Performance of clinical signs to detect Bluetongue virus serotype 8 outbreaks in cattle and sheep during the 2006-epidemic in The Netherlands

HAL Id: hal-00532354

https://hal.archives-ouvertes.fr/hal-00532354

Submitted on 4 Nov 2010

HAL

is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire

destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Performance of clinical signs to detect Bluetongue virus serotype 8 outbreaks in cattle and sheep during the

2006-epidemic in The Netherlands

A.R.W. Elbers, A. Backx, H.M. Ekker, A.N. van der Spek, P.A. van Rijn

To cite this version:

A.R.W. Elbers, A. Backx, H.M. Ekker, A.N. van der Spek, P.A. van Rijn. Performance of clinical signs to detect Bluetongue virus serotype 8 outbreaks in cattle and sheep during the 2006-epidemic in The Netherlands. Veterinary Microbiology, Elsevier, 2008, 129 (1-2), pp.156.

�10.1016/j.vetmic.2007.10.034�. �hal-00532354�

Accepted Manuscript

Title: Performance of clinical signs to detect Bluetongue virus serotype 8 outbreaks in cattle and sheep during the

2006-epidemic in The Netherlands

Authors: A.R.W. Elbers, A. Backx, H.M. Ekker, A.N. van der Spek, P.A. van Rijn

PII: S0378-1135(07)00538-X

DOI: doi:10.1016/j.vetmic.2007.10.034

Reference: VETMIC 3874

To appear in: VETMIC Received date: 16-4-2007 Revised date: 26-10-2007 Accepted date: 31-10-2007

Please cite this article as: Elbers, A.R.W., Backx, A., Ekker, H.M., van der Spek, A.N., van Rijn, P.A., Performance of clinical signs to detect Bluetongue virus serotype 8 outbreaks in cattle and sheep during the 2006-epidemic in The Netherlands, Veterinary Microbiology (2007), doi:10.1016/j.vetmic.2007.10.034

This is a PDF file of an unedited manuscript that has been accepted for publication.

As a service to our customers we are providing this early version of the manuscript.

The manuscript will undergo copyediting, typesetting, and review of the resulting proof

before it is published in its final form. Please note that during the production process

errors may be discovered which could affect the content, and all legal disclaimers that

apply to the journal pertain.

Accepted Manuscript

PERFORMANCE OF CLINICAL SIGNS TO DETECT BLUETONGUE VIRUS SEROTYPE 8 1

OUTBREAKS IN CATTLE AND SHEEP DURING THE 2006-EPIDEMIC IN THE 2

NETHERLANDS 3

4

A.R.W. Elbers¹, A. Backx¹, H.M. Ekker², A.N. van der Spek², and P.A. van Rijn¹ 5

6

1 Department of Virology, Central Institute for Animal Disease Control of Wageningen UR, Lelystad, 7

The Netherlands; ² Food and Consumer Product Safety Authority, The Hague, The Netherlands 8

9

Key words: Bluetongue – serotype 8 - clinical signs - sensitivity – specificity - ROC 10

11

ABSTRACT 12

The performance of clinical signs as a diagnostic test for the detection of BTV-8 outbreaks during the 13

2006-epidemic in The Netherlands was evaluated by constructing and analysing receiver operating 14

characteristic (ROC) curves. The area under the ROC curve of the BT-associated clinical signs in 15

cattle was 0.77. An optimal efficient test (maximising both Sensitivity and Specificity)in cattle herds 16

combined a sensitivity (Se) of 67% with a specificity (Sp) of 72%, comprising the following clinical 17

signs: ulcerations and/or erosions of oral mucosa or erosions of lips/crusts in or around nostrils or 18

oedema of the nose or hyperaemic/purple coloration of tongue, tongue protrusion or coronitis or 19

apathy/tiredness or muscle necrosis, stiffness of limbs or loathing or refusal to move, prostration or 20

torticollis or anoestrus. The area under the ROC curve of the BT-associated clinical signs in sheep was 21

0.81. The optimal efficient test in sheep flocks combined a Se of 76 % with a Sp of 72 %, comprising 22

the following clinical signs: ulcerations of oral mucosa or serous nasal discharge or erosions/ulceration 23

of tongue mucosa or hypersensitivity of the skin or muscle necrosis, stiffness of limbs or coronitis or 24

grinding of teeth or salivation or weakness/paresis.

25 26

INTRODUCTION 27

Bluetongue (BT) is an arthropod-borne viral non-contagious disease of domestic and wild ruminants, 28

particularly affecting sheep with severe clinical disease including mortality. At present 24 different BT 29

virus (BTV)-serotypes have been identified and the disease is transmitted by biting midges 30

Manuscript

Accepted Manuscript

(Culicoides). BT is endemic in southern member-states of the European Union (EU), and several new 31

incursions have been seen in Italy, Greece, Turkey, the French island of Corsica, the Spanish islands 32

of Menorca and Mallorca and Portugal. Up to now, BTV-serotypes 1, 2, 4, 9 and 16 were involved in 33

epidemics in the EU member-states (Mellor and Wittmann, 2002).

34

Starting August 2006 a major epidemic of BT was diagnosed in the North-Western part of Europe, 35

affecting The Netherlands, Belgium, Germany, Luxemburg and the North of France (Elbers et al., 36

2007). BTV may infect many different species of ruminants, but clinical disease signs are generally 37

associated with sheep and consequently most descriptions of the disease apply to sheep (Erasmus, 38

1975). There have been earlier reports on BT outbreaks of serotype 8 (Elbers et al., 2007), but there 39

have not been detailed accounts on the clinical signs associated with BTV-8.

40

The performance of the clinical diagnostic procedure to detect BT-infected livestock herds is crucial 41

for early detection. The diagnosis of BT includes early recognition of a suspect clinical situation by 42

the farmer and the veterinary practitioner, clinical inspection by veterinary specialists and laboratory 43

tests on blood to detect the virus or specific antibodies. Clinical signs associated with BTV can vary 44

considerably between animal species and are sometimes non-specific, especially in cattle.

45

Furthermore, farmers and veterinary practitioners in Western Europe are unfamiliar with the disease.

46

The quality of clinical diagnosis determines whether an infection with BTV will be recognized shortly 47

after infection of livestock animals. In this paper we evaluate the performance of clinical signs for the 48

detection of BTV-8 in cattle herds and sheep flocks by creating and analyzing receiver operating 49

characteristic (ROC) curves. This will help farmers and veterinary practitioners within Western 50

Europe and other countries to be better prepared for clinical recognition of BT.

51 52

MATERIALS AND METHODS 53

Clinical inspection of BTV-8 suspect situations 54

A clinical inspection was executed after a clinical suspicion of BT was reported to the veterinary 55

authorities by, for example, a livestock farmer or veterinary practitioner. Furthermore, a clinical 56

suspicion of BT could originate from a clinical inspection in the framework of a regular screening or 57

Accepted Manuscript

tracing visit in the affected area after the start of outbreaks. If a clinical suspect BT situation was 58

encountered during a clinical inspection, all clinically sick animals were identified and sampled 59

(EDTA blood and serum). An infected herd was defined as a herd where one or more animals tested 60

positive in the reverse transcriptase real-time PCR-test or virus isolation, conducted according to the 61

OIE guidelines (OIE, 2004).

62 63

Clinical inspection data 64

The following data were collected during clinical inspection by the veterinary authorities as part of the 65

official follow-up visit of suspect cases using a standardized questionnaire : herd type (sheep or 66

cattle); demographic information of the farmer (name and address); unique herd number; name of the 67

veterinary inspector that investigated the suspect situation; the date of investigation; the number of 68

animals with BT-associated clinical signs; the number of animals that died with BT-associated clinical 69

signs; date that first clinical signs were seen; the number and identification of animals sampled for 70

laboratory confirmation; an anamnesis (case history): describing the clinical signs seen by the farmer 71

and veterinary practitioner before the official clinical inspection.

72 73

Cattle herds and sheep flocks 74

There were 156 BTV-8 infected cattle herds and 266 BTV-8 infected sheep flocks with clinical 75

disease included in the analysis (positive submission). No BT-associated clinical signs were reported 76

from goat herds, and no BT-associated clinical signs were observed in goats located in mixed herds in 77

which BT was reported in cattle or sheep. A total of 65 cattle herds and 89 sheep flocks were included 78

in the analysis that were clinically suspected of a BTV8-infection but were concluded to be from non- 79

infected herds (negative submission). A negative submission was defined as all sampled animals 80

within the submission being negative both in reverse transcriptase real-time PCR and a blocking 81

ELISA (ID-VET, Montpellier, France) test (Toussaint et al., 2007).

82 83 84

Accepted Manuscript

Data analysis 86

Sensitivity (Se) was defined as the fraction of BT-positive submissions that showed a specific clinical 87

sign in one or more of the animals within the herd and specificity (Sp) as the fraction of BT-negative 88

submissions that did not show a specific clinical sign in one or more animals within the herd. Optimal 89

specific (matches maximum Sp to highest available Se), optimal sensitive (matches maximum Se to 90

highest available Sp), and optimal efficient (maximises both Se and Sp) combination of clinical signs 91

at a herd-level to detect a BT outbreak in the Netherlands in 2006 were determined. The 95%

92

confidence intervals of Se and Sp at these combinations of clinical signs were calculated (Fleiss, 93

1981). The legitimacy of the diagnostic tests at these cut-off values was determined by a two-by-two 94

²- test, or if appropriate by Fisher’s exact test (Statistix, 2003).

95

A ROC curve is a graph of the relationship between Se and Sp of a certain diagnostic test for different 96

values of the discriminator variable assumed as a decision threshold. These cut-off values were set by 97

different combinations of clinical signs. A parallel interpretation approach was used, indicating that 98

animals that had negative test results on different clinical signs were considered negative in the 99

diagnostic test. Se and Sp of the clinical signs were calculated for the study population of infected and 100

non-infected herds (Table 1). For clinical signs and combination of clinical signs, Se was plotted on 101

the ordinate as a function of Sp and, by connecting these points, a ROC curve was constructed. The 102

area under the ROC curve is a quantitative measure of the performance of the clinical signs and was 103

calculated according to Hanley and McNeil (1982). The area under the random ROC curve is 0.5, 104

whereas the area under the ROC of a perfect diagnostic test is 1. The area under the ROC curve was 105

compared with the area under the random ROC curve using the critical ratio as a test statistic (McNeil 106

and Hanley, 1984), and the critical ratio was compared with the Normal distribution (Statistix, 2003).

107 108

RESULTS 109

BTV-8 associated clinical signs appeared much more prominent in sheep than in cattle, and clinical 110

signs were expressed differently in sheep compared to cattle. The most prominent clinical signs in 111

BTV-8 affected cattle were: crusts/lesions of the nasal mucosa, erosions of lips/crusts in or around the 112

nostrils, erosions of the oral mucosa, salivation, fever, conjunctivitis, coronitis, muscle necrosis, and 113

Accepted Manuscript

stiffness in limbs (Table 1). Erosions of the oral mucosa, fever, salivation, facial and mandibular 114

oedema, apathy and tiredness, oedema of the lips, lameness, and dysphagia were among the most 115

frequent clinical signs recorded in BTV-8 affected sheep flocks (Table 1). However, some of these 116

clinical signs (e.g. fever both in sheep and cattle, salivation in cattle, facial oedema in sheep) were 117

both seen in the same degree in infected as in non-infected herds, resulting in a rather low specificity.

118

The area under the ROC curve of the clinical signs in cattle was 0.77 (standard error (se): 0.0335), 119

which was significantly (Z =39.8, P < 0.0001) larger than the area under the random ROC curve 120

(Figure 1). The optimal efficient test in cattle combined a Se of 67% with Sp of 72%, comprising the 121

following clinical signs (Table 2): ulcerations and/or erosions of oral mucosa or erosions of lips/crusts 122

in or around nostrils or oedema of the nose or hyperaemic/purple coloration of tongue, tongue 123

protrusion or coronitis or apathy/tiredness or muscle necrosis, stiffness of limbs or loathing or refusal 124

to move, prostration or torticollis or anoestrus.

125

The area under the ROC curve of the clinical signs in sheep was 0.81 (standard error (se): 0.0262), 126

which was significantly (Z =40, P < 0.0001) larger than the area under the random ROC curve (Figure 127

2). The optimal efficient test in sheep combined a Se of 76 % with a Sp of 72 %, comprising the 128

following clinical signs (Table 2): ulcerations of oral mucosa or serous nasal discharge or 129

erosions/ulceration of tongue mucosa or hypersensitivity of the skin or muscle necrosis, stiffness of 130

limbs or coronitis or grinding of teeth or salivation or weakness/paresis.

131 132

DISCUSSION 133

The results of the ROC-analysis indicate that BT-associated clinical signs in cattle and sheep have a 134

significantly higher performance as a diagnostic test for detection of BTV-8 infection than can be 135

attributed to chance. The performance of BT-associated clinical signs to detect BTV-8 infected herds 136

is much higher compared to the performance of clinical signs of Classical Swine Fever (CSF) to detect 137

CSF-infected pig herds with an area under the curve of 0.66 (Elbers et al., 2002) and is also higher 138

compared to the performance of gross lesions in individual pigs at post-mortem examination to detect 139

CSF-outbreaks with an area under the curve of 0.72 (Elbers et al., 2003), but lower compared to the 140

Accepted Manuscript

performance of gross lesions in poultry associated with High Pathogenic Avian Influenza subtype 141

H7N7 to detect HPAI-outbreaks with an area under the curve of 0.86 (Elbers, unpublished data).

142

It can be concluded that clinical investigation is an important step in the recognition of BT, but this 143

diagnostic tool is not very specific: clinical signs, pointing at a BTV infection, frequently occur 144

without a real BTV infection (caused by other diseases), both in sheep and cattle. The fact that BT was 145

absent in this part of Europe complicated a fast recognition of the disease by farmers’ and 146

veterinarians’ because of the low level of awareness. As early as late June, Belgian veterinary 147

practitioners observed an unusual number of bovine cases primarily attributed at that time to 148

photosensitisation or exposure to mycotoxins. These first cases were not attributed to BT at that time 149

because samples from these animals were not tested with a BT-confirmatory diagnostic test. However, 150

later on BT was confirmed in these bovine cases and the clinical signs recorded could explain a 151

possible misdiagnosis at the end of June (Thiry et al., 2006). In the first 5 BT outbreaks in the 152

Netherlands (all in sheep), the owners indicated that the first clinical signs started 13 – 15 days before 153

a suspicion was reported to the veterinary authorities via a veterinary practitioner. In addition, there is 154

some reluctance to report suspect clinical situations in general by farmers and veterinary practitioners 155

to the veterinary authorities in fear of anticipated social and economic consequences (Elbers et al., 156

2006). This is a dangerous situation, because if the above described clinical problems would have been 157

caused by e.g. FMD, disease would have spread beyond control to a considerable number of other 158

farms in the two weeks between first appearance of clinical signs and reporting to the veterinary 159

authorities. Since clinical signs of BT can resemble foot-and-mouth disease (FMD), this disease and 160

other vesicular diseases must be included in the differential diagnosis for cattle and sheep. Other 161

diseases that should be included for differentiation in cattle include bovine viral diarrhoea (BVD), 162

infectious bovine rhinotracheitis (IBR), malignant catarrhal fever (MCF), photosensitization, and 163

miscellaneous causes of bovine stomatitis (Luedke and Jones, 1984). Diseases that should be included 164

for differentiation in sheep include photosensitization, polyarthritis, foot rot, white muscle disease, 165

contagious ecthyma, haemonchosis, sheep pox, necrobacillosis, PPR, heartwater and pulpy kidney 166

disease (Verwoerd and Erasmus, 2004).

167 168

Accepted Manuscript

ACKNOWLEDGEMENTS 169

Edwin Vries and Dennis Bol (Food and Consumer Product Safety Authority, The Hague, The 170

Netherlands) are gratefully acknowledged for help with the farm visit reports. We thank Harm van 171

Wees (University of Professional Agricultural Education, Dronten, The Netherlands) for preparing 172

part of the data for analysis. This research was partly sponsored by the European Food Safety 173

Authority (EFSA), contract CT/EFSA/SCAD/2006/01.

174 175

REFERENCES 176

177

Elbers, A.R.W., Bouma, A., Stegeman, J.A., 2002. Quantitative assessment of the quality of clinical 178

signs for the detection of classical swine fever outbreaks. Vet. Microbiol. 85, 323-332.

179

Elbers, A.R.W., Vos, J.H., Bouma, A., van Exsel, A.C.A., Stegeman, J.A. Assessment of the significance of 180

gross lesions at post mortem for the detection of classical swine fever outbreaks. Vet. Microbiol. 2003; 96:

181

345-356.

182 183

Elbers, A.R.W., Loeffen, W.L.A., Dekker, A., Koch, G., Van Rooij, E.M.A., 2006. Substantial 184

improvement of early detection of notifiable animal diseases: a call for unorthodox changes. In: Proc.

185

11^th Symp. Int. Soc. Vet. Epid. Econ., Cairns, Australia: ISVEE 11, 36.

186 187

Elbers, A.R.W., Mintiens, K., Staubach, C., Gerbier, G., Meiswinkel, R., Hendrickx, G., Backx, A., 188

Conraths, F.J., Meroc, E., Ducheyne, E., Gethmann, J., Heesterbeek, J.A.P., De Clercq, K., Unger, F., 189

Stegeman, J.A., 2007. Bluetongue virus serotype 8 epidemic in North-Western Europe in 2006:

190

preliminary findings. In: Proceedings of the Society for Veterinary Epidemiology and Preventative 191

Medicine (SVEPM), Helsinki, Finland (28-30 March 2007), 231-245.

192 193

Fleiss, 1981. Statistical methods for rates and proportions. Second Edition, John Wiley and Sons, New 194

York, USA.

195

Accepted Manuscript

Hanley, J.A., McNeil, B.,J., 1982. The meaning and use of the area under a receiver operating 196

characteristic (ROC) curve. Radiology 143, 29-36.

197

Luedke, A.J., Jones, R.H., 1984. Bluetongue: diagnosis and significance in the bovine animal. The 198

Bovine Practitioner 19, 79-86.

199

McNeil, B.J., Hanley, J.A., 1984. Statistical approaches to the analysis of receiver operating 200

characteristic (ROC) curves. Med. Decis. Making 4, 137-150.

201

Mellor, P.S., Wittmann, E.J., 2002. Bluetongue virus in the Mediterranean Basin 1998-2001. Vet. J.

202

164, 20-37.

203

OIE, 2004. Bluetongue. In : Manual of diagnostic tests and vaccines for terrestrial animals. Fifth 204

Edition, 2004 (http://www.oie.int/eng/normes/mmanual/A_00032.htm, accessed 25 October 2007).

205

Statistix, 2003. Statistix version 8, User’s Manual. Analytical Software, Tallahassee, FL, USA.

206

Thiry, E., Saegerman, C., Guyot, H., Kirten, P., Losson, B., Rollin, F., Bodmer, M., Czaplicki, G., 207

Toussaint, J-F., De Klercq, K., Dochy, J-M., Dufey, J., Gillesman, J-L., Messeman, K., 2006.

208

Bluetongue in northern Europe. Vet. Rec. 159, 327.

209

210

Toussaint, J-F., Sailleau, C., Mast, J., Houdart, P., Czaplicki, G., Demeestere, L., VandenBussche, F., 211

van Dessel, W., Goris, N., Breard, E., Bounaadja, L., Thiry, E., Zientara, S., De Clercq, K., 2007.

212

Bluetongue in Belgium, 2006. Emerg. Infect. Dis. 13, 614-616.

213 214

Verwoerd, D.W., Erasmus, B.J., 2004. Bluetongue. In: Coetzer, J.A.W., Tustin, R.C. (eds.), Infectious 215

Diseases of Livestock, Vol. 2, Oxford University Press, pp. 1201-1220.

216

Accepted Manuscript

Table 1. Sensitivity (Se) and specificity (Sp) of clinical signs in cattle and sheep to detect Bluetongue (BT) outbreaks during the BT serotype 8 epidemic in 2006 in the Netherlands.

Cattle Sheep Clinical signs observed within cattle and sheep herds

(one or more animals) Se %

(156 herds) Sp %

(65 herds) Se %

(266 flocks) Sp % (89 flocks) General

Fever 24 66 44 47

Apathy, tiredness 16 98 35 82

Dysphagia, difficulty in swallowing 16 89 29 88

Oedema of nose 8 97 17 86

Facial and mandibular oedema 8 92 38 68

Oedema of lips 3 98 31 79

Oedema scrotum/vulva 1 94 0.4 98

Oedema of ears 0 100 3 93

Anorexia, weight loss, emaciation 3 95 8 92

Mortality 3 94 15 76

Grinding of teeth 3 98 3 98

Locomotion disorders

Muscle necrosis, stiffness in limbs 22 95 15 94

Coronitis 22 91 18 91

Lameness 14 92 29 82

Weakness, paresis 3 98 15 98

Loathing or refusal to move, prostration 1 100 3 93

Sloughing of the hooves 0 100 0 100

Nervous disorders

Torticollis 1 100 0 99

Mucosal membranes

Lesions and crusts nasal mucosa 55 66 14 84

Erosions of oral mucosa 28 77 49 68

Salivation 27 78 41 79

Conjunctivitis 23 72 3 91

Erythema, redness of nasal mucosa 21 75 3 97

Serous nasal discharge 20 86 10 98

Hyperaemic/purple coloration, lesions of teats 15 74 0.4 100

Ulcerations of oral mucosa 14 97 12 100

Erythema, redness of oral mucosa 12 91 30 89

Purulent nasal discharge 11 91 10 94

Erosions/ulceration of tongue mucosa 1 95 12 95

Petechial haemorrhages oral mucosa 1 97 2 95

Skin disorders

Erosions of lips/crusts in or around nostrils 29 89 24 68 Hyperaemic/purple coloration of tongue,

tongue protrusion 6 97 20 89

Hypersensitivity of skin 3 97 0.4 100

Broken skin around mouth 1 98 4 93

Erythema, inflammations, redness of skin 0 98 1 97 Reproduction disorders

Abortion 2 97 0 99

Stillbirth 1 98 0 100

Anoestrus 1 100 0 100

Teratogenic fetus 0 100 0 100

Congenital deformities 0 100 0 100

Respiratory disorders

Dyspneu 5 95 7 88

Production disorders

Depression in milk production 7 91 0 100

Broken Wool/alopecia - - 0 99

Digestive disorders

Diarrhoea 2 91 2 95

Regurgitation, vomiting 0 100 0.4 100

Table

Accepted Manuscript

Table 2. Sensitivity (Se), specificity (Sp) and accessory 95% confidence intervals (CI) of combinations of clinical signs in cattle herds and sheep flocks for the detection of a Bluetongue serotype 8 outbreak at optimal sensitive, optimal specific and optimal efficient cut-off levels.

Clinical signs observed in one or more

affected animals within ……. Optimal Se in %

(95 % CI) Sp in % (95% CI)

²- test or Fisher exact test (p-value) Cattle herds

Ulcerations/erosions of oral mucosa or lesions and crusts nasal mucosa or oedema of the nose or erosions of lips/crusts in or around nostrils or serous nasal discharge or salivation or conjunctivitis or Coronitis or lameness or Apathy/tiredness or muscle necrosis, stiffness of limbs or loathing or refusal to move, prostration or torticollis or dysphagia, no or less drinking or fever or mortality

sensitive

99 ( 95–100 )

11 ( 5 – 22 )

0.003

Loathing or refusal to move, prostration or torticollis or

anoestrus or apathy/tiredness specific 17

( 12 – 24 ) 99

(91 - 100) 0.001 Ulcerations/erosions of oral mucosa or erosions of

lips/crusts in or around nostrils or oedema of the nose or Hyperaemic/purple coloration of tongue, tongue protrusion or coronitis or apathy/tiredness or

muscle necrosis, stiffness of limbs or loathing or refusal to move, prostration or torticollis or anoestrus

efficient

67 ( 60 – 75 )

71 (60 – 82 )

< 0.0001

Sheep flocks

Ulcerations of oral mucosa or lesions and crusts nasal mucosa or erythema, redness of nasal mucosa or oedema of the lips or nose or serous or purulent nasal discharge or hyperaemic/purple coloration of tongue, tongue protrusion or erosions/ulceration of tongue mucosa or hypersensitivity of skin or

Salivation or grinding of teeth or conjunctivitis or dyspneu or apathy, tiredness or weakness, paresis or muscle necrosis, stiffness of limbs or loathing or refusal to move, prostration or coronitis or

lameness or dysphagia, no or less drinking or anorexia, emaciation, weight loss or mortality

sensitive 99 ( 97–100 )

14 ( 8 – 23 )

< 0.0001

Ulcerations of oral mucosa or hypersensitivity of skin specific 12 ( 8 – 17 )

100 (95-100)

0.0004 Ulcerations of oral mucosa or serous nasal discharge or

erosions/ulceration of tongue mucosa or

hypersensitivity of skin or muscle necrosis, stiffness of limbs or coronitis or grinding of teeth or salivation or weakness, paresis

efficient 76 ( 70 – 81 )

72 (63 – 81 )

< 0.0001

Accepted Manuscript

0 10 20 30 40 50 60 70 80 90 100

0 10 20 30 40 50 60 70 80 90 100

Specificity (in %)

Sensitivity (in %)

Figure 1. ROC curve of combinations of clinical signs in cattle herds (solid line) and random ROC curve (dotted line) for the detection of Bluetongue serotype 8 outbreaks during the epidemic in 2006 in the Netherlands.

Figure

Accepted Manuscript

0 10 20 30 40 50 60 70 80 90 100

0 10 20 30 40 50 60 70 80 90 100

Specificity (in %)

Sensitivity (in %)

Figure 2. ROC curve of combinations of clinical signs in sheep flocks (solid line) and random ROC curve (dotted line) for the detection of Bluetongue serotype 8 outbreaks during the epidemic in 2006 in the Netherlands.