DSpace at VNU: An Exploration of How Perceptions of the Risk of Avian Influenza in Poultry Relate to Urbanization in Vietnam

DSpace at VNU: An Exploration of How Perceptions of the Risk of Avian Influenza in Poultry Relate to Urbanization in Vie...

An Exploration of How Perceptions of the Risk of Avian

Influenza in Poultry Relate to Urbanization in Vietnam

Melissa L Finucane,1,2 Tuyen Nghiem,3Sumeet Saksena,1Lam Nguyen,4Jefferson Fox,1

James H Spencer,5and Trinh Dinh Thau4

1 East-West Center, 1601 East-West Rd, Honolulu, HI 96848

2 RAND Corporation, 4570 Fifth Ave, Pittsburgh, PA 15213

3 Vietnam National University, Hanoi, Vietnam

4 Hanoi University of Agriculture, Hanoi, Vietnam

5 Clemson University, Clemson, SC 29634

Abstract: This research examined how perceptions of outbreaks of highly pathogenic avian influenza (HPAI)

subtype H5N1 in poultry are related to urbanization Via in-depth interviews with village leaders, household

farmers, and large farm operators in modern, transitional, and traditional communes in the north of Vietnam,

we explored behaviors, attitudes, cultural values, and traditions that might amplify or attenuate HPAI

out-breaks We also explored conceptualizations of urbanization and its impacts on animal husbandry and disease

outbreaks Qualitative theme analyses identified the key impacts, factors related to HPAI outbreaks, and disease

prevention and management strategies The analyses also highlighted how urbanization improves some aspects

of life (e.g., food security, family wealth and health, more employment opportunities, and improved

infra-structure), but simultaneously poses significant challenges for poultry farming and disease management

Awareness of qualitative aspects of HPAI risk perceptions and behaviors and how they vary with urbanization

processes may help to improve the prevention and management of emerging infectious diseases

Keywords: risk perception, urbanization, avian influenza, HPAI, poultry, Vietnam

INTRODUCTION

Highly pathogenic avian influenza (HPAI) subtype H5N1

emerged in Southeast Asia in 2003, with repeated outbreaks

causing poultry and economic losses on a large scale

Anthropogenic and ecological studies suggest that the spread

of HPAI may be related to the ongoing process of replacing

traditional farming methods (e.g., multispecies livestock

husbandry) with industrial, mass-production-oriented

operations (e.g., Mallin and Cahoon2003); especially when the number of poultry houses and other connected estab-lishments increases without proper planning (Capua and Marangon 2000) Opportunities for disease transmission increase with industrial poultry production because of in-creases in livestock pools, intensive rearing of multiple spe-cies, and semi-vertical integration systems where, for instance feed trucks, abattoir delivery, and other vehicles, may visit several farms daily (Dent et al.2008) Gilbert et al (2008) demonstrated that a few key factors, such as human popu-lation density, rice cropping intensity, and poultry density, explain a large proportion of the spatial variation in HPAI

Correspondence to: Melissa L Finucane, e-mail: Melissa.Finucane@EastWestCenter.

Original Contribution

Ó 2014 International Association for Ecology and Health

disease risk The same study also highlighted that a

consid-erable variation remains unexplained and factors, such as

poultry production, marketing systems, or other dimensions

of human behavior, should be considered

Only a few studies have examined the perceived risk of

HPAI in poultry Studies in Thailand (Takeuchi2006), Laos

(Barennes et al.2007), and Cambodia (Ly et al.2007) suggest

that few people are aware of HPAI symptoms in poultry and

that many think it is unlikely to occur in their flocks Few

people report that they would notify authorities of poultry

deaths, despite believing that it is important to report the

deaths The most common reasons for non-reporting include

lack of knowledge about or inaccessibility of the reporting

procedure, fears of culling, and fears of being unable to sell the

remaining poultry The above studies were sampled from

areas reflecting different levels of urbanization but did not

report differences among urban, semi-urban, and rural

sub-samples, even though the level of urbanization may be related

to avian influenza outbreaks (Spencer2013) Thus, we do not

know (1) whether perceived risk varies with levels of

urban-ization and (2) how people conceptualize urbanurban-ization and its

impacts on animal husbandry and disease outbreaks

Vietnam provides an excellent site for this research because

as a result of its rapidly growing economy (5% GDP real growth

rate for 2012; Central Intelligence Agency 2013) it is

experi-encing rapid environmental transformations Urbanization in

Vietnam has substantially changed farmers’ interactions with

natural systems For example, transportation has become

motorized, increasing the potential for disease to spread further

and more easily from one farm or commune to another via

feces on moped or truck tires Understanding perceptions of

urbanization and its relationship with HPAI outbreaks in

poultry is important because urbanization and the resulting

infrastructure create a filter which distances people from the

environment and affects their ability to perceive changes in

natural resources (Lilian et al.2007) Failing to perceive change

potentially threatens a community’s ability to respond

appro-priately To fill the gap in knowledge about whether and how

HPAI outbreaks are related to urbanization, we conducted two

exploratory studies with village leaders, household farmers, and

large farm operators in the north of Vietnam

Research Design

A qualitative research design was used to allow participants

to describe beliefs and experiences about HPAI in poultry

and urbanization in their own words, rather than as a choice between the predetermined survey responses (Morgan 1998; O’Brien 1993; Pope and Mays 1995) We established sampling frames (see below) that included tra-ditional settings (customary agricultural practices, housing, and water sources), modern settings (agricultural mod-ernization, industrial cities, and modern water sources), and transitional settings (a mix of traditional and modern agricultural practices, housing, and water sources) The variables we used to identify these settings were reported previously as metrics of urbanization potentially associated with outbreaks of HPAI in poultry (e.g., Spencer 2013; Pfieffer et al.2007; Fang et al.2008; Gilbert et al.2008)

Participants

For both Studies 1 and 2, a research assistant from Hanoi University of Agriculture arranged permissions and scheduled people to attend the interviews Individuals were recruited using a purposive sampling technique with the goal of obtaining diverse samples regarding age, gender, and farming activities Participants were informed about the purpose of the study, the nature of the questions, and that they could refuse to answer questions or withdraw from the study at any time Each participant was paid 30,000–200,000 VND (approximately $1.44–$9.60 USD) to compensate for their time and travel expenses

The sample for Study 1 included 43 people recruited from four communes1in Bac Ninh Province, Vietnam The communes were selected as representative of different levels

of urbanization based on an index of water source coher-ence (Spcoher-encer2013) calculated from variables in the 1999 Vietnam census The index reflects the amount of ‘‘mixing’’

of different types of household-level water supply (well, rain, bottled, or surface water) Table1 shows the demo-graphics and other characteristics of the sample

To identify participants for Study 2, we created a sampling frame using the 2006 Vietnam agricultural census following the methods developed by Saksena et al (2013) All the communes were classified into three groups reflecting different levels of urbanization (traditional, transitional, or modern) using hierarchical cluster analysis with the following criteria: (1) percent of houses whose main income is from agricultural, fishing, or forestry activities and (2) percent of houses that have a modern

1 The smallest administrative unit in Vietnam is usually the commune, but in more urban areas they may be called a ward or town.

toilet system The resulting characteristics of these three groups are shown in Table2

The Study 2 sample included a total of 33 participants from communes at all the three levels of urbanization, but with an emphasis on transitional areas because of our interest in understanding perceptions of change Table3

shows the demographics and other characteristics of the sample

Materials and Procedures

For both studies, individual and small-group interviews followed a semi-structured protocol Study 1 participants were asked first to comment on their experience of avian influenza in poultry (e.g., ‘‘Tell me what you know about avian influenza in birds and how it affects you, your family,

Table 1 Study 1 Participant Characteristics (n = 43)

Commune, Province (urbanization level), n (%)

Thai Bao, Bac Ninh (modern) 14 (32.6)

Quang Pho Thanh, Bac Ninh (transitional) 15 (34.9)

Ninh Xa, Bac Ninh (transitional) 10 (23.3)

Dai Dong, Bac Ninh (traditional) 4 (9.3)

Participant type, n (%)

Gender, n (%)

Age, years, range (mean) 29–61 (46.57)

Ethnicity, n (%)

Education, n (%)

Vocational school or some college 11 (25.6)

Completed 4-year college or more 5 (11.6)

Primary profession, n (%)

Manager/Deputy Manager 3 (7.0)

Professional (e.g., lawyer, doctor, teacher) 1 (2.3)

Other (Head of Village

and Agricultural Extension Officer)

2 (4.7)

Annual household income,

VND, range (mean)

150,000–

300,000,000 (49,335,135) Housing, n (%)

Solid (concrete walled) 19 (44.2)

Semisolid (bamboo or concrete) 6 (14.0)

Wood frame with cottage roof

(bamboo, thatch)

11 (25.6) Temporary one-floor house 5 (11.6)

Two- or three-floor strong house 2 (4.7)

Type of latrine, n (%)

Household septic latrine 31 (72.1)

Pour-flush filtered latrine 2 (4.7)

Ventilated improved pit latrine 8 (18.6)

Table 1 continued

Main source of fresh water, n (%)

Purchased water (tank, jar, bottle, etc.) 0 (0)

River, lake, pond water 0 (0)

Animals farmed in the previous 12 months, (n participants raising animal),

livestock count range (mean) Chicken (n = 36) 5–10,000 (972.6)

Other bird (n = 3) 10–1,200 (28.6)

Fish (n = 13) 300–30,000 (1757.9) Other animals (e.g., dog, goose) (n = 4) 1–50 (2.6) Crops, n farming crop (percent)

Other cereals (e.g., corn) 18 (21.2)

and your community?’’) Then more specific probes asked

about their understanding of risk exposure processes, HPAI

effects, risk assessment and management, and Vietnamese

values and traditions relevant to animal husbandry and

disease management (e.g., ‘‘What role do spiritual beliefs

play in preventing or managing disease outbreaks among

your animals?’’) Study 2 participants were asked to

com-ment on their understanding of ‘‘rural’’ and ‘‘urban’’ (e.g.,

‘‘What does rural mean to you?’’) and how urbanization

affects: the built and natural landscapes, farming practices,

livelihoods, quality of life, waste management systems, and

HPAI in poultry

The interviews were facilitated jointly by the first

au-thor (asking questions in English) and the second auau-thor

(translating questions and responses) During individual

and small-group interviews, the facilitators and two

assis-tants recorded by hand the information about participants’

remarks and nonverbal responses Audio recordings were

not made because the topic was deemed too sensitive Each

interview took 1–2 h

For Study 1, nine small-group interviews (with 3–6

individuals per group) were held with 43 participants over

3 days (May 25–27, 2010) in meeting rooms at the

com-mune headquarters of each of the three comcom-munes in Bac

Ninh Province: Thai Bao (Day 1), Quang Pho Thanh (Day

2), and Ninh Xa (Day 3) Three groups were interviewed

each day One group comprised village leaders, one group

comprised household farmers, and one group comprised

large farm operators

For Study 2, three small-group interviews (with 5–12

individuals per group) were held with 23 participants over

3 days (July 25–27, 2011) at each of the three communes in

Quang Ninh Province: Viet Hung (Day 1), Bang Ca (Day

2), and Cam Tay (Day 3) Two groups were held in meeting

rooms at the commune headquarters and one was held at a farmer’s house Individual interviews were held with ten participants

Data Analysis

For Study 1, qualitative theme analysis (Bernard and Ryan 1998; Crabtree and Miller1999) using ATLAS.ti (version 5) was used to distinguish the salient constructs and to identify words commonly used to describe attitudes and experiences Following the guidelines presented by Fereday and Muir-Cochrane (2006), the coding template was developed using a hybrid of the data-driven inductive ap-proach of Boyatzis (1998) and the deductive a priori tem-plate of codes approach outlined by Crabtree and Miller (1999) Core themes that repeatedly appeared in notes of the interview discussions were identified by the lead author and confirmed by a research assistant Consensus was achieved via discussion For Study 2, the key-words-in-context technique (Tesch1990) was used to find instances

of key words in field notes Themes were then identified by sorting the examples into groups of similar meaning, retaining information about the sources of the examples Given that this research focused largely on identifying perceptions that do not necessarily have a correct answer, respondents’ trustworthiness was assessed by asking knowledgeable locals (e.g., village veterinarian) or members

of the local research team to corroborate the plausibility of responses Confirmability was assessed, where possible, by asking respondents to verify some statements (e.g., that there is no place to take dead animals) via a tour of their vicinity Credibility was assessed frequently during inter-views by checking respondents’ agreement with paraphra-sings of their comments

FINDINGS

Study 1: HPAI Risk Perceptions

Key Impacts

Overall, participants were very concerned about HPAI outbreaks, referring to the disease as an ‘‘epidemic,’’ ‘‘a scary disease,’’ ‘‘a big worry,’’ and ‘‘very dangerous,’’ par-ticularly because of the relatively high fatality rate of HPAI

in poultry One participant described the death of chickens from HPAI as ‘‘equal to robbing because it takes away our

Table 2 Characteristics of three groups (traditional,

transi-tional, and modern) of communes

Type of

commune

Percent of houses whose main income

is from agricultural, fishing, or forestry activities

Percent of houses with modern toilets

livelihood with no warning or reason.’’ Participants from

all the communes agreed that the most worrisome aspect of

HPAI was at the household level when substantial

invest-ments and their main income source were lost

Factors Related to HPAI Outbreaks

Modernization and Pollution Participants varied in the

factors they emphasized as important in the spread of

HPAI in poultry, but many focused on the process of

modernization and associated increase in pollution prob-lems Participants from all the communes observed that rapid population increase and economic development had

a detrimental impact on their environment The increased volume of waste from livestock and household garbage, which was perceived to be poorly managed and hard to control, was viewed as a key source of air, land, and water pollution (e.g., dead chickens are thrown onto the street or into fish ponds) Many participants noted the importance

of keeping a clean farm to prevent disease One participant (from a transitional commune) indicated that he kept his farm clean; but that he could not control the whole system, which he observed to be increasingly interdependent Par-ticipants from the traditional commune emphasized the role of transportation in spreading the virus (e.g., trucks spread disease via their tires when transporting feed and international movement of livestock)

Farming Practices The trade-offs of new versus old farming practices were emphasized For instance, some participants (modern commune) commented that the new, industrial way is good because ‘‘the feed is more relaxing’’ (i.e., farmers do not have to collect or prepare the feed) and the livestock have more economic value However, they also noted that the industrial feed may use too many chemicals, be moldy, or cause chickens to put on weight too fast, causing the livestock to be more prone to disease Farm Size Participants had differing opinions on the role

of the farm size Some participants (modern and transi-tional communes) tended to believe that small and large operations were equally likely to experience an outbreak of disease Some participants in transitional communes, however, emphasized that the high-volume farming oper-ations caused environmental problems because the chicken density was too high and that small-scale operations posed fewer problems Others noted that large operators were more cautious and thus more likely to vaccinate their livestock, but that small-scale farmers were more able to check each chicken and to detect a problem early

Poultry Varieties Most participants agreed that traditional Vietnamese chickens have a higher resistance (and do not require much care) than industrial broiler or layer chickens, but that the latter were more profitable Some noted that compared with broiler chickens, layer chickens were probably less vulnerable because they were kept on a high floor (for easy egg retrieval)

Table 3 Study 2 Participant Characteristics (n = 33)

Commune, Province (urbanization level), n (%)

Mo Lao, Ha Dong (modern) 1 (3.0)

Cam Tay, Quang Ninh (modern) 8 (24.2)

Viet Hung, Quang Ninh (transitional) 13 (39.4)

Phu Luong, Ha Dong (transitional) 1 (3.0)

Quoc Oai, Ha Tay (transitional) 1 (3.0)

Phung Xa, Thach That (transitional) 2 (6.1)

Bang Ca, Quang Ninh (traditional) 7 (21.2)

Participant type, n (%)

Gender, n (%)

Age, years, range (mean) (n = 29) 25–69 (47.9)

Ethnicity, n (%)

Education, n (%)

Vocational school or some college 3 (9.1)

Completed 4-year college or more 0 (0)

Primary profession, n (%)

Non-agricultural wage laborer 3 (9.1)

Professional (e.g., lawyer, doctor, teacher) 0 (0)

Other (e.g., student, homemaker) 8 (24.2)

Annual household income,

VND, range (mean) (n = 22)

10,000,000–

168,000,000 (49,154,545)

Seasonal Variation Most participants noted that

out-breaks were more likely around Teˆt, the Vietnamese New

Year, which usually falls in the western calendar months of

January or February and is a period when more chickens

are raised and exposure is more likely One participant

noted that the powder of xoan (Melia azedarach) flower

(which flowers in spring) causes HPAI

Wild Birds When asked about the role of wild birds, some

said that they had heard about this issue in the media, but

thought it was unlikely to be an important factor because of

a lack of wild birds in their area (in part because villagers

shoot them) Others suggested that wild birds may play a

role in disease outbreaks

Disease Management Strategies

Vaccination Vaccination was generally thought to be

effective, though problematic when poultry are already sick

Some participants from the traditional commune indicated

that they might not buy the vaccine until their poultry was

sick because it was too expensive Both the household

farmers and large operators in transitional and traditional

communes expressed reservations about vaccine

effective-ness Some participants stressed the problems in obtaining

high-quality vaccines (e.g., distributors might dilute it, not

keep it cool during transportation, or sell expired bottles)

and that the availability of too many brands of unknown

quality was confusing

Waste and Disease Management Systems Participants

dis-cussed the need for better waste and disease management

systems Several noted that there was no place to take dead

animals (so they were buried in the paddy fields near the

communes with lime) and there was little systematic

guid-ance about how to deal with waste or disease One participant

in a transitional commune stressed that there was noone to

identify and quarantine the source of an outbreak and that

sometimes quarantine restrictions were implemented too

late One participant (modern commune) observed that

people do not always report poultry deaths Another

par-ticipant (transitional commune) stressed that reporting

deaths brought little extra support Some participants

indi-cated that they were not willing to make reports because they

did not want to precipitate in new culling campaigns

Uncertainty, Knowledge, and Training In general,

partici-pants expressed a desire for more knowledge and training

about disease management options and more consultation opportunities (e.g., about which vaccine to use) One participant (transitional commune) suggested that an early warning system would be helpful (e.g., providing forecasts about when each province will be most likely to experience

an outbreak) Others reflected on the importance of mass media (television and newspapers) as a source of infor-mation about HPAI origins and impacts

Support Systems Key sources of support for understand-ing and managunderstand-ing HPAI included village veterinarians, village leaders, and faculty at the Hanoi University of Agriculture When asked about the role of spirituality, some participants said it did not play an important role in disease management, but others noted that they thought it was important to put food on their altars or to pray internally to ask their ancestors for good health and pro-duction

Figure 1schematically depicts these qualitative themes and their relationships as described in the interview dis-cussions

Study 2: Perceptions of Urbanization and its Impacts

Characteristics of Rural and Urban Settings

Participants from all the communes agreed that the key characteristics of a rural setting include: good or support-ive, sharing relationships and emotional ties among resi-dents; a feeling of coolness and clean, pure air; and agricultural activities (using the land for farming to pro-duce food and raise livestock) They also noted that rural houses were spacious with old-style (pit) toilets In con-trast, key characteristics of an urban setting include: weaker emotional ties among residents; large-scale industrial development; better quality houses with more furniture; a closed sewage system; piped water for cooking; less space (e.g., to raise children); and more closed doors and private spaces Participants noted that income in urban areas typically comes from trading and selling services, that there are a lot of jobs, and that it is easier to make money

Process of Urbanization

Participants from the traditional commune of Bang Ca said that the process of urbanization had been mostly

worth-while, but that it came with both positive and negative

aspects (e.g., the modernization of agriculture increased the

spraying of chemicals to improve crop productivity but

posed a human health risk) Participants in the traditional

settings noted that there has been no change in the type of

toilets or home gardens, but that forested areas have

de-creased (the land is being used for roads, houses, and

mines) and environmental pollution has increased One

man noted that the environment was ‘‘cooler’’ in the past

when their houses were made from wood (rather than

ce-ment) and were placed lower in the valley (rather than

higher up the hillside) Residents in the traditional areas

can now take advantage of more and different (nonfarm)

income sources (e.g., making baskets), but some people still

did not have enough money to cover the basic living

ex-penses They reported that ‘‘civilization is more accessible’’

(e.g., through television, which is available in every house),

but that the traditional customs (praying, ceremonies, and

shamans) remained the same

Participants in the transitional communes (Viet Hung,

Phu Luong, Quoc Oai, and Phung Xa) noted rapid change

over the past 20 years They highlighted not only

improv-ing food security and food diversity, but also new envi-ronmental and health problems (e.g., pesticides and diabetes) One female farmer said that urbanization brought more needs (e.g., her children want to go out for breakfast instead of staying at home like they used to) and that it is more difficult to earn money Another participant noted that the number and types of houses were changing (‘‘no more bamboo fences or simple tile roof houses’’) and that the environment felt ‘‘hotter.’’ Participants described how land value has decreased because it was rezoned for industry; but then industries were not developed, leaving them in ‘‘suspense,’’ not knowing how long they can keep farming their land Participants mentioned that there are more nonfarm income sources and that people now worked for individuals or households rather than the cooperative Despite the negative aspects of transition, participants generally agreed that transition is worthwhile because their living standard and well-being has improved and they have more time to rest

Participants in the modern communes (Cam Tay and

Ha Dong) thought that in general urbanization is worth-while, but that sometimes people cannot keep up with a Figure 1 A mental model of highly pathogenic avian influenza (HPAI) outbreaks as reflected in interview discussions with village leaders, household farmers, and large farm operators in modern, transitional, and traditional communes in Bac Ninh Province, Vietnam

rapid rate or high number of changes They noted how

urbanization has changed the ‘‘coolness’’ of their

envi-ronment, primarily by the removal of trees Other

partici-pants reported that the amount and type of housing was

changing in their area (e.g., from thatched or tile roofs to

brick roofs), but not too fast Every house now has its own

toilet and reliable electricity Stream flow has decreased so

residents buy trucked water for cooking and use catchment

water for washing Participants noted that food security has

improved and there is less hardship Human health is also

better because of improved access to health care and

awareness of disease prevention strategies Families have

fewer children Also, relationships among neighbors have

changed, with more private spaces and less visits because

people are spending more time working

Impact of Urbanization on Poultry and Disease Outbreaks

Participants in the traditional commune reported difficulty

in controlling poultry disease, especially when poultry are

bought from outside the commune and have an unknown

vaccination record They said that their poultry’s health was

not affected by water pollution or human waste, but was

affected by increasing animal density and chemicals in the

environment Modern treatments were available to deal

with disease, but there was uncertainty about what is most

effective One farmer said that they cannot prepare

effec-tively because they do not know what is needed

In contrast, participants in transitional communes

commented that their ability to raise poultry has improved

because they have more medicines and training from the

government on how to use those medicines and deal with

disease They reported vaccinating frequently, so their

exposure to disease is less They felt that it is easier to

quarantine sick animals with the increases in fences and

walls

Participants in the modern communes noted that

animal and human waste can cause disease, but that it does

not happen in their commune Air and water pollution

were identified as causing poultry disease One participant

said that it was easier for poultry to get sick now because

the weather was unstable (changing between extremely hot

to extremely cold) and food is now freely exchanged among

areas (e.g., sickness spreads across the commune easily

when a sick chicken is brought in from another commune)

Animal density was also thought to cause disease because

there was less air for the poultry to breathe

DISCUSSION

HPAI is Perceived as Risky and Disease Manage-ment is Linked with Urbanization Processes

Across the communes, HPAI in poultry was viewed by the participants as a dangerous epidemic with worrying eco-nomic consequences The role of change in agriculture in the spread of HPAI was recognized Both new and old farming practices were perceived as having advantages and disadvantages Participants were sensitive to the roles of farm size, type of chicken, seasonal variation, and wild birds in the spread of HPAI The most prominent disease management strategies focused on vaccination, quarantine, and farm and commune hygiene; with calls for better information and training programs to assist farmers These findings are not only consistent with Fielding

et al.’s (2009) description of lay explanations for the cause

of HPAI in humans, but also extend our understanding of risk mental models to explain the perceived causes of HPAI

in poultry The present results are consistent with studies finding a reluctance to report poultry deaths to authorities because of fears of culling and economic impacts (Barennes

et al 2007; Ly et al 2007; Sultana et al 2012; Takeuchi 2006) However, participants in our study seemed to ex-press more motivation for improving waste and disease management practices compared with the previous studies The results of the present research support the notion that urbanization changes people’s ability to respond appropriately to variations in their environment The inability to respond is not necessarily because of an inability to perceive change (cf Lilian et al.2007) Rather, our participants described a rapid and extensive change that poses different challenges for poultry management as communes move from rural to transitional to urban set-tings

Limitations

The first limitation of this research is that samples were drawn from people willing and able to participate in a 2-h interview and may not represent the general population of farmers in Vietnam in terms of sociodemographic variables such as age, education, and income There is also a gender imbalance in each study, potentially biasing the perspec-tives obtained toward issues most relevant to the majority gender However, the consistency of the findings from this

study with those reported elsewhere encourages the

confi-dence that many of the observations are generalizable

A second limitation concerns the small size of samples

in these studies It is possible that additional information

would have been obtained with a broader sample of

indi-viduals Within this sample, however, we believe that the

questions posed were thoroughly addressed at least by

modern and transitional commune participants because

little new information was obtained in the final interviews

This phenomenon is called saturation in qualitative

re-search and indicates that a topic has been adequately

sampled (i.e., enough interviews conducted) Another

limitation of the small sample sizes is that they do not

permit statistical differences to be analyzed and thus

con-clusions are restricted to the descriptive statements elicited

Finally, the topic of avian influenza is a sensitive issue

that participants may have been reluctant to discuss openly,

especially in a group setting Other research suggests that

farmers are reluctant to report cases of animal illness

be-cause of fears of culling and loss of income (e.g., Sultana

et al 2012) This is a limitation of most investigations

relying on self-report, but was minimized in the present

research by building good rapport and selecting a diverse

sample to increase the chance of including some more open

participants Nonetheless, constrained responses limit our

capacity to draw conclusions

FUTURE DIRECTIONS

The exploratory nature of this research established preliminary

qualitative information about a topic poorly documented to

date, namely how perceptions of HPAI risk in poultry relate to

urbanization in Vietnam The research findings offer a first step

toward understanding mental models of disease risk and

urbanization in a developing country context The findings

raise several interesting questions for future research: What

research methods (e.g., anonymous surveys) are effective at

capturing representative perspectives of a sensitive issue in a

rapidly changing environment? How well do risk perceptions

correspond with the actual outbreaks of disease in poultry and

does the relationship vary with setting (traditional, transitional,

modern)? What analytic methods are needed to identify the

components of urbanization that best predict the perceived risk

and the actual outbreaks of avian influenza in poultry?

Addressing these and other questions will help policymakers be

better informed about the optimal risk prevention and

man-agement strategies in diverse settings

The authors are grateful to colleagues at the Hanoi University of Agriculture for assistance in conducting interviews; to Boonyarit Itsara and Charles Nguyen for assistance in verifying and analyzing data The authors also thank Michael DiGregorio, Nancy Lewis, and Bruce Wilcox for assistance in developing the interview protocol This work was funded by the National Science Foundation Grant # DEB-0909410

REFERENCES Barennes H, Martinez-Aussel B, Vongphrachanh P, Strobel M (2007) Avian influenza risk perceptions, Laos (letter) Emerging Infectious Diseases 13(7):1126

Bernard HR, Ryan GW (1998) Text Analysis: Qualitative and Quantitative Methods In: Handbook of Methods in Cultural Anthropology, Bernard HR (editor), Walnut Creek: Altamira Boyatzis R (1998) Transforming qualitative information: thematic analysis and code development, Thousand Oaks: Sage

Capua I, Marangon S (2000) The avian influenza epidemic in Italy, 1999–2000: a review Avian Pathology 29:289–294 Central Intelligence Agency (2013) The World Factbook

https://www.cia.gov/library/publications/the-world-factbook/ geos/vm.html Accessed 25 July 2013

Crabtree B, Miller W (1999) A template approach to text analysis: developing and using codebooks In: Doing Qualitative Research, Crabtree B, Miller W (editors), Newbury Park: Sage, pp 163–177 Dent JE, Kao RR, Kiss IZ, Hyder K, Arnold M (2008) Contact structures in the poultry industry in Great Britain: exploring transmission routes for a potential avian influenza virus epi-demic BMC Veterinary Research 4:27

Fang L, de Vlas SJ, Liang S, Looman CWN, Gong P, Xu B, et al (2008) Environmental factors contributing to the spread of H5N1 avian influenza in mainland China PLoS One 3(5):e2268–e2268 Fereday J, Muir-Cochrane E (2006) Demonstrating rigor using thematic analysis: a hybrid approach of inductive and deductive coding and theme development International Journal of Qual-itative Methods 5(1):1–11

Fielding R, Leung GM, Lam WWT, Jiang CQ, Sitthi-Amorn C, Ahn LV, et al (2009) A pan-Asian survey of risk perception, attitudes and practices associated with live animal markets Hong Kong Medical Journal 15(5):17–20

Gilbert M, Xiangming X, Pfeiffer DU, Epprecht M, Boles S, Czarnecki C, et al (2008) Mapping H5N1 highly pathogenic avian influenza risk in Southeast Asia Proceedings of the Na-tional Academy of Sciences 105:4769–4774

Lilian A, Kliskey A, Williams P (2007) The distancing effect of modernization on the perception of water resources in Arctic communities Polar Geography 30(3–4):175–191

Ly S, Van Kerkhove MD, Davun Holl D, Froehlich Y, Vong S (2007) Interactions between humans and poultry, rural Cam-bodia Emerging Infectious Diseases 13(1):130–132

Mallin MA, Cahoon LB (2003) Industrialized animal production?

A major source of nutrient and microbial pollution to aquatic ecosystems Population and Environment 24:369–385

Morgan D (1998) The Focus Group Guidebook, Thousand Oaks:

Sage

O’Brien K (1993) Improving Survey Questionnaires Through

Focus Groups In: Successful Focus Groups: Advancing the State

of the Art, Morgan D (editor), Newbury Park: Sage, pp 105–117

Pfieffer DU, Minh PQ, Martin V, Epprecht M, Otte MJ (2007) An

analysis of the spatial and temporal patterns of highly

patho-genic avian influenza occurrence in Vietnam using national

surveillance data The Veterinary Journal 174:302–309

Pope C, Mays N (1995) Researching the parts other methods

cannot reach: an introduction to qualitative methods in health

and health services research British Medical Journal 311:42–45

Saksena S, Trinh DT, Pham NL, Sultana N, Finucane ML, Fox JM

(2013) Animal husbandry and farm-level bio-security practices

across the rural to urban gradient in Vietnam and their

rela-tionship to avian influenza Paper presented at the American Association of Geographers, Los Angeles, CA, April 11, 2013 Spencer JH (2013) The urban health transition hypothesis: empirical evidence of an avian influenza Kuznets curve in Viet Nam? Journal of Urban Health 90(2):343–357

Sultana R, Rimi NA, Azad S, Islam MS, Khan MSU, Gurley ES,

et al (2012) Bangladeshi backyard poultry raisers’ perceptions and practices related to zoonotic transmission of avian influ-enza Journal of Infection in Developing Countries 6(2):156–165 Takeuchi MT (2006) Avian influenza risk communication, Thai-land (letter) Emerging Infectious Diseases 12(7):1172–1173 Tesch R (1990) Qualitative research: analysis types and software tools New York: Falmer