An Analysis of the Adoption and Implementation of Breastfeeding Policies in Washington State Clinics

An Analysis of the Adoption and Implementation of Breastfeeding Policies in Washington State Clinics


I. Background
The advantages of breastfeeding for the mother-infant-dyad are well documented. [1][2][3][4][5] Notably, these benefits bear a positive, dose-response relationship with breastfeeding duration and exclusivity. 3 Beyond the health benefits for individuals, breastfeeding yields economic and environmental returns for society. 2,6 In the United States, breastfeeding initiation rates continue to improve, from 71.4% in 2002 to 81.1% in 2013. 7,8 Meanwhile, rates for breastfeeding duration and exclusive breastfeeding remain below Healthy People 2020 (HP2020) targets and World Health Organization (WHO) recommendations. 7,9,10 U.S. familiesparticularly those from communities of color, low income households, and rural regionsface substantial obstacles to achieving their breastfeeding goals. [11][12][13][14][15] Though 80% of expectant mothers intend to breastfeed, 60% do not reach the breastfeeding goals they hoped to achieve. 11,12 This disparity between intent and outcome underscores the need to exercise "collective societal responsibility" to eliminate barriers to breastfeeding including economic obstacles, sociopolitical factors, and inconsistent healthcare access. [16][17][18]

II. Breastfeeding Policies in the Continuum of Care
Enabling interventions, including evidence-based policies, remove structural barriers to breastfeeding. [19][20][21][22] In the healthcare setting, breastfeeding policies influence organizational norms, create consistent messaging amongst providers, and promote relevant knowledge and training for staff. [22][23][24] Collectively, these practices can create a supportive environment for families to achieve their breastfeeding intentions. 16 The United Nations Children's Fund (UNICEF) and World Health Organization's (WHO) Baby-Friendly Hospital Initiative (BFHI) 25 and Ten Steps to Successful Breastfeeding (Ten Steps) 26 are internationally-recognized examples of supportive breastfeeding policies. Evidence that supports these policies and other quality-of-care practices continues to grow. 12,[27][28][29] For instance, DiGiloramo et al. (2008) reports that mothers who give birth in hospitals that implement at least six of the Ten Steps are 13 times more likely to report that they are still breastfeeding at six weeks postpartum. 27 Breastfeeding support within the healthcare sector extends beyond the hospital setting; support throughout a continuum of care involves coordination with clinics and other community health partners. 18,30 Care coordinationdescribed by WHO as integrated care 31is an evolving concept intended to "facilitate the appropriate and effective delivery of health care services both within and across services…". 32 With enhanced coordination between hospitals and clinics, families receive improved continuity and quality of care for breastfeeding support. 2,21,23,33 Though clinics play an integral role in breastfeeding support during prenatal and postpartum care, limited studies indicate that clinic breastfeeding policies are inconsistent. 33,34 The current body of literature surrounding breastfeeding policies is largely focused upon the hospital setting with limited information regarding clinics. [35][36][37] It is important to examine the unique influences on policy adoption in clinics so that we can improve breastfeeding care coordination for families.

III. Study Purpose
Breastfeeding policy research that is grounded in theoretical frameworks of organizational change can inform future efforts to improve the adoption and implementation of effective policies. 35,38 By studying the adoption and implementation of breastfeeding policies through an organization-level lens, we can improve our understanding of contextual factors that facilitate or inhibit the process. 19,35,37 Therefore, the purpose of this study is to investigate how the development of evidence-based breastfeeding policies and practices is supported or hindered in clinics using an organization-level diffusion of innovations theory as our framework.

I. General
This article describes the secondary analysis of qualitative data derived from a larger breastfeeding policy study in Washington State. During the planning process for the initial study, the research team sought input from an advisory board composed of sector-specific stakeholders and the Washington State Department of Health (WA DOH) breastfeeding workgroup. The advisory board provided input regarding the study design, interview guide development, and participant recruitment. The team utilized the consolidated criteria for reporting qualitative research (COREQ) 39 and Greenhalgh's Diffusion of Innovations in Service Organizations Framework 40,41 to inform the study design.

II. Analytic Framework
Given the complex nature of policy adoption, the research team applied an existing framework for study design and analysis. The selected framework -Diffusion of Innovation in Service Organizations 40highlights the process and relationships in the diffusion, dissemination and implementation of innovations through an organizational lens. Greenhalgh et al. 40 , derived the framework from a systematic review of diffusion of innovation literature. This framework has been applied to illuminate the policy implementation process for electronic medical records, integrated care pathways and telemedicine. 40,41 In the present study the "innovation" was supportive breastfeeding policies and practices. The model contains nine areas of influence in organizational adoption and implementation of an innovation. Figure 1 illustrates the basic framework and interactions between the model's components.

III. Interview Development & Recruitment
The research team incorporated the Greenhalgh 40 framework into interview questions to highlight facilitators and barriers to policy adoption and implementation. Additionally, the WA DOH's Ten Steps for Clinics 36 was used to develop the interview guidebook for the clinic sector. Interviewees were recruited using purposive sampling. Members of the advisory board initially reached out to potential participants via email. For the clinic sector, potential interviewees included front-line staff and providers, breastfeeding coordinators, nurse managers, and administrators. The team contacted interested participants via phone and email; recruitment continued until the pre-determined participation target was achieved. In total, 19 interviewees representing 17 clinics participated in the interview process.

IV. Data Collection & Analysis
After obtaining consent, trained interviewers conducted 45-60 minute phone interviews. The interviews were audio-recorded and professionally transcribed verbatim (Proof Positive Transcriptions, Garland, Texas). The interviewers offered each participant a $35 gift card for their participation. Interviews were conducted from August 2014 through January 2015.
The research team used qualitative analysis software (Atlas.ti Scientific Software Development GmbH) to organize and synthesize data. Using a selected sample of transcripts (n=16 out of 130, across all sectors), the team used a blinded, thematic analysis process and the Greenhalgh 40 framework to develop a codebook. Once the initial codebook was formed, two or more independent coders applied the initial codebook to 10% of the total interviews (across sectors). The team then reconciled codes and revised the codebook, forming the final guide. Using the finalized codebook, the team double-coded one-third of the transcripts and reached consensus on all coding discrepancies.
Once coded, the team reviewed and summarized transcripts to identify themes. The team then merged these key themes into the Greenhalgh framework diagram and presented the preliminary findings to the study's advisory board for confirmation. For the secondary analysis, we replicated the initial summarization process and translation of information into the Greenhalgh diagram. Using the diagram, we were able to visualize and summarize the cross-cutting themes as facilitators and barriers to breastfeeding policy adoption and implementation.
Using the criteria of the University of Washington Institutional Review Board (IRB), this study was exempt from IRB approval.

I. Sample
Our sample represented 17 organizations including pediatric and neonatal clinics, WIC clinics, and Federally Qualified Health Clinics (FQHC). The sample was mostly based in the Northwestern Region of Washington, the most populous area of the state. See Table 1 for participant characteristics.

II. Factors that Supported or Hindered Policy Adoption & Implementation
The Greenhalgh framework components that were most influential in breastfeeding policy adoption and implementation for clinics were the perceptions of the innovation, system antecedents, organizational readiness, communication & influence, and outer context.

A. Perceptions of the Innovation
Greenhalgh's framework suggests that to be successfully and widely adopted, an innovation must demonstrate relative advantage, compatibility, low complexity, and limited task issues to the relevant organization. 40 Because the innovation(s) for our study was supportive breastfeeding policies and practices, interviewees described their perceptions based upon personal experiences and clinic-specific characteristics.

A.i. Positive Attributes of Breastfeeding Policies and Practices
Breastfeeding policies improved staff and provider accountability and consistency of care for clients. When asked to describe the advantages of a written policy, one interviewee noted,

"I think it's very important, because then you can expect follow-through…Even though it's definitely a main focus here at this WIC clinic and our satellite clinics as well and with the WIC travel team…I think that when you have policies, then it holds people accountable."
Notably, interviewees considered staff and provider accountability an advantage regardless of the level of support by medical providers and staff. Regarding consistency of care, one interviewee captured the importance of written policies as a starting point for consistent messaging. Though their clinic did not have a breastfeeding policy in place, the interviewee noted, "I think [a written policy] is the starting point or jump-off point where we all get on the same page. We can then begin to speak the same language to our patients with regards to breastfeeding."

A.ii. Concerns Regarding Breastfeeding Policies and Practices
Alternatively, some interviewees in "policy-laden" work environments described negative sentiments towards "another policy" and time constraints. When asked how their colleagues felt about breastfeeding policies, an interviewee noted, "I think that they probably just feel like it's one more thing, because at XXX Hospital we have so many policies and procedures about everything. It's like one more thing that we have to fit in." Other concerns included the perception of policies as inflexible, thus limiting one's clinical judgment. An interviewee noted, "I would not be opposed to writing a policy that speaks to the encouragement, training, and advocacy for, but when I think of policies sometimes they can be created in black and white." Lastly, interviewees expressed concerns regarding compatibility between the innovation and the clinic's population. An interviewee highlighted the challenge of adopting a formula-free policy for their facility given the specific community served: "… I don't see us ever taking the role of a formula-free environment. Like I said, our patients are struggling financially and many of them do feed their children formula. To say that we wouldn't have any formula samples available, that would require a much higher buy-in, and I actually don't think that we'd get it." Overall, the interviewees' perceptions of breastfeeding policies and procedures were grounded in the adopter's personal experiences and clinic's specific characteristics.

B. System Antecedents for Innovation
System antecedents for innovation illustrate the organizational contexts that influence the adoption of an innovation. System antecedents include the organization's structure, absorptive capacity for new knowledge, and receptivity for change.

B.i. Organizational Structure
Key factors of organizational structure, including the clinic's size and differentiation, staff makeup, and availability of resources, greatly influenced breastfeeding adoption and implementation. The size of the clinic appeared to influence the level of prioritization for breastfeeding policies. In particular, respondents of non-WIC clinics affiliated with larger healthcare systems often expressed difficulty prioritizing breastfeeding policies with other critical organization responsibilities or interests. When asked to describe the level of prioritization for breastfeeding policies in their clinic, one interviewee noted, "Well, that's tricky. I feel like XXX Hospital is such a humongous organization with so many different priorities, that even I guess within the OB department I guess I would put it somewhere in the middle priority." Notably, interviewees from clinics that offered WIC services and clinics affiliated with the Baby-Friendly Hospitals Initiative (BFHI) did not express this concern regarding prioritization. Interviewees also described the relationship between scarce resources and the level of prioritization for breastfeeding policies. While reflecting upon the level of organizational support for the innovation, an interviewee stated,

C. Organizational Readiness
Interviewees described the innovation-system-fit, power balance, and tension for change, as three major factors which influenced a clinic's readiness to adopt and implement breastfeeding policies and practices.

C.i. Innovation-System-Fit
Innovation-system-fit captures the compatibility between an innovation and the organization's norms, values, mission and overarching goals.
When discussing innovation-system-fit, interviewees described key elements including a clinic's overarching goals, flow of tasks, and specific needs of the clinic population served. These components could either facilitate and/or be a barrier to breastfeeding policy adoption and implementation. For example, one interviewee highlighted the compatibility between breastfeeding policies and the organization's overarching mission as a facilitator:

"I think because this is WIC, [breastfeeding policy support] is pretty high up there. It's a pretty big part of just kind of every appointment, even though it's not necessarily a breastfeeding appointment and just because of the population that we see and what WIC's mission statement is." C.ii. Power-Balance
The power balance within an organization is about the quantity and position of those in support of the innovation versus persons in opposition.
Interviewees often alluded to a positive power balance for breastfeeding support within their clinic, particularly in clinics with young, female staff and providers. Yet, the balance required to shift the policy adoption process forward was influenced by the position of individuals or groups in the organization. In particular, support or opposition from medical providers and administrators greatly influenced the clinic's momentum for adoption and implementation of breastfeeding policies. For champions who were not in positions of power, leveraging leadership was a key facilitator. When asked about the leadership's level of support, one interviewee noted:

"I think that it would be pretty high. They just need to have very compelling evidence. …For the infant feeding policy, I think what really moves things is when you have the medical director buy-in. If the medical director is onboardwhich he is for this, and so I would say
that I have high support from him -then things move through faster, too."

C.iii. Tension for Change
Tension for change describes the circumstances that staff and leadership perceive a cause for changesuch as an undesirable or even "intolerable" situation.
Tension for change was a critical facilitator for the adoption of policies, especially in organizations with a negative power balance or neutral support from leadership. Interviewees often described tension for change as motivation spurred by champions in positions of power and external mandates or incentives. As noted previously, clinics affiliated with larger health care organizations that were pursuing Baby-Friendly status were motivated to adopt and implement breastfeeding policies by the situation of potential policy discrepancies.

D. Communication & Individual Influence for Adoption & Implementation
The method in which the innovation is diffused or disseminated is defined as communication and influence in the Greenhalgh model. Modes of communication may vary in formality but are present throughout the policy process.
Boundary spanners, change agents, and champions were critical influences for the adoption and implementation of breastfeeding policies. The level of impact by specific persons was related to their social network, prior knowledge, and level of power within the organization. As noted above, providers and staff who rotated between hospitals and clinics were often identified as boundary spanners and facilitators for clinical breastfeeding policies and practices. An interviewee captured their role as both a champion, expert, and boundary spanner within the regional healthcare community when they stated,

"Well, one of the things that I do, because everybody knows that I'm interested in breastfeeding, I'm sort of just the person, you know? If the word breast comes up in a conversation, they all look at me! In some ways, just being me brings breastfeeding into the conversation… I do presentations and journal clubs and teach the fellows. I do a lot of education, formal and informal."
Champions not only laid the groundwork for organizational adoption, but they also supported implementation by providing expert opinion, technical support, and a bridge for healthcare community partnerships.
Furthermore, while champions often laid the groundwork for their clinic's progression towards adoption, the organization's structure and level of readiness also contributed to this complex process. For instance, an interviewee described the intersection of their pre-existing knowledge and passion with the organization's interest to pursue the Baby-Friendly status:

E. Outer Context and Influence in Breastfeeding Policy Adoption
The outer context includes the sociopolitical climate, incentives and mandates, and interorganizational norm setting that influence an organization.
Mandates and incentives from accrediting bodies, professional organizations and federal/state laws strongly influenced the adoption of breastfeeding policies in clinics. The Joint Commission (JCAHO) and the Special Supplemental Program for Women, Infants and Children (WIC) were frequently referenced as levers that shifted the momentum for breastfeeding policy adoption.

"I think that every JCAHO measure is at the very top of the department, the relevant departments' priority list and probably everything else falls under that. I think that will be what it takes for everyone to feel like breastfeeding is super important. Right now I think that everyone feels like it's probably of average importance."
As noted throughout the results, the Baby-Friendly Hospital Initiative (BFHI) alone acted as a catalyst for policy adoption in clinics affiliated with larger healthcare organizations. Of note, this facilitator for adoption did not ensure effective implementation. To that end, technical-and financial support were also key facilitators during the policy adoption process. For example, one interviewee from a participating WIC clinic described the impact of financial-and technical assistance during policy development:

I. Summary of Barriers and Facilitators
Our study illuminated factors that support or hinder breastfeeding policy adoption and implementation in clinics. These elements ranged from individual perceptions about the need for breastfeeding policies in general to organizational characteristics and external, societal, and governmental influences. Interviewees expressed hesitation towards policies they perceived to be "too rigid" or too much of an interference with a mother's choice. At the organization level, clinics often faced competing interests when prioritizing breastfeeding policies, thus limiting momentum for policy adoption. This was especially true in organizations with limited motivation for change. A negative or neutral power balanceparticularly from leadershipdelayed breastfeeding policy adoption and successful implementation. Insufficient resources were often cited as a major drawback for leadership support, policy development, and staff training. Furthermore, lack of adequate training and technical assistance hindered effective breastfeeding policy implementation.
Breastfeeding policies that reflected an organization's mission, goals, and values were viewed positively by leadership and staff. Breastfeeding policy adoption was facilitated in clinics connected with larger organizations when breastfeeding was prioritized from a top-down approach (i.e. WIC or BFHI). Other enabling factors included support form persons or groups in power, considerable pre-existing staff knowledge, and active networks for knowledge-sharing. To that end, champions and boundary spanners facilitated breastfeeding policy adoption and provided technical expertise for effective implementation. Tension for changeresulting from positive power balances and/or external influenceswas a tipping point for leadership buy-in. Moreover, external influences, including mandates and incentives from governing or accrediting bodies, formed a catalyst for clinic policy adoption because of the strong incentives for compliance.

II. Critical Engagement with Literature
Our findings compliment previous breastfeeding policy literature, including research in both the clinic and hospital setting. Similar to previous studies, we identified limited financial resources 35 , inadequate staff training 19,37 , and provider concern for limiting maternal choice 35 as barriers to policy adoption and implementation. Established facilitators included adequate staff training 28 , engaged champions 35 , collaborative healthcare networks 19 , and sufficient leadership buy-in. 37 Our findings expand upon the unique role of external influences as an impetus for policy adoption. This was particularly influential for clinics that were part of larger healthcare systems and/or the WIC program. Given the competing priorities in these organizations, mandates or incentives acted as a catalyst to move the policy adoption process forward. 42 Notably, sufficient resources for development and training, often provided by external funding, were also important to facilitate effective policy implementation. 19,28

III. Greenhalgh Theory
The Greenhalgh framework is a useful tool to understand the complex dynamics of breastfeeding policy adoption. By focusing upon factors at the organization-level, we were able to draw a more complete picture of the complex, varied policy process in healthcare settings. Furthermore, the framework highlighted the intricate links between an organization's structure, knowledge-base, administrative priorities, and external influences and can be used to gain a deeper understanding of leverage points to support comprehensive breastfeeding policy adoption.

A. Research
Our study offers a descriptive account of the current breastfeeding policy process in Washington State clinics. These findings could inform future policy research, including intervention models that strengthen facilitators and reduce barriers outlined in our study. By focusing upon these factors, research could further refine strategies to expand breastfeeding policy adoption and implementation in clinics. Studies could also incorporate a larger sample size and blend policy process research with quantitative breastfeeding outcomes to advance understanding of the relationship between the development of supportive breastfeeding policies and breastfeeding exclusivity and duration rates.

B. Practice
In the U.S., some families face substantial obstacles achieving their breastfeeding goals. 19,21,43 This often perpetuates disparities in breastfeeding rates and health inequities. [11][12][13][14] Supportive breastfeeding policies and practices ameliorate structural barriers and improve consistency of breastfeeding care. This is especially effective when healthcare systems coordinate supportive breastfeeding practices (i.e. coordinated care) 18 30 have each called for increased collaboration across the continuum of breastfeeding care. Nevertheless, there have been limited efforts to enhance breastfeeding care coordination amongst hospitals, primary care clinics and community organizations. 18 The following are recommendations directed towards a) the healthcare sector and b) public agencies and mandating organizations to improve the adoption and implementation of breastfeeding policies both within clinics and across the healthcare system.

B.i. Healthcare Sector
The healthcare sector can facilitate breastfeeding policy adoption by engaging in current policy initiatives, ensuring adequate staff training, and championing knowledge-sharing networks. Our study highlights the unique impact of the BFHI upon clinic-based services. Though the U.S. Baby-Friendly initiative 46 emphasizes hospitals and birth centers, we observed opportunities to leverage BFHI-type policies for improved breastfeeding care coordination and provider training. By pursuing the Baby-Friendly status, hospitals can support neighboring clinics and improved care coordination. Clinic administrators can apply specific recommendations and protocolssuch as the Academy of Breastfeeding Medicine Protocols 44,47,48 to advance breastfeeding support within their clinics. Clinic leadership can also facilitate breastfeeding policy adoption and implementation by prioritizing staff training and engagement in knowledge-sharing networks. 36,43 Given the discontinuity of breastfeeding care, collaboration across health care professions and settings contributes to successful policy implementation. 18 To that end, clinic and hospital healthcare providers can act as boundary spanners to bridge gaps in communication and knowledge-sharing. They can also act as champions to promote supportive breastfeeding practices and provide expert opinion during policy development and implementation. [49][50][51]

B.ii. Public Agencies & Accrediting Organizations
Public agencies and accrediting bodies can enable policy adoption and implementation through mandates, incentives, and technical assistance. Initiatives such as the Affordable Care Act's mandate for comprehensive lactation support 52,53 and the Pediatric Quality Measure Program's 54 quality indicator for postpartum care coordination act as catalysts for leadership buy-in and help to bridge the gap in breastfeeding care. As noted previously, leaders in larger organizations are particularly responsive to mandates and incentives. 42 Leveraging these points of motivation could improve care coordination. 18 Nevertheless, breastfeeding interventions in the primary care setting have been described as labor intensive and poorly reimbursed. 55 Healthcare reform models, 56 such as accountable care organizations (ACOs) 53 and patient-centered medical homes (PCMHs) 57 provide opportunities to improve the quality and continuity of care between prenatal, perinatal, and postpartum settings.
In addition, the U.S. BFHI and similar state-led initiatives 58 could be expanded to incorporate clinics and other sectors of the breastfeeding continuum of care. 38,59 Though mandates and incentives are key motivators for policy adoption and implementation, these levers do not ensure successful policy implementation. Adequate infrastructure and technical assistance are also key components of policy implementation. Feldman-Winter et al., captured the importance of technical assistance for leadership when describing the CDC National Institute for Children's Health Quality "Best Fed Beginnings" (BFB). 42,60 This quality improvement initiative included a threeyear collaboration to provide technical assistance to leaders in 89 hospitals pursuing Baby-Friendly designation. Though the intervention was designed for the hospital setting, this strategy could be translated to the clinical setting. 35,36 This and similar initiatives highlight the value of technical support for organization administrators throughout the adoption and implementation process. 35,36,42 V.

Limitations & Strengths
There are several limitations that impact the generalizability of this study. First, the initial study used purposive sampling to recruit participants. Since participants were referred by the study's advisory board members, participants likely had previous exposure to supportive breastfeeding policies. To that end, our study was limited by the geographic spread of the clinics, many of which were located in an urban setting; experiences in other clinical settings may not be identical to our findings. Despite these limitations, there is still great value in applying a diffusion of innovations framework to better understand specific organizational factors that help or hinder breastfeeding policy adoption and implementation in clinic settings. Furthermore, findings presented here capture a variety of clinical settings and organizational structures -ranging from rural, public health or WIC clinics to urban, multi-level healthcare organizations.

CONCLUSION
The systematic evaluation of dynamic breastfeeding policy adoption and implementation provides valuable insight into the drivers and obstacles of policy development. Clinics play an integral role in the breastfeeding continuum of care and would benefit from further policy process research, inclusion in ongoing breastfeeding initiatives, and adequate financial and technical support. .