G

V

S1

Ci

2

3

GQ14

D5

6

7

A8

R9

R10

A11

A

12

K13

L14

F15

E16

117

18

b19

l20

m21

m22

i23

l24

R25

o26

27

t28

p29

a30

o31

a32

p33

t34

t35

US

0h

ARTICLE IN PRESS Model

IRMET 12078 1–4

Journal of Virological Methods xxx (2013) xxx– xxx

Contents lists available at SciVerse ScienceDirect

Journal of Virological Methods

jou rn al h om epage: www.elsev ier .com/ locate / jv i romet

hort communication

omparison of conventional lateral-flow assays and a new fluorescentmmunoassay to detect influenza viruses

ary P. Leonardi ∗, Adele M. Wilson, Alejandro R. Zurettiepartment of Pathology & Laboratories, Nassau University Medical Center, East Meadow, NY, United States

rticle history:eceived 9 August 2012eceived in revised form 14 February 2013ccepted 18 February 2013vailable online xxx

eywords:ateral-flow immunoassayluorescence

a b s t r a c t

Sofia, a novel, fluorescent lateral-flow immunoassay was compared with two conventional colorimetricassays, Quickvue Influenza A+B and Directigen FLU A+B, to identify influenza viral antigen from patientnasopharyngeal specimens. A total of 118 frozen original influenza-positive specimens and 57 prospec-tive specimens were examined. Using rt-PCR as a referee assay, sensitivity values (%) for influenza A/B of80.0/74.8, 73.3/59.3 and 73.3/40.7 were obtained using the Sofia, Quickvue and Directigen assays, respec-tively. All assays demonstrated reduced sensitivity for influenza B as compared with influenza A virus.With respect to the Sofia assay, the sensitivity of influenza B for the Directigen assay was significantlydiminished.

nhanced influenza sensitivity False positive results were not observed in the Sofia and Directigen assays. The Quickvue assay produced3 false-positive results (2 influenza A and 1 influenza B) resulting in a specificity (%) of 96 and 98 forinfluenza A and B, respectively. Cross-reactivity to other respiratory viruses was not observed amongimmunoassays. A sensitivity rank (highest to low) of rt-PCR > culture > Sofia> Quickvue > Directigen wasestablished using dilutions of influenza A and B. Sofia provides enhanced sensitivity and objective result

ntion

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

interpretation over conve

. Introduction

Diagnostic immunoassays rely on specific antigen–antibodyinding for accurate results. Although variations in format exist,

ateral flow immunoassays generally use a hydrophobic reactionembrane strip containing bound, anti-target antibodies. Speci-en, treated with reagents to produce target antigen-conjugates,

s allowed to migrate along the membrane, producing a coloredine if the antigen-conjugate is captured by the bound antibodies.esult interpretation is subjective and dependent on the strengthf the color signal.

Advantages demonstrated by lateral flow assays have madehem a cornerstone for influenza virus detection. Rapid result timeermits point-of-care testing to occur in emergency departmentsnd physician office settings on a 24 h, 7-day-a-week basis. Thus,ptimal health care with respect to the administration of antiviralgents or the institution of infection control measures, includingatient isolation, can occur readily. Immunoassays are also rela-

Please cite this article in press as: Leonardi, G.P., et al., Comparison of condetect influenza viruses. J. Virol. Methods (2013), http://dx.doi.org/10.1016

ively inexpensive, require little or no equipment, little technicalraining, and produce results in about 15 min.

∗ Corresponding author at: Virology & Molecular Pathology Laboratory, Nassauniversity Medical Center, 2201 Hempstead Tpke, East Meadow, NY 11554, Unitedtates. Tel.: +1 516 572 8711; fax: +1 516 572 6409.

E-mail address: leonardi@numc.edu (G.P. Leonardi).

54

55

56

166-0934/$ – see front matter © 2013 Published by Elsevier B.V.ttp://dx.doi.org/10.1016/j.jviromet.2013.02.008

al colorimetric immunoassays.© 2013 Published by Elsevier B.V.

Despite these advantages, the sensitivity of rapid influenza

immunoassays is lacking with respect to that obtainable using cul-ture or molecular assays. This was exemplified during the recent

influenza A/2009 H1N1 pandemic. Reports of low, unacceptable

sensitivity values, ranging from 18 to 70% depending on which ref-

eree assay was chosen for comparison, tarnished the perception

of immunoassay worth as diagnostic tools (CDC, 2009; Ginocchio

et al., 2009; Leonardi et al., 2010; Leonardi, 2011).

A novel lateral-flow immunoassay, Sofia (Quidel; San Diego,

CA, USA) has recently received CLIA-waived status for diagnos-

tic testing. Instead of subjective, colorimetric results, this assay

employs immunofluorescent technology to detect influenza nucle-

oprotein, producing objective data. Sofia was compared with two

other lateral-flow immunoassays, Quickvue Influenza A+B (Quidel)

and Directigen FLU A+B (Becton Dickinson, Sparks, MD, USA), to

identify influenza viruses from original nasopharyngeal specimens.

The Directigen and Quickvue immunoassays have been utilized

widely for diagnostic testing and research evaluations (Chartrand

et al., 2012). The latter assay has also been CLIA-waived for diag-

nostic use.

2. Materials and methods

ventional lateral-flow assays and a new fluorescent immunoassay to/j.jviromet.2013.02.008

2.1. Specimens 57

A total of 118 original patient nasopharyngeal swab speci- 58

mens, identified previously as influenza-positive by immunoassay, 59

ING Model

V

2 irologi

c60

i61

d62

t63

264

s65

f66

(67

a68

69

D70

w71

q72

r73

w74

o75

w76

s77

i78

w79

80

a81

m82

p83

u84

(85

a86

287

88

t89

i90

o91

u92

293

94

W95

m96

t97

n98

t99

s100

r101

p102

2103

104

v105

3106

3107

108

s109

f110

w111

b112

i113

B114

b115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

ARTICLEIRMET 12078 1–4

G.P. Leonardi et al. / Journal of V

ulture and/or rt-PCR were used to evaluate immunoassays. Spec-mens were stored at either −20 or −70 ◦C and were obtaineduring the respiratory virus seasons from 2006 to 2011. In addi-ion, 57 nasopharyngeal specimens, obtained during the current011–2012 influenza season, were evaluated prospectively. Thesepecimens were stored at 2–8 ◦C prior to testing which was per-ormed within 72 h post-collection. Determination of patient agechildren 12 and under vs. adults) and influenza A subtype werelso made on study specimens.

Aliquots of consensus-positive influenza A and B virus (NYS-OH, proficiency testing event # 121, sample nos. 1312 & 1314)ere used to prepare sample dilutions which were tested subse-

uently by immunoassay, shell vial culture and rt-PCR methods. Aank order of sensitivity was established. Different dilution seriesere used for the influenza A and B samples because of the high titer

f the influenza A specimen. Dilutions were chosen and tabulatedhen an immunoassay failed to identify virus. At each dilution, a

ample was tested by each immunoassay once. When a negativemmunoassay results occurred at a given dilution, a second sample

as tested to confirm the result.Other common respiratory viruses (respiratory syncytial virus,

denovirus, parainfluenza types 1–3 and human metapneu-ovirus) were also tested by the immunoassays using original

atient specimens. Presence of these viruses was confirmedsing shell-vial culture and immunofluorescent antibody methodsD3Ultra DFA respiratory virus and D3 DFA Metapneumovirus IDssays; DHI, Athens, OH, USA).

.2. Immunoassays

All immunoassays were performed according to manufac-urer’s procedures. Specimens were tested simultaneously by allmmunoassays. Comparison of immunoassays with respect to easef use, result reporting, result time, and equipment needs was eval-ated.

.3. Resolution of discrepant immunoassay results by rt-PCR

A molecular rt-PCR assay (ProFLU+, GenProbe/Prodesse;aukesha, WI, USA) served as the “gold standard” confirmatoryethod to decipher discordant or negative results in specimens

ested among the immunoassays. Shell vial culture (R-mix; Diag-ostic Hybrids, Athens, OH, USA) was also inoculated (100 �L) ifhe remnant specimen volume allowed, so that comparative sen-itivity of the immunoassays and culture could be evaluated withespect to gold standard rt-PCR. Cultures were harvested at 48 host-inoculation and examined using fluorescent antibodies (DHI).

.4. Statistics

95% confidence intervals were computed for assay sensitivityalues using a modified method of Wald (Agresti and Coull, 1998).

. Results

.1. Specimens

In 118 influenza-positive specimens (92 type A and 26 type B), 3pecimens negative by all immunoassays and rt-PCR were excludedrom the study. Of the remaining 115 retrospective specimens, 20ere negative among all immunoassays (16 type A and 4 type B)

Please cite this article in press as: Leonardi, G.P., et al., Comparison of condetect influenza viruses. J. Virol. Methods (2013), http://dx.doi.org/10.1016

ut were confirmed as positive using rt-PCR. Virus was identifiedn 5 of the 57 prospective specimens (4 influenza A and 1 influenza). One prospective specimen was negative by all immunoassaysut confirmed as A/H3-like using rt-PCR.

PRESScal Methods xxx (2013) xxx– xxx

When retrospective and prospective specimens are separately

analyzed, similar sensitivity/specificity values (%) for influenza A

and B of approximately 80/100 in were obtained for the Sofia assay.

Quickvue and Directigen assay sensitivity for influenza A and B in

the retrospective samples were also essentially the same as when

all samples were combined. However, influenza A sensitivity val-

ues (%) in the Quickvue and Directigen assays were respectively

reduced in the prospective samples, producing values of 60 (3 of 5

positive) and 40 (2 of 5 positive samples). All assays produced 100%

sensitivity for influenza type B because all correctly identified the

one positive prospective sample.

Specimens from children, 12 and under accounted for 84 of

the total 175 (48%) samples studied. Of these, 58 of 84 (70.2%)

were influenza influenza-positive (51 type A, 7 type B). The pan-

demic influenza A/2009/H1N1subtype accounting for 38 of the

51 (74.5%) influenza A-positive specimens among children. Of thetotal 97 influenza A specimens studied, 48, 31 and 7 were identi-fied as pandemic A/2009/H1N1, seasonal A/H3N2-like and seasonalA/H1N1-like agents, respectively. Subtype identity was not deter-

mined for 11 of the 97 influenza A specimens.

3.2. Immunoassay performance

Discrepant results (10 influenza A and 12 influenza B specimens)

were produced among the immunoassays (Table 1). Using this data,

sensitivity values (%) for influenza A/B of 80.0/74.8, 73.3/59.3 and

73.3/40.7 were obtained with the Sofia, Quickvue and Directigen

assays, respectively. All immunoassays were more sensitive for

influenza A rather than influenza B viruses. No false positive results

were obtained for the Sofia and Directigen assays; the Quickvue

assay exhibited 3 false-positive results, producing specificity (%)

values of a 96.0/98.0 for influenza A/B, respectfully.

Among 33 cases of discrepant or negative immunoassay results,

enough specimen volume remained to permit inoculation of R-

mix shell-vial cultures. With respect to rt-PCR, culture identified

influenza in 27/33 samples (81.8% sensitivity). All immunoassays

produced negative results when tested with other common respi-

ratory viruses.

Dilutions of known positive-influenza A and B specimens,

obtained from remnant proficiency testing samples, compared

viral detection among immunoassay, culture and molecular assays

(Table 2). A rank order for sensitivity was established. From high

to low: rt-PCR > culture > Sofia > Quickvue > Directigen. The Directi-

gen and Quickvue assays performed poorly in identifying influenza

B samples, failing to identify antigen at dilutions of 1:30 and 1:75,

respectively. Sofia produced positive influenza B results at a dilu-

tion of 1:800. Although the use of a single sample to presently test

the various diagnostic assays at each dilution precludes the ability

to calculate statistical significance, the results clearly demonstrate a

trend toward increased sensitivity for the Sofia assay and warrants

further, more extensive investigation.

In comparing these immunoassays from a laboratory-use per-

spective, the Quickvue assay was simplest to perform, required

less procedural steps, offered the fastest result time and did not

require any operational equipment. However, result interpreta-

tion for Quickvue was cumbersome as the technician sometimes

labored to visualize a pink result line imposed on a pink mem-

brane background. Although the background of the Directigen

cassette membrane was less pink in color, visualization of faintly-

pink positive results were also sometimes difficult to interpret. The

Directigen assay did not require equipment in its operation and

produced results in approximately 15 min.

ventional lateral-flow assays and a new fluorescent immunoassay to/j.jviromet.2013.02.008

The Sofia assay required a fluorescent analyzer to perform the 176

assay. It also had the most procedural steps in assay operation. 177

Despite these limitations, Sofia demonstrated enhanced sensitiv- 178

ity over its competitors, especially with respect to influenza B 179

ARTICLE IN PRESSG Model

VIRMET 12078 1–4

G.P. Leonardi et al. / Journal of Virological Methods xxx (2013) xxx– xxx 3

Table 1Performance characteristics of 3 direct antigen detection assays using rt-PCRa as a referee assay to evaluate discrepant and negative immunoassay results.b

ITEM Influenza type A specimens Influenza type B specimens

Sofia Quickvue Directigen Sofia Quickvue Directigen

TP 72 66 66 22 16 11FN 18 24 24 5 11 16TN 51 49 51 51 50 51FP 0 2 0 0 1 0SENSITIVITY (%) 80.0 73.3 73.3 74.8 59.3 40.7CI (95%)c 70.5–87.0 63.3–81.4 63.3–81.4 62.8–91.0 40.7–75.5 24.5–59.3SPECIFICITY (%) 100 96 100 100 98 100CI (95%) 91.6–100 86–99.7 91.6–100 91.6–100 88.7–99.9 91.6–100PPV (%) 100 97 100 100 94 100NPV(%) 74 67 68 91 82 76.

TP: true positive; FN; false negative; TN: true negative; FP: false positive. Sensitivity was calculated as TP/TP + FN × 100%. Specificity was calculated as TN/TN + FP × 100%.PPV: positive predictive value, calculated as TP/TP + FP; NPV: negative predictive value, calculated as TN/TN + FN. CI: 95% confidence interval.

a ProFlu+ and ProFAST+ assays (GenProbe-Prodesse; Waukesha, WI).b In 33 cases of discrepant or negative results among the antigen detection assays, enough specimen remained for inoculation of R-mix shell vials. Viral antigen was

culture-confirmed in 27 of 33 cases (81.8% sensitivity).c Confidence Interval (95%). Statistical significance demonstrated between Sofia and Directigen influenza B sensitivity.

Table 2Ability of diagnostic assays to identify influenza A and B viruses using dilutions of positive proficiency testing samples (New York State Department of Health-CLEP Program).

Specimen Dilution Diagnostic assay resultsa

Antigen detection Cultureb rt-PCRc

Sofia Quickvue Directigen

Influenza A1:100 + + + + ND1:400 + − − + ND1:3000 + − − + ND1:5000 − ND ND + +1:10,000 − ND ND − +

Influenza B1:25 + + + + ND1:30 + + − + ND1:75 + + − + ND1:100 + − ND + ND1:800 + − ND + ND1:3200 − ND ND + +1: 12,800 − ND ND − +

ND: no data.a All assays were performed according to manufacturer’s recommended procedures.b 180 �

IoFlu+

v180

p181

s182

a183

w184

4185

186

i187

m188

p189

i190

i191

i192

i193

194

b195

m196

o197

p198

2199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

Shell vial cultures (R-mix, DHI, Athens, OH) were inoculated with specimens (nc.).

c Nucleic acids extracted using an EasyMag (Biomerieux, Inc.) and tested using Pr

iruses. In addition, Sofia allowed objective results to be obtained,ermitted the potential integration into a laboratory informationystem, could store 500 results in the analyzer memory, provided

hard-copy result and could be operated in either a read now oralk-away testing mode.

. Discussion

Although the role of lateral flow immunoassays for rapidnfluenza testing was well established, the reliability of these

ethods came into question during the 2009 influenza A/H1N1andemic. This outbreak necessitated an accurate diagnosis of

nfluenza. Further, the ability to distinguish seasonal and pandemicnfluenza A/H1N1 subtypes was necessary because of the differ-ng antiviral drug sensitivity profiles of these viruses. Lateral-flowmmunoassays could not perform either function effectively.

In subsequent influenza seasons, enzyme immunoassays haveeen utilized in a limited capacity. Adequate assay specificity per-

Please cite this article in press as: Leonardi, G.P., et al., Comparison of condetect influenza viruses. J. Virol. Methods (2013), http://dx.doi.org/10.1016

its its use as a first screening tool to rule in disease. Meta-analysisf 159 studies comparing rapid influenza immunoassays reported aooled specificity (%) of 98.2, supporting this role (Chartrand et al.,012). However, highly heterogeneous sensitivity values (95% CI of

l), harvested after 24 h and stained with appropriate monoclonal antibodies (DHI,

and ProFAST+rt-PCR assays (Gen Probe/Prodesse; Waukesha, WI).

57.9% to 66.6%) clearly suggest that negative results cannot stand

alone and require confirmation by DFA, cell culture or molecularassay.

Direct fluorescent-antibody (DFA) immunoassays demonstrate

enhanced sensitivity over their enzymatic counterparts with

one investigation reporting similar performance characteristics

between DFA and PCR in identifying pandemic influenza (Pollack

et al., 2009). An influenza A/2009 H1N1-specific fluorescent

immunoassay was also developed, permitting laboratories lacking

molecular diagnostic capability the ability to definitively identify

this new pathogen (Leonardi, 2010). Unlike enzyme immunoassays,

DFA requires extensive technical training, equipment, and greater

assay turn-around-times. These pitfalls would thus preclude DFA

testing in any point-of-care or emergency room setting.

A decrease in sensitivity for influenza B as compared to influenza

type A was demonstrated among all immunoassays. This finding

parallels the meta-analysis study which reported pooled sensitiv-

ity values (%) of 64.6/52.2 for influenza A/B, respectively Chartrand

ventional lateral-flow assays and a new fluorescent immunoassay to/j.jviromet.2013.02.008

et al., 2012). A plausible explanation for this finding may rest with 218

the comparative “virulence” of the influenza types. Influenza A 219

leads to more severe disease and higher rates of hospitalization 220

and death than infection with type B (Fiore et al., 2011). Differences 221

ING Model

V

4 irologi

i222

w223

s224

s225

s226

p227

s228

s229

r230

231

o232

s233

b234

i235

d236

w237

238

s239

s240

s241

c242

t243

p244

A245

o246

o247

i248

249

w250

u251

f252

t253

f254

u255

t256

3257

s258

u259

260

S261

t262

p263

p264

w265

b266

g267

268

s269

f270

t271

d

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

Leonardi, G.P., 2011. Immunodiagnostic assays, influenza and the H1N1 pandemic. 320

ARTICLEIRMET 12078 1–4

G.P. Leonardi et al. / Journal of V

n virulence also exist among individual sub types of influenza A,here A/H3N2 demonstrates far greater disease severity than sea-

onal A/H1N1 infection (Chartrand et al., 2012). Greater diseaseeverity usually means greater viral load and thus, greater observedensitivity in studies that predominantly examine influenza A sam-les, particularly the A/H3N2-like subtype. Among influenza Apecimens evaluated presently, 7, 31 and 48 were identified aseasonal H1-like, seasonal H3-like and 2009 H1N1-like subtypes,espectively.

Variations in assay sensitivity can also be influenced by the agef patients submitting specimens for study. Influenza immunoas-ays reportedly perform better in children than adults, presumablyecause higher viral loads and longer duration of shedding occur

n the former (Fiore et al., 2011). Meta-analysis has reportedlyemonstrated an approximate 13% increase in assay sensitivityhen children were studied (Chartrand et al., 2012).

The Quickvue and Directigen assays produced modestly greaterensitivity values than obtained in the meta-analysis of pooledtudies examining these products. In addition, influenza A/Bensitivity for Sofia was even higher than observed among itsompetitors. A study bias based on the selection of specimensested may account for the increased sensitivity values observedresently. Factors including a high percentage of positive-influenza

to B specimens (ratio of approximately 4:1), the high percentagef children’s specimens studied (approximately 50%) and a paucityf seasonal H1N1 samples would all favor increased assay sensitiv-ty.

A second source of bias may exist with respect to samples whichere positive by all immunoassays but not confirmed presentlysing gold standard rt-PCR. Positive confirmation was not per-ormed because the amount of sample left following immunoassayesting was often insufficient to permit such analysis. Further, theact that these specimens were previously identified as positivesing methods considered more sensitive than immunoassay (cul-ure and/or rt-PCR) and were currently identified as positive by

immunoassays which detect differing antigens and or antigenicites seem to make the potential for false positive results highlynlikely.

In head-to-head, same-specimen, simultaneous evaluation, theofia outperformed its competitors providing greater accuracy overraditional colorimetric immunoassays. Sofia also produced com-arable sensitivity to culture, however, the frozen-retrospectiveatient specimens may have contained non-viable virus whichould enhance antigen immunoassay detection but negatively

ias culture results. In the dilution studies, culture demonstratedreater sensitivity over the Sofia assay.

Sofia may represent a next step in the evolution of immunoas-

Please cite this article in press as: Leonardi, G.P., et al., Comparison of condetect influenza viruses. J. Virol. Methods (2013), http://dx.doi.org/10.1016

ays, combining lateral flow and fluorescent antibody detectionormats. The choice of fluorescent rather than colorimetric detec-ion is at least partially responsible for the improved sensitivityemonstrated by Sofia because the former is known to increase

PRESScal Methods xxx (2013) xxx– xxx

sensitivity while also widening the dynamic assay detection range

over the latter (Gibbs, 2001). Even greater sensitivity in future

assays may be achieved with the development of luminescent

immunoassays, which presumably may amplify results similar to

that seen in PCR.

With respect to sensitivity, the Sofia assay demonstrated a

statistically significant increase in influenza B viruses over the

Directigen assay. Although a trend toward higher sensitivity for

influenza A was observed for Sofia, the results among assays were

equivalent statistically. A larger clinical study may be needed to

demonstrate any potential significance for influenza A viruses

among these assays.

Features offered by Sofia, including production of objective

results, operation in a run-now or walk-away mode, and having

hard copy and memory stored results make it valuable to the diag-

nostic laboratory. The enhanced sensitivity demonstrated by the

Sofia assay helps reestablish the value of these methods for rapiddiagnostic testing, allowing meaningful patient management deci-sions to be made at the point-of-care.

Acknowledgements

The authors are grateful to Dr. Paul Olivo, Quidel Corporation forhis assistance in the statistical analysis of the data.

Sofia and Quickvue assay kits used in this study were donated

by Quidel Corporation.

References

Agresti, A., Coull, B.A., 1998. Approximate is better than exact for interval estimationof binomial proportions. Am. Stat. 52, 119–126.

Centers for Disease Control and Prevention, 2009. Evaluation of rapid influenza diag-

nostic tests for detection of novel influenza A (H1N1) virus-United States, 2009.

Morb. Mortal. Wkly. Rep. 58, 826–829.

Chartrand, C., Leeflang, M.G., Minion, J., Brewer, T., Madhukar, P., 2012. Accuracy of

rapid influenza diagnostic tests. A meta-analysis. Ann. Int. Med. 156, 500–511.

Fiore, A.E., Fry, A., Shay, D., Gubareva, D., Breese, L., Uyeki, T.M., 2011. Centers for

Disease Control and Prevention (CDC). Antiviral agents for the treatment andchemoprophylaxis of influenza-recommendations of the Advisory Committee

on Immunization Practices (ACIP). Morb. Mortal. Wkly. Rep. 60, 1–24.

Gibbs, J. 2001. Selecting the Detection System – Colorimetric, Fluorescent, Lumines-

cent Methods. ELISA Technical Bulletin No 5. Corning Life Sciences, Corning, Inc.Corning, NY. Revised 1/2001.

Ginocchio, C., Zhang, F., Manji, R., Arora, S., Bornfreund, M., Falk, L., Lotikar, M.,

Kowaska, M., Becker, G., Korologos, G., de Gerinomo, M., Crawford, J.M., 2009.

Evaluation of multiple test methods for the detection of novel influenza A (H1N1)during the New York City outbreak. J. Clin. Virol. 45, 191–195.

Leonardi, G.P., 2010. Rapid identification of 2009 H1N1 influenza A virus using flu-

orescent antibody methods. Am. J. Clin. Pathol. 134, 910–915.

Leonardi, G.P., Mitrache, I., Pigal, A., Freedman, L., 2010. Public hospital-based expe-

rience during an outbreak of pandemic influenza A (H1N1) virus infections. J.

Clin. Microbiol. 48, 1189–1194.

ventional lateral-flow assays and a new fluorescent immunoassay to/j.jviromet.2013.02.008

Clin. Laboratory Int. 35 (April and May), 32–34. 321

Pollack, N.R., Doung, S., Cheng, A., Han, L., Smole, S., Kirby, J.E., 2009. Ruling out 322

novel H1N1 influenza virus infection with direct fluorescent antigen testing. 323

Clin. Infect. Dis. 49, 66–68. 324