SEARCH

SEARCH BY CITATION

Keywords:

  • H5N1 HPAI virus;
  • transmission rate;
  • reproductive number;
  • Nigeria

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We quantified the between-village transmission rate, β (the rate of transmission of H5N1 HPAI virus per effective contact), and the reproductive number, Re (the average number of outbreaks caused by one infectious village during its entire infectious period), of H5N1 highly pathogenic avian influenza (HPAI) virus in Nigeria using outbreak data collected between December 2005 and July 2008. We classified the outbreaks into two phases to assess the effectiveness of the control measures implemented. Phase 1 (December 2005–October 2006) represents the period when the Federal Government of Nigeria managed the HPAI surveillance and response measures, while Phase 2 (November 2006–July 2008) represents the time during which the Nigeria Avian Influenza Control and Human Pandemic Preparedness project (NAICP), funded by a World Bank credit of US$ 50 million, had taken over the management of most of the interventions. We used a total of 204 outbreaks from 176 villages that occurred in 78 local government areas of 25 states. The compartmental susceptible-infectious model was used as the analytical tool. Means and 95% percentile confidence intervals were obtained using bootstrapping techniques. The overall mean β (assuming a duration of infectiousness, T, of 12 days) was 0.07/day (95% percentile confidence interval: 0.06–0.09). The first and second phases of the epidemic had comparable β estimates of 0.06/day (0.04–0.09) and 0.08/day (0.06–0.10), respectively. The Re of the virus associated with these β and T estimates was 0.9 (0.7–1.1); the first and second phases of the epidemic had Re of 0.84 (0.5–1.2) and 0.9 (0.6–1.2), respectively. We conclude that the intervention measures implemented in the second phase of the epidemic had comparable effects to those implemented during the first phase and that the Re of the epidemic was low, indicating that the Nigeria H5N1 HPAI epidemic was unstable.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

An outbreak of H5N1 highly pathogenic avian influenza (HPAI) occurred in Nigeria between December 2005 and July 2008. It affected a total of 25 states including the Federal Capital Territory (Henning et al., 2012). The disease mostly affected backyard commercial layer farms and a few free-range poultry. Overall, approximately 711 000 birds were reported to have died from the disease, and a further 1 264 191 were culled in the course of implementing the emergency control measures (Fasina et al., 2009). A phylogenetic analysis of the H5N1 HPAI viruses isolated during the 2006–2007 outbreak indicated that the viruses clustered in distinct sublineages designated I, II and IV (Ducatez et al., 2007; Monne et al., 2008). Analysis of virus isolated from a duck sampled at a live-bird market in the city of Gombe, Gombe state classified the virus under sublineage III similar to the H5N1 HPAI viruses isolated in Europe and Middle East in 2007 (Fusaro et al., 2009). These results indicate that the H5N1 virus was introduced into the country repeatedly. Ducatez et al. (2006) and Brandenburg (2008) suggest that the virus could have been introduced into the country via (i) migratory birds and (or) (ii) poultry trade linking Nigeria with international poultry markets.

The Federal Government of Nigeria (FGN), with support from international partners implemented numerous emergency response measures that included culling of the sick and in-contact poultry in the affected areas, enforcement of movement restrictions, banning importation of poultry and poultry products and enhancing surveillance using the 170 surveillance points, which had been established under the earlier Pan African Programme for the Control of Epizootics (PACE) throughout the country (UNICEF, 2007). In addition, the national HPAI contingency plan was updated with time. The FGN managed most of the emergency responses until the Nigeria Avian Influenza Control Project was founded in April 2006 to coordinate the HPAI interventions. The FGN received a loan of US$ 50 million from the World Bank in September 2006 to strengthen its response capacity, specifically to prevent further spread of the disease to unaffected areas (Perry et al., 2011). Nigeria Avian Influenza Control Project's ability to manage the epidemic was boosted after this loan was received.

In this study, which was part of an impact assessment of the NAICP interventions (Perry et al., 2011), we quantify the transmission rate of the H5N1 HPAI virus between villages using data from the whole outbreak period and use it to assess the effectiveness of the HPAI control measures implemented at the various stages of the epidemic. We also estimate the reproductive number of the disease as the product of the transmission rate and the duration of infectiousness. The analytical methods we have used have been described and applied previously by Mannelli et al. (2007) to study the effectiveness of HPAI measures taken in industrial poultry in Northern Italy, by Stegeman et al. (2004) to assess the effectiveness of HPAI control measures in the Netherlands and by Stegeman et al. (1999) to study the effectiveness of classical swine fever control measures in the Netherlands.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

Sources of the data used in the study

We obtained the outbreak data from avian influenza desk officers who managed HPAI control activities under the states’ Departments of Veterinary Services in Nigeria. They supervised surveillance, field response and diagnostic teams during the HPAI outbreaks (Perry et al., 2011; Henning et al., 2012). A request was sent to all the officers in each state to share outbreak data collected between December 2005 and August 2008. These data indicated dates when (i) each outbreak was noticed and reported by farmers, (ii) samples were shipped to the laboratory, (iii) laboratory results were received and (iv) depopulation was implemented. The data also named the states, local government areas (LGA) and villages affected. From the farmers’ perspective, an HPAI outbreak comprised a peracute to acute disease of poultry that resulted in a high mortality and spreads rapidly between households in a village.

Confirmatory diagnoses were carried out at the National Veterinary Research Institute (NVRI), Vom, using reverse transcription-polymerase chain reaction (PCR) on tissue samples. Tissue samples used included tracheas, lungs, livers, spleens, brains, hearts, intestines as well as intestinal contents. The samples were collected aseptically during post-mortem examinations of freshly killed, moribund or dead birds acquired from the outbreak sites. Other tests such as virus isolation, haemagglutination inhibition and AGID tests were occasionally conducted to reduce the costs of running PCR tests. The NVRI laboratory periodically sent positive samples to the FAO/OIE Reference Laboratory at Istituto Zooprofilattico Sperimentale delle Venezie (IZSVe), Padua, Italy, for verification. Between 2006 and 2008, a total of 80 samples were sent there (Perry et al., 2011).

Analytical model and assumptions

A simple compartmental Susceptible-Infectious (SI) model was used to describe the transmission of H5N1 HPAI between villages, assuming that all newly infected villages (C) were infected through an indirect contact with infectious villages (I) during the same wave of the epidemic. Contact between villages could have occurred via purchase or movement of infected birds or transfer of infectious material (e.g. faeces) via fomites from infected to uninfected village (FAO, 2008). Representing the number of susceptible villages per day as S, the number of infectious villages per day as I and the total number of villages as N, the number of C is given by inline image, where β is the transmission rate parameter representing the rate of transmission of H5N1 HPAI virus per effective contact.

The parameter β is therefore given by inline image, and the reproductive number Re is calculated as the product of β and T, the duration of infectiousness (Ward et al., 2009).

This model ignores spatial transmission dynamics as it assumes that villages had similar epidemiological characteristics (defined by the same HPAI risk factors, contact patterns and recovery rates, etc.) and were equally at risk of infection at the start of the epidemic. We used MS Excel (Microsoft Corporation, 2010) to estimate β for each day (t) of the epidemic. The daily β estimates were analysed using bootstrapping techniques as described later in the 'Analysis of data'; Table 1 summarises the parameters used in, or produced by, this analysis.

Table 1. Definition of the input parameters used and output parameters obtained from the simple compartmental susceptible-infectious model used to analyse the rates of transmission of H5N1 HPAI in Nigeria (2005–2008)
ParameterDefinitionValue/Range
  1. a

    The minimum and maximum mean estimates of T derived from the sensitivity analyses to show the full range of T.

  2. b

    This assumes that the duration of infectiousness, T is 12 days.

Input
NNumber of villages in Nigeria48 000–90 000
CNumber of newly infected villages/dayVaries/day
SNumber of susceptible villages/dayVaries/day
TDuration of infectiousnessa7.58–17.24
Output
βTransmission rateb0.07 (0.06–0.09)
ReReproductive numberb0.9 (0.7–1.1)

Parameter estimation

The total number of villages (N) and the number of susceptible villages per day, S

The exact number of villages in Nigeria at the time of the epidemic could not be determined. However, the National Geospatial-Intelligence Agency (2012) http://geonames.nga.mil/ggmagaz/ suggests that there more than 48 000 populated places (assumed to represent villages) in Nigeria, while a report by the Nigeria Guinea Worm Project indicates that there were 90 000 villages in the country in 1988–1989 (CDC, 1991). We used these estimates to define a range of values that were used in a sensitivity analysis assuming that the true population of villages in Nigeria was bounded by this range (48 000–90 000).

The number of villages susceptible at the start of each day decreased as new infections occurred. The number of villages susceptible at the start of each day was therefore derived by subtracting the number that became infected on day i from the number that was susceptible at the start of day i.

Number of newly infected (C) and infectious villages (I) per day

A village was classified as being newly infected on day i if an outbreak commenced in that village on that day. On the basis of this classification, an aggregate number of newly infected villages (C) were calculated for each day of the epidemic, and its frequency distribution illustrated graphically. Repeat outbreaks within a village were collapsed into single outbreaks if they occurred within a period of 21 days of preceding ones (i.e. based on the difference between the date of depopulation of a preceding outbreak and the date when the current outbreak is noticed). This interval was based on the fact that the incubation period of HPAI ranges between 2 and 14 days, and in rare cases, it can be as high as 21 days (Henning et al., 2009). For all these analyses, the duration of infectiousness was assumed to range from the date when an outbreak was noticed to the date when depopulation was carried out. The analysis was conducted after sorting outbreak data by village name and date when an outbreak was noticed by a farmer.

To assess whether the duration of the outbreak interval had an effect on β and Re, a sensitivity analysis was conducted assuming that a repeat outbreak occurred after a period of 4, 10 and 21 days from the previous one.

The number of infectious villages (I) in a given day (day i) comprised villages that had had an infection before day i and had not been depopulated by the end of that day.

Mean duration of infectiousness (T) by village

Three estimates of the duration of infectiousness (T) were derived and used in a sensitivity analysis to assess the effect of T on β and Re. The first was the period between reporting an outbreak to the veterinary authorities and depopulation, the second was the time between noticing an outbreak and depopulation, while the third assumed that the infectious period started 4 days before an outbreak was noticed and continued until depopulation was completed. The 4-day period is expected to cover the initial period of infectiousness when birds infected with H5N1 HPAI virus would be shedding the virus but not manifesting any clinical signs. A similar approach was also used by Stegeman et al. (2004). For cases that had multiple dates of reporting and (or) depopulation, the earliest date of reporting and the latest date of depopulation were always used.

Phases of the H5N1 HPAI epidemic

On the basis of the data collected by NADIS, Nigeria experienced three outbreak waves of H5N1 HPAI that occurred successively between December 2005 and October 2006, November 2006 and November 2007 and in July 2008 (Ahmed, 2008; Henning et al., 2012).We considered this timeline while assessing the effectiveness of the measures implemented through the NAICP and therefore classified the epidemic into two phases including:

  1. Phase 1 – the period between December 2005 through October 2006 when the FGN and NAICP could implement limited range of intervention because of budgetary constraints;
  2. Phase 2 – the period between November 2006 and July 2008 when NAICP had received financial support from the World Bank to improve its capacity. This period excludes a 2-month lag period (September–October 2006) to allow for the time when NAICP was developing its infrastructure.
Phase 1

Emergency intervention measures implemented at the start of Phase 1 included a ban on the importation of poultry and poultry products from HPAI-infected countries, implementation of an effective HPAI surveillance system, improvement of the functionality of the quarantine services and commencement of community-based training involving backyard farmers focussing on ways of identifying HPAI. It also began refining the contingency plan. Depopulation was also implemented during this period, but compensation was provided at a fixed rate of N250/bird (USD $1.6/bird). Towards the end of the phase, the NAICP commenced working on an integrated response plan – this was completed during the second phase. The project also started developing an administrative infrastructure, which was used more effectively in the second phase of the epidemic.

Phase 2

In November – December 2006, NAICP recruited avian influenza desk officers in each state, trained surveillance and response officers and set up a policy for compensation based on the market value of the culled poultry. This compensation policy was implemented as from January 2007. Other activities that the NAICP undertook as from February 2007 included the training of farmers and fowl sellers on biosecurity practices, the registration of farms in a bid to encourage disease reporting and financing the development and airing of public awareness jingles about HPAI through the mass media.

Analysis of data

Means and 95% percentile confidence intervals for β, T and Re were estimated using bootstrapping method for all the scenarios used for sensitivity analysis described above (Efron and Tibshirani, 1993); GenStat (GenStat® Release 14, 2011). First, 1000 new data were sampled from the existing β and T data with replacement. Re was then derived by multiplying the simulated β and T arrays. The resulting distributions of the estimated mean values were used to evaluate overall means and respective 95% percentile confidence intervals. Means that had overlapping percentile intervals were considered as not being significantly different.

Bootstrap samples were first generated for an entire outbreak period. This was followed by phase-specific resampling to permit bootstrapping of the data to take place independently within groups (phases) that were being compared.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The number of new and infectious cases over time

We considered a total of 204 outbreaks from 176 villages in 78 LGAs in 25 states. A few records that did not carry dates of when depopulation occurred were excluded from the analysis. Figure 1 shows the distribution of the new outbreaks on each month of the epidemic. The graph also demarcates the epidemic into the two phases identified above. The number of new outbreaks peaked in February during the first phase of the epidemic and in January–February during the second phase. The first phase covered a shorter time period of 322 days and had a total of 103 outbreaks, while the second covered a longer duration of 644 days and had 101 outbreaks. The second phase has a longer duration because it includes the 3rd wave of the outbreak that occurred much latter in 2008 (Fig. 1).

image

Figure 1. Monthly distribution of newly infected villages, C, based on the dates when the outbreaks were noted or reported by farmers during the first and the second phases of the H5N1 HPAI epidemic in Nigeria in 2005–2008.

Download figure to PowerPoint

The duration of infectiousness, T, transmission rate, β and reproductive number, Re

Figure 2 shows the distribution of the number of days between the dates when outbreaks were assumed to have commenced to the dates when they were assumed to have ended. These periods represent alternative assumptions for the duration of infectiousness of H5N1 HPAI virus in Nigeria. Figure 3 gives the mean values of T, β and Re; 95% percentile intervals obtained from the bootstrapping technique are given below each figure.

image

Figure 2. Distribution of the number of days between (i) reporting an outbreak to depopulation, (ii) noticing an outbreak to depopulation and (iii) 4 days before noticing an outbreak to depopulation; these periods represent alternative assumptions for the duration of infectiousness of the H5N1 HPAI virus during the 2005–2008 outbreak in Nigeria.

Download figure to PowerPoint

image

Figure 3. Analysing the effect of varying the duration of the interval between repeat outbreaks of H5N1 HPAI observed in Nigeria in 2005–2008 (4, 10 and 21 days) on the transmission rate (β), the duration of infectiousness (T) and reproductive number, Re. Outbreaks are defined at the village level.

Download figure to PowerPoint

The results in Fig. 3 indicate that an increase in the mean interval between outbreaks, from 4 to 21 days, leads to an insignificant reduction in the mean value of β estimates (by about 2% points) and marginal increase in the T. These changes do not affect Re. The increase in the duration of infectiousness is modest because the percentage of repeat outbreaks, which form the basis of this sensitivity analysis, is only 28.9% (n = 66), 22.3% (n = 47) or 13.8% (n = 28) when the outbreak interval is 4, 10 or 21 days, respectively.

Figure 4 shows that doubling the duration of T (comparing Figs 4a and c whose mean T values were 8 and 16, respectively) leads to a significant reduction in β and a marginal increase in Re. Generally, phase 2 had higher β estimates compared to phase 1, and in fact, this difference tends to become significant at high values of T (Fig. 4c).

image

Figure 4. Analysing the effect of varying the duration of infectiousness of H5N1 HPAI virus based on the scenarios presented in Fig. 2 on the transmission rate (β) and reproductive number (Re) of the virus during the 2005–2008 outbreak in Nigeria.

Download figure to PowerPoint

The transmission parameters estimated were not responsive to the variation in the number of villages (N) considered. For example, β estimates obtained by assuming village population sizes of 50 000 and 90 000 were 0.0719 (0.0620–0.0820) and 0.0719 (0.0619–0.081), respectively. Re estimates remained constant at 0.88.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

This is the first study to evaluate transmission parameters of the HPAI epidemic, which occurred in Nigeria between December 2005 and July 2008. We used a compartmental SI model (Stegeman et al., 1999, 2004; Ward et al., 2009; Penny et al., 2010) with various simplifying assumptions given the limited understanding of the dynamics that supported the disease transmission in the country, as well as the lack of data that could be used to build more informative models. One of these is the assumption that villages were equally susceptible and could randomly come into contact, for example through the movement of live birds and other farm-bridge species, service trucks, people and fomites. The clustered nature of the outbreaks as illustrated by Ikpi et al. (2010), however, suggests that not all villages had equal levels of HPAI risk. Henning et al. (2012) also highlighted in a spatial analysis of the outbreak patterns that only very few LGAs within each state and only few villages within affected LGAs experienced HPAI outbreaks.

We could not obtain information on the number of villages in Nigeria at the time of the epidemic. Nonetheless, the sensitivity analysis indicated that both β and Re are insensitive to the number of susceptible villages used in the analysis. This is because the number of infectious villages at any one time is much smaller compared to the susceptible ones (Ward et al., 2009).

We assumed that all outbreaks were detected and recorded. However, underreporting of cases often occurs during H5N1 HPAI epidemics because of a number of factors including lack of sensitive and specific diagnostic tests. Few studies have assessed the impact of underreporting on disease transmission parameters; White and Pagano (2010) indicate that if the fraction of cases reported does not change dramatically throughout the course of an epidemic, then epidemiological parameters (such as Re) estimated from such data are not substantially impacted. It is unlikely that the fraction of cases detected by the surveillance system employed in Nigeria substantially varied over time because throughout the epidemic period, farmers freely volunteered information on the outbreaks to qualify for compensation.

Generally, β and Re estimates were more or less similar between the two phases of the epidemic. This indicates that the effectiveness of the measures implemented by NAICP was comparable to those of the measures that had been implemented in phase 1 (i.e. before the project took over the coordination of the H5N1 HPAI control efforts). This is in accordance with an earlier conclusion that the speed of reporting and depopulation did not improve between the 2006 and 2007 epidemics (Henning et al., 2012). Ikpi et al. (2010) have highlighted some of the challenges Nigeria faced while managing the outbreaks. These include (i) the low level of compliance by live-bird transporters and traders with movement control measures, (ii) the poor capacity of the government to enforce movement control measures, (iii) poor funding at the state and LGA level to support the activities of the field workers and desk officers and (iv) insufficient veterinary capacity in all the states to cope with the demand for active disease surveillance and control. Furthermore, Fasina et al. (2009) identified practices that could have promoted the spread of the disease between poultry farms as poor biosecurity standards, trade in live poultry, visitors gaining access to poultry premises and farm workers living outside poultry farms. In an earlier study, Fasina et al. (2009) speculated that the outbreak occurrence might have been associated with legal or illegal trade, use of live-bird markets, inappropriate disposal of infected carcasses and poor implementation of disease control measures.

Effective control of HPAI could potentially be achieved by well-structured depopulation in affected areas (Stegeman et al., 2004). This was effectively carried out in the Netherlands, Italy, Hong Kong, South Korea and Japan (Sims et al., 2003), but was not achieved in Southeast Asia and Egypt. This was probably associated with the underreporting of cases, poor detection, and challenges associated with the implementation of the measure among small-scale poultry producers (Pfeiffer et al., 2007; Sims, 2007). In Nigeria, the interventions introduced by the NAICP, for example, paying of compensation funds based on the market value of the culled birds, the registration of poultry farms (which was expected to incentivize disease reporting) and the training response teams, were all aimed at enhancing the effectiveness of depopulation, and other measures used to combat the epidemic. With respect to underreporting, Ikpi et al. (2010) indicated that most of the actors in poultry value chains, especially retailers, traders and backyard chicken producers, exhibited poor reporting practices, yet the FGN and NAICP had developed a well-structured compensation scheme that was expected to provide an incentive for case reporting.

Our analysis suggests that the Re of H5N1 HPAI epidemic in Nigeria ranged between 0.7 and 1.1 considering an infectious period of 12 days and an outbreak interval of 21 days. The low Re reported in this study is consistent with a fairly long period (December 2005–July 2008) over which the epidemic occurred in Nigeria (over >600 days in 25 states). Higher reproductive numbers have been associated with outbreaks that occur over a relatively shorter period. Ward et al. (2009) estimated a R0 of 1.68–2.95 in a H5N1 HPAI outbreak that occurred in Romania where 110 outbreaks occurred over 25 days. Marquetoux et al. (2012) estimated an R0 varying between 1.27 and 1.60 in Thailand using 1208 outbreaks that occurred between July 2004 and April 2005. We could not find relevant studies to use for comparative assessment from Africa even though a total of 11 African countries reported confirmed outbreaks of H5N1 HPAI in the period 2006–2008. Most of these countries however had transient outbreaks.

Anecdotal reports suggest that the epidemic could have burnt out because Nigeria has a low population of ducks and other water fowls that are considered as being the reservoirs for the virus (Leigh Perkins and Swayne, 2002; Henning et al., 2010). Waterfowls account for only about 10% of the total poultry population (of about 150 million birds) in the country (Akinwumi et al., 2011). The premise that the epidemic could have burnt out is supported by the information given above suggesting that the H5N1 HPAI virus was introduced to the country repeatedly and that Nigeria is the only country in Africa that had multiple sublineages of the virus (Ducatez et al., 2006; Fusaro et al., 2009). We hypothesize, therefore, that had the re-introductions of the virus not occurred, the duration of the epidemic in Nigeria might have taken a shorter time than observed. That period, however, could have been slightly longer than that reported for the other countries that were affected in the region such as Togo and Benin, whose H5N1 HPAI epidemics burnt out after about 1 year (http://web.oie.int/wahis/public.php?page=disease_timelines). This is because Nigeria has higher poultry population density than these other West African countries.

Conclusion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

The study quantified important epidemiological parameters β, T and Re, of the H5N1 HPAI epidemic in Nigeria and demonstrated that the effectiveness of the measures implemented by FGN between December 2005 and October 2006 was comparable to those that were implemented by the NAICP from November 2006. The results therefore suggest that the initial responses by the FGN reduced the incidence of the disease, and NAICP's interventions had to produce much higher impacts for such an assessment to show any significant effect. In general, this study demonstrated that the Reof the disease was very low, indicating that the Nigeria H5N1 HPAI epidemic would have been unlikely to remain endemic even with minimal interventions.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References

We are grateful to avian influenza desk officers of each state and officials of the NAICP for sharing data and information on HPAI.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Acknowledgements
  9. References
  • Ahmed, I. G., 2008: Principal activities and situation of avian influenza in Nigeria. Proceedings of a meeting of a steering committee of the sub-regional mechanism for the prevention and control of highly pathogenic avian influenza held at Abidjan on 19–24 September 2008, Epidemiology Unit, FDL, Nigeria.
  • Akinwumi, J., I. Okike, B. Bett, T. Randolph, and K. Rich, 2011: Analysis of the poultry value chain and its linkages and interactions with HPAI risk factors in Nigeria. Value Chain Study Report, Pro-poor HPAI Risk Reduction in Africa and Asia. ILRI/IFPRI, Nairobi.
  • Brandenburg, B., 2008: The animal health surveillance system and avian influenza in Nigeria. EDRS-AIA Consultancy Report, International Livestock Research Institute, Nairobi, Kenya. Available at http://www.docstoc.com/docs/20141055/The-Animal-Health-Surveillance-System-and-Avian-Influenza-in. (accessed March 23, 2011).
  • CDC, 1991: International notes update: Dracunculiasis eradication – Ghana and Nigeria, 1990. 40, 245247. Available at http://www.cdc.gov/mmwr/preview/mmwrhtml/00001960.htm (accessed March 23, 2011).
  • Ducatez, M. F., C. M. Olinger, A. A. Owoade, S. De Landtsheer, W. Ammerlaan, H. G. M. Niesters, A. D. M. E. Osterhaus, R. A. M. Fouchier, and C. P. Muller, 2006: Avian Flu: multiple introductions of H5N1 in Nigeria. Nature 442, 37.
  • Ducatez, M. F., C. M. Olinger, A. A. Owoade, Z. Tarnagda, M. C. Tahita, A. Sow, S. De Landtsheer, W. Ammerlaan, J. B. Ouedraogo, A. D. Osterhaus, R. A. Fouchier, and C. P. Muller, 2007: Molecular and antigenic evolution and geographical spread of H5N1 highly pathogenic avian influenza viruses in western Africa. J. Gen. Virol. 88, 22972306.
  • Efron, B., and R. J. Tibshirani, 1993: An Introduction to the Bootstrap. Chapman and Hall, London.
  • FAO, 2008: Biosecurity for highly pathogenic avian influenza – issues and options. Technical Paper Series ISSN 0254 – 6019. FAO, Rome. Available at http://www.oie.int/doc/ged/D5896.PDF (accessed July 19, 2012).
  • Fasina, F. O., S. P. Bisschop, T. M. Joannis, L. H. Lombin, and C. Abolnik, 2009: Molecular characterization and epidemiology of the highly pathogenic avian influenza H5N1 in Nigeria. Epidemiol. Infect. 137, 456463.
  • Fasina, F. O., A. L. Rivas, S. P. Bisschop, A. J. Stegeman, and J. A. Hernandez, 2011: Identification of risk factors associated with highly pathogenic avian influenza H5N1 virus infection in poultry farms in Nigeria during the epidemic of 2006 – 2007. Prev. Vet. Med. 98, 204208.
  • Fusaro, A., T. Joannis, I. Monne, A. Salviato, B. Yakubu, C. Meseko, T. Oladokun, S. Fassina, I. Capua, and G. Cattoli, 2009: Introduction into Nigeria of a distinct genotype of avian influenza virus (H5N1). Emerg. Infect. Dis. 15, 445447.
  • GenStat® Release 14, 2011: The Guide to GenStat® Release 14. Part 2: Statistics. VSN International, UK.
  • Henning, J., D. U. Pfeiffer, and L. Tri Vu, 2009: Risk factors and characteristics of H5N1 highly pathogenic avian influenza (HPAI) post-vaccination outbreaks. Vet. Res. 40, 15.
  • Henning, J., H. Wibawa, J. Morton, T. Bhakti Usman, A. Junaidi, and J. Meers, 2010: Scavenging ducks and transmission of highly pathogenic avian influenza, Java, Indonesia. Emerg. Infect. Dis. 16, 12441250.
  • Henning, J., B. Bett, I. Okike, P. Abdu, and B. Perry, 2012: Incidence of highly pathogenic avian influenza H5N1 in Nigeria, 2005–2008. Transbound Emerg. Dis. In press, published online: 25 April 2012. DOI: 10.1111/j.1865-1682.2012.01331.x.
  • Ikpi, A., I. Okike, B. Bett, and T. Randolph, 2010: Alignment of Poultry Sector Actors with Avian Influenza Control Measures in Nigeria. Draft Pro-poor Risk Reduction Project Report. International Livestock Research Institute, Nairobi.
  • Leigh Perkins, L. E., and D. E. Swayne, 2002: Pathogenicity of a Hong Kong-Origin H5N1 highly pathogenic avian influenza virus for emus, geese, ducks, and pigeons. Avian Dis. 46, 5363.
  • Mannelli, A., L. Busani, M. Toson, S. Bertolini, and S. Marangon, 2007: Transmission of highly pathogenic avian influenza (H7N1) among industrial poultry farms in northern Italy in 1999–2000. Prev. Vet. Med. 81, 318322.
  • Marquetoux, N., M. Paul, S. Wongnarkpet, C. Poolkhet, W. Thanapongtharm, F. Roger, C. Ducrot, and K. Chalvet-Monfray, 2012: Estimating spatial and temporal variations of the reproductive number for highly pathogenic avian influenza H5N1 epidemic in Thailand. Prev. Vet. Med. In press, doi:10.1016/j.prevetmed.2012.01.021.
  • Microsoft Corporation, 2010. Microsoft Excel® 2010. Available at http://www.microsoft.com (accessed August 16, 2012).
  • Monne, I., T. M. Joannis, A. Fusaro, P. De Benedictis, L. H. Lombin, H. Ularamu, A. Egbuji, P. Solomon, T. U. Obi, G. Cattoli, and I. Capua, 2008: Reassortant avian influenza virus (H5N1) in poultry, Nigeria, 2007. Emerg. Infect. Dis. 14, 637640.
  • Penny, M. A., J. Saurina, I. Keller, L. Jenni, H.-G. Bauer, W. Fiedler, and J. Zinsstag, 2010: Transmission dynamics of highly pathogenic avian influenza at Lake Constance (Europe) during the outbreak of winter 2005–2006. EcoHealth 7, 275282.
  • Perry, B., T. Randolph, B. Bett, D. Grace, A. Globig, M. Fadiga, K. Sones, J. Henning, P. Pali, J. Poole, H. Hannah, E. Ekanem, P. Abdu, and C. Molokwu, 2011: Independent impact assessment of the World Bank-funded Nigeria Avian Influenza Control and Human Pandemic Preparedness and Response Project (NAICP). Project report. Available at http://www.aicpnigeria.org/documents/NAICP%20Independent%20Evaluation%20_Main%20Report%20plus%20Annexes_21%20April%202011_1-1.pdf (accessed November 10, 2011).
  • Pfeiffer, D. U., P. Q. Minh, V. Martin, M. Epprecht, and M. J. Otte, 2007: An analysis of the spatial and temporal patterns of highly pathogenic avian influenza occurrence in Vietnam using national surveillance data. Vet. J. 174, 302309.
  • Sims, L. D., 2007: Lessons learned from Asian H5N1 outbreak control. Avian Dis. 51, 174181.
  • Sims, L. D., T. M. Ellis, K. K. Liu, K. Dyrting, H. Wong, M. Peiris, Y. Guan, and K. F. Shortridge, 2003: Avian influenza in Hong Kong 1997–2002. Avian Dis. 47, 832838.
  • Stegeman, A., A. R. W. Elbers, Smak. Jan, and M. C. de Jong, 1999: Quantification of the transmission of classical swine fever virus between herds during the 1997–1998 epidemic in The Netherlands. Prev. Vet. Med. 42, 219234.
  • Stegeman, A., A. Bouma, A. R. W. Elbers, M. C. M. de Jong, G. Nodelijk, F. de Klerk, G. Koch, and M. van Boven, 2004: Avian influenza A virus (H7N7) epidemic in the Netherlands in 2003: course of the epidemic and effectiveness of control measures. J. Infect. Dis. 190, 20882095.
  • UNICEF, 2007: Avian influenza information sheet. Available at http://www.unicef.org/wcaro/WCARO_Nigeria_Factsheets_AvianInfluenza.pdf (accessed March 23, 2011).
  • Ward, M. P., D. Maftei, C. Apostu, and A. Suru, 2009: Estimation of the basic reproductive number (Ro) for epidemic highly pathogenic avian influenza subtype H5N1 spread. Epidemiol. Infect. 137, 219226.
  • White, L. F., and M. Pagano, 2010: Reporting errors in infectious disease outbreaks, with an application to pandemic influenza A/H1N1. Epidemiol. Perspect. Innov. 7, 12.