Introduction and Purpose
Bangladesh is a lower-middle-income country, with an agricultural-based economy (World Bank
2016a and
b). Its poultry sector, both commercial and family (particularly, small extensive scavenging and extensive scavenging) poultry production, is growing rapidly (Dolberg
2008; FAO
2014). Bangladesh is one of the most densely populated countries, for both human (1072 people/km
2) and poultry populations (1194 birds/km
2) (World Bank
2013). Consistent with the practice in other South Asian countries, Bangladesh continues the culture of keeping animals together with people within the same house (household farms) and of live bird markets (Dolberg
2008; Gerloff et al
2016). A high proportion of all poultry products go through these live bird markets, and most of the products are sold unprocessed (Dolberg
2008). Thus, multiple factors, like rapidly growing poultry production, densely populated state, the culture of household farms and live bird markets, and selling unprocessed birds, make Bangladesh prone to zoonotic disease outbreaks, such as avian influenza. Avian influenza spreads within poultry populations, but, in some cases (such as the highly pathogenic avian influenza (HPAI) H5N1), may affect human beings and cause severe illness and death (WHO
2016a and
b). Humans who have had avian influenza generally developed the infection after coming into close contact with infected birds, dead or alive. In Bangladesh, the first poultry outbreak of H5N1 was reported in 2007 (OIE
2013). Some subtypes, such as H5N1 and H9N2 which are endemic in Bangladesh, are zoonotic, and their potential for recombination and re-assortment raises concerns about their pandemic potential (Li et al
2010; Russell et al
2012; Monne et al
2013; OIE
2013). Thus, avian influenza is a major animal and public health concern in Bangladesh.
Policy development and implementation is a complex process which frequently takes place in an unstable and rapidly changing context, subject to unpredictable internal and external factors (Brugha et al
2000; Varvasovszky et al
2000). In 2006, the Government of Bangladesh adopted the first national avian influenza and human pandemic influenza preparedness and response plan to cover the period 2006–2008 (DGHS
2006). The plan provided a strategic framework for coordinating activities within and between different stakeholders (and sectors) for preparedness and response to avian and pandemic influenza in Bangladesh. The second plan was drafted in 2008, to cover the period 2009–2011, but the highest level did not approve this version (DGHS
2009). As the epidemiology of avian and pandemic influenza is evolving, it was mentioned that the plans would be periodically reviewed and revised whenever deemed necessary. However, even after a decade of development and implementation of the first plan, a stakeholder analysis was never performed, neither of the policy development process nor the actual policy. A stakeholder analysis is a systematic process of gathering and analysing information from different stakeholders in relation to the policy development process and/or the actual policy, which in turn allows the identification of opportunities for improvement/reforms (Brugha et al
2000; Schmeer
2000; Varvasovszky et al
2000). The use of qualitative stakeholder analysis as a tool has become increasingly popular in the health policy field, and this popularity reflects and recognises the central role of stakeholders in the context of policy development and implementation (Mason and Mitroff
1981; Crosby
1992; Walt
1994; Zeng et al
2017). Thus, the specific objectives of this qualitative stakeholder analysis study were to identify: (1) the reasons behind the development of the avian influenza policy in Bangladesh, (2) the sectors involved in the development of this policy, (3) the factors considered while developing this policy, (4) the future policy options to prevent and control avian influenza and other poultry-related zoonotic diseases and (5) the future policy environment that is suitable for developing and implementing such policies.
Methods
This qualitative stakeholder analysis study consisted of two stages. At the first stage, semi-structured interviews were conducted with 23 key stakeholders from the government, international multilateral organisations, non-governmental organisations (NGOs) and trade associations in Bangladesh. Those with high-level authority and/or special knowledge were purposively selected for this purpose. Data collection and analysis proceeded simultaneously until the saturation of themes was reached (Green and Thorogood
2014). Interviews were conducted either face-to-face or through video-conferencing using a pre-developed interview guide. Questions were asked in relation to the five objectives of this study (Lindenberg and Crosby
1981; Freeman
1984). These interviews were conducted in February and March 2016, in either English or Bengali, depending on the interviewee’s preference. Each interview lasted for around 30–45 min.
At the second stage, a Chatham House roundtable was convened in Dhaka, Bangladesh, on 15 May 2016. The aim of the roundtable group discussion was to validate the findings from the first stage. The roundtable was attended by 40 key stakeholders from the government, international multilateral organisations, NGOs and trade associations in Bangladesh. Those with high-level authority and/or special knowledge were purposively selected for this purpose. The roundtable was conducted in English and under the Chatham House Rule, which states that “when a meeting, or part thereof, is held under the Chatham House Rule, participants are free to use the information received, but neither the identity nor the affiliation of the speaker(s), nor that of any other participant, may be revealed” (Chatham House
2002). This rule aims to provide anonymity to speakers in order to encourage openness and the sharing of information and is now used throughout the world as an aid to a free discussion (Chatham House
2002).
The interviews and the group discussion were noted and digitally recorded with consent. Then, they were transcribed (verbatim), translated (if necessary), anonymised and checked for accuracy. An interpretive analysis was conducted using thematic analysis (Boyatzis
1998). Transcripts were read and re-read. Initial codes were developed and applied initially to a small number of transcripts, enabling iteration of the thematic index. Coding of each transcript against the index was undertaken. Microsoft Excel 2013 was utilised in the analytic process. The study was approved by the Chittagong Veterinary and Animal Sciences University Research Ethics Committee, Bangladesh.
Discussion
A decade after development and implementation of the first national avian influenza and human pandemic influenza preparedness and response plan in Bangladesh, a qualitative stakeholder analysis was performed in relation to the policy development process and the actual policy. This study led to the recommendation of policy options to prevent and control avian influenza and other poultry-related zoonotic diseases in Bangladesh. It was recommended that the policy should be based on the One Health concept, be evidence-based, sustainable, reviewed and updated as necessary. The future policy environment that is suitable for developing and implementing these policies should take into account the following points: the need to formally engage multiple sectors, the need for clear and acceptable leadership, roles and responsibilities and the need for a common pool of resources and provision for transferring resources. Most of these recommendations are directed towards the Government of Bangladesh. However, other sectors, including research and poultry production stakeholders, also have a major role to play to inform policy making, and actively participate in the multi-sectoral approach.
Major recommendations that came up during the study were: to base the policy on the One Health concept and to establish a physical One Health Secretariat in Bangladesh. The implementation of the One Health approach to prevent and control zoonotic diseases has been advocated globally (Okello et al
2015). The term ‘One Health’ has evolved to acknowledge the close relationship between humans, animals and the natural, political and socioeconomic environments in which they coexist (One Health Initiative Task Force
2008). One Health advocates maintain that disease control as a result of inter-sectoral collaboration between the veterinary, medical and environmental sectors results in added benefits to each individual sector (Schelling et al
2005; Zinsstag et al
2005). The One Health movement has gained momentum in recent years, evolving from its roots in One Medicine to promote inter-sectoral collaboration and a ‘whole of society’ approach to global health governance (Zinsstag et al
2005 and
2007; Scoones
2010; CDC and EU
2011; Lee et al
2013). Political vigour towards zoonotic disease control under One Health appeared to escalate in 2005 after David Nabarro’s dire warning that essentially reframed the HPAI policy debate ‘from a problem of chicken farmers and hygienically inadequate markets in East and Southeast Asia to one that could affect everyone’ (BBC
2005; Scoones et al
2007). The 2008 multi-partner ‘One World, One Health’ strategic framework is a noticeable example of the ‘significant policy shift’ that took place globally at the time (FAO et al
2008; Chien
2013).
In spite of such well-intentioned initiatives, apprehension remains that the ‘big politics’ of stamping out intermittent disease outbreaks continues to dominate international health policy dialogue, neglecting different livelihoods and systems-based approaches that are arguably more relevant to developing countries or endemic situations (de Savigny et al
2004; Scoones
2010; Mwacalimba
2012). It should be noted that village poultry (small extensive scavenging and extensive scavenging family poultry, including their owners and traders) was frequently implicated in the disease transmission in the early days of the HPAI H5N1 pandemic. The pandemic had a negative impact on village poultry (small extensive scavenging and extensive scavenging family poultry) in many countries. However, with the improved understanding of the disease epidemiology, it was recognised that village poultry (small extensive scavenging and extensive scavenging family poultry) raised under extensive conditions pose relatively less of a threat than intensively raised poultry of homogenous genetic stock with poor bio-security (Alders et al
2014). Now, there is a growing awareness of the importance of facilitating socially and culturally sensitive dialogue to develop avian influenza prevention and control options. Although it is important to consider local perspectives while developing such policies, the ease of transmission of the virus should not be downplayed (Alders et al
2014). This happened in Bangladesh where the radius (of an affected farm) for the destruction of poultry farms was reduced over time making it no longer consistent with the international guidelines.
In 2012, the Ministry of Health and Family Welfare, Ministry of Fisheries and Livestock and Ministry of Environment and Forests, jointly with FAO, WHO and UNICEF, developed and endorsed a One Health strategic framework in Bangladesh (IEDCR
2012). This framework provided direction for the implementation of the One Health approach for preventing and controlling emerging, re-emerging and high-impact zoonotic diseases in Bangladesh (IEDCR
2012). However, the complete implementation of such an important framework remains questionable, even after four years of its publication. On a better note, the inter-ministerial One Health steering committee in Bangladesh, a newly established coordination body, has recently (after completion of this study) decided to establish the One Health Secretariat in Bangladesh. It has been agreed that each of the three ministries (Ministry of Health and Family Welfare, Ministry of Fisheries and Livestock and Ministry of Environment and Forests) will send one officer to the Secretariat on secondment. The terms of reference of the Secretariat have been developed through a workshop, as agreed by all the major government actors. It has been unequivocally agreed to locate the Secretariat at IEDCR, considering the fact that IEDCR has been informally playing the role of a coordinator since the inception of the One Health concept in Bangladesh. The functions of the Secretariat could be similar to that of the UK’s Human Animal Infections and Risk Surveillance (HAIRS) group, such as horizon scanning, risk assessment, risk management and risk communication (Morgan
2014), and could go beyond these tasks, such as to develop, review and update policies, as recommended by the stakeholders. The HAIRS group is a collaboration between a number of human and animal health organisations within the UK government, has been described as a major One Health initiative and has been cited as an example of how the fields of animal and human health can effectively work together on risk assessment for emerging health threats (Lightfoot et al
2013; Heymann et al
2014). Internationally, the One Health agenda tends to be dominated by veterinarians, and one of the features of the HAIRS group is the equal roles played by human and animal health practitioners. Some of the potential reasons behind the success of the HAIRS group are: a systematic approach to its work and early on agreed terms of reference, assiduously recording of discussions and decisions, effective communication and transparent risk assessment processes. Collaborative relationships are best developed ahead of a crisis and the frequent exchanges between the members of the group ensure that these relationships are maintained. It is always a balance between members being senior enough to be able to represent an organisation and having sufficient time to contribute effectively to the group. Egos, hidden agendas and professional superiority are often features of working groups, but such behaviours are rarely exhibited in the HAIRS group (Morgan
2014).
This study has a number of strengths and weaknesses. As far as we are aware, this is the first stakeholder analysis study in Bangladesh in the context of avian influenza policy. Almost all the purposively selected stakeholders participated in the study, which included a range of key stakeholders and some of them were busy, senior officials. Some of the study authors are senior officials in Bangladesh, and this helped us to gain access to these key stakeholders. In terms of generalisability, the study findings could be valid in countries having similar political, economic, social and cultural systems (such as in other South Asian countries). A single, bilingual researcher (KC) analysed the qualitative data. Thus, the data analysis could be quite subjective. However, a two-stage process was followed to validate the study findings. Single researchers can also ensure a more uniform approach in collecting qualitative data, ensuring higher reliability and more internally valid cross-comparisons of data (Brugha et al
2000; Varvasovszky et al
2000). Moreover, he was an external (non-Bangladeshi) and independent researcher, which reduced the chances of individual biases. Finally, this was a cross-sectional study, conducted over a limited period of time. The whole context of the study is subject to change over time, including stakeholders’ perception of the past, their interest or position. The political context of policy development and implementation is frequently complex and unstable, especially in many developing countries like Bangladesh, and can be subject to sudden, unexpected transformations (Brugha et al
2000; Varvasovszky et al
2000).