Image texture, low contrast liver lesion detectability and impact on dose: Deep learning algorithm compared to partial model-based iterative reconstruction.

OBJECTIVES: To compare deep learning (True Fidelity, TF) and partial model based Iterative Reconstruction (ASiR-V) algorithm for image texture, low contrast lesion detectability and potential dose reduction. METHODS: Anthropomorphic phantoms (mimicking non-overweight and overweight patient), containing lesions of 6 mm in diameter with 20HU contrast, were scanned at five different dose levels (2,6,10,15,20 mGy) on a CT system, using clinical routine protocols for liver lesion detection. Images were reconstructed using ASiR-V 0% (surrogate for FBP), 60 % and TF at low, medium and high strength. Noise texture was characterized by computing a normalized Noise Power Spectrum filtered by an eye filter. The similarity against FBP texture was evaluated using peak frequency difference (PFD) and root mean square deviation (RMSD). Low contrast detectability was assessed using a channelized Hotelling observer and the area under the ROC curve (AUC) was used as figure of merit. Potential dose reduction was calculated to obtain the same AUC for TF and ASiR-V. RESULTS: FBP-like noise texture was more preserved with TF (PFD from -0.043mm-1 to -0.09mm-1, RMSD from 0.12mm-1 to 0.21mm-1) than with ASiR-V (PFD equal to 0.12 mm-1, RMSD equal to 0.53mm-1), resulting in a sharper image. AUC was always higher with TF than ASIR-V. In average, TF compared to ASiR-V, enabled a radiation dose reduction potential of 7%, 25 % and 33 % for low, medium and high strength respectively. CONCLUSION: Compared to ASIR-V, TF at high strength does not impact noise texture and maintains low contrast liver lesions detectability at significant lower dose.

Prospective multicenter study on personalized and optimized MDCT contrast protocols: results on liver enhancement.

OBJECTIVE: To determine a personalized and optimized contrast injection protocol for a uniform and optimal diagnostic level of liver parenchymal enhancement, in a large patient population enrolled in a multicenter study. METHODS: Six hundred ninety-two patients who underwent a standardized multi-phase liver CT examination were prospectively assigned to one contrast media (CM) protocol group: G1 (100 mL fixed volume, 37 gI); G2 (600 mgI/kg of total body weight (TBW)); G3 (750 mgI/kg of fat-free mass (FFM)), and G4 (600 mgI/kg of FFM). Change in liver parenchyma CT number between unenhanced and contrast-enhanced images was measured by two radiologists, on 3-mm pre-contrast and portal phase axial reconstructions. The enhancement histograms were compared across CM protocols, specifically according to a target diagnostic value of 50 HU. The total amount of iodine dose was also compared among protocols by median and interquartile range (IQR). The Kruskal-Wallis and Mann-Whitney U tests were used to assess significant differences (p < 0.005), as appropriate. RESULTS: A significant difference (p < 0.001) was found across the groups with liver enhancement decreasing from median over-enhanced values of 77.0 (G1), 71.3 (G2), and 65.1 (G3) to a target enhancement of 53.2 HU for G4. Enhancement IQR was progressively reduced from 26.5 HU (G1), 26.0 HU (G2), and 17.8 HU (G3) to 14.5 HU (G4). G4 showed a median iodine dose of 26.0 gI, significantly lower (p < 0.001) than G3 (33.9 gI), G2 (38.8 gI), and G1 (37 gI). CONCLUSIONS: The 600 mgI/kg FFM-based protocol enabled a diagnostically optimized liver enhancement and improved patient-to-patient enhancement uniformity, while significantly reducing iodine load. KEY POINTS: • Consistent and clinically adequate liver enhancement is observed with personalized and optimized contrast injection protocol. • Fat-free mass is an appropriate body size parameter for correlation with liver parenchymal enhancement. • Diagnostic oncology follow-up liver CT examinations may be obtained using 600 mgI/kg of FFM.

Assessment of the cost-effectiveness of alternative bovine tuberculosis surveillance protocols in French cattle farms using the mixed interferon gamma test.

Periodic screening in farms, using intradermal cervical comparative tuberculin test (ICCT), is a component of the French ante mortem surveillance of bovine tuberculosis (bTB). Previous studies have estimated the cost-effectiveness of the French mandatory bTB screening protocols. In these protocols, a second ICCT (ICCT2) is performed 42 days after the first one (ICCT1), either on the entire herd (strict protocol) or in series on animals with non-negative results (reactors) to ICCT1 (compliant protocol). The 42-days interval reduced protocols' cost-effectiveness. To minimize this interval, we suggested two alternative protocols, in which a mixed interferon gamma test (IFNMIX), with better sensitivity than ICCT2 and comparable specificity, replaces the ICCT2, and is carried out directly after the ICCT1. In the strict alternative protocol, reactors to ICCT1 are culled to perform laboratory analyses (PCR, bacteriology). Negative results to these analyses imply the IFNMIX testing of the entire herd. In the compliant alternative protocol, only reactor(s) to ICCT1 are tested with IFNMIX, and animals with positive results to IFNMIX are culled for laboratory analyses. We evaluated these protocols through scenario tree modelling. The estimated cost-effectiveness indexes showed that the compliant alternative protocol was the most efficient. The strict protocols (mandatory and alternative) were never the most efficient, but were the most effective. Therefore, using IFNMIX instead of ICCT2 may be useful in reducing the costs of the compliant protocol used when the probability of infection is considered low. The strict alternative protocol may become more attractive would IFNMIX's price decreased.

Cost-effectiveness of two different protocols for animal tracing investigations of bovine tuberculosis outbreaks in France.

In the French bovine tuberculosis (bTB) surveillance program, tracing-on and back investigations have a major importance as, in 2016, they represented about 21 % of the detected outbreaks. Building on our previous work on the other surveillance system components (Poirier et al., 2019), we evaluated for the first time the sensitivity and the cost of the two existing protocols of bTB's tracing-on investigations trough scenario tree modelling with a stochastic approach. We used French databases (national database for bovine identification and database recording all bTB surveillance and control results) and direct and indirect costs collected in a previous study. These assessments allowed us to calculate the cost-effectiveness index (cost/sensitivity) of each tracing-on protocol. In the first protocol (trace-and-cull protocol), the animal(s) linking the farm to an outbreak are systematically culled for bacteriology, PCR and histology testing. In the second protocol (trace-and-test protocol), the traced animal is culled only if it had non-negative result to an intradermal cervical comparative tuberculin test (ICCT). We estimated herd sensitivity of the two tracing-on protocols for 12 herd types defined by their production type, size and herd turnover. For the trace-and-cull protocol, mean herd sensitivity was estimated between 67.3 % [66.8-67.7]CI95 % and 89.2 % [88.7-89.7]CI95 % and between 51.2 % [50.8-51.5]CI95 % and 73.1 % [72.6-73.6]CI95 % for the trace-and-test protocol, depending on herd type. The trace-and-cull protocol was between 278 €/herd and 717 €/herd more expensive than the trace-and-test protocol, depending on herd type. Regardless of herd type, the trace-and-cull protocol had the smaller cost/sensitivity ratio and was therefore the most cost-effective protocol. That work showed that systematically culling traced animals to perform bacteriology and PCR on them (trace-and-cull protocol) is associated with a better herd sensitivity and is more cost-effective for all herd types. That is consistent with French veterinary authorities' recommendations but does not account for sociological aspects such as the bond between the farmer and his animals. Yet, cost-effectiveness difference was minor in small dairy and beef herds with a low turnover, suggesting the protocol could be chosen depending on the epidemiological context in those herds.

Integrating data of veterinarians' practices in assessing the cost effectiveness of three components of the bovine tuberculosis surveillance system by intradermal tuberculin testing in French cattle farms through a scenario-tree approach.

Disease surveillance systems' effectiveness relies on participants following prescribed practices. We developed a general method to improve a previous cost-effectiveness evaluation of three French screening program protocols for bovine tuberculosis (bTB) to account for the practices of participants by scenario tree modelling. This method relies on: 1) semi-directive interviews of participants to identify the variability of practices and potentially influential factors, and to understand the sociological context; 2) a quantitative survey, based on multiple-choice questions, to quantify various practices and identify significantly influential factors by multivariable regression analyses; 3) addition of the scenario-tree nodes corresponding to the practices and their influential factors and configuration of the new limbs according to the data of the quantitative survey. We used this approach to integrate data concerning veterinary practices and identify some failures to conform to regulatory guidelines regarding intradermal cervical comparative tuberculin test (SICCT) (testing and notification of non-negative results). Such nonconformities appeared to be mainly caused by cattle restraint issues and the perception of veterinarians of the bTB control program. Indeed, their perception of that program significantly influenced veterinarians' practices. We modelled the influence of the SICCT practices on the SICCT results. The incorporation of these data led to a major decrease of the herd sensitivity estimations relative to the previous assessments that did not incorporate data of practices (15% to 42% decrease). This result shows the important impact of veterinarians' practices and their influencing factors (such as perception of the bTB control program) on the effectiveness of the surveillance system.

Cost-effectiveness assessment of three components of the bovine tuberculosis surveillance system by intradermal tuberculin testing in French cattle farms by a scenario tree approach.

In most officially bovine tuberculosis (bTB)-free countries, bTB has not been fully eradicated. Costly and time-consuming surveillance and control measures are therefore still in place to control this infection. An officially bTB-free status, both at the national and at the herd level, influences whether and when animals can be sold. Thus, this infection is still an economic issue, justifying measures towards its eradication. An evaluation of the cost-effectiveness of such measures would be highly useful, especially to optimise the costs of control measures and their adaptation to a local epidemiological context. We evaluated the cost-effectiveness of three mandatory surveillance protocols currently used in France by herd type (type of production, size, and turnover of the herd) under French field conditions. The first protocol ("strict") implies the direct slaughter and post-mortem analyses of any intradermal cervical tuberculin test (ICT) reactor, and negative results to a second intradermal cervical comparative tuberculin test (ICCT) to regain bTB-free status of the herd. In the second protocol ("compliant quick-path") bTB-free status can be regain if post-mortem analyses of reactors to the first ICT are negative. In the third protocol ("compliant slow-path"), ICCT-reactive animals are tested using the interferon gamma assay; the results of this test influence the path of further investigation. We built scenario trees for each of these protocols at the animal level. They allowed us to estimate herd sensitivity and the total cost of each protocol by herd type. The protocols could be ordered by decreasing herd sensitivity and cost, regardless of the herd type, as follows: strict protocol, compliant quick-path protocol, and compliant slow-path protocol. We calculated a cost-effectiveness index to evaluate the cost-effectiveness of each protocol. The strict protocol was never the most cost-effective, regardless of herd type, due to higher costs relative to the other protocols, despite better herd sensitivity. We found the compliant quick-path to be the most cost-effective protocol for big beef, big dairy, and mixed herds. The compliant slow-path was the most cost-effective for small-scale beef and dairy herds. All differences were significant. This comparison of the cost-effectiveness of the protocols by herd type could help authorities to choose the most suitable protocol in the investigation of suspected cases, depending on the herd type, but could be improved by accounting for important sociological data, such as the acceptability of the protocols.

The singular insemination status of Hypothenemus hampei (Ferrari) (Coleoptera: Curculionidae: Scolytinae) females during the inter-harvest season of a coffee crop.

The coffee berry borer, Hypothenemus hampei (Ferrari), can survive in residual coffee berries during the inter-harvest period, while new fructification only appears 2-3 months after the last harvest. The dispersal of colonizing females is an adaptation that enables the life cycle of the species to go ahead whenever his flight aptitude allows. This paper focuses on accurately determining the rate of inseminated females ready to reproduce when emerging from residuals berries to colonize new ones, which constitutes a characteristic of the live cycle far from common in Curculionidae. We dissected females caught in traps baited with a mixture of alcohols during the inter-harvest season, females from infested residual berries collected from branches, and virgin females obtained from pupae reared individually in the laboratory. After microscopic preparation with Giemsa stain, spermathecae were observed to identify the physiological status of each specimen. Out of the females found in the traps, 98.4% displayed recent and abundant insemination and 1.6% sporadic insemination. In contrast, in residual berries, most of females were recently inseminated (84.5%), followed by virgin females (10.5%) and older inseminated females (5%). In addition, the flight tests of the virgin females were negative. These results indicate that all colonizing females were inseminated, ready for flying and oviposition, females inside residual berries showed different physiological status, and virgin females could not migrate since they could not flight. The large number of inseminated females inside the residual berries, and the capacity of migrating females to colonize and reproduce, suggest that it is necessary to control residual berries and use traps to stop the dispersal and reproduction of this pest.

Short-distance dispersal of Hypothenemus hampei (Ferrari) females (Coleoptera: Curculionidae: Scolytidae) during the coffee tree fruiting period.

The coffee berry borer (CBB), Hypothenemus hampei (Ferrari), is a multivoltine species closely associated with coffee crops worldwide, causing severe damage to the bean. In Mexico, as in all tropical regions, CBB survives during the inter-harvest period in residual berries on the ground or in dry berries remaining on the branches, and then disperses in search of the first suitable berries. In this study, we investigated how CBB dispersed from the first infested nodes during the fruiting period of Coffea canephora Pierre, which provides a favourable trophic level for this insect. Forty-five branches equally distributed in 15 coffee trees, with one infested node and four uninfested nodes, were selected. The branches were subjected to three treatments over nine weeks: 1) glue between nodes with full protection, 2) glue between nodes without protection, and 3) no glue and no protection. In addition, 45 CBB-free branches were selected and subjected to the same three treatments. CBB colonization can occur in three ways: 1) from an infested node to an uninfested node on the same branch, 2) from infested berries to uninfested berries within the nodes, 3) from branches to other branches. We also found that CBB dispersal between nodes of the same branches never occurred by walking but by flying. Thus, in this context of coffee berry development and ripening, and unlike the phenological situation of the inter-harvest period, CBB continuously travels very short distances, thus limiting its control.

Estimating the incidence of equine viral arteritis and the sensitivity of its surveillance in the French breeding stock.

Equine viral arteritis (EVA) may have serious economic impact on the equine industry. For this reason, it is monitored in many countries, especially in breeding stock, to avoid its spread during breeding activities. In France, surveillance is mainly based on serological tests, since mares are not vaccinated, but difficulties in interpreting certain series of results may impair the estimation of the number of outbreaks. In this study, we propose specific rules for identifying seroconversion in order to estimate the number of outbreaks that were detected by the breeding stock surveillance component (BSSC) in France between 2006 and 2013. A consensus among multidisciplinary experts was reached to consider seroconversion as a change in antibody titer from negative to at least 32, or as an eight-fold or greater increase in antibody level. Using these rules, 239 cases and 177 outbreaks were identified. Subsequently, we calculated the BSSC's sensitivity as the ratio of the number of detected outbreaks to the total number of outbreaks that occurred in breeding stock (including unreported outbreaks) estimated using a capture-recapture model. The total number of outbreaks was estimated at 215 (95% credible interval 195-249) and the surveillance sensitivity at 82% (CrI95% 71-91). Our results confirm EVA circulation in French breeding stock, show that neutralizing antibodies can persist up to eight years in naturally infected mares and suggest that certain mares have been reinfected. This study shows that the sensitivity of the BSSC is relatively high and supports its relevance to prevent the disease spreading through mating.

Comparative evaluation of three surveillance systems for infectious equine diseases in France and implications for future synergies.

It is necessary to assess surveillance systems for infectious animal diseases to ensure they meet their objectives and provide high-quality health information. Each system is generally dedicated to one disease and often comprises various components. In many animal industries, several surveillance systems are implemented separately even if they are based on similar components. This lack of synergy may prevent optimal surveillance. The purpose of this study was to assess several surveillance systems within the same industry using the semi-quantitative OASIS method and to compare the results of the assessments in order to propose improvements, including future synergies. We have focused on the surveillance of three major equine diseases in France. We have identified the mutual and specific strengths and weaknesses of each surveillance system. Furthermore, the comparative assessment has highlighted many possible synergies that could improve the effectiveness and efficiency of surveillance as a whole, including the implementation of new joint tools or the pooling of existing teams, tools or skills. Our approach is an original application of the OASIS method, which requires minimal financial resources and is not very time-consuming. Such a comparative evaluation could conceivably be applied to other surveillance systems, other industries and other countries. This approach would be especially relevant to enhance the efficiency of surveillance activities when resources are limited.

Assessment of the sensitivity of the gamma-interferon test and the single intradermal comparative cervical test for the diagnosis of bovine tuberculosis under field conditions.

In some French départements, the eradication of bovine tuberculosis is incomplete and usual skin tests [single intradermal tuberculin test (SIT) and single intradermal comparative cervical test (SICCT)] have poor specificity due to cross-reactions with non-pathogenic mycobacteria, causing economic losses. In Côte d'Or (Burgundy, France), an experimental serial testing scheme based on the combination of SICCT and gamma-interferon (IFN-γ) tests has been initiated in order to shorten the interval between suspicion and its invalidation in herds with false-positive results to skin tests. Our aim was to assess the scheme's sensitivity and to compare it to the sensitivity of the screening scheme recommended by the European Commission. Our study included 1768 animals from Côte d'Or. The sensitivities of both schemes were estimated using a Bayesian approach. The individual sensitivity of the IFN-γ test [88·1%, 95% credibility interval (CrI) 72·8-97·5] was not significantly different from individual SICCT sensitivity (80·3%, 95% CrI 61·6-98·0) and individual SIT sensitivity (84·2%, 95% CrI 59·0-98·2). The individual specificity of the IFN-γ test was 62·3% (95% CrI 60·2-64·5). No significant difference could be demonstrated between the sensitivities of the serial testing scheme used in Côte d'Or (73·1%, 95% CrI 41·1-100) and the European Union serial testing scheme (70·1%, 95% CrI 31·5-100·0).

Bovine tuberculosis surveillance in cattle and free-ranging wildlife in EU Member States in 2013: a survey-based review.

Bovine tuberculosis (TB) is a common disease in cattle and wildlife, with animal health, zoonotic and economic impacts. Most of the TB data for the European Union (EU) concern the epidemiological situation, but comprehensive descriptions of the way in which surveillance is conducted in each country are rare, despite being essential for cross-Europe comparisons. A European survey was conducted in the 28 Member States and in three other neighboring countries (Norway, Macedonia and Switzerland), to review TB surveillance in cattle and wildlife. EU legislation currently requires TB surveillance solely in cattle. Considerable differences between the surveillance systems of the 26 responding countries were observed, according to the official TB-freedom status of the country and the local prevalence of TB. These differences related principally to the combination of surveillance components (routine screening test in herd and/or movement testing and/or slaughterhouse surveillance), the tests used and their interpretation, and the definition of an infected herd or animal. For wildlife TB surveillance, only 8 on 21 respondent countries have declared to have implemented passive and/or active surveillance, with marked differences concerning the species and the geographical scale of the surveillance. The choice of the combination of surveillance components depends on the national or regional epidemiological situation, the species involved in TB epidemiology and epidemiological risk factors, although various surveillance systems have been recorded for countries with similar epidemiological status. Assessments of the cost-effectiveness of each surveillance system would be useful, to confirm the advantages of implementing one or more components.

Structural vulnerability of the French swine industry trade network to the spread of infectious diseases.

The networks generated by live animal movements are the principal vector for the propagation of infectious agents between farms, and their topology strongly affects how fast a disease may spread. The structural characteristics of networks may thus provide indicators of network vulnerability to the spread of infectious disease. This study applied social network analysis methods to describe the French swine trade network. Initial analysis involved calculating several parameters to characterize networks and then identifying high-risk subgroups of holdings for different time scales. Holding-specific centrality measurements ('degree', 'betweenness' and 'ingoing infection chain'), which summarize the place and the role of holdings in the network, were compared according to the production type. In addition, network components and communities, areas where connectedness is particularly high and could influence the speed and the extent of a disease, were identified and analysed. Dealer holdings stood out because of their high centrality values suggesting that these holdings may control the flow of animals in part of the network. Herds with growing units had higher values for degree and betweenness centrality, representing central positions for both spreading and receiving disease, whereas herds with finishing units had higher values for in-degree and ingoing infection chain centrality values and appeared more vulnerable with many contacts through live animal movements and thus at potentially higher risk for introduction of contagious diseases. This reflects the dynamics of the swine trade with downward movements along the production chain. But, the significant heterogeneity of farms with several production units did not reveal any particular type of production for targeting disease surveillance or control. Besides, no giant strong connected component was observed, the network being rather organized according to communities of small or medium size (<20% of network size). Because of this fragmentation, the swine trade network appeared less structurally vulnerable than ruminant trade networks. This fragmentation is explained by the hierarchical structure, which thus limits the structural vulnerability of the global trade network. However, inside communities, the hierarchical structure of the swine production system would favour the spread of an infectious agent (especially if introduced in breeding herds).

Glomerular filtration rate estimated by Cockcroft-Gault formula better predicts anti-Xa levels than modification of the diet in renal disease equation in older patients with prophylactic enoxaparin.

OBJECTIVES: Older people have an increased risk of low molecular weight heparin accumulation leading to an increased bleeding risk. The objective of this study was to assess whether reduced glomerular filtration rate (GFR), estimated by the Cockcroft-Gault or modification of the diet in renal disease (MDRD) equations, indicates drug accumulation by increased anti-Xa levels in older subjects receiving prophylactic enoxaparin treatment. DESIGN: Cohort study. SETTING: Acute geriatric units in Nancy Hospital. PARTICIPANTS: Ninety-two consenting consecutive patients, 65 and older, confined to bed for an acute medical condition requiring enoxaparin for prevention of venous thromboembolism, and hospitalized for at least six days were enrolled. MEASUREMENTS: Serum creatinine and peak plasma anti-Xa levels 3 to 4 hours after the daily injection of enoxaparin were measured at days 3, 6, 9 and 12 (first dose of enoxaparin at day one). Analyses of variance for repeated measures were used to evaluate significant predictors of peak anti-Xa activity in univariate and multivariate analyses. RESULTS: A significant correlation was observed between anti-Xa activity and GFR estimated with the Cockcroft formula r=0.43. Following univariate analysis, the three factors associated with higher anti-Xa levels were a lower Cockcroft-Gault GFR (p=0.0002), female gender (p=0.0003) and a lower bodyweight (p<.0001). No significant association between anti-Xa levels and MDRD GFR (p=0.33) was observed. Following multivariate analysis, female gender (p=0.02), bodyweight (p=0.04) and Cockcroft GFR (p=0.05) remained independent determinants of anti-Xa levels. CONCLUSION: In hospitalized patients older than 65 years old, the Cockcroft-Gault equation, in contrast to the MDRD equation, is able to predict the risk of higher anti-Xa levels.

Network of contacts between cattle herds in a French area affected by bovine tuberculosis in 2010.

France attained 'Officially Tuberculosis-Free' status in 2000. However, the Côte d'Or department (a French administrative unit) has since seen an increase in bovine tuberculosis (bTB) cases, with 35% of cases attributed to neighbourhood contamination. The aim of this study was to investigate the characteristics of neighbourhood contacts in an area affected by bTB in 2010, through the use of social network methods. We carried out a survey to determine the frequency and distribution of between-herd contacts in an area containing 22 farms. Contacts were weighted, as not all types of contact carried the same risk of bTB transmission. Cattle movement was considered to be associated with the highest risk, but was not observed within the studied area during the study period. Contact with wild boars was the most frequent type of contact, but was associated with a very low risk. Direct cattle-to-cattle contacts in pasture and contacts with badger latrines were less frequent, but entailed a greater risk of M. bovis transmission. Centrality values were heterogeneous in these two networks. This would enable the disease to spread more rapidly at the start of epidemics than in a perfect randomly mixed population. However, this situation should also result in the total number of infected herds being smaller. We attributed 95% of the contacts to direct contact in pasture or contact with wild boars or badger latrines. Other kinds of contact occurred less frequently (equipment sharing, cattle straying) or did not occur at all (attendance at a show). Most of the contact types were correlated, but none was sufficient in itself to account for all contacts between one particular farm and its neighbours. Contacts with neighbours therefore represent a challenge for the implementation or improvement of control measures.

Structuring the passive surveillance network improves epizootic detection and control efficacy: a simulation study on foot-and-mouth disease in France.

Rapid detection of infection is critical to the containment and control of contagious pathogens. Passive surveillance, based on the detection of clinical signs through farmers' observations and subsequent veterinarian notification, is the primary means of initially detecting an epizootic and for implementing control measures. The objective of this study was to analyse how the composition and structure of passive surveillance networks may impact epizootic spread and control. Three compositions of passive surveillance network were considered: (i) A veterinarian-based surveillance network composed of farmers and veterinarians (the common passive surveillance network where each veterinarian follows up a group of holdings), (ii) a farmer-based surveillance network composed of farmers only (the farmer plays the same role as in the preceding network as well as that of the veterinarian but his point of view is limited to his animals) and (iii) a hierarchical surveillance network composed of farmers, veterinarians and district-level veterinarian specialists (in case of doubt, the local veterinarian calls the specialist veterinarian). We compared the efficacy of these different network types where actors have successively a structurally wider perspective than the preceding ones using a specific stochastic model for the spread of foot-and-mouth disease (FMD). The model was forced by actual data to generate realistic simulated FMD epizootics in France. Our results show that maintaining the presence of field veterinarians following-up several holdings in breeding areas is fundamental and adding veterinarian specialists to passive surveillance networks could greatly enhance surveillance network efficacy.