Enhancing the use of stakeholder analysis for policy implementation research: towards a novel framing and operationalised measures

Abstract

Background
Policy is shaped and influenced by a diverse set of stakeholders at the global, national and local levels. While stakeholder analysis is a recognised practical tool to assess the positions and engagement of actors relevant to policy, few empirical studies provide details of how complex concepts such as power, interest and position are operationalised and assessed in these types of analyses. This study aims to address this gap by reviewing conceptual approaches underlying stakeholder analyses and by developing a framework that can be applied to policy implementation in low-and-middle income countries.

Methods
The framework was developed through a three-step process: a scoping review, peer review by health policy experts and the conduct of an analysis using key informant interviews and a consensus building exercise. Four characteristics were selected for inclusion: levels of knowledge, interest, power and position of stakeholders related to the policy.

Result
The framework development process highlighted the need to revisit how we assess the power of actors, a key issue in stakeholder analyses, and differentiate an actor’s potential power, based on resources, and whether they exercise it, based on the actions they take for or against a policy. Exploration of the intersections between characteristics of actors and their level of knowledge can determine interest, which in turn can affect stakeholder position on a policy, showing the importance of analysing these characteristics together. Both top-down and bottom-up approaches in implementation must also be incorporated in the analysis of policy actors, as there are differences in the type of knowledge, interest and sources of power among national, local and frontline stakeholders.

Conclusion
The developed framework contributes to health policy research by offering a practical tool for analysing the characteristics of policy actors and tackling the intricacies of assessing complex concepts embedded in the conduct of stakeholder analyses.

Full text

1
BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
Enhancing the use of stakeholder
analysis for policy implementation
research: towards a novel framing and
operationalised measures
Marysol Astrea Balane ,1 Benjamin Palafox ,2 Lia M Palileo- Villanueva,1
Martin McKee ,2 Dina Balabanova3
Original research
To cite: BalaneMA, PalafoxB,
Palileo- VillanuevaLM,
etal. Enhancing the use
of stakeholder analysis
for policy implementation
research: towards a novel
framing and operationalised
measures. BMJ Global Health
2020;5:e002661. doi:10.1136/
bmjgh-2020-002661
Handling editor Stephanie M
Topp
Received 15 April 2020
Revised 26 August 2020
Accepted 27 August 2020
1College of Medicine, University
of the Philippines Manila,
Manila, Philippines
2Centre for Global Chronic
Conditions, London School of
Hygiene and Tropical Medicine,
London, UK
3Department of Global Health
& Development, London School
of Hygiene & Tropical Medicine,
London, UK
Correspondence to
Marysol Astrea Balane;
marysol. balane@ gmail. com
© Author(s) (or their
employer(s)) 2020. Re- use
permitted under CC BY.
Published by BMJ.
ABSTRACT
Background Policy is shaped and inuenced by a diverse
set of stakeholders at the global, national and local levels.
While stakeholder analysis is a recognised practical
tool to assess the positions and engagement of actors
relevant to policy, few empirical studies provide details
of how complex concepts such as power, interest and
position are operationalised and assessed in these types of
analyses. This study aims to address this gap by reviewing
conceptual approaches underlying stakeholder analyses
and by developing a framework that can be applied to
policy implementation in low- and- middle income countries.
Methods The framework was developed through a three-
step process: a scoping review, peer review by health
policy experts and the conduct of an analysis using key
informant interviews and a consensus building exercise.
Four characteristics were selected for inclusion: levels of
knowledge, interest, power and position of stakeholders
related to the policy.
Result The framework development process highlighted
the need to revisit how we assess the power of actors,
a key issue in stakeholder analyses, and differentiate an
actor’s potential power, based on resources, and whether
they exercise it, based on the actions they take for or
against a policy. Exploration of the intersections between
characteristics of actors and their level of knowledge can
determine interest, which in turn can affect stakeholder
position on a policy, showing the importance of analysing
these characteristics together. Both top- down and
bottom- up approaches in implementation must also be
incorporated in the analysis of policy actors, as there are
differences in the type of knowledge, interest and sources
of power among national, local and frontline stakeholders.
Conclusion The developed framework contributes to
health policy research by offering a practical tool for
analysing the characteristics of policy actors and tackling
the intricacies of assessing complex concepts embedded
in the conduct of stakeholder analyses.
INTRODUCTION
Researchers seeking to influence policy must
engage with relevant stakeholders. But whom
and how? Stakeholder analysis can identify
key actors in the policy process and develop
strategies to engage with them.1 Stakeholders
are defined by Varvasovszky and Brugha as
“actors who have an interest in the issue under
consideration, who are affected by the issue,
or who – because of their position – have or
could have an active or passive influence on
the decision- making and implementation
processes.”2
Stakeholder analysis can be used for
different purposes in policy research,
Key questions
What is already known?
►Stakeholder analyses require assessment of the lev-
els of power, position and interest of actors relative
to health policies, but few empirical studies provide
details of how the complex concepts embedded
within these analyses are operationalised.
What are the new ndings?
►There is no universally agreed way of conducting
stakeholder analyses, as different studies followed
diverse guidelines and frameworks and operation-
alised key concepts such as stakeholder power and
interest in various ways.
►The developed framework proposes to assess stake-
holder power as potential power in terms of access
to resources, and position to reect an actor’s actu-
al exercise of power based on actions taken for or
against a policy. Intersections between knowledge,
interest, power and position must also be taken into
account in stakeholder analyses.
►Both top- down and bottom- up approaches in policy
implementation must be considered when analysing
policy actors at the global, national and local levels.
What do the new ndings imply?
►The developed framework addresses a gap in health
policy research by offering a practical tool for an-
alysing the characteristics of policy actors and by
tackling the intricacies of assessing complex con-
cepts while conducting stakeholder analyses.

2BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
BMJ Global Health
retrospectively to assess stakeholder roles in policy
processes, or prospectively to inform future policy direc-
tions.2 It covers the entire policy cycle, from agenda
setting to policy formulation, adoption, implementation
and evaluation.3 Stakeholders with competing ideologies
or interests can influence formulation of policies4 and
can reshape adopted policies by contesting and negoti-
ating their implementation.5
Stakeholders are typically analysed by their interests,
position and, especially, their power.6 Interest refers to
their concerns about how a particular policy will affect
them7; position reflects their level of support for or oppo-
sition to the policy1; and power is their ability to affect
policy, reflecting their resources and ability to mobilise
them.1 While interest and position can be straightfor-
ward to ascertain, assessing power is more complex as
it impacts on all steps of the policy process.8 Yet, power
is often poorly characterised in empirical research on
implementation of disease management policies, espe-
cially in low- and- middle income countries (LMICs).9 10
Sriram et al describe power as conceptually fluid,
viewed on different levels, political angles and sociocul-
tural lenses.8 In health policy, Lukes’ three faces of power
include a first, visibly played out in the formal political
arena; a second involving formal and informal processes
underpinning development of political agendas; and a
third, invisible but shaping the narrative on measures
considered acceptable.11 Gaventa expands this approach,
introducing the concept of a ‘power cube’ with one
dimension represented as visible, hidden and invisible,
as in Luke’s model; and a second categorising power
as local, national and global. A third divides spaces for
engagement into closed or decision- making by an elite
group of actors; invited spaces that allow participation by
citizens or beneficiaries; and claimed or created spaces
emerging from social mobilisation or natural gatherings
outside formal policy arenas. Gaventa argues that signif-
icant changes are possible by aligning strategies to axes,
like a Rubik’s cube.12 VeneKlasen and Miller distinguish
four expressions of power. ‘Power over’ is the capability
of those who hold power to exert influence on those
without, ‘power with’ involves synergy with different
actors, ‘power to’ pertains to one’s own ability to act,
while ‘power within’ refers to self- awareness and recogni-
tion of self- worth leading to action.13
In policy implementation, power relates to the distri-
bution of authority in a system. The traditional top- down
model sees actors deriving power from their place in a
de facto hierarchy. Policies are formulated at national or
international levels and cascaded downwards,14 those
in higher tiers setting objectives to be accomplished by
implementers.10 Bottom- up implementation focuses
on the active role of implementers and their ability to
modify or react to policies based on local context.14 It
views implementation as interactive, involving negotia-
tion and conflict.14
Stakeholder analyses face several challenges. Fast-
changing policy environments can shift stakeholder
positions so findings are time- bound.15 Having many
potential stakeholders may pose difficulties, as does the
ability to delineate personal and role- driven opinions of
those in organisations. Other challenges include sensi-
tivities around asking about power and interest, and
potential bias arising from the position of the analysts,
often immersed in the policy process themselves.6 These
can be addressed using longitudinal studies that capture
changing positions, limiting analyses to main stake-
holders, capturing personal views and organisational
positions, triangulating primary data against secondary
sources, and self- awareness and diversity of analysts in
research.6
Another challenge is that many existing analyses fail
to fully describe the process by which power, position
and interest are operationalised. Given the complexity
involved, understanding how researchers assessed stake-
holder characteristics becomes crucial to guide later
stakeholder engagement and interventions. The process
starts with defining concepts, choosing variables or
domains to represent these concepts and operational-
ising them using relevant indicators. Where several ques-
tions measure a single concept, scales or indices can be
composite measures.16
This paper contributes to the methodology of stake-
holder analysis, especially the operationalisation of
stakeholder characteristics in health policy implemen-
tation research. We reviewed conceptual approaches in
stakeholder analyses, proposed an analytical framework
that drew on those conceptual approaches and included
domains and value scales used to operationalise stake-
holder characteristics. The framework was reviewed by
experts and field- tested by diverse policy actors involved
in a study of hypertension in the Philippines.17 The field-
testing of the framework with implementation of the Phil-
ippine Package of Essential Non- communicable Disease
(NCD) Interventions, to identify and manage NCD risk
factors in primary care,18 provides a rich backdrop for
framework development, involving interaction with
diverse policy actors within a pluralistic health system.
METHODS
Scoping review
We performed several interlinked steps (figure 1). First,
a scoping review of stakeholder analyses mapped key
concepts and identified how stakeholder characteristics
were operationalised, defined and measured in health
research literature.
Our scoping review process was guided by Arksey and
O’Malley’s framework, with five stages: identifying the
research question, identifying relevant studies, study
selection, data charting and collating, and reporting
results.19 We followed the Preferred Reporting Items
of Systematic Reviews and Meta- Analysis extension
for Scoping Reviews (PRISMA- ScR) (figure 2 and
online supplemental appendix 1).20 Papers between
January 2009 and July 2019 that included definitions of

BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661 3
BMJ Global Health
stakeholder characteristics in the analysis were eligible.
Studies that did not provide definitions of characteristics
were excluded.
Characteristics of power and influence were central
to the analytical process. PubMed was searched with the
key words ‘Stakeholder analysis AND power’ and ‘Stake-
holder analysis AND influence’ which together gener-
ated 433 records. Duplicates were filtered using EndNote
V.X9, leaving 411 unique records. A researcher screened
titles and abstracts for eligibility. Sixteen were deemed
relevant and analysed (online supplemental appendix 2).
A manual search of selected references gathered relevant
guidelines on methodologies for operationalising stake-
holder characteristics, and were reviewed as part of the
framework development.
Our data charting form extracted the following from
each paper: year of publication, authors, study location,
purpose of analysis, framework and guidelines used,
definitions of stakeholder characteristics, and domains
used to assess characteristics. Using a narrative synthesis
approach, the scoping review generated an initial frame-
work synthesising definitions and domains of stakeholder
characteristics such as knowledge, interest, position and
power, and methods used to assess them.
Review by experts
The initial framework was sent for peer- review by interna-
tional experts on health policy and health systems lead-
ership, purposively selected due to (1) their expertise in
conducting policy and stakeholder analyses in LMICs,
(2) self- identifying as policy analysts and applying their
skills in different projects, and (3) work highly regarded
by peers or frequently cited. We initially contacted five via
email, with three providing detailed reviews. They were
from academia with doctoral degrees in health policy
and politics and at least 10 years of experience in health
policy and systems research. Specifically, the experts
were asked to comment on the definitions, domains and
scales used to assess stakeholder characteristics related
to knowledge, power, interest and position. This was
Figure 1 Methodology for framework development.
Figure 2 Flow diagram of scoping review on stakeholder characteristics.

4BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
BMJ Global Health
followed by a review of literature provided by experts,
who guided us to particular theories and empirical work
on power,11 12 21 22 which we supplemented with a manual
search of selected references. We included references
based on their fit with our study and revised our frame-
work based on expert comments and the literature.5 8 23
Field-testing of the framework in the Philippine context
We field- tested the revised framework in a study on
implementation of an NCD policy in the Philippines to
ascertain its appropriateness for assessing the levels of
knowledge, interest, position and power of stakeholders.
Field- testing was via key informant interviews from August
2019 to March 2020, and a 1- day consensus workshop in
November 2019.
Eighteen key informants were selected via purposive
sampling, which were identified either through docu-
ment review or a snowball technique wherein respon-
dents were requested to identify stakeholders with whom
they usually collaborate. Identified stakeholders were
eligible for interview based on their engagement with
policy implementation and views of other stakeholders.
This technique was particularly useful in narrowing down
interviewees as it identified those currently wielding
power in implementation processes, and those that can
potentially influence other implementers.
A semistructured interview topic guide was developed
based on the framework (online supplemental appendix
3). The interviews took between 30 and 90 min and were
conducted in a mix of Tagalog and English depending on
the preference of the interviewee. Written consent forms
were requested, and participant numbers were assigned
to interviewees to protect their identities.
A 1- day consensus- building workshop was conducted
to further field- test the framework by assessing key stake-
holders according to their level of knowledge, interest,
power and position, and to evaluate the framework’s
performance. Nine stakeholders from academic institu-
tions, non- government organisations, professional soci-
eties, local governments and frontline health workers
participated in the consensus workshop and engaged in
discussions (table 1).
Given the varied experiences and perspectives, and the
potential professional hierarchies in the assembled group,
the workshop was designed to solicit expert views using
elements of focus group discussion to allow participants
and facilitators to freely share knowledge and evidence
and discuss each topic; and elements of the Delphi
process, allowing participants to provide anonymous
input.24–27 The workshop sought to achieve consensus on
four topics: (1) identification of key stakeholder groups;
(2) understanding of definitions of framework concepts,
domains, indicators and scoring; (3) scoring based on
stakeholder characteristics as defined in the analytical
framework; and (4) determining strategies on engaging
organisations identified as important and influential in
implementation.
From an initial long list of stakeholders, participants
narrowed the list down to 14 key actors. Following a
framework orientation presentation, participants then
scored the key actors on a scale of 0–3 based on their
level of knowledge, interest and power; while stakeholder
position was scored 1–5 based on the strength of support
for the policy. After a first round of scoring, a summary
was shown to the participants and a facilitated discus-
sion ensued which clarified participants’ understanding
and achieved consensus on the definitions of framework
concepts, domains, indicators and scores, and to gather
information on their experiences of assessing levels of
stakeholder characteristics in a Delphi exercise. This was
followed by further focus group discussion to achieve
consensus on the scoring of each key stakeholder’s level
of knowledge, interest, power and position. Our general
consensus processes used a conventional group problem-
solving approach involving (1) problem clarification, (2)
agreement on deliberation procedures, (3) information
and perspective sharing, (4) option development, (5)
group selection of the preferred option,26 which was
operationalised using simple majority vote. Key points
from the discussion were recorded by the research team
and integrated into the final version of the framework.
A power- position matrix was then shown to the partici-
pants based on the results and a consensus- oriented
focus group discussion ensued to determine strategies
for engaging with stakeholders based on their location
in the matrix.
Key informant interviews and the consensus building
workshop were recorded, with audio files transferred to
an encrypted laptop and transcribed verbatim. Feedback
from the interviews and consensus building exercise were
then taken into account in the revision of the framework
to ensure its appropriateness in analysing stakeholders in
the Philippine context and the topic area (CVD). The
Table 1 Organisations that participated in the stakeholder
analysis study
Interview respondents
Consensus workshop
participants
Organisation No Organisation No
Department of
Health (Central and
Regional)
5 Academia 2
Local government
units
3Local government
unit
3
Professional society 2 Frontline health
worker
2
Multilateral
organisation
4 International
organisation
1
International and
national NGOs
2 Professional
society
1
Academia 1
Media 1
NGOs, non- government organisations.

BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661 5
BMJ Global Health
framework was reviewed and revised at every step by the
researchers who were health professionals and/or faculty
members from well- known universities.
RESULTS
Scoping review
Sixteen articles were analysed, of which 12 applied to
policy processes, 2 to health interventions, 1 was set in an
organisation and 1 was a methodological paper. There
was a mix of 9 prospective analyses and 7 retrospective
analyses included in the review. In the 15 empirical
studies included, analyses were applied across different
contexts with 3 studies conducted in low- income econo-
mies, 4 in lower- middle income economies, 4 in upper-
middle income economies and 4 in high- income econ-
omies.
The review revealed the absence of a standard way of
operationalising characteristics in a stakeholder anal-
ysis, with different empirical studies over the past decade
applying a variety of frameworks and guidelines. Some
studies followed a single guideline when analysing stake-
holder characteristics. Table 2 presents an overview of
the different frameworks, including the stakeholder
characteristics assessed in each study. Among those
that followed a single framework, most used Schmeer’s
guidelines, an approach recommended by WHO,1 which
proposed to analyse stakeholders based on seven char-
acteristics. Studies that adapted this, however, selected
anywhere from 2 to 7 of the proposed characteristics,
depending on which were deemed appropriate to the
context being analysed.28–31 Six of the 16 papers did not
follow a particular guideline, but drew on several sources
to define and operationalise stakeholder characteristics.
An example of this is Abiiro and McIntyre’s study on the
premium payment policy in Ghana which adapted defi-
nitions of characteristics from various methodological
papers and studies, and determined domains of power
from stakeholder insights.32
When defining stakeholder characteristics, power and
influence were used interchangeably in several studies
and were defined, variously, as the ability of a stakeholder
to affect policy formulation or implementation, their
access to resources and ability to mobilise them.6 33–35
Three studies defined power as the stakeholder’s ability
to influence policy, or assessed power and influence sepa-
rately.29 32 36
Interest was defined in terms of the advantages or
disadvantages that policy implementation conferred on a
stakeholder, or their political stake or degree of involve-
ment in an issue.1 37 The definition of position was more
or less consistent across different studies, and indicates
the level of support or opposition towards a policy or a
programme.
There seems to be a general agreement across studies
on the definitions of power, interest and position,
suggesting that there is an agreement at the conceptual
level. There is, however, considerable variability in how
these characteristics were operationalised in guidelines
and empirical work.
Table 3 shows domains that have been proposed in
guidelines and methodological papers to assess stake-
holder characteristics as well as domains that have been
applied in empirical studies. We examined five guidelines
and frameworks cited in the studies shown in table 2, and
one methodological paper identified in our search. Out
of these six, we included four that provided domains for
assessing power, interest or position in the table.
For studies that conducted analyses in specific policy
contexts, only 7 out of 15 papers provided details on what
specific aspects of stakeholder characteristics were evalu-
ated. Among the seven papers that included such details,
six conducted prospective analyses with stakeholders
at the agenda setting,36 38 policy formulation32 33 39 and
intervention planning stages.35 One paper conducted a
retrospective analysis applied to policy implementation.28
It should be noted that while the guidelines offered
well- defined characteristics, some offered flexibility
for researchers to use their own assessment criteria
depending on their aim or specific contexts. Even in
studies that proposed domains, some of these were
defined broadly, such as ownership or access to resources,
which still allows for researchers to ultimately define
Table 2 Approaches and characteristics used in the
stakeholder analysis studies
Source
Stakeholder
characteristics
used in studies Studies (n)
Adapted from
Schmeer’s Stakeholder
Analysis Guidelines1
Power
Interest
Position
Leadership
Knowledge
Alliances
Attitude
428–31
Adapted from
methodology papers
by Varvasovzky and
Brugha2 42
Position
Inuence
138
Stakeholder Salience
Theory43
Power
Legitimacy
Urgency
144
Contextual Interaction
Theory45
Power
Motivation
Information
146
Combination of different
guidelines and previous
studies
Understanding
Interest
Power
Position
Inuence
66 32–34 47 48
Developed from
stakeholder views
Power 136
No specic guidelines
or approach mentioned
Power
Interest
Position
Inuence
235 39

6BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
BMJ Global Health
Table 3 Domains used in stakeholder analysis guidelines, methodological papers and studies to assess power and inuence,
interest, and position
Domains for characteristics
No of guidelines/method
papers (N=4)
No of empirical studies
(N=7)
Domains for power or inuence
Technical/professional knowledge/skills 147 328 32 36
Decision- making 328 33 38
Political/inuential position 21 47 2 28 36
Financial power/money 31 43 47 2 28 36
Legal mandate 147 228 32
Human 11
Technological 11
Ability to place the issue on the agenda 233 38
Legislative power for policy approval 232 38
Inuence over policy outcomes 232 38
Attribution of power (actor’s power as perceived by themselves and others) 149
Ability to mobilise on the issue 11235 38
Coercive: physical resources of force, violence or restraint 243 47
Normative: symbolic inuences 143
Connections to inuential stakeholders 147
Ownership/control of resources 149 232 35
Voting power/inuence over voters 132
Involvement in policy formulation 132
Willingness to engage in policy discussions 139
Ability to be heard in discussions 139
Ability to inuence other actors 139
Ability to inuence public opinion 132
Directly or indirectly take action for or against the policy 135
Control over implementation at the local level 132
Determine policy success and sustainability 132
Possession of privileges 135
Ability to organise members 132
No of votes 147
Domains for interest
Overall perceived impact 11132
Key interest/concerns 11132
Professional afliation 133
Stakeholder agendas 133
Status within the community 133
Pursue benets for stakeholder 147
Achieve equitable treatment for player’s group 147
Advance player’s view of common good 147
Garner more resources 147
Preserve power 147
Domains for position
Level of support or opposition 21 47 628 32 33 35 38 39

BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661 7
BMJ Global Health
which particular resources to assess. Examination of
empirical studies, therefore, provides helpful insights on
which domains can be applied to actual policy contexts.
Table 3 reflects common domains used for power or
influence, interest and position revealed by our review.
Based on the scoping review findings, four character-
istics were selected for inclusion in our adapted frame-
work: knowledge, interest, power and position. Power,
interest and position were the most frequently assessed
characteristics in the stakeholder analysis literature,
while knowledge was included in recognition that a stake-
holder’s understanding of a policy may determine their
level of interest and perception of how it can potentially
affect them.32
Aside from the results of the scoping review, a meth-
odological paper and several studies informed the devel-
opment of the initial analytical framework; these include
Schmeer’s definitions of characteristics, value scales from
Caniato et al, and domains to assess power from Abiiro
and McIntyre’s study.1 30 32 Table 4 shows a summarised
version of the initial stakeholder analysis framework
developed for policy implementation research and the
subsequent modifications made after the expert review,
key informant interviews and the consensus building
exercise.
Feedback from expert review
The health policy experts provided feedback on the defi-
nitions of stakeholder characteristics and the domains
used to assess them in the initial version of the framework.
For knowledge, one expert suggested that the domains
and value scales included seemed to imply that the policy
is clearly laid out when, in reality, policies can be vague or
not publicised. To address this comment, source of infor-
mation was added as a domain in the framework to iden-
tify gatekeepers and determine the process and potential
gaps in the transfer of knowledge.
With regard to interest, an expert argued that the
concept refers to the concerns and driving motivations
of stakeholders and how policies impact their organisa-
tion. Based on this feedback, the definition of interest
was revised and domains were changed to reflect whether
the policy is considered as a priority or perceived to affect
the stakeholder in any way.
All three experts provided comments on the charac-
teristic of power and suggested a review of relevant theo-
ries and empirical work analysing power to see how these
can be incorporated in the framework. Following this
feedback, additional literature on power was reviewed to
determine its different dimensions and how these can be
assessed in practice. A study by Dalglish et al applied to
the policy for integrated community case management
of childhood illness in Niger was particularly relevant
as it examined a range of power theories and opted to
select three dimensions of power deemed to be rele-
vant to the country’s context: political authority, finan-
cial resources and technical expertise.23 These three
domains also emerged in the scoping review and were
thus incorporated in the revised analytical framework to
be field- tested in the Philippine setting.
Still pertaining to power, one of the experts commented
on how the domains seemed to be underpinned by a
top- down view of implementation and may not take into
account bottom- up approaches. Accordingly, the domain
of ‘Leadership’ was added to describe a stakeholder’s
ability to convene partners and mobilise them to work
together to implement a policy. The additional domain
was drawn from Lehmann and Gilson’s study on the
micro- practices of power of community health workers
in South Africa.5
Finally, one expert highlighted how position and
interest are linked, as a stakeholder’s perception of how
policy will impact their organisation can affect their level
of support. As a result, another domain, ‘actions taken
to demonstrate support or opposition’, was added to
the framework to draw out ways in which stakeholders
express their positions during policy implementation.
Field-testing the appropriateness of the framework to CVD
policy in the Philippines
Overall, the framework was found to be acceptable and
appropriate to the policy context in the Philippines
during the interviews and consensus building exercise.
There were clarifications and discussions, however, about
some of the characteristics and on how to evaluate levels
of knowledge, interest, power and position of stake-
holders, which led to further refinement of the frame-
work.
With regard to knowledge, participants in the group
consensus exercise differentiated between operational
knowledge of the policy and understanding the overall
policy goal, as some stakeholders may know how to imple-
ment the policy’s components, without necessarily being
aware of what the policy seeks ultimately to achieve.
Following discussions, four domains were included in
the final framework to reflect awareness of policy, oper-
ational knowledge of policy, understanding of policy
rationale and source of information. Value scales were
likewise revised to categorise extensive knowledge as
understanding both policy rationale and implementa-
tion issues, general knowledge as operational know- how
in implementation, while limited knowledge refers to
awareness about the policy without knowing specific
details about it.
Another point of discussion during the group exercise
was the link between awareness and interest, and how
knowledge of the policy can determine level of interest.
Those unaware of the policy may thus appear to have
low interest, highlighting the need to assess interest in
conjunction with knowledge. Interview findings suggest
that asking about interest can also be potentially sensitive
as it delves into the underlying motivations of different
stakeholders. The direct question about policy impact on
the organisation seemed to be unclear for some stake-
holders and follow- up questions exploring opportunities
and costs of the policy, as well as providing examples,

8BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
BMJ Global Health
helped stakeholders think more concretely about it. In
the framework, this translates into specifying perceived
impact as opportunities and costs to the stakeholder.
An important issue tackled during the consensus work-
shop was whether to rate stakeholders based on their
potential power or actual exercise of power. A stake-
holder, for example, may have resources and, as such, the
potential to be involved in implementation, but for one
reason or another does not fully exercise their potential
power. To resolve the issue, the participants reached an
agreement to rate the power characteristic as potential
power, and to reflect the actual exercise of power when
rating overall position.
A final issue on the link between personal attributes and
stakeholder power arose during the interview process.
Within organisations there may be charismatic and
Table 4 Summary of stakeholder characteristics, denitions, domains and value scales in the initial framework, and key
changes made after expert review, key informant interviews and consensus building exercise
Initial framework based on scoping review Key changes after expert review
Key changes after key informant interviews and
consensus building exercise
Knowledge
Denition: Stakeholder’s level of knowledge and
understanding of the policy
Domains:
►Knowledge of policy (awareness and ability to
describe key features)
►Understanding of policy purpose
Value scales: 1—No or minimum knowledge, 2—
General knowledge, 3—Extensive knowledge
Denition: Retained
Changes in domains:
►Knowledge of policy and its implementation
►Source of information (added)
Value scales: Retained
Denition: Retained
Changes in domains:
►Operational knowledge of policy
►Awareness of policy (added)
Changes in value scales: 0—No knowledge, 1—
Limited knowledge, 2—General knowledge, 3—
Extensive knowledge
Note: General knowledge dened as operational
knowledge on policy, while extensive knowledge
includes both operational knowledge and
understanding of policy rationale
Interest
Denition: Extent to which stakeholders perceive
policy implementation as relevant and likely to affect
them
Domains:
►Relevance of policy to stakeholder
►Willingness to participate in implementation
►Likelihood to affect stakeholder
Value scales: 1—No or minimum interest, 2—Limited
interest, 3—General interest, 4—High interest
Changes in denition: Stakeholder’s motivations
and perceived impact of policy implementation to
their own organisation
Changes in domains:
►Policy objective core to organisation’s mission
►Policy is a priority for organisation
►Perceived impact of policy implementation to
own organisation
Value scales: Retained
Denition: Retained
Changes in domains:
►Perceived policy impact in terms of opportunities
and costs to the stakeholder
Changes in value scales: 0—No Interest, 1—
Limited interest, 2—General interest, 3—High
interest
Power
Denition: The ability of the stakeholder to affect
policy implementation
Domains:
►Capacity to design policies
►Capacity to fund policy implementation
►Capacity to implement policy
►Ability to lead and gather support from
stakeholders
►Ability to inuence public opinion
Value scales: 1—Low, 2—Medium, 3—High
Note: Stakeholders rated based on possession and
control of resources
Denition: Retained
Changes in domains:
►Political authority
(a) Direct: Derived from hierarchy, legal mandate,
regulatory regimes.
(b) Indirect: Ability to create incentives and
constraints for other act.
►Financial capacity
Possession and control of financial resources
►Technical expertise
Technical capacity to produce,interpret and
disseminate knowledge and information.
►Leadership
Ability to build partnerships and motivate other
stakeholders for or against policy implementation.
Changes in value scales: 1—Low, 2—Medium,
3—High
Note: Stakeholders rated based on possession and
control of resources and ability to make decisions in
policy implementation
Changes in denition: The potential ability of the
stakeholder to affect policy implementation based
on resources
Changes in domains:
►Leadership
(a)Ability to build partnerships, motivate other
stakeholders and/or shape opinion for or against
policy implementation.
(b)Personal attributes of individuals within the
organisation which can include charismatic
authority, personal commitment and motivation
Changes in value scales: 1—Low, 2—Medium,
3—High
Note: Stakeholders rated based on possession and
control of resources and potential to affect policy
implementation
Position
Denition: Whether the stakeholder supports,
opposes or is neutral about policy implementation
Domains
►Degree of support or opposition to policy
Value scales:
1—Opponent, 2—Moderate opponent, 3—Neutral,
4—Moderate supporter, 5—Supporter
Denition: Retained
Changes in domains:
►Actions taken to demonstrate support or
opposition to policy (added)
Value scales: Retained
Denition: Retained
Changes in domains:
►Degree of support or opposition to policy
expressed through use of potential power or
resources
►Actions taken to demonstrate support or
opposition to policy
Changes in value scales: 1—Strong opponent,
2—Moderate opponent, 3—Neutral, 4—Moderate
support, 5—Strong support
Source: Denitions, domains and value scales for the frameworks were adapted from elements in the methodological papers and studies of Varvasovsky and
Brugha (2000), Schmeer (2000), Abiiro and McIntyre (2013), Lehmann and Gilson (2013), Caniato et al (2014), Dalglish et al (2015) and Sriram et al (2018) and
feedback from health policy experts and stakeholders.

BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661 9
BMJ Global Health
motivated individuals that help move policy implemen-
tation forward. Thus, personal attributes were included
in the final framework under the domain of leadership
to reflect the role of individuals in policy implementa-
tion. Table 5 shows the final revised framework based on
combined feedback from expert consultations, interviews
and group consensus exercise.
DISCUSSION
In this study, we present an adapted framework for stake-
holder analysis that draws on empirical research, theory
and advice from health policy experts, specifically devel-
oped for application to health policy contexts in LMICs.
Its key domains and characteristics are fully operational-
ised. From our experience of taking a flexible and iter-
ative approach to develop and field- test the framework
in the Philippines, we believe it is a practical tool that is
able to assess the stakeholder landscape in which a health
policy is implemented and examine complex stakeholder
characteristics in a rigorous, transparent, yet straightfor-
ward manner—a process that is seldom described well in
empirical stakeholder analyses.6
While existing guidelines and frameworks for stake-
holder analyses clearly define stakeholder characteris-
tics, the process of operationalising or measuring these
concepts is often left to the discretion of researchers to
ensure they are fit for purpose and are adapted to their
particular context. It is, however, important that empir-
ical studies explicitly state the criteria for assessing char-
acteristics to minimise bias,2 reduce ambiguity and allow
the analysis to be replicated by other scholars intending
to do similar studies. Our study contributes to the stake-
holder analysis literature by describing our process, and
the intricacies of identifying domains to include and
putting a value on abstract concepts. This is a critical step
required in analyses, but posing challenges that are not
discussed adequately in the literature. By synthesising
the most frequently used domains in studies, bringing
insights from studies on power of actors outside of stake-
holder analyses and going through an iterative process of
operationalisation, the study offers a framework that can
more easily be adapted and applied by other researchers.
The framework also contributes to the overall discussion
on power of actors and how to assess this, especially at the
policy implementation stage.
The development process identified multiple intersec-
tions between stakeholder characteristics. Level of knowl-
edge was linked to level of interest, as stakeholders unaware
of the policy may be perceived as having low interest in its
implementation, suggesting the need to analyse interest
in conjunction with knowledge. This finding is consis-
tent with Abiiro and McIntyre’s study, which postulated
that a stakeholder’s understanding of a policy affects
its perceived impact or interest in it.32 Interest was also
linked to position, as the perceived impact of the policy
on the stakeholder determines whether or not they will
support or oppose its implementation. This finding is
Table 5 Finalised framework for stakeholder analysis
applied to the PhilPEN policy implementation context
Knowledge
Denition: Stakeholders’
knowledge and understanding of
the policy
Domains:
►Awareness of policy
►Operational knowledge of
policy
►Understanding of policy
rationale
►Source of information
Value scales:
0—No knowledge
Stakeholder is not aware of policy
1—Limited knowledge
Stakeholder is aware of policy but
have minimal knowledge about its
purpose or implementation
2—General knowledge
Stakeholder has operational
knowledge about policy
3—Extensive knowledge
Stakeholder understands policy
rationale and has operational
knowledge of policy
Interest
Denition: Stakeholder’s
motivations and perceived impact
of policy implementation to their
own organisation.
Domains:
►CVD control core to
organisation’s mission
►Policy is a priority for
organisation
►Perceived policy impact in
terms of opportunities and
costs to the stakeholder
Value scales:
0—No interest
Policy is not considered a priority
and not perceived to impact
stakeholder
1—Limited interest
Policy is not considered a priority
and has minimum impact on
stakeholder
2—General interest
Policy is a priority and has
moderate impact on stakeholder
3—High interest
Policy is part of the stakeholder’s
core mission and has high
perceived impact on stakeholder
Power
Denition: The potential ability
of the stakeholder to affect policy
implementation
Domains:
►Political authority
a. Direct: Derived from
hierarchy, legal mandate,
regulatory regimes.
b. Indirect: Ability to create
incentives and constraints
for other actors.
►Financial capacity
Possession and control of nancial
resources
►Technical expertise
Technical capacity to produce,
intrepet and disseminate
knowledge and information
►Leadership
a. Ability to build partnerships,
motivate other stakeholders
and/or shape opinion
for or against policy
implementation.
b. Personal attributes of
individuals within the
organisation which can
include charismatic
authority, personal
commitment and
motivation.
Value scales:
1—Low power
Stakeholder possesses and has
control over use of one to two
sources of power, low potential to
affect policy implementation
2—Medium power
Stakeholder possesses and
has control over use of two to
three sources of power, has
moderate potential to affect policy
implementation
3—High power
Stakeholder possesses and
has control over use of three
to four sources of power, has
high potential to affect policy
implementation
Continued

10 BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
BMJ Global Health
consistent with the definition of interest in other studies
as ‘positive and negative impact’40 or ‘advantages and
disadvantages’ of the policy to the stakeholders.1
The link between power and position was also explored,
as the question of whether to measure potential power
based on resources versus actual exercise of power deter-
mined through stakeholder actions became pertinent.
Discussions revealed that there is value in looking at
these two separately in order to identify stakeholders with
high potential power, but who are not fully exercising
this power in their implementation efforts. Such differ-
entiation is also helpful when determining appropriate
stakeholder engagement strategies. As a result, power,
as used in the framework, meant potential power, while
position reflected the exercise of power in terms of the
actual use of available resources and actions taken by the
stakeholder to support or oppose policy implementation.
The operationalisation process highlighted the diffi-
culty of assessing stakeholder power. With the abundance
of theories on power and its implicit and explicit mani-
festations in stakeholder interactions,8 it was challenging
to determine which particular domains to include in the
framework. Domains of power found in the stakeholder
analysis literature typically identify access to sources of
power, but discussions with stakeholders revealed that
access to these sources is only one aspect of power,
and effective use of potential power to achieve policy
outcomes is also key but more challenging to assess. Eval-
uating the exercise of power by different actors in a stake-
holder analysis involves the examination of policy actors
interacting at the global, national and local levels. At the
international level, some studies have shown that donors
can control implementation outcomes through condi-
tions stipulated in funding agreements41 and can also
influence different stages of the policy process through
leverage of financial resources, technical expertise and
intersectoral pressure.21 Frontline providers, on the other
hand, can exercise ‘micro- practices of power’ through
day- to- day decision- making that can either support or
subvert intended policy outcomes.5 Therefore, a compre-
hensive assessment of power requires sampling that
provides sufficient representation of perspectives from
global down to local levels. This task becomes even more
complex and onerous in highly heterogeneous settings,
which may result from factors, such as health system
decentralisation, as in the Philippines.
Furthermore, when analysing stakeholders involved in
policy implementation, it is important to consider incor-
porating both top- down and bottom- up approaches to
account for the important role that frontline workers
play in the implementation process. Implementing
actors, often considered as having low power, can actually
exercise very high levels of discretionary power (eg, by
withholding labour), which, when done as a group, can
undermine a policy’s goals.5
The new framework has several limitations. Since it was
applied in the context of health policy implementation
in an LMIC context, discussions were mostly focused on
assessing characteristics of actors in implementing the
policy as opposed to their ability to advocate or design
policies, which would be more relevant at the policy
formulation stage. While the domains can also be seen as
relevant for assessing actors across different policy stages,
high- income economy settings or different fields outside
health policy research, its application to these contexts is
beyond the scope of the study.
Although the study touched on the concept of power,
it focused more on practical domains for assessing power
among stakeholders, which was mainly sourced from
previous empirical studies and feedback from experts
and stakeholders. Domains of power identified in the
framework were drawn mostly from more visible sources
of power, or those that can be verified through document
review, interviews and consensus among stakeholders.
While the results of the interviews can provide some
insights on the less visible forms of power at play during
implementation, an additional layer of analysis may be
needed to situate power dynamics between actors within
the broader macro- political context and societal struc-
tures, such as those along the lines of gender, class or
race.10 An example of this process of contextualisation
can be seen in Gilson et al’s study in South Africa and
Tanzania which identified situational, structural, exoge-
nous and cultural factors affecting stakeholder interac-
tions in universal health coverage debates.6 Sources of
power should also be treated as context dependent and
time bound as the power of stakeholders may shift over
time and may only be applicable in certain contexts.10
Another limitation is that the scoping review only used
one database, PubMed. While we reviewed additional
Position
Denition: Whether the
stakeholder supports, opposes
or is neutral about policy
implementation
Domains:
►Degree of support or opposition
to policy expressed through
use of potential power (sources
of power)
►Actions taken to demonstrate
support or opposition to policy
Value scales:
1—Strong opponent
Stakeholder uses potential power
to strongly act against policy
implementation
2—Moderate opponent
Stakeholder uses potential power
to moderately act against policy
implementation
3—Neutral
Stakeholder does not use potential
power and does not act for or
against policy implementation
4—Moderate support
Stakeholder uses potential power
to moderately act in support of
policy implementation
5—Strong support
Stakeholder uses potential power
to act strongly in support of policy
implementation
Source: Denitions, domains and value scales for the framework were
adapted from elements in the methodological papers and studies of
Varvasovsky and Brugha (2000), Schmeer (2000), Abiiro and McIntyre
(2013), Lehmann and Gilson (2013), Caniato et al (2014), Dalglish et al
(2015) and Sriram et al (2018) and feedback from health policy experts
and stakeholders.
Table 5 Continued

BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661 11
BMJ Global Health
literature provided to us by the experts we consulted,
allowing us to draw insights from other relevant fields
such as political science, it was beyond the scope of the
project to attempt a comprehensive review of literature
from all the fields that might have something to say. We
felt given the aim of the study, this approach enabled
us to include seminal papers on stakeholder analysis.
Also, the stakeholders who took part in the consensus
building exercise were mostly frontline implementers,
and higher- level actors were unable to participate despite
repeated efforts to reach them. However, as noted above,
the involvement of frontline implementers ultimately
strengthened the framework refinement by representing
crucial bottom- up perspectives during the development
process, while insights from higher- level policy makers
were captured during key informant interviews from
international and national stakeholders.
CONCLUSION
While there is a wealth of theories, guidelines and
approaches, empirical works providing details on how
stakeholder characteristics are assessed remain scarce.
We offer an adapted framework for stakeholder analysis
that builds on key advances in the field and has been
shown to be applicable to health policy implementation
research in an LMIC context. The paper presents prac-
tical guidance on how to develop the framework domains
and its specific characteristics, emphasising the impor-
tance of revisiting how complex concepts such as knowl-
edge, interest, power and position have been defined and
operationalised in stakeholder analysis studies.
While the framework was developed in the context
of the Philippine health system, it is likely to be highly
relevant to researchers conducting stakeholder analyses
in other LMIC contexts. This is especially important
for comparisons of stakeholders across countries,
which require consistency in the definition of concepts,
domains, indicators and scoring. Our experience empha-
sises the need for researchers conducting stakeholder
analyses to include details and accounts of how they have
operationalised and assessed the concepts, as they seek to
arrive at an overall understanding of the diverse ways in
which actors relate and interact with each other to shape
and influence policy processes.
Acknowledgements The authors would like to thank the Wellcome Trust/
Newton Fund- MRC Humanities & Social Science Collaborative Award scheme
(200346/Z/15/Z) for providing funding for this research. MAB would like to thank
Chevening Scholarships, the UK government’s global scholarship programme,
funded by the Foreign and Commonwealth Ofce (FCO) and partner organisations
for her scholarship to study in the UK, and the London School of Hygiene & Tropical
Medicine Trust Fund for the travel grant support to conduct the study.
Contributors DB, BP, MM, LMP- V and Christine Candari conceptualised the work
described. MAB, DB, BP and LMP- V contributed to the development of the study
design and data collection. MAB produced the rst draft of the manuscript which
was critically revised by DB, BP, LMP- V and MM. All authors revised and approved
the nal version.
Funding This study was funded by Wellcome Trust/Newton Fund- MRC Humanities
& Social Science Collaborative Award scheme 200346/Z/15/Z.
Competing interests All authors report grants from Wellcome Trust, during the
conduct of the study; LPV reports grants from Philippine College of Physicians,
non- nancial support from Sano and Cadila and grants from WHO Asia Pacic
Observatory, outside the submitted work.
Patient and public involvement Patients and/or the public were not involved in
the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Ethics approval Ethical approval was obtained from the ethics research boards of
the University of the Philippines Manila- Panel 2 and London School of Hygiene and
Tropical Medicine.
Provenance and peer review Not commissioned; externally peer reviewed.
Data availability statement All data relevant to the study are included in the
article or uploaded as supplementary information.
Open access This is an open access article distributed in accordance with the
Creative Commons Attribution 4.0 Unported (CC BY 4.0) license, which permits
others to copy, redistribute, remix, transform and build upon this work for any
purpose, provided the original work is properly cited, a link to the licence is given,
and indication of whether changes were made. See:https:// creativecommons. org/
licenses/ by/ 4. 0/.
ORCID iDs
Marysol AstreaBalane http:// orcid. org/ 0000- 0002- 7466- 5764
BenjaminPalafox http:// orcid. org/ 0000- 0003- 3775- 4415
MartinMcKee http:// orcid. org/ 0000- 0002- 0121- 9683
REFERENCES
1 Schmeer K. Stakeholder analysis guidelines, 2000. Available: https://
www. who. int/ workforcealliance/ knowledge/ toolkit/ 33. pdf [Accessed
3 Mar 2019].
2 Varvasovszky Z, Brugha R. How to do (or not do). A stakeholder
analysis. . Oxford Univ Press, 2000: 15. 338–45. http:// dess. fmp. ueh.
edu. ht/ pdf/ Zsuzsa_ Varvasovsky_ 2000_ stakeholder_ analysis. pdf
3 Howlett M, Ramesh M. Studying public policy: policy cycles and
policy subsystems. second ed. Toronto: Oxford University Press,
2003.
4 Surjadjaja C, Mayhew SH. Can policy analysis theories predict and
inform policy change? reections on the battle for legal abortion in
Indonesia. Health Policy Plan 2011;26:373–84.
5 Lehmann U, Gilson L. Actor interfaces and practices of power in
a community health worker programme: a South African study of
unintended policy outcomes. Health Policy Plan 2013;28:358–66.
6 Gilson L, Erasmus E, Borghi J, etal. Using stakeholder analysis to
support moves towards universal coverage: lessons from the shield
project. Health Policy Plan 2012;27 Suppl 1:i64–76.
7 Roberts M, Hsiao W, Berman P, etal. Getting health reform right.
Oxford: Oxford University Press, 2008.
8 Sriram V, Topp SM, Schaaf M, etal. 10 best resources on power
in health policy and systems in low- and middle- income countries.
Health Policy Plan 2018;33:611–21.
9 Gilson L, Raphaely N. The terrain of health policy analysis in low and
middle income countries: a review of published literature 1994-2007.
Health Policy Plan 2008;23:294–307.
10 Erasmus E, Gilson L. How to start thinking about investigating power
in the organizational settings of policy implementation. Health Policy
Plan 2008;23:361–8.
11 Lukes S. Power: a radical view. Second ed. Palgrave Macmillan,
2005: 14–38.
12 Gaventa J. Finding the spaces for change: a power analysis. IDS
bull. , 2006: 37, 23–33. https://www. powercube. net/ wp- content/
uploads/ 2009/ 12/ nding_ spaces_ for_ change. pdf
13 VeneKlasen L, Miller V. Power and Empowerment. In: A New Weave
of Power, People & Politics: The Action Guide for Advocacy and
Citizen Participation, 2002: 39–58. https:// justassociates. org/ sites/
justassociates. org/ les/ 07chap3_ power_ nal. pdf
14 Walt G. Health policy: an introduction to process and power.
London; New Jersey: Zed Books, 1994: 153–77.
15 Hyder A, Syed S, Puvanachandra P, etal. Stakeholder analysis
for health research: case studies from low- and middle- income
countries. Public Health 2010;124:159–66.
16 Engel R, Schutt R. Chapter 4: Conceptualization and Measurement.
In: Fundamentals of social work research, 2014: 67–71. https:// us.
sagepub. com/ sites/ default/ les/ upm- assets/ 61666_ book_ item_
61666. pdf

12 BalaneMA, etal. BMJ Global Health 2020;5:e002661. doi:10.1136/bmjgh-2020-002661
BMJ Global Health
17 Palafox B, Seguin ML, McKee M, etal. Responsive and equitable
health Systems- Partnership on non- communicable diseases
(respond) study: a mixed- methods, longitudinal, observational
study on treatment seeking for hypertension in Malaysia and the
Philippines. BMJ Open 2018;8:e024000.
18 Department of Health Philippines. Administrative Order No.
2012-0029: Implementing guidelines on the institutionalization of
Philippine Package of Essential NCD Interventions (Phil Pen) on the
integrated management of hypertension and diabetes for primary
health care facilities, 2012. Available: https:// dmas. doh. gov. ph: 8083/
Rest/ GetFile? id= 336917
19 Arksey H, O'Malley L. Scoping studies: towards a methodological
framework. Int J Soc Res Methodol 2005;8:19–32.
20 Tricco AC, Lillie E, Zarin W, etal. PRISMA extension for scoping
reviews (PRISMA- ScR): checklist and explanation. Ann Intern Med
2018;169:467.
21 Khan MS, Meghani A, Liverani M, etal. How do external donors
inuence National health policy processes? experiences of domestic
policy actors in Cambodia and Pakistan. Health Policy Plan
2018;33:215–23.
22 Walls H, Liverani M, Chheng K, etal. The many meanings of
evidence: a comparative analysis of the forms and roles of evidence
within three health policy processes in Cambodia. Health Res Policy
Syst 2017;15:95.
23 Dalglish SL, Surkan PJ, Diarra A, etal. Power and pro- poor policies:
the case of iCCM in niger. Health Policy Plan 2015;30 Suppl
2:ii84–94.
24 Meyer MA, Booker JM. Eliciting and analyzing expert judgment:
a practical guide. Society for industrial and applied mathematics,
2001.
25 Perera AH, Drew CA, Johnson CJ. Expert knowledge and its
application in landscape ecology. Springer New York, 2011.
26 Susskind LE, McKearnen S, Thomas- Lamar J. The consensus
building Handbook: a comprehensive guide to reaching agreement.
Sage publications, 1999.
27 McMillan SS, King M, Tully MP. How to use the nominal group and
Delphi techniques. Int J Clin Pharm 2016;38:655–62.
28 Heydari M, Seyedin H, Jafari M, etal. Stakeholder analysis of Iran's
health insurance system. J Educ Health Promot 2018;7:135.
29 Durham J, Warner M, Phengsavanh A, etal. Stakeholder analysis
of community distribution of misoprostol in Lao PDR: a qualitative
study. PLoS One 2016;11:e0162154.
30 Caniato M, Vaccari M, Visvanathan C, etal. Using social network
and stakeholder analysis to help evaluate infectious waste
management: a step towards a holistic assessment. Waste Manag
2014;34:938–51.
31 Haidari AM, Zaidi S, Gul R. Prospects for the sustainability of
delivering the basic package of health services in Afghanistan: a
stakeholder analysis. 20, 2014. http:// applications. emro. who. int/
EMHJ/ V20/ 05/ EMHJ_ 2014_ 20_ 5_ 300_ 308. pdf
32 Abiiro GA, McIntyre D. Universal nancial protection through
national health insurance: a stakeholder analysis of the proposed
one- time premium payment policy in Ghana. Health Policy Plan
2013;28:263–78.
33 Ancker S, Rechel B. Hiv/Aids policy- making in Kyrgyzstan: a
stakeholder analysis. Health Policy Plan 2015;30:8–18.
34 Franco- Trigo L, Marqués- Sánchez P, Tudball J, etal. Collaborative
health service planning: a stakeholder analysis with social network
analysis to develop a community pharmacy service. Research in
Social and Administrative Pharmacy 2020;16:216–29.
35 Namazzi G, N KS, Peter W, etal. Stakeholder analysis for a maternal
and newborn health project in eastern Uganda. BMC Pregnancy
Childbirth 2013;13:58.
36 Fischer SE, Strandberg- Larsen M. Power and Agenda- Setting in
Tanzanian health policy: an analysis of Stakeholder perspectives. Int
J Health Policy Manag 2016;5:355–63.
37 Eden C, Ackermann F. Making strategy: the journey of strategic
management. London: SAGE Publications Ltd, 1998: 113–35.
38 Knai C, McKee M, Pudule I, etal. Soft drinks and obesity in Latvia: a
stakeholder analysis. Eur J Public Health 2011;21:295–9.
39 Gil A, Polikina O, Koroleva N, etal. Alcohol policy in a Russian
region: a stakeholder analysis. Eur J Public Health 2010;20:588–94.
40 Thomas S, Gilson L. Actor management in the development of
health nancing reform: health insurance in South Africa, 1994-1999.
Health Policy Plan 2004;19:279–91.
41 Kapilashrami A, McPake B. Transforming governance or reinforcing
hierarchies and competition: examining the public and hidden
transcripts of the global fund and HIV in India. Health Policy Plan
2013;28:626–35.
42 Varvasovszky Z, Brugha R. Stakeholder analysis: a review. Health
Policy and Planning2000;15:338–45.
43 Mitchell RK, Agle BR, Wood DJ. Toward a theory of Stakeholder
identication and salience: dening the principle of who and what
really counts. Acad Manage Rev 1997;22:853–86.
44 Reyes- Alcázar V, Casas- Delgado M, Herrera- Usagre M, etal.
Stakeholder analysis. Health Care Manag 2012;31:365–74.
45 Bressers H, Klok P- J, O'Toole L. Explaining policy action: a
deductive but realistic theory. Available: http:// citeseerx. ist. psu.
edu/ viewdoc/ download? doi= 10. 1. 1. 534. 1901& rep= rep1& type= pdf
[Accessed July 12 2019].
46 Schang L, Thomson S, Czypionka T. Explaining differences in
stakeholder take up of disease management programmes: a
comparative analysis of policy implementation in Austria and
Germany. Health Policy 2016;120:281–92.
47 Bryson JM, Patton MQ, Bowman RA. Working with evaluation
stakeholders: a rationale, step- wise approach and toolkit. Eval
Program Plann 2011;34:1–12.
48 Behzadifar M, Gorji HA, Rezapour A, etal. Hepatitis C virus- related
policy- making in Iran: a stakeholder and social network analysis.
Health Res Policy Syst 2019;17.
49 Owens KA. Understanding how actors inuence policy
implementation, 2008. Available: https:// ris. utwente. nl/ ws/
portalles/ portal/ 6084898/ thesis_ Owens. pdf [Accessed July 12
2019].