Abstract
The FilmArray Respiratory Panel (RP) multiplexed nucleic acid amplification test (Idaho Technology, Inc., Salt Lake City, UT) was compared to laboratory-developed real-time PCR assays for the detection of various respiratory viruses and certain bacterial pathogens. A total of 215 frozen archived pediatric respiratory specimens previously characterized as either negative or positive for one or more pathogens by real-time PCR were examined using the FilmArray RP system. Overall agreement between the FilmArray RP and corresponding real-time PCR assays for shared analytes was 98.6% (kappa = 0.92 [95% confidence interval (CI), 0.89 to 0.94]). The combined positive percent agreement was 89.4% (95% CI, 85.4 to 92.6); the negative percent agreement was 99.6% (95% CI, 99.2 to 99.8). The mean real-time PCR threshold cycle (CT) value for specimens with discordant results was 36.46 ± 4.54. Detection of coinfections and correct identification of influenza A virus subtypes were comparable to those of real-time PCR when using the FilmArray RP. The greatest comparative difference in sensitivity was observed for adenovirus; only 11 of 24 (45.8%; 95% CI, 27.9 to 64.9) clinical specimens positive for adenovirus by real-time PCR were also positive by the FilmArray RP. In addition, upon testing 20 characterized adenovirus serotypes prepared at high and low viral loads, the FilmArray RP did not detect serotypes 6 and 41 at either level and failed to detect serotypes 2, 20, 35, and 37 when viral loads were low. The FilmArray RP system is rapid and extremely user-friendly, with results available in just over 1 h with almost no labor involved. Its low throughput is a significant drawback for laboratories receiving large numbers of specimens, as only a single sample can be processed at a time with one instrument.
INTRODUCTION
A variety of viruses and bacteria are responsible for acute upper and lower respiratory tract infections in children and adults worldwide. Severe and even fatal disease can occur in the very young, the elderly, the immunocompromised, and those individuals with underlying conditions that affect cardiopulmonary function. While certain clinical syndromes are classically associated with specific pathogens, in reality, the syndromes caused by these organisms are often indistinguishable, and rapid and accurate detection is important for appropriate patient management and prevention of nosocomial spread (2, 5, 6, 7, 13, 23, 26, 31, 35). Through the years, PCR has shown excellent clinical utility over the more traditional laboratory methods of virus culture and direct antigen tests for the sensitive detection of respiratory viruses (8, 15, 24, 25, 34), and both laboratory-developed and commercial molecular assays are now available (22). However, because of the complexity of these molecular tests, implementation has been a challenge for the routine clinical laboratory.
More recently, significant advances in microfluidics, microelectronics, and microfabrication have paved the way for the development of simplified molecular systems with the possibility of sample-to-result automation, facilitating their implementation in laboratories that previously lacked the capacity or expertise to perform molecular testing and potentially even at the point of care (16, 30, 36). With this study, we compared the FilmArray Respiratory Panel (RP) (Idaho Technology, Inc., Salt Lake City, UT), which is performed on a fully automated instrument using a self-contained pouch for sample preparation, nested multiplex PCR, and result analysis, to laboratory-developed real-time PCR assays for the rapid detection of a number of respiratory viruses and certain bacterial pathogens.
MATERIALS AND METHODS
Specimens.
A total of 215 archived respiratory specimens from pediatric patients were used, including 196 (91.2%) nasopharyngeal aspirates, 16 (7.4%) nasopharyngeal swabs, 1 (0.5%) tracheal aspirate, 1 (0.5%) bronchoalveolar lavage sample, and 1 (0.5%) autopsy lung tissue. The samples were originally submitted to the Clinical Virology Laboratory at the Children's Hospital of Philadelphia between January 2006 and May 2011, and multiple single-use aliquots of each specimen were immediately stored frozen at −70°C following processing for clinical testing. Patients from whom specimens had been collected ranged in age from 19 days to 22.2 years with a median age of 2.24 years; 70.2% of patients were under 5 years of age. All samples were previously characterized as either negative or positive for one or more viruses using a panel of laboratory-developed real-time PCR assays for adenovirus; respiratory syncytial virus types A and B; influenza virus type A (including subtype determination for seasonal H1 and H3 and 2009 H1) and influenza virus type B; parainfluenza virus types 1, 2, and 3; metapneumovirus; and rhinovirus. A subset of 65 samples originally determined to be positive or negative by real-time PCR for bocavirus (n = 23), Mycoplasma pneumoniae (n = 21), Bordetella pertussis (n = 20), or both M. pneumoniae and B. pertussis (n = 1) were also tested on the FilmArray platform. The individual threshold cycle (CT) values for positive analytes spanned the reportable range of the real-time PCR assays.
In addition to the clinical specimens tested, 20 characterized adenovirus serotypes representing groups A (type 12), B (types 3, 7, 11, 14, 21, and 35), C (types 1, 2, 5, and 6), D (types 10, 19, 20, 29, 32, and 37), E (type 4), and F (types 40 and 41) were obtained from either the American Type Culture Collection (ATCC), Manassas, VA, or Adriana Kajon, Infectious Diseases Program, Lovelace Respiratory Research Institute, Albuquerque, NM. The different serotypes were grown in lung adenocarcinoma epithelial cells (A549; ATCC CCL-185), diluted to achieve targeted real-time PCR CT values between 25.00 and 27.99 and between 35.00 and 37.99 that corresponded to high and low viral loads, respectively, and then tested by the FilmArray RP.
This work was deemed not to be human subject research and was declared to be exempt by the institutional review board at the Children's Hospital of Philadelphia.
Real-time PCR assays.
Nucleic acids were extracted from 200 μl of each clinical specimen by standard procedures using either the MagNA Pure LC automated instrument (Roche Diagnostics, Indianapolis, IN) and corresponding Roche total nucleic acid isolation kit for viral DNA and RNA or the QIAamp DNA mini kit (Qiagen, Germantown, MD) to extract bacterial DNA. Individual real-time PCR assays were performed in 50-μl volumes on a 7500 real-time PCR system (Applied Biosystems, Foster City, CA) using 5 μl of eluted nucleic acid; universal master mixes for either RNA (Ambion AgPath-ID One-Step RT-PCR master mix; Applied Biosystems) or DNA (TaqMan universal master mix; Applied Biosystems); universal amplification conditions consisting of 1 cycle for 10 min at 45°C and 1 cycle for 10 min at 95°C, followed by 45 two-step cycles of 15 s at 95°C and 45 s at 60°C; and TaqMan fluorogenic chemistry for detection (14). Positive and negative controls were processed with each batch of clinical specimens from extraction of nucleic acids through the detection of amplified products. Negative controls consisted of 1.0 × 106 cells/ml of an uninfected human lung carcinoma cell line (A549 cells; ATCC CCL-185), and positive controls were prepared as a mixture of clinical material from previously positive patients. No-template controls were included in each reaction plate for all sets of primers and probes. Primer and probe sequences targeted conserved regions of the genome for each organism (Table 1) and were based on the published literature for adenovirus (10), respiratory syncytial virus types A and B (33), influenza virus types A and B (Centers for Disease Control and Prevention [CDC], personal communication), influenza A virus subtypes H1N1 and H3N2 (32), metapneumovirus (21), rhinovirus (20), bocavirus (19), Mycoplasma pneumoniae (9), and B. pertussis (17) or were designed by the Clinical Virology Laboratory for influenza A virus subtype 2009 H1 and parainfluenza virus types 1, 2, and 3 HN genes (11) using the default parameters of Primer Express software (ABI, Foster City, CA). A human albumin gene primer and probe set was used in separate PCRs as an internal control to ensure that samples contained nucleic acid and to exclude the presence of inhibitors. Specimens and controls were considered positive when the generated fluorescence signal at the threshold cycle (CT) exceeded a defined threshold limit. Specimens that reached the threshold before 38 cycles were considered positive without further testing, and those that reached the threshold at or after 38 cycles but before the last of 45 cycles were considered positive only if, upon duplicate repeat testing of separate aliquots of stored original specimen, at least one of the two repeat tests also reached the threshold before 45 cycles. For certain experiments, the quantity of adenovirus DNA was determined by real-time PCR from a standard curve generated using a set of five nucleic acid standards ranging from 108 to 104 copies/ml.
Table 1.
Nucleotide sequences of primers and probes used for real-time TaqMan PCRa
| Assay | Forward primer (5′ to 3′) | Reverse primer (5′ to 3′) | TaqMan probe (5′ to 3′) | Gene target |
|---|---|---|---|---|
| Adv | GCC ACG GTG GGG TTT CTA AAC TT | GCC CCA GTG GTC TTA CAT GCA CAT C | (FAM)-TGC ACC AGA CCC GGG CTC AGG TAC TCC GA-(TAMRA) | Hexon |
| RSV A | AGA TCA ACT TCT GTC ATC CAG CAA | TTC TGC ACA TCA TAA TTA GGA GTA TCA AT | (FAM)-CAC CAT CCA ACG GAG CAC AGG AGA T-(TAMRA) | N |
| RSV B | AAG ATG CAA ATC ATA AAT TCA CAG GA | TGA TAT CCA GCA TCT TTA AGT ATC TTT ATA GTG | (VIC)-TTC CCT TCC TAA CCT GGA CAT AGC ATA TAA CAT ACC T-(TAMRA) | N |
| IV A | CAT GGA RTG GCT AAA GAC AAG ACC | AGG GCA TTT TGG ACA AAK CGT CTA | (FAM)-TGC AGT CCT CGC TCA CTG GGC ACG-(TAMRA) | Matrix |
| IV A H1 | ACT ACT GGA CTC TGC TKG AA | AAG CCT CTA CTC AGT GCG AA | (FAM)-TTG AGG CAA ATG GAA ATC TAA TAG C-(TAMRA) | HA |
| IV A H3 | TGC TAC TGA GCT GGT TCA GAG T | AGG GTA ACA GTT GCT GTR GGC | (FAM)-AGA T+GC +TC+T A+TT +GG+G AGA CC-(BHQ-1) | HA |
| IV A 2009 H1N1 | TAG AGC CGG GAG ACA AAA | CGT GGA CTG GTG TAT CTG AA | (FAM)-TCT AGT GGT ACC GAG ATA TGC ATT CGC-(TAMRA) | HA |
| IV B | TCC TCA ACT CAC TCT TCG AGC G | CGG TGC TCT TGA CCA AAT TGG | (VIC)-CCA ATT CGA GCA GCT GAA ACT GCG GTG-(TAMRA) | NSP |
| PIV type 1 | ATT GGG CGC CTC ACA AAG | CTC TCG GAC ATT TGT TGA ACC A | (FAM)-CCT GTC ACG ACC AGG AAA CCC AGA CT-(TAMRA) | HN |
| PIV type 2 | AGT CAT ATC TCT TCC GAA CAC AAC AG | AGA TGA TAG ATC CCG CTT CCA A | (FAM)-CAA TGG GCC ACA ATC AAT CCT GCA-(TAMRA) | HN |
| PIV type 3 | CAG ACA AAT CCA AAT YCG AGA TG | YGT YTC CAG CTC ATT ACC AGC AT | (FAM)-AAT ACT GGA AGC AYA CCA ATC ACG GGA A-(TAMRA) | HN |
| MPV | CAT ATA AGC ATG CTA TAT TAA AAG AGT CTC | CCT ATT TCT GCA GCA TAT TTG TAA TCA G | (FAM)-TGY AAT GAT GAG GGT GTC ACT GCG GTT G-(TAMRA) | NP |
| RhV | CY+A GCC +TGC GTG GC | GAA ACA CGG ACA CCC AAA GTA | (FAM)-TCC TCC GGC CCC TGA ATG YGG C-(TAMRA) | 5′ NCR |
| Bocavirus | TGC AGA CAA CGC YTA GTT GTT T | CTG TCC CGC CCA AGA TAC A | (FAM)-CCA GGA TTG GGT GGA ACC TGC AAA-(TAMRA) | NS-1 |
| AGA GGC TCG GGC TCA TAT CA | CAC TTG GTC TGA GGT CTT CGA A | (FAM)-AGG AAC ACC CAA TCA RCC ACC TAT CGT CT-(TAMRA) | NP-1 | |
| B. pertussis | CCC ATA AGC ATG CCC GAT TGA C | CGC ACA GTC GGC GCG GTG AC | (JOE)-TGC CTG AAG CGG CCC GCG C-(BHQ-1) | IS481 |
| M. pneumoniae | CCA ACC AAA CAA CAA CGT TCA | ACC TTG ACT GGA GGC CGT TA | (JOE)-TCA ACT CGA ATA ACG GTG ACT TCT TAC CAC TG-(BHQ-1) | P1 AP |
| Internal control | GCT GTC ATC TCT TGT GGG CTG T | AAA CTC ATG GGA GCT GCT GGT T | (FAM)-CCT GTC ATG CCC ACA CAA ATC TCT CC-(BHQ-1) | Albumin |
Abbreviations: Adv, adenovirus; RSV, respiratory syncytial virus; IV, influenza virus; PIV, parainfluenza virus; MPV, metapneumovirus; RhV, rhinovirus; FAM, 6-carboxyfluorescein; VIC, proprietary formulation, Applied Biosystems; JOE, 6-carboxy-4′,5′-dichloro-2′,7′dimethoxyfluorescein; TAMRA, 6-carboxytetramethylrhodamine; BHQ-1, black hole quencher 1; N, nucleocapsid; HA, hemagglutinin; NSP, nonstructural protein; HN, hemagglutinin-neuraminidase; NP, nucleoprotein; NCR, noncoding region; NS, nonstructural; IS, insertion sequence; AP, adhesion protein. International Union of Biochemistry base codes: Y = C or T, K = G or T, R = A or G. A plus sign in front of a base indicates a locked nucleic acid.
FilmArray instrument and respiratory panel.
Specimens were tested using the FilmArray system and corresponding 21-member premarket Respiratory Panel kit (Idaho Technology, Inc.) according to the manufacturer's instructions. This panel includes assays for adenovirus; bocavirus; coronavirus types 229E, HKU1, OC43, and NL63; influenza A virus (including subtype determination); influenza B virus; metapneumovirus; parainfluenza virus types 1, 2, 3, and 4; respiratory syncytial virus; rhinovirus; Bordetella pertussis; Chlamydophila pneumoniae; and Mycoplasma pneumoniae. A self-contained, disposable, plastic pouch is used to perform the respiratory panel and houses all of the chemicals and enzymes needed to isolate, amplify, and detect nucleic acids from a specimen (Fig. 1). The pouch is divided into discrete segments (blisters) where the specific steps are carried out. Prior to a run, the pouch is inserted into a loading station and individual syringes are used to inject 1.0 ml of molecular-reagent-grade water (hydration solution) into the reagent port to rehydrate the freeze-dried reagents and 300 μl of prediluted, clinical sample (300-μl sample mixed in 500 μl sample buffer) into the sample port. The pouch is then loaded into the instrument, and the assay is started. The instrument then interacts with the pouch to mix and move reagents and heat and cool them to perform cell lysis, magnetic bead-based nucleic acid isolation, reverse transcription, first-stage multiplex PCR, and an array of second-stage nested PCRs. The FilmArray software uses endpoint DNA melting-curve analysis to automatically generate a result for each target. Each sample run contains internal process controls for RNA extraction, DNA extraction, reverse transcription-PCR (RT-PCR), PCR 1, dilution of first-stage PCR product, and PCR 2. A single specimen is run at a time, and results are generated in approximately 70 min.
Fig 1.
Illustration of the FilmArray RP pouch and the steps involved in processing a specimen for testing using the FilmArray system.
Resolution of discordant results.
When initial results were discordant for a particular organism when a specimen was tested by the FilmArray RP and laboratory-developed real-time PCR assays, the two tests were repeated concurrently on two separate aliquots of the relevant original sample that had been stored frozen at −70°C. The final determination of whether a specimen was concordant or discordant was based on an analysis of the initial and repeat test results obtained by both methods. If the initial test result was positive for a given assay and one or both of the duplicate retests were positive, the final result was reported as positive. Conversely, if the initial test result was negative for a given assay and one or both of the duplicate retests were negative, the final result was reported as negative. Organisms not included in the panel of real-time PCR assays but positive by the FilmArray RP, i.e., coronavirus types HKU1, NL63, and OC43 and parainfluenza virus type 4, were not considered discordant between the two methods.
Statistical analysis.
The kappa statistic and all confidence intervals (CIs) were calculated using the VassarStats website for statistical computation, http://faculty.vassar.edu/lowry/VassarStats.html. This site was accessible as of 30 September 2011.
RESULTS
Among pathogens tested for by both methods, the overall agreement between the FilmArray RP and the corresponding real-time PCR assays was 98.6% (kappa = 0.92 [95% confidence interval (CI), 0.89 to 0.94]). The combined positive percent agreement for analytes shared by the two methods was 89.4% (95% CI, 85.4 to 92.6), and the combined negative percent agreement was 99.6% (95% CI, 99.2 to 99.8). Results by pathogen are presented in Table 2. Real-time PCR-positive and FilmArray RP-negative discordant results were more frequently obtained for specimens with higher CT values (i.e., lower pathogen loads) for the real-time PCR assays (Table 3). The mean real-time PCR CT value of all tests with discordant results was 36.46 ± 4.54 (95% CI, 34.70 to 38.22). The greatest comparative difference in sensitivity was observed for adenovirus; only 11 out of 24 (45.8%; 95% CI, 27.9 to 64.9) adenovirus real-time PCR-positive clinical samples were also positive by the FilmArray RP. Compared to the other analytes that were evaluated, the detection of adenovirus by the FilmArray RP was less closely related to the amount of organism present in the PCR; the mean CT value for the 13 specimens that were adenovirus real-time PCR positive and FilmArray RP negative was 33.31 (95% CI, 30.61 to 36.01), while the mean CT value of the 15 discordant tests for analytes other than adenovirus was 39.19 (95% CI, 37.89 to 40.48). In addition, upon testing 20 characterized adenovirus serotypes from groups A to F that were prepared at high and low concentrations of diluted virus, the FilmArray RP did not detect serotypes 6 and 41 at either level and also failed to detect serotypes 2, 20, 35, and 37 when viral loads were low (Table 4).
Table 2.
Detection of respiratory pathogens by FilmArray RP and real-time PCR assays
| Organism | No. detected: FilmArray RP/PCR |
Positive % agreement |
Negative % agreement |
|||||
|---|---|---|---|---|---|---|---|---|
| +/+ (a) | +/− (b) | −/+ (c) | −/− (d) | a/(a + c) (%) | 95% CI | d/(b + d) (%) | 95% CI | |
| Adenovirus | 11 | 0 | 13 | 191 | 11/24 (45.8) | 27.9–64.9 | 191/191 (100) | 98.0–100 |
| Respiratory syncytial virus A | 16 | 0 | 1 | 198 | 16/17 (94.1) | 73.0–99.0 | 198/198 (100) | 98.1–100 |
| Respiratory syncytial virus B | 18 | 0 | 1 | 196 | 18/19 (94.7) | 75.4–99.1 | 196/196 (100) | 98.1–100 |
| Influenza A virus H1 | 17 | 0 | 0 | 198 | 17/17 (100) | 81.6–100 | 198/198 (100) | 98.1–100 |
| Influenza A virus H3 | 15 | 0 | 2 | 198 | 15/17 (88.2) | 65.7–96.7 | 198/198 (100) | 98.1–100 |
| Influenza A virus 2009 H1 | 16 | 0 | 1 | 198 | 16/17 (94.1) | 73.0–99.0 | 198/198 (100) | 98.1–100 |
| Influenza B virus | 15 | 0 | 2 | 198 | 15/17 (88.2) | 65.7–96.7 | 198/198 (100) | 98.1–100 |
| Parainfluenza virus type 1 | 16 | 0 | 1 | 198 | 16/17 (94.1) | 73.0–99.0 | 198/198 (100) | 98.1–100 |
| Parainfluenza virus type 2 | 15 | 0 | 2 | 198 | 15/17 (88.2) | 65.7–96.7 | 198/198 (100) | 98.1–100 |
| Parainfluenza virus type 3 | 16 | 0 | 1 | 198 | 16/17 (94.1) | 73.0–99.0 | 198/198 (100) | 98.1–100 |
| Metapneumovirus | 16 | 0 | 4 | 195 | 16/20 (80) | 58.4–91.9 | 195/195 (100) | 98.1–100 |
| Rhinovirus | 30 | 7 | 0 | 178 | 30/30 (100) | 88.7–100 | 178/185 (96.2) | 92.4–98.6 |
| Bocavirus | 17 | 3 | 0 | 3 | 17/17 (100) | 81.6–100 | 3/6 (50) | 18.8–81.2 |
| B. pertussis | 9 | 0 | 0 | 12 | 9/9 (100) | 70.1–100 | 12/12 (100) | 75.8–100 |
| M. pneumoniae | 9 | 0 | 0 | 13 | 9/9 (100) | 70.1–100 | 13/13 (100) | 77.2–100 |
| Total | 236 | 10 | 28 | 2,372 | 236/264 (89.4) | 85.1–92.6 | 2,372/2,382 (99.6) | 99.2–99.8 |
Table 3.
Performance of FilmArray RP for detection of 15 respiratory pathogens based on real-time PCR CT values
| Organism | No. positive (%) by FilmArray RP for different ranges of real-time PCR CT values |
||||||
|---|---|---|---|---|---|---|---|
| <20 | 20.00–24.99 | 25.00–29.99 | 30.00–34.99 | 35.00–39.99 | 40.00–44.99 | All | |
| Adenovirus | 1/1 (100) | 4/5 (80) | 2/4 (50) | 2/7 (29) | 2/6 (33) | 0/1 (0) | 11/24 (46) |
| Respiratory syncytial virus A | 2/2 (100) | 5/5 (100) | 4/4 (100) | 3/3 (100) | 2/3 (67) | NAa | 16/17 (94) |
| Respiratory syncytial virus B | 1/1 (100) | 4/4 (100) | 5/5 (100) | 4/4 (100) | 3/3 (100) | 1/2 (50) | 18/19 (95) |
| Influenza A virus H1 | 1/1 (100) | 5/5 (100) | 4/4 (100) | 4/4 (100) | 3/3 (100) | NA | 17/17 (100) |
| Influenza A virus H3 | 1/1 (100) | 5/5 (100) | 4/4 (100) | 4/4 (100) | 1/2 (50) | 0/1 (0) | 15/17 (88) |
| Influenza A virus 2009 H1 | 1/1 (100) | 5/5 (100) | 4/4 (100) | 4/4 (100) | 2/3 (67) | NA | 16/17 (94) |
| Influenza B virus | 2/2 (100) | 4/4 (100) | 5/5 (100) | 3/3 (100) | 1/3 (33) | NA | 15/17 (88) |
| Parainfluenza virus type 1 | 1/1 (100) | 5/5 (100) | 4/4 (100) | 4/4 (100) | 2/3 (67) | NA | 16/17 (94) |
| Parainfluenza virus type 2 | NA | 4/4 (100) | 5/5 (100) | 4/4 (100) | 2/4 (50) | NA | 15/17 (88) |
| Parainfluenza virus type 3 | 2/2 (100) | 6/6 (100) | 4/4 (100) | 3/3 (100) | 1/2 (50) | NA | 16/17 (94) |
| Metapneumovirus | 1/1 (100) | 7/7 (100) | 4/4 (100) | 2/3 (67) | 2/2 (100) | 0/3 (0) | 16/20 (80) |
| Rhinovirus | 2/2 (100) | 8/8 (100) | 7/7 (100) | 6/6 (100) | 2/2 (100) | 5/5 (100) | 30/30 (100) |
| Bocavirus | 2/2 (100) | 2/2 (100) | 3/3 (100) | 3/3 (100) | 4/4 (100) | 3/3 (100) | 17/17 (100) |
| B. pertussis | 6/6 (100) | 2/2 (100) | 1/1 (100) | NA | NA | NA | 9/9 (100) |
| M. pneumoniae | NA | 3/3 (100) | 4/4 (100) | 2/2 (100) | NA | NA | 9/9 (100) |
| Total | 23/23 (100) | 69/70 (99) | 60/62 (97) | 48/54 (89) | 27/40 (68) | 9/15 (60) | 236/264 (89) |
NA, specimens not available at this concentration of organism.
Table 4.
Performance of FilmArray RP for detection of characterized adenovirus serotypes at high and low viral loadsa
| Adenovirus type (ATCC no.) | High viral load |
Low viral load |
||||
|---|---|---|---|---|---|---|
| PCR CT value | PCR DNA copies/ml (log10) | FilmArray RP result | PCR CT value | PCR DNA copies/ml (log10) | FilmArray RP result | |
| Group A | ||||||
| ATCC HAdv-12 (VR-863) | 26.05 | 3.41 × 107 (7.53) | Positive | 36.24 | 2.12 × 104 (4.33) | Positive |
| Group B | ||||||
| LRRI 3p | 26.16 | 1.75 × 107 (7.24) | Positive | 37.43 | 1.46 × 104 (4.16) | Positive |
| ATCC HAdv-7 (VR-7) | 25.70 | 4.24 × 107 (7.63) | Positive | 35.34 | 9.89 × 104 (5.00) | Positive |
| ATCC HAdv-11 (VR-12) | 26.47 | 2.61 × 107 (7.42) | Positive | 35.90 | 2.63 × 104 (4.42) | Positive |
| LRRI 14p | 27.20 | 1.65 × 107 (7.22) | Positive | 36.51 | 3.62 × 104 (4.56) | Positive |
| ATCC HAdv-21 (VR-256) | 23.77 | 5.47 × 107 (7.74) | Positive | 37.24 | 3.03 × 104 (4.48) | Positive |
| ATCC HAdv-35 (VR-718) | 28.30 | 6.02 × 106 (6.78) | Positive | 37.97 | 8.25 × 103 (3.92) | Negative |
| Group C | ||||||
| ATCC HAdv-1 (VR-1) | 27.87 | 5.98 × 106 (6.78) | Positive | 35.91 | 7.03 × 104 (4.85) | Positive |
| ATCC HAdv-2 (VR-846) | 25.88 | 1.51 × 107 (7.18) | Positive | 37.45 | 9.35 × 103 (3.97) | Negative |
| LRRI 2 | 27.16 | 9.29 × 106 (6.97) | Positive | 37.24 | 1.64 × 104 (4.22) | Negative |
| LRRI 5p | 27.92 | 1.05 × 107 (7.02) | Positive | 35.15 | 1.12 × 105 (5.05) | Positive |
| ATCC HAdv-6 (VR-6) | 26.84 | 2.07 × 107 (7.32) | Negative | |||
| Group D | ||||||
| LRRI 10 | 27.18 | 9.16 × 106 (6.96) | Positive | 36.50 | 3.64 × 104 (4.56) | Positive |
| ATCC HAdv-19 (VR-254) | 25.24 | 2.18 × 107 (7.34) | Positive | 34.91 | 4.97 × 104 (4.70) | Positive |
| LRRI 20p | 26.37 | 1.06 × 107 (7.03) | Positive | 37.14 | 2.45 × 104 (4.39) | Negative |
| ATCC HAdv-29 (VR-272) | 26.77 | 8.27 × 106 (6.92) | Positive | 36.27 | 1.04 × 104 (4.02) | Positive |
| ATCC HAdv-32 (VR-625) | 27.70 | 7.13 × 106 (6.85) | Positive | 36.30 | 1.04 × 104 (4.02) | Positive |
| ATCC HAdv-37 (VR-929) | 26.41 | 1.57 × 107 (7.20) | Positive | 36.70 | 1.60 × 104 (4.20) | Negative |
| Group E | ||||||
| ATCC HAdv-4 (VR-4) | 24.46 | 3.54 × 107 (7.55) | Positive | 35.03 | 4.55 × 104 (4.66) | Positive |
| Group F | ||||||
| ATCC HAdv-40 (VR-931) | 26.89 | 1.17 × 107 (7.07) | Positive | 36.61 | 3.39 × 104 (4.53) | Positive |
| ATCC HAdv-41 (VR-930) | 27.20 | 9.61 × 106 (6.98) | Negative | |||
Abbreviations: ATCC, American Type Culture Collection; LRRI, Lovelace Respiratory Research Institute. Real-time PCR CT values and viral loads represent means of multiple replicates. Repeat testing was performed on all discordant results.
The FilmArray RP was also highly specific compared to the real-time PCR assays with differences observed for only 10 out of 2,382 negative determinations; rhinovirus/enterovirus was identified in 7 specimens and bocavirus in 3 specimens by FilmArray that were not detected by real-time PCR. With respect to subtype determination for influenza A virus, the FilmArray generated the same subtype result as identified by our laboratory-developed real-time PCR subtyping assays in 44 out of 48 (91.7%) specimens positive for influenza A virus. Four specimens that were subtyped as influenza A seasonal H1 virus by our laboratory-developed assay generated a result of “influenza A virus detected, no subtype detected” by FilmArray.
Forty-four of the 215 tested clinical specimens were positive for more than one pathogen when tested with our laboratory-developed real-time PCR assays. The performance of the FilmArray RP in the detection of these coinfections is summarized in Table 5. The FilmArray RP identified all expected pathogens in 29 out of these 44 specimens (65.9%). In the remaining 15 samples, one of the analytes detected by real-time PCR was not detected by FilmArray. The mean real-time PCR CT value for analytes not detected by the FilmArray RP when present as part of coinfections (n = 15) was 37.55 (95% CI, 35.62 to 39.49), which was not significantly different from the mean CT value (35.19; 95% CI, 31.94 to 38.45) obtained for undetected analytes that were not part of coinfections (n = 13).
Table 5.
Comparative detection of coinfections by FilmArray RP and real-time PCR assaysc
| Total no. of infections | Distinct coinfection combination detected by laboratory-developed real-time PCR assays |
Total no. of coinfections | No. of discrepant coinfectionsa | Discrepant analyte(s)a | |||
|---|---|---|---|---|---|---|---|
| Analyte 1 | Analyte 2 | Analyte 3 | Analyte 4 | ||||
| Adenovirus | MPV | 2 | 2 | Adenovirus (two) | |||
| Adenovirus | Rhinovirus | 1 | 1 | Adenovirus | |||
| Adenovirus | IV A H1 | 2 | 1 | Adenovirus | |||
| Adenovirus | IV A H3 | 1 | 1 | Adenovirus | |||
| Adenovirus | PIV type 3 | 1 | 1 | Adenovirus | |||
| Adenovirus | RSV A | 1 | 0 | ||||
| Adenovirus | RSV B | 1 | 1 | Adenovirus | |||
| MPV | Rhinovirus | 1 | 1 | MPV | |||
| MPV | IV A 2009 H1 | 1 | 0 | ||||
| MPV | PIV type 1 | 1 | 1 | MPV | |||
| MPV | PIV type 3 | 1 | 0 | ||||
| MPV | M. pneumoniae | 1 | 0 | ||||
| Rhinovirus | IV A H1 | 1 | 0 | ||||
| Rhinovirus | IV A H3 | 1 | 0 | ||||
| Rhinovirus | IV A 2009 H1 | 3 | 0 | ||||
| Rhinovirus | PIV type 2 | 1 | 1 | PIV type 2 | |||
| Rhinovirus | PIV type 3 | 1 | 1 | PIV type 3 | |||
| Rhinovirus | RSV B | 2 | 1 | RSV B | |||
| Rhinovirus | Bocavirus | 2 | 0 | ||||
| Rhinovirus | B. pertussis | 6 | 0 | ||||
| IV A H1 | RSV B | 2 | 0 | ||||
| IV A H3 | IV B | 1 | 1 | IV B | |||
| IV B | Bocavirus | 1 | 0 | ||||
| PIV type 1 | Bocavirus | 1 | 0 | ||||
| PIV type 3 | RSV A | 1 | 0 | ||||
| Bocavirus | M. pneumoniae | 1 | 0 | ||||
| Adenovirus | MPV | PIV type 3 | 1 | 0 | |||
| Adenovirus | MPV | RSV B | 1 | 0 | |||
| Rhinovirus | PIV type 2 | RSV B | 1 | 1 | PIV type 2 | ||
| Rhinovirus | PIV type 2 | Bocavirus | 1 | 0 | |||
| Rhinovirus | RSV A | Bocavirus | 1 | 0 | |||
| Adenovirus | Rhinovirus | PIV type 3 | Bocavirus | 1 | 1 | Adenovirus | |
| Dual | 38 | 13 | 13/76b | ||||
| Triple | 5 | 1 | 1/15b | ||||
| Quadruple | 1 | 1 | 1/4b | ||||
| Coinfections | 44 | 15 | 15/95 | ||||
A discrepant coinfection or discrepant analyte was defined as one that was detected by real-time PCR but not detected by the FilmArray Respiratory Panel.
The denominator represents the expected number of detections for specimens containing two (i.e., 38 × 2), three (i.e., 5 × 3), and four (i.e., 1 × 4) pathogens, respectively.
Abbreviations: MPV, metapneumovirus; IV, influenza virus; PIV, parainfluenza virus; RSV, respiratory syncytial virus.
A total of 22 detections of organisms unique to the FilmArray RP and not routinely tested for in our laboratory were observed, including coronavirus types HKU1, NL63, and OC43 in 4, 10, and 5 samples, respectively, and parainfluenza virus type 4 in 3 specimens. These 22 detections occurred in 22 (10.2%) different respiratory specimens. There were no detections of coronavirus 229E or Chlamydophila pneumoniae.
During our evaluation of the FilmArray RP, there were a total of 15 invalid runs out of 359 performed over the 5-month study period. This resulted in a failure rate of 4.2% and necessitated repeat testing of the affected sample in each case. Of the 15 invalid runs, 3 were due to failure of the RNA process control, 1 was the result of a failed PCR 2 control, 4 were due to software error, 2 were due to instrument failure, and 5 were the result of undefined errors where the runs appeared to proceed normally through all stages of testing but the instrument did not generate an analysis or provide final results at the end of the run. In addition, the required vacuum seal failed for 3 (0.8%) pouches provided within different kit lots purchased from the manufacturer, causing the pouches to not rehydrate properly; these pouches were discarded and not used for testing.
DISCUSSION
With this study, we found the FilmArray RP to be comparable in sensitivity and specificity to laboratory-developed real-time TaqMan PCR assays for the detection of a variety of viral and bacterial pathogens. Discordant results between the FilmArray and comparator PCR assays were minimal and mostly involved organisms that were present in low quantity within tested specimens. The one major exception was adenovirus, for which a significant comparative difference in sensitivity was noted with a decreased performance observed for the FilmArray system. With respect to functionality, the FilmArray instrument is a fully integrated, closed system that minimizes contamination concerns while providing complete specimen-to-result automation. The platform has the distinct advantages of being rapid and exceedingly easy to use, and the multiplexed design of the FilmArray RP allows for a single extracted specimen to be efficiently and simultaneously interrogated for multiple organisms. The system requires only 3 to 5 min of total hands-on time to process one sample, and results are available in just over 1 h with almost no labor involved. In this regard, the FilmArray system shows considerable promise to extend the availability of molecular diagnostics to every clinical laboratory and even to point of care. A significant drawback of the system is its low throughput, as only a single sample can be processed on the instrument at one time. Without additional instrumentation at an added expense, this limits its overall utility in laboratories with moderate to high numbers of specimens to be tested. The noted failed runs and vacuum seal failures, which in our study occurred at rates of 4.2% and 0.8%, respectively, are also potential problems that could increase the overall cost of performing the assay.
Our results for the decreased detection of adenovirus are consistent with those reported by others in unpublished evaluations of the FilmArray RP (3, 18) and with data recently provided by the manufacturer (1, 12). The lower limit of detection (LOD) of the FilmArray RP is reported by the manufacturer to be 100 times higher for adenovirus serotype 2 and 10,000 times higher for serotype 6 than the LOD of 300 50% tissue culture infectious doses (TCID50)/ml recorded for most other members of species C and serotypes within species A, B, and D to F. Though we were unable to type the adenoviruses detected from clinical samples by PCR but missed by the FilmArray RP in this study, we do know that 69.3% (480/693) of adenoviruses identified to species level over 7.5 years from 2001 to 2008 in our pediatric population were from species C (Adriana Kajon, personal communication). Knowing that 6 adenovirus serotypes, including 2 out of the 4 members of species C, were not efficiently detected when we examined well-characterized adenovirus strains, we can hypothesize that this most likely accounts for the fact that the FilmArray RP detected only about half of the adenovirus-positive clinical samples in our evaluation. While the other missed types (i.e., serotypes 20, 35, 37, and 41) may be of lesser importance in the pathogenesis of human respiratory disease, adenovirus serotypes 2 and 6 of species C are both among the common serotypes responsible for respiratory syndromes in children (4, 29). Acknowledging the reduced sensitivity for adenovirus serotypes 2 and 6, the manufacturer does recommend that specimens with negative results for adenovirus by FilmArray RP be confirmed by an alternate method (12).
A recently published comparison of the FilmArray RP with another multiplex molecular system for the detection of respiratory viruses, the Luminex xTAG Respiratory Viral Panel (RVP) (Luminex Corporation, Toronto, Canada), demonstrated comparable detections of adenovirus between the two systems (28). One possible explanation for this finding is that neither of the commercial molecular assays is altogether efficient in detecting the many different types of human adenoviruses; the Luminex xTAG RVP has a reported sensitivity of 58.1% for adenovirus in a published comparison with a laboratory-developed molecular test (27).
The FilmArray RP detected rhinovirus/enterovirus in 7 samples that were negative for rhinovirus by our laboratory-developed real-time PCR assay. The FilmArray RP utilizes a combination of PCRs that can detect rhinovirus or that may more broadly detect rhinovirus/enterovirus without distinguishing between the two (12). While there may be some cross-reactivity when enterovirus is present at very high titers, our laboratory-developed real-time PCR assay, using primers and probes originally designed by the CDC (20), is biased toward an enhanced specificity for rhinovirus. It is possible that the FilmArray RP is more sensitive than our laboratory-developed assay for some rhinovirus serotypes, but it may also be the case that some of the real-time PCR-negative but FilmArray RP-positive specimens contained enterovirus rather than rhinovirus.
Correct identification of influenza A virus subtypes using the FilmArray RP was comparable to real-time PCR. There were only 4 (8.3%) of the 48 influenza A virus-positive samples for which the FilmArray RP detected the presence of influenza A virus but did not make a subtype determination; all of these samples had been previously characterized as seasonal H1N1 by our laboratory-developed typing assay. The ability of the FilmArray system to simultaneously detect and accurately subtype influenza A virus with a very rapid turnaround time may be a particularly attractive feature to clinical laboratories and may impact on the clinical decision to prescribe appropriate antiviral medications. Our laboratory uses a two-tiered approach to influenza A virus diagnosis, first using a broadly reactive PCR assay to identify influenza A virus followed by three separate typing assays for seasonal H1, seasonal H3, and 2009 H1. This takes much more time, labor, and resources than does the FilmArray RP.
In our laboratory, small batches of 6 specimens are continuously processed and extracted throughout the day and evening shifts and tested with our 10-member panel of laboratory-developed real-time PCR assays. Our standard procedure incorporates the use of robotic systems, including the epMotion 5075 Liquid Handling Workstation (Eppendorf North America, Hauppauge, NY) for the preparation and routine dispensing of master mix formulations into 96-well plates and the MagNA Pure LC instrument (Roche Diagnostics, Indianapolis, IN) for the postelution pipetting of nucleic acids extracted from clinical specimens. The total run time is approximately 3 h 10 min with a hands-on time of 45 min, and the cost of testing 6 specimens for all 10 respiratory viruses is $334.90, including reagents, supplies, and labor. Assuming the use of a single FilmArray instrument, it takes about 6 h 20 min to test 6 specimens, with a total hands-on time of about 18 to 30 min. The list price for a 30-test FilmArray RP kit is $3,870 ($129.00/test), and no added reagents or supplies are needed. Assuming that 2 tests from each kit lot will be used to run external negative and positive controls and incorporating the cost of labor, the total cost to run 6 samples by FilmArray RP in our laboratory would be $838.36. For lower-volume laboratories and those with different capabilities and workflows, however, the combination of accuracy, ease of use, rapid turnaround time, and cost offered by the FilmArray RP may be more favorable than that of other available methods.
The U.S. Food and Drug Administration recently licensed the FilmArray instrument and FilmArray RP for the U.S. market, and the approved panel includes assays for adenovirus; influenza A virus (including H1, H3, and 2009 H1 subtype determination); influenza B virus; metapneumovirus; parainfluenza virus types 1, 2, 3, and 4; respiratory syncytial virus; rhinovirus/enterovirus; and coronavirus types HKU1 and NL63. The manufacturer is in the process of obtaining a CE marking for international customers and anticipates that the CE-marked version of the FilmArray RP will also include bocavirus, coronaviruses 229E and OC43, B. pertussis, M. pneumoniae, and C. pneumoniae as part of the comprehensive panel (W. Stevenson, Idaho Technology, Inc., personal communication).
ACKNOWLEDGMENT
We thank Adriana Kajon (Infectious Diseases Program, Lovelace Respiratory Research Institute, Albuquerque, NM) for providing characterized adenovirus strains and typing information.
Footnotes
Published ahead of print 23 November 2011
REFERENCES
- 1. Amiott E, et al. 2011. Analytical evaluation of the FilmArray Respiratory Panel, abstr 2586. Abstr. 111th Gen. Meet. Am. Soc. Microbiol. American Society for Microbiology, Washington, DC [Google Scholar]
- 2. Bonner AB, Monroe KW, Talley LI, Klasner AE, Kimberlin DW. 2003. Impact of the rapid diagnosis of influenza on physician decision-making and patient management in the pediatric emergency department: results of a randomized, prospective, controlled trial. Pediatrics 112:363–367 [DOI] [PubMed] [Google Scholar]
- 3. Couturier MR, et al. 2011. Validation of the Idaho Technology Inc. FilmArray Respiratory Panel for clinical use at Primary Children's Medical Center, Salt Lake City, UT, abstr M-60. Abstr. 27th Annu. Clin. Virol. Symp. [Google Scholar]
- 4. Demmler GJ. 2008. Adenoviruses, p 1052–1053 In Long SS. (ed), Principles and practice of pediatric infectious diseases. Churchill Livingstone Elsevier, Philadelphia, PA [Google Scholar]
- 5. D'Heilly SJ, Janoff EN, Nichol P, Nichol KL. 2008. Rapid diagnosis of influenza in older adults: influence on clinical care in a routine clinical setting. J. Clin. Virol. 42:124–128 [DOI] [PubMed] [Google Scholar]
- 6. Esposito S, Marchisio P, Morelli P, Crovari P, Principi N. 2003. Effect of a rapid influenza diagnosis. Arch. Dis. Child. 88:525–526 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Falsey AR, Murata Y, Walsh EE. 2007. Impact of rapid diagnosis on management of adults hospitalized with influenza. Arch. Intern. Med. 167:354–360 [DOI] [PubMed] [Google Scholar]
- 8. Freymuth F, et al. 2006. Comparison of multiplex PCR assays and conventional techniques for the diagnostic of respiratory virus infections in children admitted to hospital with an acute respiratory illness. J. Med. Virol. 78:1498–1504 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Hardegger D, Nadal D, Bossart W, Altwegg M, Dutly F. 2000. Rapid detection of Mycoplasma pneumoniae in clinical samples by real-time PCR. J. Microbiol. Methods 41:45–51 [DOI] [PubMed] [Google Scholar]
- 10. Heim A, Ebnet C, Harste G, Pring-Akerblom P. 2003. Rapid and quantitative detection of human adenovirus DNA by real-time PCR. J. Med. Virol. 70:228–239 [DOI] [PubMed] [Google Scholar]
- 11. Hibbitts S, Rahman A, John R, Westmoreland D, Fox JD. 2003. Development and evaluation of NucliSens basic kit NASBA for diagnosis of parainfluenza virus infection with “end-point” and “real-time” detection. J. Virol. Methods 108:145–155 [DOI] [PubMed] [Google Scholar]
- 12. Idaho Technology, Inc 2011. FilmArray Respiratory Panel (RP) package insert. Idaho Technology, Inc, Salt Lake City, UT [Google Scholar]
- 13. Jernigan DB, et al. 2011. Detecting 2009 pandemic influenza A (H1N1) virus infection: availability of diagnostic testing led to rapid pandemic response. Clin. Infect. Dis. 52:S36–S43 [DOI] [PubMed] [Google Scholar]
- 14. Joshi YB, Elkan M, Hodinka RL. 2005. Real-time PCR for detection of seven major respiratory viruses from clinical specimens, abstr M-55. Abstr. 22nd Annu. Clin. Virol. Symp. [Google Scholar]
- 15. Kuypers J, et al. 2006. Comparison of real-time PCR assays with fluorescent-antibody assays for diagnosis of respiratory virus infections in children. J. Clin. Microbiol. 44:2382–2388 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Lee WG, Kim Y, Chung BG, Demirci U, Khademhosseini A. 2010. Nano/microfluidics for diagnosis of infectious diseases in developing countries. Adv. Drug Deliv. Rev. 62:449–457 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lind-Brandberg L, et al. 1998. Evaluation of PCR for diagnosis of Bordetella pertussis and Bordetella parapertussis infections. J. Clin. Microbiol. 36:679–683 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Loeffelholz M, et al. 2011. Comparison of the FilmArray Respiratory Panel and Prodesse assays for the detection of respiratory pathogens, abstr M-75. Abstr. 27th Annu. Clin. Virol. Symp. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Lu X, et al. 2006. Real-time PCR assays for detection of bocavirus in human specimens. J. Clin. Microbiol. 44:3231–3235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lu X, et al. 2008. Real-time reverse transcription-PCR assay for comprehensive detection of human rhinoviruses. J. Clin. Microbiol. 46:533–539 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Maertzdorf J, et al. 2004. Real-time reverse transcriptase PCR assay for detection of human metapneumoviruses from all known genetic lineages. J. Clin. Microbiol. 42:981–986 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mahony J. 2008. Detection of respiratory viruses by molecular methods. Clin. Microbiol. Rev. 21:716–747 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Mahony J, et al. 2009. Cost analysis of multiplex PCR testing for diagnosing respiratory virus infections. J. Clin. Microbiol. 47:2812–2817 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Mahony J, et al. 2007. Development of a respiratory virus panel test for detection of twenty human respiratory viruses by use of multiplex PCR and a fluid microbead-based assay. J. Clin. Microbiol. 45:2965–2970 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Miernyk K, et al. 2011. Performance of a rapid antigen test (Binax NOW RSV) for diagnosis of respiratory syncytial virus compared with real-time polymerase chain reaction in a pediatric population. J. Clin. Virol. 50:240–243 [DOI] [PubMed] [Google Scholar]
- 26. Mills JM, Harper J, Broomfield D, Templeton KE. 2011. Rapid testing for respiratory syncytial virus in a paediatric emergency department: benefits for infection control and bed management. J. Hosp. Infect. 77:248–251 [DOI] [PubMed] [Google Scholar]
- 27. Pabbaraju K, Tokaryk KL, Wong S, Fox JD. 2008. Comparison of the Luminex xTAG Respiratory Viral Panel with in-house nucleic acid amplification tests for diagnosis of respiratory virus infections. J. Clin. Microbiol. 46:3056–3062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Rand KH, Rampersaud H, Houck HJ. 2011. A comparison of two multiplex methods for the detection of respiratory viruses: FilmArray RP and xTAG RVP. J. Clin. Microbiol. 49:2449–2453 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Rhee EG. 2009. Adenoviruses, p 2028–2029 In Mandell GL, Bennett JE, Dolin R. (ed), Mandell, Douglas, and Bennett's principles and practice of infectious diseases, 7th ed, vol 2 Churchill Livingstone Elsevier, Philadelphia, PA [Google Scholar]
- 30. Riedlinger J, et al. 2010. Multicenter evaluation of the BD Max GBS assay for detection of group B streptococci in prenatal vaginal and rectal screening swab specimens from pregnant women. J. Clin. Microbiol. 48:4239–4241 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Sharma V, Dowd MD, Slaughter AJ, Simon SD. 2002. Effect of rapid diagnosis of influenza virus type A on the emergency department management of febrile infants and toddlers. Arch. Pediatr. Adolesc. Med. 156:41–44 [DOI] [PubMed] [Google Scholar]
- 32. Suwannakarn K, et al. 2008. Typing (A/B) and subtyping (H1/H3/H5) of influenza A viruses by multiplex real-time RT-PCR assays. J. Virol. Methods 152:25–31 [DOI] [PubMed] [Google Scholar]
- 33. van Elden LJ, et al. 2003. Applicability of real-time quantitative PCR assay for diagnosis of respiratory syncytial virus infection in immunocompromised adults. J. Clin. Microbiol. 41:4378–4381 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Weinberg GA, et al. 2004. Superiority of reverse-transcription polymerase chain reaction to conventional viral culture in the diagnosis of acute respiratory tract infections in children. J. Infect. Dis. 189:706–710 [DOI] [PubMed] [Google Scholar]
- 35. Woo PC, Chiu SS, Seto W, Peiris M. 1997. Cost-effectiveness of rapid diagnosis of viral respiratory tract infections in pediatric patients. J. Clin. Microbiol. 35:1579–1581 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Yager P, et al. 2006. Microfluidic diagnostic technologies for global public health. Nature 442:412–418 [DOI] [PubMed] [Google Scholar]