|Home | About | Journals | Submit | Contact Us | Français|
The Electronic Communications and Home Blood Pressure Monitoring trial (e-BP) demonstrated that team care incorporating a pharmacist to manage hypertension using secure E-mail with patients resulted in almost twice the rate of blood pressure (BP) control compared with usual care. To translate e-BP into community practices, we sought to identify contextual barriers and facilitators to implementation.
Interviews were conducted with medical providers, staff, pharmacists, and patients associated with community-based primary care clinics whose physician leaders had expressed interest in implementing e-BP. Transcripts were analyzed using qualitative template analysis, incorporating codes derived from the Consolidated Framework for Implementation Research (CFIR).
Barriers included incorporating an unfamiliar pharmacist into the health care team, lack of information technology resources, and provider resistance to using a single BP management protocol. Facilitators included the intervention’s perceived potential to improve quality of care, empower patients, and save staff time. Sustainability of the intervention emerged as an overarching theme.
A qualitative approach to planning for translation is recommended to gain an understanding of contexts and to collaborate to adapt interventions through iterative, bidirectional information gathering. Interviewees affirmed that web pharmacist care offers small primary care practices a means to expand their workforce and provide patient-centered care. Reproducing e-BP in these practices will be challenging, but our interviewees expressed eagerness to try and were optimistic that a tailored intervention could succeed.
Almost 1 in 3 adults in the United States has hypertension (HTN),1,2 a sustained blood pressure (BP) of ≥140/90 mm Hg. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High BP cites that a population-wide 5-mm Hg reduction in systolic BP would result in a 14% overall reduction in stroke, a 9% reduction in coronary heart disease mortality, and a 7% decline in all-cause mortality.3 Despite these treatment benefits, BP remains uncontrolled in almost half of those with HTN.4
There is strong evidence that including a nurse or pharmacist as a team member in the management of HTN improves BP control.5–9 The Electronic Communications and Home BP Monitoring (e-BP) study demonstrated that team care incorporating a pharmacist who intensified BP management strategies (eg, medications) through secure web messaging with patients resulted in almost twice the rate of BP control compared with usual care.10 However, e-BP was implemented in a large integrated group practice with in-house pharmacists and an existing patient-shared electronic health record (EHR), conditions that are not typical in community practices. In this study we sought to identify barriers and facilitators to implementing web-based pharmacist team care (web pharmacist care) in community practice settings so we could plan for an adaptation of the e-BP intervention.
Context matters when transferring evidence-based practices from one setting to another and when translating research evidence into everyday practice.11–13 Accordingly, planning for the translation of e-BP required assessment of the resources, needs, and preferences of the communities that had expressed interest in implementing this model of care.14 We interviewed stakeholders associated with interested community practices to better understand the challenges and opportunities of implementing e-BP in settings with contextual features very different from those of the integrated group practice for which it was originally designed. This article provides a rationale for our approach and a summary of what we learned.
Interviews were conducted in 1 Washington and 2 Idaho communities with populations between 50,000 and 90,000. These communities include practices that are members of the WWAMI (Washington, Wyoming, Alaska, Montano, Idaho) region Practice and Research Network (WPRN), a practice-based research network in the Pacific Northwest. Leaders from the WPRN member practices recommended 4 community-based primary care practices using EHRs that they believed might be willing to provide input on the translation of e-BP into community settings. All 4 practices agreed to participate. They each had between 5 and 10 medical providers, the majority of whom were family physicians; the remainder were advanced registered nurse practitioners and physician assistants. Provider practices were supported largely by medical assistants (MAs) or licensed practice nurses (LPNs). Only one clinic had a registered nurse. Two practices had clinical pharmacists who provided some clinical services such as diabetes education.
The research team interviewed medical providers (n = 8), clinical staff (LPNs and MAs; n = 9), pharmacists (1 hospital pharmacist, 3 independent pharmacists, 1 chain pharmacist, and 1 practice-based pharmacist), and patients (n = 12) associated with the 4 community-based primary care clinics whose physician leaders had expressed interest in implementing e-BP. Patients with HTN were identified through flyers placed in clinics and public places (eg, senior centers); 4 patients were 24 to 54 years old and 8 were 55 to 68 years old; there were 5 men and 7 women. The institutional review boards of the Group Health Cooperative, the University of Washington, and Idaho State University approved all study procedures.
We developed semistructured interview guides specific to the type of interviewee (eg, medical providers, pharmacists, patients) based on the Chronic Care Model (CCM), the implementation model on which the e-BP study was based.15 The CCM provides a conceptual framework for improving chronic health care by optimizing 6 key elements: improved health system and organizational support, self-management support, delivery system design, decision support, clinical information systems, and integration with community resources.16 CCM-related questions explored interviewees’ experiences with and attitudes about, for example, (1) the design of the delivery system (eg, How do you feel about working with other health care providers [eg, registered nurses, MAs] using protocols that you approve to adjust antihypertensive types and dosing, as was done in the e-BP study? How can you envision collaborating with these health care providers in managing your patients’ hypertension?); (2) self-management support (eg, What are your thoughts about patients taking their own blood pressures?); and (3) clinical information systems (eg, What are your thoughts about the idea of pharmacists using E-mail or web sites to communicate with patients about HTN management?) Follow-up probes were tailored to different professional groups and patients and were updated as emergent themes informed subsequent interviews.
We created a pictograph that was used at the beginning of each interview to orient participants to the general concept of delivering web pharmacist care (Figure 1). Interviews lasted between 60 and 90 minutes for medical providers, staff, and pharmacists; patient interviews were approximately 30 to 40 minutes in duration.
Template analysis, as defined and developed by King,17 was employed to code the interview transcripts. This qualitative approach to analyzing text results in the development of a template, or a list of codes, comprising broad to more narrowly focused thematic categories. It is distinguishable from other types of text analysis in that it is not rooted in any single epistemology, and it is common to define a priori codes strongly expected to be relevant to a study’s aims and objectives.18 Like other types of text analysis, the analytic process is iterative. Investigators carefully read the text, separating data from the original context of individual cases and assigning codes to units of meaning in the text, then examine the codes for patterns and reorganize the data around central themes and relationships drawn across all the cases.19
Two of the investigators (LSR, EJ) independently coded a subset of interview transcripts, meeting periodically to adjudicate coding differences and create a consensus template (Table 1). LSR completed coding using the template, which incorporated categories drawn from 2 conceptual frameworks: the CCM and the Consolidated Framework for Implementation Research (CFIR).20 The CCM categories, which had guided the construction of interview protocols, provided a starting point for coding interviewee’s descriptions of the current state of HTN management in their communities, including, for example, how providers and patients were currently managing and communicating about patients’ BPs. Additional codes were needed, however, to capture interviewees’ descriptions of how HTN management might look in the future—that is, if e-BP were implemented—and to categorize the types of “barriers” and “facilitators” to implementation associated with stakeholders’ beliefs about the intervention, the different clinics’ readiness to implement, and the influence of governmental policies and regulations on the ability to implement e-BP in communities. For this, the CFIR, which “specifies a list of constructs within general domains that are believed to influence (positively or negatively) implementation,” proved useful.20 The CFIR provided a means for organizing interviewees’ perceptions of (1) the characteristics of the intervention, including the relative advantages of implementing web pharmacist care versus an alternative solution; (2) the “outer setting” factors that might affect implementation, including each clinic’s economic, political, and social context and relevant governmental policies and regulations; (3) the “inner setting” factors that might influence implementation, including each clinic’s norms and values, and the level of resources dedicated for implementation and ongoing operations, including money, training, education, physical space, and time; and (4) the characteristics of individuals affecting implementation, including knowledge and beliefs about the intervention and beliefs in their own capabilities to perform the actions necessary to achieve implementation goals. AtlasTi, a qualitative data management program (ATLAS.ti Scientific Software Development GmbH, Berlin, Germany), was used to facilitate the organization, management, and coding of qualitative data.
Participants freely shared opinions about the short-comings of the current health care delivery system, the fit between web pharmacist care and their aspirational views of patient-centered health care generally and BP management specifically, and their perceptions of the barriers and facilitators to implementing web pharmacist care. The coding template illustrates how e-BP components mapped onto CCM categories and how CFIR-defined domains provide a means of classifying where in the sociocultural milieu barriers and facilitators arise (eg, “outer setting”, or societal “inner setting”, or clinic level, or “individual level”) (Table 1). Table 2 provides illustrative examples of stakeholders’ quotes verbatim, arranged thematically and by CFIR domains.
Home BP monitoring was attractive to everyone interviewed. According to the CFIR framework, this intervention characteristic would be considered a facilitator to implementation because stakeholders valued its potential to improve health care quality and empower patients. (Table 2). However, providers expressed concerns about the affordability of BP cuffs, the accuracy of patients’ home BP readings, and the perceived lack of competence with information technology among older patients in their communities. These potential barriers would be categorized as CFIR outer setting factors related to patient resources (as a whole) in the community. Patients described concerns about their ability to accurately take or interpret home BP readings (a CFIR individual-level barrier related to self-efficacy), although all expressed willingness to perform home BP monitoring and most were already doing so.
Because few community-based primary care clinics employ a clinic-based pharmacist and because patients fill their prescriptions at many different pharmacies, we solicited stakeholders’ opinions about using clinic staff (eg, LPNs and MAs) rather than pharmacists to collaboratively manage BP. Staff reported they had worked with medication protocols (eg, for urinary tract infections and strep throat) in the past but voiced discomfort about taking on the additional responsibility of collaborative BP management because they were already stretched too thin with current duties (CFIR individual-level and inner-level clinic resource barriers).
Providers, especially those with access to an in-house pharmacist, felt pharmacists were well suited to managing HTN because of their in-depth knowledge of medications and medication management. They indicated that pharmacists would be the best choice for team BP management if they could be as readily accessible as clinic staff. Providers specified trust as the essential criterion for selecting a pharmacist team member, expressing reticence to work with a pharmacist with whom they had no established relationship.
Providers and pharmacists universally described many attractive features of web pharmacist care, including patient convenience, saving staff time, additional attention and support for patients, and long-term cost savings (facilitators related to CFIR intervention characteristics). Pharmacists described themselves as uniquely qualified to be members of the health care team because patients trust them and because they are more accessible than physicians.
Both pharmacists and providers highlighted 2 barriers to implementing web pharmacist care: the lack of financial incentive to practice collaboratively (CFIR outer setting factor), and the difficulties of implementing collaborative care among physicians who believe they need to make all decisions about patient care on their own (CFIR individual-level factor).
Patients expressed enthusiasm for web pharmacist care, noting that the team approach is empowering, affords convenient access to a medical professional, provides closer BP monitoring, and promotes long-term cost savings (CFIR intervention characteristics). A couple of patients voiced discomfort about working with a pharmacist they had not met face to face, and several expressed concerns about team care (CFIR individual-level factors). These latter concerns were mitigated by the trust they put in their physicians to incorporate a trustworthy team member.
One of the clinics had a web-based patient portal (a secure site where patients could log on to communicate with their health care providers or use other functionalities such as making an appointment). Others were planning to add this as one mechanism for meeting meaningful use. Providers and pharmacists viewed secure E-mail very positively, noting its potential to increase efficiency of communication with patients, facilitate documentation of information exchanges, and improve quality of care by extending interactions beyond the limits of the clinic visit (CFIR intervention characteristics). Opinions were mixed about whether patients’ lack of access or ability to use the web would present a major barrier. From the providers’ perspective, one important barrier to implementation is the current lack of financial incentive to use secure E-mail (CFIR outer setting). Some clinic staff expressed reservations about patient E-mail because of worries that “needy” patients would take advantage of the ease of accessibility and substantially increase their workload as well as that of the providers (CFIR individual characteristics). Others felt that E-mailing was no more intrusive than telephoning and would have a neutral effect on workload. Many regarded secure E-mail as a means of reducing patient costs, increasing accessibility to patients, and promoting patient empowerment and adherence to treatment (CFIR intervention characteristics).
Pharmacists and clinic staff generally endorsed the use of protocols to guide medication management, although one pharmacist expressed nervousness about adding calcium channel blockers specifically (CFIR individual self-efficacy). Pharmacists noted that protocols provide protection against liability (CFIR outer setting), while staff commented on the advantages of using protocols to standardize care (CFIR intervention characteristics). Providers wanted assurance they would have substantial input into the development of the protocol and that any protocol developed would protect their patients from medication errors (CFIR individual characteristics).
All the clinics had EHRs, although none were positioned to use them to electronically identify, track, and manage patients with uncontrolled HTN as proposed (CFIR inner setting barriers). However, clinics were interested in using their EHRs as registries to manage chronic conditions like HTN and saw this as one mechanism for meeting meaningful use.
Concern about sustaining the proposed intervention once extramural funding had ended emerged as an overarching theme. One provider explained, “I’m just projecting ahead that if I’ve got 5 clinics … and each clinic has devoted a quarter time person … then I’ve got to suddenly defend $60,000 in expenditures or more; that poses a challenge.” Stakeholders repeatedly expressed frustration about sustaining web-based team care in a fee-for-service environment. They were critical of reimbursement policies, noting that services most beneficial for patients, such as working with pharmacists who can provide education about how medications work to control their conditions, are not eligible for reimbursement. However, at least one medical provider was hopeful that reimbursement for this kind of care will be the “the wave of the future.”
Despite pharmacists’ enthusiasm about web pharmacist care, retail pharmacists wondered how they would find time and money to support or sustain its implementation. A pharmacist who reported filling up to 535 prescriptions a day stated this “doesn’t lend itself to extra time.” At least one pharmacist hoped that participating in the community-based intervention would yield evidence demonstrating the value of web pharmacist care and serve as a basis for reforms to sustain this evidence-based, patient-centered practice.
Translating evidence into practice requires investigating real-world settings to identify the many contextual variables that will influence the implementation process. Our qualitative approach to planning for implementation aligns with recommendations to anticipate and work to address likely barriers, identify and work with key community stakeholders to gain a thorough understanding of their contexts, and collaborate to design/adapt interventions based on iterative, bidirectional information gathering.21,22 This qualitative case study was intentionally limited in scope to enable in-depth investigation of stakeholders’ readiness to collaborate in the implementation of an evidence-based intervention. Because of the small samples of key staff, providers, pharmacists, and patients interviewed, our findings may not generalize across all practice settings.
Collecting qualitative contextual data about the primary care practices and communities interested in implementing a CCM-based model of BP control, however, illuminated intervention design features requiring adaptation. For example, not all practices had an in-house pharmacist who could be easily incorporated into a practice team to support BP management—a key feature of the e-BP delivery system. Because some clinics lacked this inner setting resource, we explored stakeholder preferences for restructuring care using different team members, such as clinic-based medical assistants, but lack of time (CFIR inner setting) and the confidence of MAs (CFIR individual level) were barriers. We learned that although most practices had little opportunity to directly engage pharmacists in patient care, stakeholders valued their expertise and were interested in exploring ways to incorporate community pharmacists into the health care team. Stakeholders did express concern about working with pharmacists they may never have met face to face or with whom they had no established relationship. As reported in previous research, facilitating the building of trust in care teams that incorporate community-based pharmacists or operate outside of an integrated system is a challenge and needs to be addressed.23–25 In addition, although an initial prerequisite for participating in the e-BP community implementation was having an existing EHR, at the time of our interviews only one site had an existing patient web portal with secure E-mail. Other technological options for supporting secure, asynchronous communication between patients and pharmacists will need to be found or technical assistance will need to be provided to help clinics set up patient web portals.
Even at this early stage of exploration and planning sustainability arose as an issue of concern. Sustaining web pharmacist care once grant funding had ended was a key concern for physician leaders. Our interview findings suggested several potential mechanisms for sustainability and future dissemination. For example, several clinics interested in participating in web pharmacist care were affiliated with a hospital that is developing a clinically integrated network (similar to an accountable care organization), which rewards team-based care models that improve quality and efficiency.26 Some were planning to participate in pay-for-performance initiatives, which financially reward physicians who meet quality targets such as controlled BP rates among patients with HTN.27 Pharmacists already can bill for Medicare Part D medication management,28 and this model could be expanded to include chronic conditions.
The incorporation of CCM and CFIR domains into our coding template provided a ready-made framework of constructs and definitions for organizing information relevant to the implementation of each intervention component and identifying and categorizing the kinds of implementation barriers we faced (eg, individual, institutional, societal). The main benefit of approaching a qualitative data set with a set of a priori constructs at hand is that it can speed up the initial coding process. The drawbacks are that by attending to predefined constructs, researchers may overlook material that does not fit neatly into them, or the predefined constructs may not be the best way of characterizing the data.29 Use of the CFIR and CCM conceptual frameworks did not constrain our ability to capture the unique observations made by our stakeholders or identify emergent themes; in fact, the approach holds promise as a model for programmatic study of the factors influencing the translation of similar health interventions into community-based practices. As observed by Feldstein and Glasgow,30 developers of the Practical, Robust Implementation and Sustainability Model (PRISM) for integrating research findings into practice, it is the absence of conceptual frameworks in implementation research that has impeded progress in improving program implementation.
The e-BP study demonstrated that team-based care could be delivered using secure E-mail connected to an EHR. Interviewees affirmed that web pharmacist care offers a promising strategy for small primary care practices to expand their workforce and provide patient-centered care, and they expressed eagerness to implement it in their settings. They characterized web-based collaborative health care as “the right thing to do” and expressed hope that participation in a successful implementation would yield hard evidence to support significant changes in the way that health care is delivered and reimbursed. Reproducing e-BP in small primary care practices will be challenging, but our interviewees expressed optimism that a tailored intervention would succeed.
The authors thank Denise Mia Lishner, MSW, for her assistance in copyediting the manuscript.
Funding: This work was completed with the support of grant 1UL1RR025014 from the Clinical and Translational Science Award program of the National Center for Research Resources, National Institutes of Health (NIH), and the NIH Roadmap for Medical Research, as well as funds from the Group Health Research Institute Director’s Fund.
Conflict of interest: none declared.