Abstract
Diarrhea due to enteric bacterial pathogens causes significant morbidity and mortality in the United States and worldwide. However, bacterial pathogens may be infrequently identified. Currently, culture and enzyme immunoassays (EIAs) are the primary methods used by clinical laboratories to detect enteric bacterial pathogens. We conducted a multicenter evaluation of the BD Max enteric bacterial panel (EBP) PCR assay in comparison to culture for the detection of Salmonella spp., Shigella spp., Campylobacter jejuni, and Campylobacter coli and an EIA for Shiga toxins 1 and 2. A total of 4,242 preserved or unpreserved stool specimens, including 3,457 specimens collected prospectively and 785 frozen, retrospective samples, were evaluated. Compared to culture or EIA, the positive percent agreement (PPA) and negative percent agreement (NPA) values for the BD Max EBP assay for all specimens combined were as follows: 97.1% and 99.2% for Salmonella spp., 99.1% and 99.7% for Shigella spp., 97.2% and 98.4% for C. jejuni and C. coli, and 97.4% and 99.3% for Shiga toxins, respectively. Discrepant results for prospective samples were resolved with alternate PCR assays and bidirectional sequencing of amplicons. Following discrepant analysis, PPA and NPA values were as follows: 97.3% and 99.8% for Salmonella spp., 99.2% and 100% for Shigella spp., 97.5% and 99.0% for C. jejuni and C. coli, and 100% and 99.7% for Shiga toxins, respectively. No differences in detection were observed for samples preserved in Cary-Blair medium and unpreserved samples. In this large, multicenter study, the BD Max EBP assay showed superior sensitivity compared to conventional methods and excellent specificity for the detection of enteric bacterial pathogens in stool specimens.
INTRODUCTION
The World Health Organization reports that, worldwide, there are nearly 1.7 billion cases of diarrheal disease every year and that diarrheal disease is the second leading cause of death of children <5 years old. Each year, diarrhea kills ∼760,000 children under the age of 5 years, and most importantly, it is preventable and treatable. Diarrhea is also a leading cause of malnutrition in this same age group. Most of this disease is related to unsafe drinking water, inadequate sanitation, and poor hygiene (1, 2).
Acute gastroenteritis is caused by a number of bacterial, viral, and parasitic agents (3). In the United States, noroviruses cause most cases of self-limited, acute gastroenteritis. However, Salmonella, Campylobacter, Shigella, and Shiga toxin-producing Escherichia coli (STEC) are the most common diarrheagenic bacteria, and they are usually associated with foodborne illness (4). Furthermore, Shigella is more frequently transmitted from person to person due to the low infectious dose. Importantly, agents of gastroenteritis may not be distinguished clinically. Identifying the cause of diarrhea is important for both the treatment of individual patients and public health intervention through outbreak management (5).
Conventional microbiological cultures remain the gold standard for identification, despite their limited sensitivity. Additionally, traditional methods are time-consuming and labor-intensive and require considerable technical skill. The application of nucleic acid amplification methods could have significant impact on diagnosis and treatment, as well as our understanding of the epidemiology of this disease (6, 7).
The BD Max enteric bacterial panel (EBP) is a multiplex nucleic acid amplification assay designed for the detection of Salmonella spp., Shigella spp., Campylobacter spp. (C. jejuni and C. coli), and Shiga-like toxin genes (stx1 and/or stx2) in stool specimens with the BD Max system (BD Diagnostics, Baltimore, MD, USA). The BD Max system is a walkaway PCR instrument that can process the specimen and amplify and detect nucleic acids in a batch of up to 24 samples in ∼3 h using only 1.25 min of hands-on time per sample (8). In contrast, routine diagnostic culture methods typically require 24 to 72 h of incubation, with additional technical time for the identification of pathogens by matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) or biochemical methods.
A number of commercially available diagnostic systems for the detection of enteric pathogens have recently been developed. These systems use a variety of approaches and differ as to the number and type of targets incorporated in the assay and the overall platform design and throughput. The Prodesse ProGastro SSCS assay (Gen-Probe Prodesse, San Diego, CA) detects the same pathogens as those detected by the BD Max EBP assay, but unlike the BD Max assay, the ProGastro SSCS assay requires external extraction on a separate instrument, increasing the hands-on time and technical complexity (9). Other systems have been designed to detect viral and parasitic enteric pathogens as well as additional bacterial targets. The Luminex xTag gastrointestinal pathogen panel (GPP) assay (Luminex Corp., Austin, TX) can detect up to 15 targets (9 bacteria, 3 viruses, and 3 parasites) (6, 10), and the FilmArray gastrointestinal (GI) panel (Biofire Diagnostics, Inc., Salt Lake City, UT) contains 23 targets, including 14 bacterial, 5 viral, and 4 parasitic targets (9). The Luminex methodology is a high-throughput assay but also requires an external extraction step. The FilmArray assay requires very little technical time, as all analyses are performed with a single instrument, with results being available in an hour. Throughput is lacking, however, as only one sample at a time can be tested on each instrument. The Verigene Enteric Pathogens (EP) test (Nanosphere, Northbrook, IL), detects 7 bacterial and 2 viral targets. Like the FilmArray assay, throughput is limited. Single stool specimens are extracted and amplified on a processing instrument and moved to the Verigene reader for final analysis (11). From the description of these assays, it is clear that laboratories can choose from a variety of test platforms that may be implemented depending on whether a focused or broad approach to pathogen detection is desired. The need for throughput and proficiency of the technical staff with molecular assays may weigh heavily on the choice of assay. The BD Max EBP assay focuses on only the most prevalent bacterial enteric pathogens, might be characterized as having medium throughput, and requires almost no previous technical expertise in molecular diagnostics.
We conducted a large, multicenter evaluation of the BD Max EBP assay in comparison to routine diagnostic culture for Salmonella spp., Shigella spp., C. coli, and C. jejuni and an enzyme immunoassay (EIA) for Shiga toxins. Stool samples were collected prospectively and included specimens preserved in Cary-Blair transport medium and unpreserved specimens.
(The results of this study were presented in part at the 114th General Meeting of the American Society for Microbiology, Boston, MA, 17 to 20 May 2014, and at the 24th European Congress of Clinical Microbiology and Infectious Diseases, Barcelona, Spain, 10 to 13 May 2014.)
MATERIALS AND METHODS
Specimens.
Soft or diarrheal stool specimens (n = 4,242) from pediatric or adult patients submitted for routine analysis of bacterial stool pathogens were included in the study. Formed stools and rectal swabs were excluded. Samples were submitted either in a clean, dry container or preserved in Cary-Blair transport medium (e.g., Para-Pak [Meridian Bioscience, Cincinnati, OH]). Prospective and retrospective specimens were collected and tested at six clinical centers in the United States and one in Canada. Additionally, specimens were provided by collection sites (2 in the United States, 1 in Canada, and 1 in Mexico). For some clinical centers, reference cultures were performed by Microbiology Specialists, Inc. (Houston, TX). Specimens were collected in compliance with site-specific Institutional Review Board (IRB) protocols. Between December 2012 and September 2013, 3,457 samples were collected prospectively: 1,345 (38.9%) were unpreserved and 2,112 (61.1%) were preserved. To increase the number of positive results, samples yielding pathogens by culture or Shiga toxin EIA collected between 2007 and 2013 and frozen at a temperature of −20°C or lower were included. The majority of previously characterized and archived specimens (84%) were collected and stored between March 2012 and August 2013. Retrospective specimens were thawed prior to testing and did not undergo other freeze-thaw cycles. When possible, positive retrospective samples were paired with one or more culture/EIA-negative specimens from the same time period. A total of 785 retrospective specimens were initially included: 321 (40.9%) were unpreserved, and 464 (59.1%) were preserved. Since targets in retrospective samples may have degraded during storage, prior to testing with the BD Max EBP assay, results for retrospective samples were confirmed by alternate PCR, according to previously reported methods (12, 13). Specimens with historical results that were not confirmed were excluded from further analysis.
Reference culture and EIA methods.
Preserved specimens were stored at 2°C to 25°C and planted for culture or tested for Shiga toxins by an EIA within 96 h of collection. Reference methods for unpreserved specimens were initiated upon receipt in the clinical laboratory, according to established laboratory protocols. Clinical Laboratory Improvement Amendment (CLIA) (31)-compliant culture methods for standard patient care were used at each site, and appropriate quality control was documented according to Clinical and Laboratory Standards Institute (CLSI) M22-A3 guidelines (14). Stool specimens were planted on primary culture medium for the detection of Salmonella spp., Shigella spp., Campylobacter spp., and E. coli O157. However, specific media varied somewhat. Media included MacConkey agar, selective campylobacter agar, Trypticase soy blood agar, xylose-lysine-deoxycholate agar, Hektoen enteric agar, salmonella-shigella agar, eosin-methylene blue agar, sorbitol MacConkey agar, CHROMagar Salmonella, and O157 CHROMagar and Gram-negative (GN) or selenite broth. Cultures were incubated for 48 h at 35°C for the detection of Salmonella spp., Shigella spp., and E. coli O157 and at 42°C under microaerophilic conditions for 72 h for the detection of Campylobacter spp. Identification of colonies suspicious for stool pathogens was performed with tube biochemicals, including API 20 (bioMérieux, Inc., Durham, NC), Vitek (bioMérieux, Marcy l'Etoile, France), Phoenix (BD, Sparks, MD), or MicroScan (Beckman Coulter, Brea, CA). Campylobacter isolates with negative hippurate and positive indoxyl acetate results were confirmed to be C. jejuni or C. coli by 16S rRNA sequence analysis. Non-sorbitol-fermenting colonies identified as E. coli were tested with O157-specific antiserum. Assessment for Shiga toxins 1 and/or 2 was performed following enrichment in GN or MacConkey broth for 18 to 24 h with either the Immunocard Stat! enterohemorrhagic E. coli (EHEC) or the Premier EHEC test according to the manufacturer's instructions (Meridian Bioscience, Cincinnati, OH).
BD Max EBP automated PCR.
Prospective samples were tested with the BD Max EBP assay within 48 h if stored at 8°C to 25°C and within 120 h if kept at 2°C to 8°C, according to the manufacturer's instructions. When thawed, retrospective samples were tested within the same time frames as those required for prospective samples. The BD Max EBP assay was performed according to the IUO (investigational use only) package insert for the BD Max instrument (15). BD Max and BD Max EBP are licensed trademarks of Becton, Dickinson and Company. By using a calibrated bacteriological loop, 10 μl of a homogenized stool specimen was placed into a sample buffer tube (SBT) containing diluent formulated to minimize inhibition associated with stool matrices. Individualized unitized reagent strips (URSs) containing reagents and consumables for DNA extraction were placed onto the BD Max instrument. A sample processing control (SPC) containing a plasmid with a proprietary synthetic target DNA sequence was incorporated into the extraction tube to monitor the effectiveness of the DNA extraction, amplification, and detection steps. Also loaded into the URSs were small tubes containing EBP-specific primers, TaqMan probes, and PCR master mix. The system performs lysis and DNA extraction and automatically pipettes the extracted sample into the PCR master mix. Tests for samples with a failed SPC result were repeated with the same SBT. If the repeat SPC test failed, a second SBT was inoculated from the original sample. Up to three repeats were allowed, provided that samples from the SBT were tested within 120 h and stored at 2°C to 8°C. Samples with repeatedly negative SPC results were reported as unresolved.
The BD Max EBP multiplex real-time PCR assay uses four primer sets: spaO for Salmonella spp., ipaH for Shigella spp. and enteroinvasive E. coli, stx1 and stx2 for Shiga toxin-producing organisms, and tuf for Campylobacter coli and C. jejuni. The two Campylobacter species and stx1 and stx2 are not differentiated from each other with these primer sets. Primers and probe sequences are proprietary and not available for publication. The BD Max instrument has five different optical channels for the detection of fluorescent signals. Discrimination of pathogen-specific amplicons and the SPC was accomplished with TaqMan probes with different fluorometric properties for each target. Amplification and detection occur in a disposable microfluidic cartridge. A software algorithm within the BD Max instrument interprets the amplification data and provides a positive or negative result. Each cartridge may be used for up to 12 samples. The total run time, including sample processing, PCR, and result reporting, was <3 h.
A positive and a negative external control were included at minimum in the first run of each day. The positive control was one of the four targets, alternated on a daily basis. Each species was tested at least once with each BD Max EBP kit lot. S. enterica subsp. enterica serovar Typhimurium (ATCC 14028), Shigella sonnei (ATCC 9290), and Shiga toxin-producing E. coli (ATCC 43890) were subcultured on 5% Trypticase soy sheep blood and incubated for 18 to 24 h. C. jejuni (ATCC 33291) was subcultured on brucella agar with 5% sheep blood, hemin, and vitamin K and incubated for 2 to 3 days in a microaerophilic environment. A 1:100 dilution of a suspension equivalent to a 0.5 McFarland turbidity standard was prepared from a fresh culture. By using a bacteriological loop, 10 μl was inoculated into an SBT prior to testing with the BD Max system. The negative control was saline.
Result interpretation and discrepant analysis.
Results for each analyte could not be reported for every sample for several possible reasons, including discrepancies in compliance with the study protocol, an unresolved SPC result for at least one target, or instrument failure. Samples with results that could not be reported after repeat testing due to system failure or instrument failure were termed indeterminate or incomplete. Results for noncompliant, unresolved, indeterminate, and incomplete samples were removed from the performance analysis.
The reference method for the detection of Shigella spp., Salmonella spp., and C. coli and C. jejuni was culture. EIA was the reference method for Shiga toxins 1 and 2. If an E. coli O157 isolate was identified by culture and the EIA result was negative, the EIA was repeated. If the result of the repeated EIA was negative, the final result was considered negative because the BD Max EBP assay detects toxin genes, irrespective of species. Culture results for E. coli O157 were used as supplemental information and did not contribute to performance statistics.
Prospective samples with discrepant results between culture or EIA and the BD Max EBP assay were tested by an alternate PCR method (10, 11) (performed at BD Diagnostics, Sparks, MD) and bidirectional sequencing of amplicons (ILS Genomics, LLC, Research Triangle Park, NC). The alternate PCR assessed different targets for Salmonella spp. (invA) and Campylobacter coli and C. jejuni (cadF) and utilized alternate primers for Shigella spp. (ipaH) and Shiga toxins (stx1 and stx2). For samples that were culture negative and BD Max EBP positive and for those with any results discrepant between the Shiga toxin EIA and BD Max EBP assay, the results of the alternate PCR were used as the final result. Growth of Salmonella spp., Shigella spp., Campylobacter coli, or C. jejuni in culture was considered a positive result, regardless of results with the alternate PCR. As retrospective samples were qualified prior to enrollment, they were not subject to additional discrepant analysis.
Statistical analysis.
The prevalence for each target was defined as the number of prospective specimens positive by standard culture or EIA reference methods divided by the total number of compliant prospective specimens. As no true reference method was available, including clinical criteria to determine whether subjects had disease or did not have disease, percent agreement calculations rather than clinical sensitivity and specificity estimates were performed (16). Positive percent agreement (PPA) is the number of samples positive by both the reference and test methods divided by the number of samples positive by both methods plus the number positive by the reference method only. Negative percent agreement (NPA) is the number of samples negative by both the reference and the test methods divided by the number of samples negative by both methods plus the number negative by the reference method only. PPA and NPA were calculated with 95% confidence intervals for all samples with compliant results for both the reference and BD Max EBP methods. The chi-square test was used for comparisons of PPA and NPA for preserved and unpreserved specimen types and prospective and retrospective data sets when the number of values in denominator for comparison was ≥5. Otherwise, Fisher's exact test was performed.
RESULTS
A total of 4,242 stool specimens were collected. The age distribution of patients was as follows: 3.7% were <1 year of age, 10.3% were 1 to 4 years of age, 11.5% were 5 to 12 years of age, 10.5% were 13 to 18 years of age, 48.6% were 19 to 65 years of age, 15.3% were >65 years of age, and 0.1% were of an unknown age. Approximately 25% of specimens were from children <12 years of age. Eighty-one percent of specimens were collected prospectively (n = 3,457), and the remainder (n = 785) were retrospective samples. As not all specimens met study criteria for compliant results with all targets, the number of results for each target varies.
Results for prospective specimens.
The results of the BD Max EBP assay compared to those of culture and Shiga toxin EIA for prospective specimens are shown in Table 1. Table 1 presents all data combined and stratified according to whether the specimen was preserved or not, prior to the resolution of discrepant results. PPA values were as follows: 97.8% for Campylobacter spp., 87.2% for Salmonella spp., 100% for Shigella spp., and 80.0% for Shiga toxins. PPA values tended to be higher for unpreserved specimens than for preserved specimens. However, differences were not statistically significant for any of the four targets. NPA values for prospective specimens were as follows: 98.2% for Campylobacter spp., 99.1% for Salmonella spp., 99.7% for Shigella spp., and 99.3% for Shiga toxins (Table 1). Although the overall NPA was 98.2% for Campylobacter spp., more false-positive results were noted for unpreserved specimens than for preserved specimens.
TABLE 1.
Results for prospective specimens prior to resolution of discrepancies
| Target and specimen type | No. of samples with BD Max EBP result with reference to culture/EIA |
Total no. of samples | PPA (95% confidence interval) | P value for PPAa | NPA (95% confidence interval) | P value for NPAa | |||
|---|---|---|---|---|---|---|---|---|---|
| True positive | False negative | False positive | True negative | ||||||
| Campylobacter | |||||||||
| Total | 44 | 1 | 51 | 2,797 | 2,893 | 97.8 (88.4–99.6) | 98.2 (97.7–98.6) | ||
| Preserved | 24 | 1 | 21 | 1,681 | 1,727 | 96.0 (80.5–99.3) | 1.000 | 98.8 (98.1–99.2) | 0.010 |
| Unpreserved | 20 | 0 | 30 | 1,116 | 1,166 | 100 (83.9–100) | 97.4 (96.3–98.2) | ||
| Salmonella | |||||||||
| Total | 34 | 5 | 26 | 2,857 | 2,922 | 87.2 (73.3–94.4) | 99.1 (98.7–99.4) | ||
| Preserved | 16 | 3 | 15 | 1,721 | 1,755 | 84.2 (62.4–94.5) | 0.951 | 99.1 (98.6–99.5) | 0.950 |
| Unpreserved | 18 | 2 | 11 | 1,136 | 1,167 | 90.0 (69.9–97.2) | 99.0 (98.3–99.5) | ||
| Shigella | |||||||||
| Total | 24 | 0 | 10 | 2,889 | 2,923 | 100 (86.2–100) | 99.7 (99.4–99.8) | ||
| Preserved | 7 | 0 | 5 | 1,743 | 1,755 | 100 (64.6–100) | 1.000 | 99.7 (99.3–99.9) | 0.732 |
| Unpreserved | 17 | 0 | 5 | 1,146 | 1,168 | 100 (81.6–100) | 99.6 (99.0–99.8) | ||
| Shiga toxins | |||||||||
| Total | 8 | 2 | 17 | 2,347 | 2,374 | 80.0 (49.0–94.3) | 99.3 (98.9–99.6) | ||
| Preserved | 6 | 2 | 12 | 1,701 | 1,721 | 75.0 (40.9–92.9) | 1.000 | 99.3 (98.8–99.6) | 1.000 |
| Unpreserved | 2 | 0 | 5 | 646 | 653 | 100 (34.2–100) | 99.2(98.2–99.7) | ||
P value for testing of the significance of the difference between preserved and unpreserved specimens.
Results for retrospective specimens.
Combined results and data for retrospective samples stratified according to whether the sample had been placed into preservative or not are shown in Table 2. PPA values were as follows: 97.0% for Campylobacter spp., 99.4% for Salmonella spp., 98.9% for Shigella spp., and 100% for Shiga toxins. Six samples were positive by culture and alternate PCR but were negative by BD Max EBP. These included four samples positive for Campylobacter, one positive for Salmonella, and one positive for Shigella. Four of the six samples were preserved specimens. There were no false-negative results for Shiga toxins among retrospective samples. The NPA was ≥99.5% for all targets. Three unpreserved samples were positive by BD Max EBP but negative by culture and alternate PCR. These samples included two samples in which Campylobacter was detected and one that was positive for Salmonella spp. Among retrospective specimens, no false-positive results were determined for Shigella spp. or Shiga toxins.
TABLE 2.
Results for retrospective specimens
| Target and specimen type | No. of specimens with BD Max EBP result with reference to culture/EIA confirmed by alternate PCR |
Total no. of specimensa | PPA (95% confidence interval) | P value for PPAb | NPA (95% confidence interval) | P value for NPAb | |||
|---|---|---|---|---|---|---|---|---|---|
| True positive | False negative | False positive | True negative | ||||||
| Campylobacter | |||||||||
| Total | 129 | 4 | 2 | 372 | 507 | 97.0 (92.5–98.8) | 99.5 (98.1–99.9) | ||
| Preserved | 64 | 2 | 0 | 151 | 217 | 97.0 (89.6–99.2) | 1.000 | 100 (97.5–100) | 0.657 |
| Unpreserved | 65 | 2 | 2 | 221 | 290 | 97.0 (89.8–99.2) | 99.1 (96.8–99.8) | ||
| Salmonella | |||||||||
| Total | 166 | 1 | 1 | 450 | 618 | 99.4 (96.7–99.9) | 99.8 (98.8–100) | ||
| Preserved | 105 | 1 | 0 | 213 | 319 | 99.1 (94.8–99.8) | 1.000 | 100 (98.2–100) | 1.000 |
| Unpreserved | 61 | 0 | 1 | 237 | 299 | 100 (94.1–100) | 99.6 (97.7–99.9) | ||
| Shigella | |||||||||
| Total | 91 | 1 | 0 | 451 | 543 | 98.9 (94.1–99.8) | 100 (99.2–100) | ||
| Preserved | 50 | 1 | 0 | 187 | 238 | 98.0 (89.7–99.7) | 1.000 | 100 (98.0–100) | 1.000 |
| Unpreserved | 41 | 0 | 0 | 264 | 305 | 100 (91.4–100) | 100 (98.6–100) | ||
| Shiga toxins | |||||||||
| Total | 66 | 0 | 0 | 90 | 156 | 100 (94.5–100) | 100 (95.9–100) | ||
| Preserved | 41 | 0 | 0 | 79 | 120 | 100 (91.4–100) | 1.000 | 100 (95.4–100) | 1.000 |
| Unpreserved | 25 | 0 | 0 | 11 | 36 | 100 (86.7–100) | 100 (74.1–100) | ||
The number of specimens varies by target due to differences in compliance with the protocol and unresolved results.
P value for testing of the significance of the difference between preserved and unpreserved specimens.
Combined results after discrepant analysis.
Prior to the resolution of discrepant results, PPA and NPA values for all sample types and prospective and retrospective samples combined were as follows: 97.2% and 98.4% for C. jejuni and C. coli, 97.1% and 99.2% for Salmonella spp., 99.1% and 99.7% for Shigella spp., and 97.4% and 99.3% for Shiga toxins, respectively. (Percentages were calculated from combined total data in Tables 1 and 2.) After discrepant analysis, PPA and NPA values (with 95% confidence intervals) were as follows: 97.5% (94.3 to 99.2%) and 99.0% (98.6 to 99.3%) for Campylobacter spp., 97.3% (94.3 to 99.0%) and 99.8% (94.5 to 99.9) for Salmonella spp., 99.2% (95.6 to 99.9%) and 100% (99.9 to 100%) for Shigella spp., and 100% (95.7 to 100%) and 99.7% (99.4 to 99.9%) for Shiga toxins, respectively (Table 3). When prospective and retrospective samples were compared for each target, PPA and NPA were similar for all targets, with the exception of PPA for Salmonella spp.
TABLE 3.
Combined results after resolution of discrepancies
| Target and specimen type | No. of specimens with BD Max EBP result with reference to culture/EIA or alternate PCRa |
Total no. of specimensc | PPA (95% confidence interval) | P value for PPAd | NPA (95% confidence interval) | P value for NPAd | |||
|---|---|---|---|---|---|---|---|---|---|
| True positive | False negative | False positive | True negative | ||||||
| Campylobacter | |||||||||
| Total | 195 | 5 | 31 | 3,170 | 3,401 | 97.5 (94.3–99.2) | 99.0 (98.6–99.3) | ||
| Prospective | 66 | 1 | 29 | 2,797 | 2,893 | 98.5 (91.3–99.7) | 0.867 | 99.0 (98.5–99.3) | 0.525 |
| Retrospectiveb | 129 | 4 | 2 | 373 | 508 | 97.0 (92.5–98.8) | 99.5 (98.1–99.9) | ||
| Salmonella | |||||||||
| Total | 219 | 6 | 8 | 3,307 | 3,540 | 97.3 (94.3–99.0) | 99.8 (94.5–99.9) | ||
| Prospective | 53 | 5 | 7 | 2,857 | 2,922 | 91.4 (81.0–97.1) | 0.005 | 99.8 (99.5–99.9) | 1.000 |
| Retrospective | 166 | 1 | 1 | 450 | 618 | 99.4 (96.7–99.9) | 99.8 (98.8–100) | ||
| Shigella | |||||||||
| Total | 124 | 1 | 1 | 3,340 | 3,466 | 99.2 (95.6–99.9) | 100 (99.8–100) | ||
| Prospective | 33 | 0 | 1 | 2,889 | 2,923 | 100 (89.3–100) | 1.000 | 100 (99.8–100) | 1.000 |
| Retrospective | 91 | 1 | 0 | 451 | 543 | 98.9 (94.1–99.8) | 100 (99.2–100) | ||
| Shiga toxins | |||||||||
| Total | 85 | 0 | 8 | 2,437 | 2,530 | 100 (95.7–100) | 99.7 (99.4–99.9) | ||
| Prospective | 19 | 0 | 8 | 2,347 | 2,374 | 100 (82.2–100) | 1.000 | 99.7(99.3–99.9) | 1.000 |
| Retrospective | 66 | 0 | 0 | 90 | 156 | 100 (94.5–100) | 100 (95.9–100) | ||
For Campylobacter spp., Salmonella spp., and Shigella spp., a positive result by either culture or alternate PCR was considered a true-positive result. For the resolution of discrepant results for Shiga toxins, results of alternate PCR were considered true-positive results.
Results for all retrospective samples were confirmed by alternate PCR and bidirectional sequencing prior to testing with the BD Max EBP assay. Discrepancies were not further resolved.
The number of prospective specimens varies by target due to differences in compliance with the study protocol and unresolved results.
P value for testing of the significance of the difference between prospective and retrospective data sets.
Discrepant results for prospective specimens are detailed in Tables 4 and 5. Six samples were positive by culture but negative by BD Max EBP. One culture grew Campylobacter and five grew Salmonella, which were not detected by either BD Max EBP or alternate PCR. Four of these six samples were preserved samples. Two samples were positive by Shiga toxin EIA and negative by both BD Max EBP and alternate PCR. There were no samples that were positive by culture and negative by the BD Max EBP assay for Shigella.
TABLE 4.
Discrepant results for prospective samples negative by BD Max EBPa
| Target | No. of BD Max EPB-negative specimens with result |
||
|---|---|---|---|
| Culture positive | EIA positive | Alternate PCR negative | |
| Campylobacter | 1 | NA | 1 |
| Salmonella | 5 | NA | 5 |
| Shiga toxin | NA | 2 | 2 |
NA, not applicable.
TABLE 5.
Discrepant results for prospective samples positive by BD Max EBPa
| Target | No. of BD Max EPB-positive specimens with result |
||
|---|---|---|---|
| Culture negative | EIA negative | Alternate PCR positive | |
| Campylobacter | 51 | NA | 22 |
| Salmonella | 26 | NA | 19 |
| Shigella | 10 | NA | 9 |
| Shiga toxin | NA | 17 | 9 |
NA, not applicable.
Prior to the resolution of discrepancies, numbers of false-positive results for prospective samples with the BD Max EBP assay were 51 for Campylobacter spp., 26 for Salmonella spp., 10 for Shigella spp., and 17 for Shiga toxins. Discrepant analysis demonstrated that many samples were likely falsely negative by culture or Shiga toxin EIA. The numbers of discrepant samples positive by alternate PCR were 22/51 (43.1%) for Campylobacter spp., 19/26 (73.1%) for Salmonella spp., 9/10 (90%) for Shigella spp., and 9/17 (52.9%) for Shiga toxins (Table 5).
PPA for culture compared to consensus results.
Specimens positive by BD Max EBP and by alternate PCR and sequencing but negative by culture or EIA (Table 5) are likely to be true-positive results that were not detected by culture due to a loss of organism viability or a low organism burden. The PPA for culture was calculated for all specimen types combined. With the definition of a true-positive result being a specimen positive by either culture/EIA or BD Max EBP (confirmed by alternate PCR), the PPAs for culture/EIA for each target were determined to be 90.1% for Campylobacter spp., 92.2% for Salmonella spp., 93.2% for Shigella spp., and 90.4% for Shiga toxins.
Prevalence and coinfections.
Based on culture and EIA reference methods, the prevalences of each target in prospective samples were as follows: 45/2,893 samples (1.5%) for Campylobacter spp., 39/2,922 (1.2%) for Salmonella spp., 24/2,923 (0.8%) for Shigella spp., and 10/2,374 (0.4%) for Shiga toxin. With the addition of samples positive by BD Max EBP and alternate PCR, the prevalences of each target in prospective specimens were as follows: 67/2,893 (2.3%) for Campylobacter spp., 58/2,922 (2.0%) for Salmonella spp., 33/2,923 (1.1%) for Shigella spp., and 19/2,374 (0.8%) for Shiga toxin.
Ten coinfections were detected with the BD Max EBP assay. None were detected by culture. Discrepant analysis confirmed four coinfections. No significant patterns of coinfection were observed. Campylobacter spp. were present in six coinfections, Salmonella spp. were present in five, Shigella spp. were present in three, and Shiga toxins were present in six.
Unresolved results.
Samples were considered unresolved if the SPC was not amplified and detected. Initially, the overall rate of unresolved results was 5.0% and was highest for testing sites B (7.7%) and C (10.0%) (Table 6). After a single repeat test from the same SBT, the overall rate of unresolved results was much lower (1.3%). Repeat testing from a redilution of original specimen in a new SBT lowered rates slightly more. A few more specimens could be resolved by a third repeat test, performed with the new SBT. The rate of unresolved results for site C remained higher (2.6%) than that for all other sites (≤0.8%). Rates of unresolved results for unpreserved and preserved samples were 7.1% and 3.7%, respectively. The final rate of unresolved results after repeat testing for unpreserved specimens (1.2%) was higher than that for preserved specimens (0.3%) (P = 0.0006).
TABLE 6.
Unresolved results
| Testing site | % of unresolved test results |
|||
|---|---|---|---|---|
| Initial | Initial SBT (1st repeat) | 2nd SBT (2nd repeat) | 2nd SBT (3rd repeat) | |
| A | 3.5 | 0.7 | 0.7 | 0.4 |
| B | 7.7 | 1.6 | 1.2 | 0.8 |
| C | 10.0 | 4.5 | 3.2 | 2.6 |
| D | 4.7 | 0.6 | 0.4 | 0.4 |
| E | 2.6 | 0.1 | 0.0 | 0.0 |
| F | 4.1 | 1.1 | 1.1 | 0.7 |
| G | 4.7 | 1.7 | 1.0 | 0.5 |
| H | 5.0 | 1.0 | 0.2 | 0.2 |
| Overall | 5.0 | 1.3 | 0.9 | 0.6 |
DISCUSSION
The BD Max EBP assay is a multiplex real-time PCR assay for the detection of Campylobacter jejuni, C. coli, Salmonella spp., Shigella spp., and Shiga toxin 1 and 2 genes. The target used for the detection of Shigella is also found in enteroinvasive E. coli (EIEC), potentially allowing the detection of this pathogen as well. However, detection of EIEC was not specifically assessed. The BD Max EBP assay is performed with the BD Max instrument, a walkaway instrument that conducts all steps of the molecular assay from extraction to interpretation of results. Compared to routine stool culture, laboratories have demonstrated significant reductions in turnaround time with the use of molecular assays (17, 18). Additionally, the CDC recommends culture-based screening for E. coli O157 and EIA or molecular methods for Shiga toxins (19). Some laboratories also use EIAs to increase the sensitivity of detection of Campylobacter species (20). While the addition of EIA methods improves the detection of pathogens, EIA methods complicate the workflow of routine stool assessment and increase hands-on time. Thus, the BD Max EBP assay can save technical effort and improve the time to reporting of results.
After the resolution of discrepant samples, the overall PPA for the BD Max EBP assay compared to culture and alternate PCR was ≥97.3% for all targets. The BD Max EBP assay was more sensitive than culture, detecting an additional 22 specimens with Campylobacter spp., 19 specimens with Salmonella spp., 9 specimens with Shigella spp., and 9 specimens with Shiga toxins from among specimens collected prospectively. A review of the literature indicates that the performance of the BD Max EBP assay is similar to those of other multiplex assays developed recently. The sensitivity and specificity of the Prodesse ProGastro SSCS assay compared to culture and EIA determined by one multicenter study of 1,244 specimens were 100% and ≥99.4%, respectively, for all four bacterial targets (9). Early studies of the Luminex xTAG GPP assay demonstrated sensitivities ranging from 83 to 93% for Salmonella spp., 90 to 97% for Campylobacter spp., 93 to 100% for Shigella spp., and 94% to 100% for STEC (6, 21). The reported sensitivities based on a recent large multicenter study were higher for some targets with prospective samples (22). The FilmArray GI panel was evaluated in a multicenter study of 1,556 samples. The reported sensitivities were as follows: 100% for Salmonella spp., 95.9% for Shigella spp., 97.1% for Campylobacter spp., and 100% for Shiga toxins (23). An assessment of 611 prospective and 839 contrived specimens was conducted with the Verigene Enteric Pathogens test. The reported sensitivities were 96.5% for Salmonella spp., 93.0% for Campylobacter spp., 100% for Shigella spp., 100% for stx1, and 97.4% for stx2 (11). These studies demonstrated that multiplex PCR assays always outperformed culture and have very good sensitivity. User preferences regarding target selection, throughput, and the technical expertise needed to perform the test will contribute largely to assay selection.
Of the 12 positive samples (5 positive for Campylobacter, 6 positive for Salmonella, and 1 positive for Shigella) not detected with the BD Max EBP assay, 1 Campylobacter and 5 Salmonella specimens were also not detected by the alternate PCR. Although a previous study of the analytical sensitivity of the BD Max EBP assay in comparison with culture for Cary-Blair medium-preserved stool specimens spiked with pathogens demonstrated a superior sensitivity of the BD Max EBP assay for all pathogens at every organism concentration (24), we hypothesize that the organism load in these study samples may have been close to the limiting dilution for detection by PCR. The limit of detection (LOD) for the BD Max EBP assay has been reported to be between 10 and 910 CFU/ml, depending on the target (15). The alternate PCR assay was based on a design reported previously by Cunningham et al. (12). The LOD for the alternate PCR was similar to that for the BD Max EBP assay and was reported to be between 16 and 990 CFU/ml for all targets with the exception of C. coli diluted in fresh stool. Although PCR assays are very sensitive and increase the overall number of positive samples detected, if low numbers of pathogens are present, the volume of the specimen tested can impact detection due to sampling bias. The BD Max EBP assay uses only 10 μl of sample. When cultures are set up, a drop (∼25 to 50 μl) is generally planted on a minimum of two agar media. However, neither the BD Max EBP assay nor the alternate PCR was performed on larger volumes of specimen to assess this hypothesis. A less likely explanation for false-negative results is the possibility that some isolates were misidentified or lacked the molecular target. The pathogens isolated in this study were not confirmed in a reference laboratory or assessed for the presence of the target by PCR. However, all laboratories were CLIA certified and used standard procedures for the reference methods.
With the exception of results for Salmonella spp. in the prospective arm, there was little difference in performance between prospective and retrospective data sets (Table 3). Reduced sensitivity for Salmonella in comparison to culture has been reported (12, 17, 21, 25). Cunningham et al. attributed reduced sensitivity to the enhancement of the culture method with a selenite enrichment broth (12). Selenite broth was used at three of the centers in our study, and three of the six false-negative results were from laboratories that routinely used selenite broth for culture of Salmonella. The use of an enrichment broth incubated overnight is recommended for improving the detection of group B Streptococcus in vaginal-rectal specimens whether PCR- or culture-based methods are used (26). Overnight enrichment may similarly enhance the detection of Salmonella from stool specimens by PCR. Evaluation of the benefit of a broth enrichment step prior to multiplex PCR for enteric pathogens may be warranted.
Two prospective specimens that were positive by the Shiga toxin EIA and negative by BD Max EBP were also negative by alternate PCR (Table 4). The analytical sensitivity of the ImmunoCard Stat! EHEC assay was reported to be 107 CFU/ml, and that of the Premier EHEC EIA was 106 CFU/ml (27). Given the limited analytical sensitivity of the EIA methods, it is unlikely that PCR was insensitive compared to the EIA. The specificity of the EIA has been reported to be 99.7% (27, 28), which is consistent with our data. These apparently false-negative BD Max EBP results may represent either false-positive EIA results or the presence of toxin in specimens in which stx1 or stx2 nucleic acids were degraded.
In the prospective arm, the PPA tended to be higher for unpreserved specimens, although differences between specimen types were not statistically significant. It is reasonable to expect specimen preservative to maintain viability and thereby enhance recovery. However, when the organism is in low numbers, preservative may dilute the target below the limit of detection. Since PCR assays do not require viable organisms, further study of the role of preservative for maintaining stool specimens submitted for PCR assays may be of value.
The NPA was very high and increased to ≥99% following assessment of discrepant prospective specimens. For all targets combined, 57% (59/104) of the apparent false-positive BD Max EBP results were confirmed by alternate PCR. Particularly striking were results for Salmonella and Shigella, where 73.1% and 90% of the apparent false-positive BD Max EBP results, respectively, were confirmed by discriminant analysis. Some BD Max EBP PCR-positive results were not supported by alternate PCR results. These included 31 specimens positive for Campylobacter spp., 8 specimens positive for Salmonella spp., 1 specimen positive for Shigella spp., and 8 specimens positive for Shiga toxins (results from Tables 2 and 5 combined). Apparent false-positive results with multiplex PCR assays may truly reflect enhanced sensitivity in the presence of targets in low copy numbers or the presence of nonviable organisms. False-positive results for Campylobacter are particularly notable and call the confirmatory method into question. The reported analytical sensitivities of the alternate PCR were as low as 16 CFU/ml for C. jejuni and between 400 and 4,000 CFU/ml for C. coli, depending on the matrix (12). Although a slightly modified version of the published assay was used in our study, the previously published results suggest that unless C. coli was responsible for the apparent false-positive BD Max results, the alternate PCR was adequately sensitive. Buchan et al. also reported that 7/13 (53.4%) specimens positive for Campylobacter spp. by the ProGastro SSCS PCR and negative by culture were not confirmed by an alternate PCR method (9). Buss et al. noted 5/24 (21%) samples false positive for Campylobacter by FilmArray that were not confirmed by alternate PCR, and Coste reported confirming 9/15 (60%) false-positive results by using an alternate PCR and EIA methods (23, 29). False-positive results could also be due to cross-reactivity of primers and probes with other targets. Stool is a vastly complex matrix full of nucleic acids from organisms that may be uncharacterized and that could potentially share sequences with enteric pathogens. Nine species of Campylobacter were evaluated in premarket specificity studies, and none were cross-reactive with the BD Max EBP (15). Other, less well-characterized or unculturable species that may or may not have a role in diarrheal disease could have cross-reactive potential. Finally, false-positive results may reflect laboratory contamination. However, prior to the start of the clinical study and each week thereafter, each testing laboratory performed a wipe test of 10 work area locations to assess and control for contamination of all targets. Of 5,520 total wipe tests, only 10 (<0.2%) were positive. All contamination events were resolved after cleaning. Thus, we do not believe that contamination contributed to our findings.
Initially, the overall rate of unresolved results for all targets and all sites combined was 5.0%. A repeat of the initial SBT lowered the rate of unresolved results to 1.3%. Although the results for some samples could be resolved by diluting the original sample in a new SBT and repeating the assay a second or third time, this process provided less benefit than did the initial repeat (Table 6). Study sites B and C had the highest rates of unresolved results. These two sites included only unpreserved specimens. Sites A and F included both preserved and unpreserved specimens. All other sites included only specimens preserved in Cary-Blair medium. The rate of unresolved results for unpreserved specimens was higher than that for specimens in preservative, providing a plausible explanation for the higher rates at sites B and C. It might be expected that dilution of the stool matrix in Cary-Blair medium would reduce the effects of inhibitory substances. Thus, after a single repeat test, an unresolved rate of slightly more than 1% may be expected if samples are not preserved, and a lower rate is expected if preserved samples are tested.
The combined prevalence of all 4 targets in prospective specimens increased from 3.9% based on culture for Salmonella spp., Campylobacter spp., and Shigella spp. and the Shiga toxin EIA to 6.2% with the additional positive results detected by the BD Max EBP assay. Large, multicenter studies conducted in North American laboratories reported prevalence rates similar to those detected by the Prodesse ProGastro SSCS assay (8.3%) (6) and the Luminex xTAG GPP assay (7.1%) (19). While prevalence rates remain relatively low, those studies all demonstrate the increased sensitivity of molecular over traditional methods.
Real-time PCR methods can bring great advances in turnaround time and analysis for multiple pathogens simultaneously. Not only are these assays more sensitive than routine culture and EIA methods, they are also more likely to detect coinfections, as evidenced in this evaluation and others (6, 10, 21). However, multiplex assays that report bacterial, viral, and parasitic targets simultaneously may generate results that were not requested by clinicians and create complexities in reporting as well as in interpretation (17, 29, 30). For example, detection of Clostridium difficile in colonized individuals who actually harbor an agent of acute bacterial enteritis may lead to overtreatment for this pathogen. Multiplex assays with many targets might be best suited to limited patient populations, such as international travelers, immunocompromised individuals, or those with highly suspect disease who have negative results from assays with more limited menus.
The strengths of this study are the large number of specimens tested and the use of multiple testing sites. Specimens were collected across the United States, in Canada, and in Mexico, providing an adequate number of positive samples and a broad population base. Although serotyping and strain typing were not performed, it is likely that large numbers of clones and serotypes are represented among the pathogens detected. Another strength was the inclusion of both preserved and unpreserved samples. The BD Max EBP assay has been cleared by the FDA for testing of both specimen types. The ability to test either specimen type may allow better flexibility in specimen collection practices than those required for some other methods. One potential weakness is the comparison to culture and EIA methods, which are known to lack sensitivity for enteric pathogens (6, 9, 12, 17). However, the use of alternate PCR and bidirectional sequencing allows a more direct comparison of similar methods for specimens that gave discrepant results. In summary, this large, multicenter clinical study demonstrated very good performance of the BD Max EBP multiplex PCR assay. The throughput and ease of use may provide advantages to many laboratories, improving the detection of bacterial stool pathogens and time to reporting of results.
ACKNOWLEDGMENTS
BD Diagnostics (BD), Québec, QC, Canada, developed the study protocol and provided supplies and financial support for this study.
A.L. and T.K. are employees of BD. B.W.B. and J.E.M. have received honoraria from BD.
We gratefully acknowledge the work of the technical personnel at each study site.
REFERENCES
- 1.Henderson P. 2013. End preventable deaths: global action plan for prevention and control of pneumonia and diarrhoea. United Nations Children's Fund, World Health Organization, New York, NY. [Google Scholar]
- 2.Johansson EW, Wardlaw T. 2009. Diarrhoea: why children are still dying and what can be done. United Nations Children's Fund, World Health Organization, New York, NY. [Google Scholar]
- 3.Kotloff KL, Nataro JP, Blackwelder WC, Nasrin D, Farag TH, Panchalingam S, Wu Y, Sow SO, Sur D, Breiman RF, Faruque AS, Zaidi AK, Saha D, Alonso PL, Tamboura B, Sanogo D, Onwuchekwa U, Manna B, Ramamurthy T, Kanungo S, Ochieng JB, Omore R, Oundo JO, Hossain A, Das SK, Ahmed S, Qureshi S, Quadri F, Adegbola RA, Antonio M, Hossain MJ, Akinsola A, Mandomando I, Nhampossa T, Acacio S, Biswas K, O'Reilly CE, Mintz ED, Berkeley LY, Muhsen K, Sommerfelt H, Robins-Browne RM, Levine MM. 2013. Burden and aetiology of diarrhoeal disease in infants and young children in developing countries (the Global Enteric Multicenter Study, GEMS): a prospective, case-control study. Lancet 382:209–222. doi: 10.1016/S0140-6736(13)60844-2. [DOI] [PubMed] [Google Scholar]
- 4.Scallan E, Hoekstra RM, Angulo FJ, Tauxe RV, Widdowson MA, Roy SL, Jones JL, Griffin PM. 2011. Foodborne illness acquired in the United States—major pathogens. Emerg Infect Dis 17:7–15. doi: 10.3201/eid1701.09-1101p1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Guerrant RL, Van Gilder T, Steiner TS, Thielman NM, Slutsker L, Tauxe RV, Hennessy T, Griffin PM, DuPont H, Sack RB, Tarr P, Neill M, Nachamkin I, Reller LB, Osterholm MT, Bennish ML, Pickering LK, Infectious Diseases Society of America. 2001. Practice guidelines for the management of infectious diarrhea. Clin Infect Dis 32:331–351. doi: 10.1086/318514. [DOI] [PubMed] [Google Scholar]
- 6.Navidad JF, Griswold DJ, Gradus MS, Bhattacharyya S. 2013. Evaluation of Luminex xTAG gastrointestinal pathogen analyte-specific reagents for high-throughput, simultaneous detection of bacteria, viruses, and parasites of clinical and public health importance. J Clin Microbiol 51:3018–3024. doi: 10.1128/JCM.00896-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Rinttila T, Kassinen A, Malinen E, Krogius L, Palva A. 2004. Development of an extensive set of 16S rDNA-targeted primers for quantification of pathogenic and indigenous bacteria in faecal samples by real-time PCR. J Appl Microbiol 97:1166–1177. doi: 10.1111/j.1365-2672.2004.02409.x. [DOI] [PubMed] [Google Scholar]
- 8.Stellrecht KA, Espino AA, Maceira VP, Nattanmai SM, Butt SA, Wroblewski D, Hannett GE, Musser KA. 2014. Premarket evaluations of the IMDx C. difficile for Abbott m2000 assay and the BD Max Cdiff assay. J Clin Microbiol 52:1423–1428. doi: 10.1128/JCM.03293-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Buchan BW, Olson WJ, Pezewski M, Marcon MJ, Novicki T, Uphoff TS, Chandramohan L, Revell P, Ledeboer NA. 2013. Clinical evaluation of a real-time PCR assay for identification of Salmonella, Shigella, Campylobacter (Campylobacter jejuni and C. coli), and Shiga toxin-producing Escherichia coli isolates in stool specimens. J Clin Microbiol 51:4001–4007. doi: 10.1128/JCM.02056-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Patel A, Navidad J, Bhattacharyya S. 2014. Site-specific clinical evaluation of the Luminex xTAG gastrointestinal pathogen panel for detection of infectious gastroenteritis in fecal specimens. J Clin Microbiol 52:3068–3071. doi: 10.1128/JCM.01393-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Novak SM, Bobenchik A, Cumpio J, Marlowe E. 2014. Evaluation of the Verigene EP IUO test for the rapid detection of bacterial and viral causes of gastrointestinal infection, abstr 1317. Abstr 30th Clin Virol Symp Annu Meet Pan Am Soc Clin Virol. [Google Scholar]
- 12.Cunningham SA, Sloan LM, Nyre LM, Vetter EA, Mandrekar J, Patel R. 2010. Three-hour molecular detection of Campylobacter, Salmonella, Yersinia, and Shigella species in feces with accuracy as high as that of culture. J Clin Microbiol 48:2929–2933. doi: 10.1128/JCM.00339-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Grys TE, Sloan LM, Rosenblatt JE, Patel R. 2009. Rapid and sensitive detection of Shiga toxin-producing Escherichia coli from nonenriched stool specimens by real-time PCR in comparison to enzyme immunoassay and culture. J Clin Microbiol 47:2008–2012. doi: 10.1128/JCM.02013-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Clinical and Laboratory Standards Institute. 2003. Quality control for commercially prepared microbiological culture media; approved standard, 3rd ed, vol M22-A3 CLSI, Wayne, PA. [Google Scholar]
- 15.Becton, Dickinson, Inc. 2014. BD Max enteric bacterial panel. GenOhm Sciences, Canada, Inc, Quebec, QC, Canada. [Google Scholar]
- 16.US Food and Drug Administration. 2007. Guidance for industry and FDA staff. Statistical guidance on reporting results from studies evaluating diagnostic tests. US Department of Health and Human Services, FDA, Silver Spring, MD. [Google Scholar]
- 17.McAuliffe GN, Anderson TP, Stevens M, Adams J, Coleman R, Mahagamasekera P, Young S, Henderson T, Hofmann M, Jennings LC, Murdoch DR. 2013. Systematic application of multiplex PCR enhances the detection of bacteria, parasites, and viruses in stool samples. J Infect 67:122–129. doi: 10.1016/j.jinf.2013.04.009. [DOI] [PubMed] [Google Scholar]
- 18.O'Leary J, Corcoran D, Lucey B. 2009. Comparison of the EntericBio multiplex PCR system with routine culture for detection of bacterial enteric pathogens. J Clin Microbiol 47:3449–3453. doi: 10.1128/JCM.01026-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Centers for Disease Control and Prevention. 2009. Recommendations for diagnosis of Shiga toxin-producing Escherichia coli infections by clinical laboratories. MMWR Recommend Rep 58(RR-12):1–14. [PubMed] [Google Scholar]
- 20.Granato PA, Chen L, Holiday I, Rawling RA, Novak-Weekley SM, Quinlan T, Musser KA. 2010. Comparison of premier CAMPY enzyme immunoassay (EIA), ProSpecT Campylobacter EIA, and ImmunoCard STAT! CAMPY tests with culture for laboratory diagnosis of Campylobacter enteric infections. J Clin Microbiol 48:4022–4027. doi: 10.1128/JCM.00486-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Claas EC, Burnham CA, Mazzulli T, Templeton K, Topin F. 2013. Performance of the xTAG gastrointestinal pathogen panel, a multiplex molecular assay for simultaneous detection of bacterial, viral, and parasitic causes of infectious gastroenteritis. J Microbiol Biotechnol 23:1041–1045. doi: 10.4014/jmb.1212.12042. [DOI] [PubMed] [Google Scholar]
- 22.Himsworth D, Buller R, Chappell J, Fader R, Mahony J, Mazzulli T, Wolk D. 2014. A multi-center clinical evaluation of the Luminex xTAG gastrointestinal panel (xTAG GPP) in patients with clinical signs and symptoms of gastrointestinal infection, abstr T31. Abstr 30th Clin Virol Symp Annu Meet Pan Am Soc Clin Virol. [Google Scholar]
- 23.Buss SN, Leber A, Chapin K, Fey PD, Bankowski MJ, Jones MK, Rogatcheva M, Kanack KJ, Bourzac KM. 2015. Multicenter evaluation of the BioFire FilmArray gastrointestinal panel for the etiologic diagnosis of infectious gastroenteritis. J Clin Microbiol 53:915–925. doi: 10.1128/JCM.02674-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Anderson NW, Buchan BW, Ledeboer NA. 2014. Comparison of the BD MAX enteric bacterial panel to routine culture methods for detection of Campylobacter, enterohemorrhagic Escherichia coli (O157), Salmonella, and Shigella isolates in preserved stool specimens. J Clin Microbiol 52:1222–1224. doi: 10.1128/JCM.03099-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Biswas JS, Al-Ali A, Rajput P, Smith D, Goldenberg SD. 2014. A parallel diagnostic accuracy study of three molecular panels for the detection of bacterial gastroenteritis. Eur J Clin Microbiol Infect Dis 33:2075–2081. doi: 10.1007/s10096-014-2177-9. [DOI] [PubMed] [Google Scholar]
- 26.Centers for Disease Control and Prevention. 2010. Prevention of perinatal group B streptococcal disease—revised guidelines from CDC, 2010. MMWR Recommend Rep 59(RR-10):1–36. [PubMed] [Google Scholar]
- 27.Vallieres E, Saint-Jean M, Rallu F. 2013. Comparison of three different methods for detection of Shiga toxin-producing Escherichia coli in a tertiary pediatric care center. J Clin Microbiol 51:481–486. doi: 10.1128/JCM.02219-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hermos CR, Janineh M, Han LL, McAdam AJ. 2011. Shiga toxin-producing Escherichia coli in children: diagnosis and clinical manifestations of O157:H7 and non-O157:H7 infection. J Clin Microbiol 49:955–959. doi: 10.1128/JCM.02119-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Coste JF, Vuiblet V, Moustapha B, Bouin A, Lavaud S, Toupance O, de Rougemont A, Benejat L, Megraud F, Wolak-Thierry A, Villena I, Chemla C, Le Magrex E, de Champs C, Andreoletti L, Rieu P, Leveque N. 2013. Microbiological diagnosis of severe diarrhea in kidney transplant recipients by use of multiplex PCR assays. J Clin Microbiol 51:1841–1849. doi: 10.1128/JCM.03366-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Khare R, Espy MJ, Cebelinski E, Boxrud D, Sloan LM, Cunningham SA, Pritt BS, Patel R, Binnicker MJ. 2014. Comparative evaluation of two commercial multiplex panels for detection of gastrointestinal pathogens by use of clinical stool specimens. J Clin Microbiol 52:3667–3673. doi: 10.1128/JCM.01637-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Clinical Laboratory Improvement Act. 1992. Code of Federal Regulations. Standards and Certification: Laboratory Requirements. 42 CFR 493. [Google Scholar]