Socializing the evidence for diabetes control to develop “mindlines”: a qualitative pilot study

Type 2 diabetes mellitus in adults is a leading contributor to morbidity and mortality in the United States and is one of the most prevalent and expensive chronic conditions managed in primary care [1]. Improvement of diabetes control, as measured by the percentage of a patient panel meeting a threshold of diabetes control, is a common measure for quality in primary care and is a Healthy People 2030 goal [2, 3]. This measure has been used on our health system’s “quality scorecard” (used to determine pay-for-performance incentives for system physicians) for the past decade but progress on significantly improving performance on this metric has been elusive.

Prior to the study, the leadership team of the department of Family & Community Medicine in our large health system in the southeast United States had developed an evidence-based intervention framework—the LIVE! Framework—to address the problem of poorly controlled diabetes in the affiliated practices. This framework emphasized four major interventions to reduce average glycosylated hemoglobin (HgbA1c or A1c) in patients with HgbA1c greater than 9%—improved healthy eating and physical activity (L—lifestyle), increased use of insulin to control blood sugars (I—insulin), increased visit frequency (to a goal of every 1–3 months) (V—visit frequency), and attention to comorbid emotional and mood disorders that can worsen diabetes control (E—emotional care). As part of development of the framework, the leadership team had gathered informal feedback from department clinicians and revised the framework accordingly. The feedback revealed clinician concerns about a gulf between the existing evidence-based recommendations and the practice and patient-centered realities – for example, time, resource awareness, and resource availability.

This gulf, or “implementation gap” has been noted and analyzed by many primary care researchers [4, 5]. Gabbay and LeMay have observed, in their qualitative study of primary care practices, that simple awareness of existing evidence and guidelines was not sufficient, but that the evidence needed to be socialized within the practice group, where it was vetted by the members of the practice and seen through the lens of the local context [6]. This “knowledge-in-practice-in-context” is postulated to form the basis of internal, tacit guidelines (called “mindlines”) used by primary care clinicians [7]. This “mindlines” concept is a model derived from careful systematic observation, but has not yet been used as an intervention to enhance evidence adoption. This intervention would consist of making clinicians aware of evidence, providing an opportunity to informally discuss and consider the evidence, and allowing for adaptation the evidence to fit local context and circumstances.

The time available to primary care clinicians to discuss and review new knowledge has decreased due to the pressures of the business of practice, increased electronic health record (EHR) documentation requirements, and issues of work-life balance [8, 9]. This development has the compound consequence of decreasing the ability of a group of clinicians to socialize the needed evidence and guidelines with their colleagues and staff, as well as decreasing the time available to form personal connections with colleagues. This latter effect could be particularly important given the rates of clinician burnout seen today [10].

This pilot project examined the feasibility of an intervention to promote the socialization of the evidence from the LIVE! framework in primary care practices that are struggling with meeting the metric of average HgbA1c < 9.0%.

The four practice visits were completed in June 2019. There was generally near complete attendance of the practice staff at the meetings, though this was not formally assessed. The principal clinicians (physicians, nurse practitioners and physician’s assistants) from each practice were present at each meeting. The practices ranged from large (20 + people, clinicians, and staff) to small (1 or 2 clinicians, < 10 staff) and included both rural and urban/suburban practices.

The practices noted many ways in which they were already engaged with improving their quality scores for the care of patients with diabetes. Some practices saw benefit from some of the technologic features that the health system was implementing. Most practices had been using both clinician-centered and clinic-centered workflow changes prompted in part by their involvement in patient-centered medical home activities in the past. Finally, practices had generally been able to access the limited community resources available to them, sometimes in creative ways. Table 2 reviews some of the details of these practice behaviors.

The practices noted many barriers to care for their patients. These fell into four main categories: patient factors, local context limitations, cost, and lack of community resources (See Table 3). In general, the discussion of barriers to care dominated the discussions, and in one practice, it was difficult to move the discussion beyond these complaints, limiting both discussion about successful behaviors as well as the mindline-associated behaviors discussed below.

To analyze whether these practice visits could generate the type of interdisciplinary practice interactions necessary for “mindlines” generation, the research team looked for 1) discussion of the evidence (the LIVE! recommendations) between practice members, 2) evidence that the practice either accepted the new evidence or engaged in adapting the recommendations to their own context, 3) if they rejected any parts of the evidence due to their practice context, and 4) if they expressed any wishes about a change in external circumstances that would allow them to achieve better quality in treating their patients with diabetes. We found examples of each of these phenomena, but the discussion of barriers to implementation of each of the aspects of the LIVE! framework dominated the discussion. The relevant excerpts of these conversations are presented in Table 4.

The practices’ performance on the diabetes control measure (percent of practice with diabetes whose HgbA1c is less than 9%) improved slightly over the course of the study (Fig. 1) but given the relatively few data points, formal trend analysis was not undertaken.

The results of this study show that a one-time, informal, practice discussion around evidence-based interventions can generate the kinds of discussions that may lead to the development of “mindlines.” Clinicians, nurses, and clerical staff engaged each other about novel ways to implement the evidence informed each other about existing practices and shared both barriers and potential solutions to implementation. Because of some recent changes in how nurse care coordination was delivered to the patients in the health system’s accountable care organization infrastructure, the discussions were dominated by some complaints and barriers referable to that change, however, each of the groups did show some evidence of “mindlines-generating” discussions of the evidence. Facilitating these sorts of conversations about the evidence routinely in practice has the potential to develop and strengthen clinicians’ internal “mindlines” around evidence-based practice.

The change in the performance on the diabetes measure was encouraging, but far from conclusive as there were too few data points to allow for meaningful trend analysis. As this was a secondary outcome, we did not compare this change to the other practices in the department, but this measure had been difficult to change in practices over the previous several years. Secular trends and/or an approaching metric deadline (the end of the fiscal year on September 30^th) may have accounted for the rise in scores.

More work is needed to operationalize the “mindlines” framework laid out by Gabbay and LeMay, and because this is a primarily qualitative study, its findings must be evaluated in a more rigorous manner before implementation at scale. Previous work in this area by Crabtree, et al., finds that “People working in practices are well educated and want to do well; however, they need support in finding ways to interact and collaborate with colleagues…most practices are resistant to protecting time for reflection…thus, “forcing” time and space for reflection may be one of the more important components of a change management strategy” [12]. A conceptual framework of barriers to evidence implementation derived from a systematic review of reviews includes many of the concepts of socialization, context and adaptability that are essential to the “mindlines” model and provide a solid theoretical foundation for deriving potential interventions such as this one [13]. In this era of increasing healthcare clinician burnout, interventions that involve interdisciplinary, relaxed, problem solving communication can re-establish and improve the social connectedness of outpatient practices, which may reduce burnout and improve practice performance [10, 14].

An intervention aimed at generating unstructured interdisciplinary conversations about evidence-based interventions to improve diabetes control is feasible in primary care practices. The resulting interdisciplinary conversations included discussions of adaptation and adoption that could influence the development of “mindlines” among the practice clinicians and staff. Future research plans in this area include a regular schedule of these meetings with incentives to maintain practice-wide participation and fine-tuning the presentation of the evidence.