Examining the feasibility and characteristics of realistic weight management support for patients: Focus groups with rural, micropolitan, and metropolitan primary care providers

The purpose of this investigation was to understand perspectives of physicians, nurses, and staff regarding the feasibility of implementing an evidence-based weight management program to support primary care practice. An exploratory aim was to examine differences in responses based on the clinic location. Ten focus groups were conducted with primary care staff from rural, micropolitan, and metropolitan clinics. The Promoting Action on Research in Health Services (PARIHS) framework was used to inform the interview guide. Transcripts were reviewed to identify common themes among PARIHS constructs (evidence, context, and facilitation). Presence of comorbidities (e.g., diabetes, hypertension) were typical prompts for provider-led discussions about patient weight. Metropolitan clinics reported the availability of health coaching, diabetes education, or dietician consultation, but no clinic reported offering a comprehensive weight management program. Participants agreed it is possible to implement a weight management program through primary care, but cited potential facilitation challenges such as costs, clinic resources, and individual patient barriers. More enthusiasm arose for a referral program with patient tracking. Program characteristics such as proven efficacy, individual tailoring, program accessibility, and patient feedback to the providers were desired. Rural focus group participants reported unique barriers (lack of local resources) and facilitators (more flexibility in practice changes) to weight management when compared to metropolitan and micropolitan focus groups. Primary care staff are interested in weight management solutions for their patients and would prefer an evidence-based program to which they could refer patients, receive feedback on patient progress, and sustainably include as part of their regular services.


Introduction
Primary care practices have been identified as a potential location for intensive weight management interventions (WMI) to reduce obesity (Katzmarzyk et al., 2020;Wadden et al., 2020;Tronieri et al., 2019). This may be particularly attractive in rural areas where resources for promoting healthful eating and physical activity are scarce, obesity rates are high (Ogden et al., 2014;Perri et al., 1993), and primary care clinics are often the only available resources to support weight management (Phillips et al., 2014). Indeed, primary care may provide an opportunity to develop practical and sustainable systems-based approaches to address obesity (Glasgow et al., 2003(Glasgow et al., , 2004 by 1) leveraging the patient-provider relationship to support program initiation and retention, 2) engaging organizational decision-makers, such as clinic medical directors and office administrators, to improve sustainability (Huang et al., 2011), and 3) using the human and technological resources of the primary care system to implement manageable change (Williams and Hummelbrunner, 2019). While there is growing evidence in the effectiveness of intensive WMI implemented through primary care, few have been systematically and consistently translated into typical clinical practice or community services (Estabrooks and Russell, 2006;Akers et al., 2010;Green et al., 2009).
The lack of translation of evidence-based WMI into primary care settings could be related to the lack of provider time to critically review the literature and identify an appropriate intervention, a mismatch between infrastructure and program intensity, and a lack of support in the initial adoption and implementation of a new intervention. Furthermore, issues related to the research to practice translation may also differ between rural and urban primary care settings where the availability of local resources differs significantly. The purpose of this research was to understand preliminary feasibility of implementing an evidence-based WMI through primary care, concentrating on factors related to the patient (e.g. patient identification and engagement, weight loss initiation and maintenance), clinical adoption (e.g. scalability, clinical uptake), and sustainability, all of which are key for the translation of evidence-based interventions into practice (Estabrooks and Russell, 2006;Leeman et al., 2017Leeman et al., , 2015Chorpita et al., 2005). Secondary aims were to explore the successes and challenges of weight management programming that may already be present in clinics and to explore differences among clinics serving rural/micropolitan (cities with a population < 50,000) and metropolitan areas.

Methods
Ten focus groups were conducted with staff from primary care clinics of the Nebraska Practice-Based Research Network (PBRN) between November 2016 and October 2017. Clinic managers and/or head physicians of each clinic were contacted by the physician leader of the PBRN and asked if they and their staff would be willing to participate. Focus groups were grouped by clinic site to facilitate candid discussion among individuals who were familiar with one another, and to discern the contextwithin a given clinicrelated to supporting patient weight management. We continued to recruit clinics until focus group responses reached saturation. All contacted clinics agreed to participate and reflected approximately two thirds of PBRN member clinics (i.e., 10 out of 15 clinics). Four clinics were located in rural areas (<10,000 residents), one clinic was located in a micropolitan area (urban core of at least 10,000, but less than 50,000 residents) (OMB, 2010) and the remaining five participating clinics were in a metropolitan area (468,262 population; U.S. Census Bureau). Of note, four of the five clinics located in a metropolitan area were members of the same health system. All other clinics came from different health systems.

Participants & conceptual framework
A total of 51 individuals participated in the 10 focus groups, with an average of five participants in each focus group (range: n = 2-12). Typically, one to two physicians and one to two nurses from each clinic participated, in addition to other personnel (Table 1). Table 2 provides an overview of the semi-structured interview guide and questions aligned with the Promoting Action on Research in Health Services (PARIHS) framework (Kitson et al., 1998;Stetler et al., 2011). This conceptual framework consists of three interacting elementsevidence, context, and facilitationthat influence the adoption and implementation of evidence-based practices (Rycroft-Malone, 2004). These elements and their sub-elements were used to create prompting questions regarding the feasibility of implementing a WMI through the clinic including patient screening and referral, WMI delivery, clinical adoption, and sustainability. Questions related to current practices and structural characteristics were included to gain a better understanding of local practice information, context, and culture of weight management in Great Plains primary care. Focus groups concluded with a brief presentation of three specific evidence-based WMI (Appendix A) (Wyatt et al., 2008;Perri et al., 2008;Almeida et al., 2014). These WMIs were selected from a systematic review of rural WMI (Porter et al., 2019) for their demonstrated effectiveness, potential for delivery in a rural area, and for their variety of delivery methods, so as to gain an understanding of clinic staff's perceptions of a wide variety of program characteristics. This study was approved by the Institutional Review Board at the University of Nebraska Medical Center and informed consent was obtained from all participants.

Data analysis
Focus groups were audio recorded and transcribed verbatim, separated into meaning units, and assigned a descriptive code by two to three independent coders (Elo and Kyngäs, 2008;Graneheim and Lundman, 2004). Once completed, each coding pair/trio met to reconcile any coding differences by consensus. Additional coders were consulted if a consensus could not be reached. Common meaning units across questions were deductively grouped together by codes and a nested hierarchy of meaning units were grouped into themes relevant to the PARIHS framework constructs (i.e., context, facilitation, and evidence). Table 3 provides a sampling of quotes and key takeaways related to these constructs. Note: Other clinic staff included medical assistants, a phlebotomist, a pharmacist, and behavioral health staff. • Role of facilitator: If you decided to make the weight loss program a permanent fixture at your facility, who would be the person in your clinic to champion a program like this?
• Role skills and attributes: What are some of the skills a facilitator would need to get a weight loss program up and running at your site?
• Other implementation interventions: What kind of changes would your organization need to make to the roles for staff so that a weight loss program could fit with the other services you provide?

Context-Current practice in primary care
3.1.1. Weight measurement & evaluation All clinics measured and tracked weight at every patient visit using an electronic medical record (EMR) system, which automatically calculated patient BMI. Most EMR systems alerted clinic staff if BMI was in the obese range or if there had been a significant change since the patient's last visit (e.g. change in font color, a pop-up window, an automatically generated email to the provider). Some systems had a BMI quality measure, which included forced entry fields for the provider to document if a high BMI (>29.9 kg/m 2 ) was addressed during that patient visit. Participants acknowledged the capabilities of their EMR systems, but responses were inconsistent across clinics regarding if and how EMR alerts or quality measures prompted action during a patient encounter. However, there was consistency on the issue of "click fatigue," (Collier, 2018) particularly related to forced pop-ups and check boxes related to weight, BMI, and many other clinical notifications. Participants stated that although well intended, the weight/BMI quality measure did not always have the intended effect in practice. As one physician noted, "I don't have more time in my day, and you just gave me more to do." No differences were observed by clinic locations regarding weight and BMI tracking, EMR alerts, or "click fatigue".

Culture of weight management support for patients
Three major themes emerged regarding current clinical culture surrounding weight managementproviders delivered health education, medication, and referral to an outside program. No participants reported having a formal WMI available at their clinic, regardless of location.
Provider health education included informal discussions led by the provider during a patient visit. Two triggers were identified for these conversations: visit type and patient comorbidities. Patient weight was generally addressed only at health maintenance visits or when weight was pertinent to the reason for the visit, regardless if an EMR prompt was given. For example, one physician described how they initiate a weight discussion at a health maintenance visit in this way: "So somebody who is diabetic, hypertensive, and especially the one thing I always look for -that have osteoarthritis symptoms…so when you see a few things going together, and it's one of those things like hey, okay, you just complained to me about your chronic knee pain, you're a hypertensive diabetic and your BMI is 32. Well I start building that whole picture." Similarly, most physicians reported using weight-related comorbidities as cues to introduce a discussion about patient weight and to illustrate to patients how weight impacts those comorbidities. Physicians reported limited use of BMI during patient discussionstypically only when the patient understood what BMI represented or when it could be used as a "road map" to draw a patient's attention to their weight as a concern.
The focus of these discussions generally centered on patient eating Preferred characteristics of a weight management program included patient and provider feedback on progress, maintaining patient accountability Program characteristics that would influence a decision to adopt a weight management program included cost, program efficacy, program objectives, frequency of program interactions, availability of program offerings, and delivery method. and physical activity behaviors and goal setting. Some providers used an informal, conversational approach to patient goal setting and overcoming barriers, while others used a formal goal setting tool that is built into the EMR system, which, in some cases, could be viewable and editable by the patient. In rare cases, physician-initiated goal setting prompted further action by clinical staff such as a health coach or care coordinator. Other clinic staff also described how they talked with patients about their health behaviors and how those behaviors affect weight. These interactions happened during the patient visit or were scheduled as a follow-up appointment with the specific staff member (e. g. health coach, coordinator). Most participants also reported they had referred their patients to a dietician. In rural/micropolitan clinics, these dieticians were employed by local grocery stores, while metropolitan clinics mentioned nutritionists or dieticians employed by the clinic or their parent healthcare organization. Metropolitan clinic participants mentioned that their nutritionists/dieticians were often covering multiple clinics at a time, which made it difficult to schedule patients and nearly impossible to monitor patient progress after the referral was made. One metropolitan clinic staff participant said that their nutritionist was the "highest to have a no-show" out of all the providers at that clinic, "probably because she's (nutritionist) forced to schedule [appointments] so far out." Prescription weight loss medication and referral to bariatric surgery were reported as additional methods to treat overweight/obesity among patients. Providers rarely used medication as monotherapy, but all reported prescribing weight loss medication to some degree, primarily after a patient requested it. All participants reported short-term success of weight loss medication among highly motivated patients, but reported scarce long-term success and showed little enthusiasm for weight loss medication, presumably due to unrealistic patient expectations. As one health coordinator said "When they realize that the pill isn't going to be magic, they didn't lose their 20 lb the first month, they go off of it and they haven't made the lifestyle changes, they gain it all back." Participants reported long-term success among patients that were referred for bariatric surgery but noted this was only a small proportion of their patients. No differences were observed in responses from participants by clinic locations regarding provider health education prompts and content, care coordinator/health coach presence, or for weight loss medication use or surgery referral.
Finally, most participants utilized external weight loss resources when available, such as referring patients to local dieticians, healthy living or nutrition programs delivered through local YMCAs or medical centers, commercial programs such as Weight Watchers, and suggesting local physical activity resources. The availability and number of patient referrals to external resources was dependent upon clinic locationmetropolitan and micropolitan clinic participants consistently cited more community resources than participants at rural clinics, who cited distance and transportation as major barriers to connect patients with weight loss resources. All focus group participants reported a lack of follow-up or monitoring of patient progress after they are referred to an external weight loss resource.

Leadership support for patient weight management
The majority of participants reported being encouraged by clinical leadership to talk to patients about weight, but with no specific directives given. In terms of adopting a new process or program, some raised the concern that while clinic leadership may vocalize support for patient weight management efforts, funding such efforts may be a challenge. When asked if clinic leadership would support physicians referring patients to a WMI, one staff participant said, "I'm sure the answer would be yes to that, but whether or not the funding would follow that, that's the question." Differences emerged by clinic locations regarding program adoption. The rural clinicsmany of which were single-standing clinicsreported having more freedom to adopt new programs and practices compared to metropolitan clinicsall of which were governed by large healthcare organizations.

Facilitation-feasibility of implementing a weight management program through primary care
While all participants agreed a WMI connected to primary care would be beneficial, significant concerns were raised regarding program implementation such as facilities, staff availability, and costs. Clinics from all areas expressed the need to weigh clinical capacity for implementation with the potential patient and clinical benefits before deciding to adopt a WMI.
All participants were more enthusiastic when asked to consider the feasibility of an external WMI to which clinics could refer patients and follow-up on progress. Patient identification and referral to an external WMI would likely occur through an EMR query of high weight or BMI, most participants noted. Discussions naturally moved from the ease of an EMR query to the challenge of gauging a patient's motivation to participate. Potential challenges raised for external programs included program costparticularly insurance coverage, staff time for patient identification, referral, and follow-up. Referral to external programs was generally recognized as a more feasible option than adopting and implementing a program in clinics, though as one physician noted insurance coverage "is my only fear of a separate entity program." When asked who would be responsible for championing a program should their clinic decide to adopt a WMI there was no consensus among clinics. Some participants suggested the idea of co-champions, and two participants from different clinics suggested sharing the responsibilities of program facilitation among all clinical staff. Most participants mentioned the importance of having buy-in from all clinical staff for program implementation and sustainability to be successful.

Clinical experience and perceptions
Very few participants reported instances of patients asking about lifestyle WMIs. However, there was consensus that both clinical staff and patients would appreciate if a WMI was available. One staff participant noted "If we had the support to do it. I think all of us would like to offer additional support for weight loss to all our patients. It's just finding the personnel or the time to do that. Or the right program. "A theme of patient accountability emergedthe majority of participants mentioned the need for a program that will keep patients accountable, such as inperson, telephone, or digital counseling support.

Local practice information & patient experience, needs, and preferences
Most participants believed their patients would have an interest in a lifestyle WMI, if it were available, but felt some patients' expectations regarding weight loss and perception of obesity was misguided and may be a barrier to their participation in a WMI. Additional barriers such as the food environment, transportation, and telephone/internet access were cited by participants across clinic locations. Notably, the issue of the food environment was rooted in the comparatively higher costs of healthy foods when discussed at the metropolitan clinics ["You walk down the (grocery store) aisle and there's a big thing of cheese balls… And it's 5 bucks! Whereas one red pepper is 2 dollars."] yet rooted in a lack of access to healthy meal options for the rural clinics ["It's just like, what can you get at McDonalds? What can you get at Taco Bell? What can you get when you go to a ball game? Or what can you pack ahead of time?"] Another potential barrier unique to rural focus group participants was patients' work schedule due to the high proportion of shift workers in rural areas. As one physician noted, "It's very typical in the community to work ten-, twelve-hour shifts, like I said, five, six, seven days a week."

Comparison of three evidence-based interventions.
Participants reviewed three evidence-based programs (see Appendix A) and were asked to consider which program might fit best with their clinics. The first program, Calcium Weighs-In, (Wyatt et al., 2008) was well liked for its simplicity, perceived ease of implementation, and short duration, which participants initially felt would contribute to greater patient interest. However, the focus on diet and provision of daily meals caused concerns about program trialability, potential effectiveness among their patients, and associated costs. The second program, TOURS, (Perri et al., 2008) was favored for its evidence of weight loss but was thought to be the least feasible to implement through primary care and the least likely to be successful due to the high number of in-person sessions. "I just don't see the longer programs and the more expensive programs working at all" (Staff, rural clinic). Finally, while participants had both positive and negative comments about each program, they responded most favorably to the third program -diaBEAT-it! (Almeida et al., 2014) -"I think the diaBEAT-it! one would be much more applicable and feasible from a population standpoint." (Physician, metropolitan clinic) DiaBEAT-it! was favored for its potential for high reach (delivered via DVD and automated telephone calls) though some participants expressed concerns that the advantages of a digitally delivered program might come at the price of participant engagement. Statements such as "It's too easy to ignore" (Physician, rural clinic) and, "I don't know that people would answer the phone calls" (Nurse, rural clinic) demonstrated these concerns.

Discussion
Within a systems-based approach to implementation, primary care practices have the structure to provide practical and sustainable methods to address weight management (Glasgow et al., 2003(Glasgow et al., , 2004 through clearly defined clinical roles and intended audience. Our findings suggest a variety of individuals, or multiple individuals, within primary care clinics in the Great Plains could serve as program champions, acknowledging the need for all staff to support the program to achieve success. Facilitation for a referral strategy appeared to receive more endorsement from the clinical staff due to the ease of implementation and evaluation of effectiveness based on number of patients that are referred. Regardless of the facilitation strategy and personnel, consistent with other literature, it is unlikely that either the implementation of an evidence-based WMI or referral strategy would be adopted without some integration of required tasks with personnel job responsibilities and duties (Hunter et al., 2020). A referral program was perceived to have a relative advantage over an in-house WMI for reasons of lack of time during patient visits and competing clinical demands, consistent with the literature (McKenna et al., 2004). The underlying reason for this was because a referral process could be completed quickly during patient visits and would not provide much disruption to competing clinical demands (higher compatibility, lower complexity). However, a discordance between this infrastructure and clinical culture to address obesity existed in that regular prompts were seen as white noise rather than a cue to action (Ew and Grove, 2016).
Differences were observed among responses from participants by clinic locations. Consistent with the literature (Porter et al., 2019;Perri et al., 1984), clinical staff from metropolitan clinics reported more community resources available for weight management compared to micropolitan or rural area clinics. Participants' perceptions of the root of individual patient barriers appeared to differ among rural and micropolitan (access, work schedule), and metropolitan patients (cost). Primary care staff reported few patients proactively asked about a WMI and made statements suggesting their overall patient panel was unmotivated to seek out a WMI through primary care. However, these perceptions of patient motivation may be underestimated, as previously documented (Befort et al., 2006).
These findings have significant implications for the potential implementation of WMI through primary care. Cost was consistently cited as a barrier to implementing interventions in clinical or community settings and was reinforced by our findings. Program costs, reimbursement concerns, and staffing costs related to time spent identifying and recruiting patients and implementing program activities are proximal barriers to adoption and implementation decisions. Low costs, technology-supported interventions provide a simple referral pathway for primary care practitioners and could help resolve these issues; specifically, it would alleviate the cost burden for clinics and participants. Contextual factors related to addressing obesity in primary care clearly include external barriers surrounding providers' ability to provide the needed resources with limited opportunities for reimbursement of services. This underscores the need for policy change related to state and insurance coverage opportunities that can vary broadly from region to region. Indeed, even for scalable and technology-supported weight loss programs that primary care providers perceive to be a good fit for their settings are not currently eligible for reimbursement through Medicare or other insurance providers. Addressing these external factors from a public health perspective could greatly increase the potential uptake of such programs in primary care.
Additionally, establishing partnerships between primary care clinics and local community organizations has great potential to improve program reach, implementation, and sustainability, and may also help distribute the cost of implementing an evidence-based WMI in a successful and sustainable way (Ackermann and Marrero, 2007;Ackermann et al., 2008). Examples of facilitated community-primary care linkages for weight management exist and have been successful not only in patient weight loss but also in maintaining the clinical-community partnership following the research study (Jebb et al., 2011;Lavin et al., 2006). Additionally, primary care referral to a commercial program may provide a long-term cost-effective solution for patients at risk of weightrelated comorbidities (Fuller et al., 2014). Undoubtably, challenges will arise at onset. However, our findings provide support that community programs are favorable options for primary care providers wishing to improve patient weight management and have the potential to be successful when strong cross-organizational partnerships and champions are involved .
There are limitations to this study that warrant consideration. While we found saturation in responses, the degree to which these findings can generalize to primary care settings outside of our sample, especially outside the Great Plains, is unclear. Due to the great variability among some clinical characteristics, such as EMR systems and functionality, we caution extrapolating findings regarding structural characteristics of the clinics. We did not collect information on clinical characteristics, which may limit the robustness of our analysis. We did, however, compare responses from clinical staff of varying geographic areas. Our findings, in concert with the literature, have practical implications for clinical practice and research, particularly regarding sustainability of interventions delivered through primary care. Primary care clinics are interested in weight management solutions for their patients and would prefer an evidence-based program to which they could refer patients, receive feedback on patient progress, and continually offer as a sustainable part of their regular services. Facilitation of WMI delivered through primary care appears to rely heavily on buy-in from clinical staff and may be improved by instituting co-champions to facilitate implementation strategies. Establishing co-champions may also address the expressed desire for patient feedback with an integrated researchpractice partnership model , and address sustainability considerations.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.