Yes |
Levofloxacin (5 days) Meropenem (5 days) |
| 120 | EPEC† (30.61) | STEC† (19.2) | Negative | 8 | Yes | No |
*Ct or Cp values are provided only for false positive cases.
†Specimens showing false positive results are indicated.
Abbreviations: PowerChek, Kogene PowerChek multiplex real-time PCR kit; Ct: cycle threshold; FilmArray GI Panel, BioFire FilmArray Gastrointestinal Panel; Cp, crossing point; EAEC, enteroaggregative E. coli; EHEC, enterohemorrhagic E. coli; EPEC, enteropathogenic E. coli; STEC: Shiga toxin-producing E. coli; NA, not available; EIEC, enteroinvasive E. coli; ETEC, enterotoxigenic E. coli.
EPEC FPs were the most common errors found with both the PowerChek Bacterial Kit (N=5) and the FilmArray GI Panel (N=10). The EPEC Ct values identified as FPs with the PowerChek Bacterial Kit ranged from 28.7 to 37.7, whereas those of the TP cases ranged from 22.9 to 37.5 (P=0.308; Fig. 1). The Cp values of 10 EPEC specimens identified as FPs with the FilmArray GI Panel ranged from 11.1 to 27.3, which were significantly different than the Cp values of 15 TP EPEC specimens (range: 5.5–22.1, P=0.003). The EAEC Cp values (range: 6.8–24.9) of the six cases confirmed as TPs and those of FP cases (range: 8.9–27.1) were determined using the FilmArray GI Panel (P=0.053). For EPEC and EAEC, no differences in the melting curves were observed between TP and FP cases. Additionally, unimodal and multi-nodal patterns were identified for both TP and FP EAEC results (Supplemental Data Fig. S1).
Fig. 1. Distributions of cycle threshold (Ct) and crossing point (Cp) values of true positive (TP) and false positive (FP) results for pathogens detected with the PowerChek Kit and FilmArray GI Panel. (A) Distributions of Ct values for TP and FP results found when detecting enteropathogenic Escherichia coli (EPEC) and adenovirus using the PowerChek Kit. (B) Distribution of Cp values for TP and FP results found when detecting EPEC, enteroaggregative E. coli (EAEC), and norovirus using the FilmArray GI Panel.
*P<0.05.
The discordant results for the viral pathogens are summarized in Table 5, with 20 specimens showing discrepancies. For viral pathogens, the PowerChek Viral Kit showed seven FP results for adenovirus, and the FilmArray GI Panel showed seven FP results (five for norovirus, one for rotavirus, and one for astrovirus) (Table 5). Among these, all FP results observed with the PowerChek Viral Kit were attributed to adenovirus, accounting for seven cases with Ct values ranging from 27.42 to 37.02, which differed significantly from the TP Ct values (P=0.015; Fig. 1). Among the five norovirus FPs detected with the FilmArray GI Panel, two did not show Cp values. The distribution of the Cp values between the TP and FP groups was significantly different (P=0.012). In addition, melting curve analysis was performed for the norovirus FP specimens, three of which showed atypical, multi-nodal peaks (Supplemental Data Fig. S1).
Table 5. Discrepancy analysis results for viral targets.
| Case No. | PowerChek (Ct value)* | FilmArray GI Panel (Cp value)* | BD MAX Viral |
|---|---|---|---|
| 6, 12, 14, 18 | Negative | Norovirus† (27.4, n.d., 27.6, n.d.) | Negative |
| 10, 31, 65 | Negative‡ | Astrovirus | Astrovirus |
| 13, 129, 130 | Negative‡ | Rotavirus | Rotavirus |
| 21 | Adenovirus (35.95)†, Astrovirus | Astrovirus | Astrovirus |
| 38 | Adenovirus (31.86)† | Norovirus† (30.0) | Negative |
| 41, 45, 48, 61 | Adenovirus† (35.45, 33.27, 32.57, 37.02) | Negative | Negative |
| 58 | Negative | Astrovirus† (24.9) | Negative |
| 102 | Negative | Rotavirus† (28.4) | Negative |
| 125 | Adenovirus (27.42)†, Norovirus | Norovirus | Norovirus |
| 133 | Negative‡ | Norovirus | Norovirus |
*Ct or Cp values are provided only for false positive cases.
†Specimens showing false positive results are indicated.
‡Specimens showing false negative results are indicated.
Abbreviations: PowerChek, Kogene PowerChek Multiplex Real-Time PCR Kit; FilmArray GI Panel, BioFire FilmArray Gastrointestinal Panel; BD MAX Viral, BD MAX Enteric Viral Panel; Ct, cycle threshold; n.d.; Cp value not detected with the FilmArray GI Panel.
DISCUSSION
Molecular techniques are generally more sensitive and less dependent on specimen quality than traditional culture methods. The Infectious Diseases Society of America recommends using molecular GI panels because they are effective diagnostic tools for detecting a broad range of enteric pathogens, given their higher sensitivity and quicker turnaround times than culture-based methods [12]. We determined the frequency of FP and FN results for each target and analyzed the Cp values and melting curve patterns for FP specimens in cases where the FilmArray GI Panel showed FP results. To our knowledge, this is the first study to evaluate both the Cp values and melting curves for TP and FP cases identified with the FilmArray GI Panel. Additionally, this is the first study to assess the clinical performance of the PowerChek Bacterial/Viral Kits.
In general, the FilmArray GI Panel detected more targets than the PowerChek Bacterial/Viral Kits. These findings are consistent with those of a previous study performed to compare the Allplex GI-Bacteria(l) Assay and FilmArray GI Panels, which demonstrated that the FilmArray GI Panel generally detected more targets than the Allplex assay. Specifically, for EPEC, the kappa value was notably low at 0.36, indicating a fair level of agreement, with the FilmArray GI Panels identifying 13 additional cases [13].
The PowerChek Bacterial/Viral Kits and FilmArray GI Panel exhibited a notably high number of FPs for EPEC. Conventional PCR assays employed escV instead of bfpA to validate discrepant results. Because both the PowerChek Bacterial Kit and the FilmArray GI Panel detect the same target genes (eaeA and bfpA), this difference in the target gene may not fully account for the high number of FPs observed with both kits. Although neither the Ct from the PowerChek Bacterial Kit nor the Cp value from the FilmArray GI Panel showed a significant difference between TP and FP cases, the differences for the TP specimens were smaller than those for the FP specimens.
Among the targets of the FilmArray GI Panel, EAEC showed the second-most common FP results. This outcome might reflect differences in the target gene in that the PowerChek Bacterial Kit and conventional PCR assays for diarrheagenic E. coli rely on a single gene (aggR), whereas the FilmArray GI Panel detects both the aggR and aatA genes for EAEC. Detecting EAEC via PCR targeting a single gene, aggR, may be effective for identifying typical EAEC; however, this approach is not applicable for detecting atypical EAEC strains lacking the gene. A previous analysis of 177 E. coli O111 strains with the EAEC pathotype showed that 116 strains (65.5%) were aggR(+)aatA(+), 18 strains (10.2%) were aggR(+)aatA(–), and 24 strains (13.6%) were aggR(–)aatA(+) [14].
The occurrence of EPEC and EAEC FPs is documented in the clinical performance section of the FilmArray GI Panel insert. According to the insert, EAEC demonstrated an NPA of 98.2%, with 27 FP results out of 1,473 specimens. Similarly, EPEC had an NPA of 97.2%, with 34 FP results reported for 1,201 specimens. The comparator method used in this clinical evaluation was PCR with bidirectional sequencing. ETEC, EPEC, and EAEC have not been routinely identified in clinical settings (with the exception of enterohemorrhagic E. coli/STEC), raising questions about their clinical relevance. These pathogens, like other enteric organisms, can either colonize asymptomatically or cause GI symptoms in affected individuals [15]. However, the Cp values for TPs were lower than those for FPs with both EPEC and EAEC. The mean Cp values for the TP/FP cases were 12.0/17.4 for EPEC and 11.4/24.9 for EAEC, respectively (Fig. 1).
Based on our results, the PowerChek Bacterial Kit showed FN results for Campylobacter spp. (1/24, 4.2%) and Salmonella spp. (2/16, 12.5%). Considering the clinical importance of these two pathogens, such high FN rates are alarming, and improvement of the PowerChek Bacterial Kit is needed. Failure to detect an infection because of an FN result could delay treatment, potentially leading to severe health consequences and increased public health costs [16]. In contrast, FilmArray GI Panel showed three FP results for Salmonella spp., consistent with a study by Jo, et al. [9], where two FP cases of Salmonella spp. were identified among 184 specimens. In a comparative study [17] between the Luminex xTAG GI Pathogen Panel (xTAG GPP; Luminex Molecular Diagnostics, Austin, TX, USA) and the FilmArray GI Panel, the respective positive rates were 65.0% and 48.6%, and the consistency was low for Salmonella spp. (kappa ≤0.40). The authors also reported a statistically significant difference in the mean Cp values between several pathogens detected with both panels versus those detected only with the FilmArray GI Panel, including adenovirus, norovirus GII, rotavirus, Salmonella spp., and ETEC. Specifically for Salmonella, the mean Cp values for cases detected with both panels (N=7) and those detected exclusively with the FilmArray GI Panel (N=11) were 15.2 and 27.0, respectively [17]. Here, the Cp values for the three FP Salmonella spp. cases were 27.3 (specimen No. 84), 16.5 (specimen No. 99), and not available (specimen No. 114). Although the BD MAX Panel was also negative for all three cases, bacterial culture results were available for two of them (specimens No. 84 and 114) and Salmonella spp. growth was confirmed for specimen No. 84 (Table 3). For specimen No. 99, which had a lower Cp value (16.5), bacterial culture was not performed. These findings suggest that the higher sensitivity of the FilmArray GI Panel might have led to misinterpretation as an FP result.
For viruses, the PowerChek Viral Kit showed seven FP results for adenovirus, and the Ct values of the FPs were significantly higher than those of the TPs. The FilmArray GI Panel showed FP results for norovirus, and the Cp values were also significantly different. Although a statistical difference in Ct/Cp values was observed, the small specimen size of the study prevents the establishment of a definitive cutoff value for distinguishing TP and FP results. However, in the case of norovirus, an atypical pattern observed via melting curve analysis may warrant retesting or the use of alternative testing methods.
This study has a limitation in that the number of specimens containing pathogens was relatively small. For certain bacterial species (Vibrio spp. and Y. enterocolitica), an insufficient number of positive specimens was available, limiting the ability to fully evaluate the performance of the kits. Among the specimens tested with the FilmArray GI Panel, only a subset underwent routine stool culture; therefore, the culture results could not be included as the gold standard in the discrepancy analysis.
In conclusion, the FilmArray GI Panel showed higher sensitivity than the PowerChek Bacterial/Viral Kits. Considering that the Cp values of the TP/FP specimens differed, reviewing the Cp values and melting curves is important for avoiding FP results, and further data are needed given the low number of cases studied. The PowerChek Bacterial/Viral Kits showed good performance for most pathogens, but improvement is required to detect several clinically relevant pathogens, such as Salmonella spp., Campylobacter spp., and rotavirus.
ACKNOWLEDGEMENTS
This work includes material that was presented as a poster at the ASM Microbe 2025 in Los Angeles, on June 21, 2025.
SUPPLEMENTARY MATERIALS
Supplementary materials can be found via https://doi.org/10.3343/alm.2025.0047.
Footnotes
AUTHOR CONTRIBUTIONS
Park YJ and Yoo IY conceptualized the study and coordinated the drafting of the manuscript. Kwon JA and Han JH participated in specimen collection and experiments. Jo S and Yoo IY conducted the data analysis and wrote the manuscript. Lee HK and Park YJ supervised the study design and reviewed the manuscript. All authors read and approved the final manuscript.
CONFLICTS OF INTEREST
None declared.
RESEARCH FUNDING
This study was supported by the Samkwang Medical Lab (grant numbers SMLCR-CMCS4-2023-001 and SMLCR-CMCS6-2023-002).
REFERENCES
- 1.Guarino A, Aguilar J, Berkley J, Broekaert I, Vazquez-Frias R, Holtz L, et al. Acute gastroenteritis in children of the world: what needs to be done? J Pediatr Gastroenterol Nutr. 2020;70:694–701. doi: 10.1097/MPG.0000000000002669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Hatchette TF, Farina D. Infectious diarrhea: when to test and when to treat. CMAJ. 2011;183:339–44. doi: 10.1503/cmaj.091495. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Anderson NW, Tarr PI. Multiplex nucleic acid amplification testing to diagnose gut infections: challenges, opportunities, and result interpretation. Gastroenterol Clin North Am. 2018;47:793–812. doi: 10.1016/j.gtc.2018.07.006. [DOI] [PubMed] [Google Scholar]
- 4.Chhabra P, Gregoricus N, Weinberg GA, Halasa N, Chappell J, Hassan F, et al. Comparison of three multiplex gastrointestinal platforms for the detection of gastroenteritis viruses. J Clin Virol. 2017;95:66–71. doi: 10.1016/j.jcv.2017.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhang H, Morrison S, Tang YW. Multiplex polymerase chain reaction tests for detection of pathogens associated with gastroenteritis. Clin Lab Med. 2015;35:461–86. doi: 10.1016/j.cll.2015.02.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Matic N, Lawson T, Young M, Jang W, Bilawka J, Gowland L, et al. Melting curve analysis reveals false positive norovirus detection in a molecular syndromic panel. J Clin Virol. 2024;173:105697. doi: 10.1016/j.jcv.2024.105697. [DOI] [PubMed] [Google Scholar]
- 7.Ramanan P, Espy MJ, Khare R, Binnicker MJ. Detection and differentiation of norovirus genogroups I and II from clinical stool specimens using real-time PCR. Diagn Microbiol Infect Dis. 2017;87:325–7. doi: 10.1016/j.diagmicrobio.2017.01.009. [DOI] [PubMed] [Google Scholar]
- 8.Buss SN, Leber A, Chapin K, Fey PD, Bankowski MJ, Jones MK, et al. Multicenter evaluation of the BioFire FilmArray gastrointestinal panel for etiologic diagnosis of infectious gastroenteritis. J Clin Microbiol. 2015;53:915–25. doi: 10.1128/JCM.02674-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jo SJ, Kang HM, Kim JO, Cho H, Heo W, Yoo IY, et al. Evaluation of the BioFire gastrointestinal panel to detect diarrheal pathogens in pediatric patients. Diagnostics (Basel) 2021;12:34. doi: 10.3390/diagnostics12010034.4ec2052777b247cf920163aafcfb825e [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Poritz MA, Blaschke AJ, Byington CL, Allen L, Nilsson K, Jones DE, et al. FilmArray, an automated nested multiplex PCR system for multi-pathogen detection: development and application to respiratory tract infection. PLoS One. 2011;6:e26047. doi: 10.1371/journal.pone.0026047.e7ad8b4b51a1448aad5d2451b0503fca [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Oh KH, Kim SB, Park MS, Cho SH. Development of a one-step PCR assay with nine primer pairs for the detection of five diarrheagenic Escherichia coli types. J Microbiol Biotechnol. 2014;24:862–8. doi: 10.4014/jmb.1312.12031. [DOI] [PubMed] [Google Scholar]
- 12.Shane AL, Mody RK, Crump JA, Tarr PI, Steiner TS, Kotloff K, et al. 2017 Infectious Diseases Society of America clinical practice guidelines for the diagnosis and management of infectious diarrhea. Clin Infect Dis. 2017;65:e45–80. doi: 10.1093/cid/cix669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jitmuang A, Lertlaksameewilai P, Poorichitiporn A, Horthongkham N, Chayakulkeeree M. Multiplex gastrointestinal panel testing in hospitalized patients with acute diarrhea in Thailand. Open Forum Infect Dis. 2024;11:ofae322. doi: 10.1093/ofid/ofae322. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schlosserová K, Daniel O, Labská K, Jakubů V, Stárková T, Bílý J, et al. Enteroaggregative Escherichia coli: frequent, yet underdiagnosed pathotype among E. coli O111 strains isolated from children with gastrointestinal disorders in the Czech Republic. Int J Med Microbiol. 2024;316:151628. doi: 10.1016/j.ijmm.2024.151628.560ccfada8444a3a9f2e4c69b4b85ca0 [DOI] [PubMed] [Google Scholar]
- 15.Abbott AN, Fang FC. Clinical impact of multiplex syndromic panels in the diagnosis of bloodstream, gastrointestinal, respiratory, and central nervous system infections. Clin Microbiol Newsl. 2017;39:133–42. doi: 10.1016/j.clinmicnews.2017.08.004. [DOI] [Google Scholar]
- 16.Chirambo AC, Nyirenda TS, Jambo N, Msefula C, Kamng'ona A, Molina S, et al. Performance of molecular methods for the detection of Salmonella in human stool specimens. Wellcome Open Res. 2020;5:237. doi: 10.12688/wellcomeopenres.16305.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Zhan Z, Guo J, Xiao Y, He Z, Xia X, Huang Z, et al. Comparison of BioFire FilmArray gastrointestinal panel versus Luminex xTAG Gastrointestinal Pathogen Panel (xTAG GPP) for diarrheal pathogen detection in China. Int J Infect Dis. 2020;99:414–20. doi: 10.1016/j.ijid.2020.08.020.abd9917cd032412ebb66cbc27b3ed744 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.