Dissonance in the discourse of the duration of diabetes: A mixed methods study of patient perceptions and clinical practice

Abstract Background Remission of diabetes can be rewarding for patients and physicians, but there is limited study of how patients perceive the timeline of a disease along the continuum of glycaemic control. Objective To explore how patients perceive the timeline of diabetes along the continuum of glycaemic control and their goals of care and to identify whether family physicians communicate the principles of regression and remission of diabetes. Design Mixed methods approach of qualitative semi‐structured interviews with purposive sampling followed by cross‐sectional survey of physicians. Participants Thirty‐three patients living with prediabetes (preDM) or type 2 diabetes mellitus (T2DM) at medical centres in Georgia and Nevada; and 387 family physicians providing primary care within the same health system. Results Patients described two timelines of diabetes: as a lifelong condition or as a condition that can be cured. Patients who perceived a lifelong condition described five treatment goals: reducing glucose‐related laboratory values, losing weight, reducing medication, preventing treatment intensification and avoiding complications. For patients who perceived diabetes as a disease with an end, the goal of care was to achieve normoglycaemia. In response to patient vignettes that described potential cases of remission and regression, 38.2% of physician respondents would still communicate that a patient has preDM and 94.6% would tell the patient that he still had diabetes. Conclusions Most physicians here exhibited reluctance to communicate remission or regression in patient care. Yet, patients describe two different potential timelines, including a subset who expect their diabetes can be ‘cured’. Physicians should incorporate shared decision making to create a shared mental model of diabetes and its potential outcomes with patients. Patient or Public Contribution In this mixed methods study, as patients participated in the qualitative phase of this study, we asked patients to tell us what additional questions we should ask in subsequent interviews. Data from this qualitative phase informed the design and interpretation of the quantitative phase with physician participants.


| INTRODUC TI ON
In the past decade, research has demonstrated the potential of the remission of type 2 diabetes mellitus (T2DM). [1][2][3][4][5][6][7] In the 2009 consensus statement, the American Diabetes Association (ADA) defined remission as serum glucose values below the diabetic range without aid of continuing pharmacologic or surgical treatment. 8 Blood glucose concentrations exist across a continuum of glucose tolerance from normoglycaemia to T2DM (hyperglycaemia), with prediabetes (preDM) labelling a liminal state of hyperglycaemia just below the diabetes threshold. 9,10 ADA cautions that preDM should not be viewed as a clinical entity but as a risk factor for diabetes and cardiovascular disease. 11 Studies demonstrate that this 'continuum of risk is curvilinear; as A1C rises, the diabetes risk rises disproportionately'. 12 Recent evidence reveals that preDM can be reversed to normoglycaemia through significant weight loss. [13][14][15][16] In the Look AHEAD (Action for Health for Diabetes) trial, intensive lifestyle intervention showed significantly greater remission of T2DM in comparison with the diabetes support and education group. 7 Moreover, patients with T2DM who completed and maintained extensive weight loss of at least 15 kg have experienced prolonged remission of diabetes to either preDM or normoglycaemia. 3,[17][18][19][20][21][22] Evidence for remission of T2DM is best documented in metabolic surgery patients. One metaanalysis showed that 2 years after metabolic surgery, remission was seen in 63% of patients. 23 Since publication of the Diabetes Prevention Program (DPP), physicians have been encouraged to counsel patients with preDM regarding effective strategies to decrease the risk of cardiovascular disease and progression to T2DM. 24 With these goals in mind, it is important to understand the role of goal setting when physicians discuss the diagnosis of preDM or T2DM and the continuum of glucose tolerance. From the time of diagnosis, physicians need to aim to achieve a shared meaningful diagnosis, in which the patient, the physician and the healthcare team have a shared mental model of the diagnosis and resulting goals of care. 25 Goal-setting, the sharing of realistic health and well-being goals by physicians and patients, is core to the theory and effective practice of personalized care planning and seen as particularly important for patients with multiple chronic and long-term health conditions. 26 Additionally, shared decision making with goal setting and motivational interviewing results in improvements in patients' knowledge regarding their diagnosis, the perception of risk and more confident decisions in their care. 27 Specifically along the continuum of glucose tolerance, literature suggests that informing patients of their diagnosis of preDM represents an opportunity to set goals and motivate patients early in the continuum of glucose tolerance. 5,7 This is important because the literature also suggests that improved clinician communication with patients can result in return to normal glucose regulation in patients with preDM. 28 Goals, such as remission of T2DM and regression of preDM, can be empowering for patients and can serve as a tool for family physicians to use to inspire patients. 29 informed the design and interpretation of the quantitative phase with physician participants.

K E Y W O R D S
diabetes remission, mixed methods, prediabetes, type 2 diabetes mellitus one of two medical centres in distinct regions of the United States.
One medical centre was in Augusta, Georgia, within a geographic region with a high prevalence of diabetes. 38 The second medical centre was in Las Vegas, Nevada. Table 1  CJWL then observed the RAs conduct two interviews at each site.
Throughout this process, verification strategies were used across the design to promote rigour, including memo-keeping and maintaining methodological coherence/congruence during the process. 42 To decrease participant burden and maintain privacy, interviews were conducted at patients' primary care centres. Before each interview, the interviewer completed the informed consent process with the patient. Each interview was audio-recorded and transcribed verbatim. CJWL and two RAs collected data from October 2016 to April 2017. Interviews lasted about 1 hour and resulted in more than 1732 pages of data.
As this was a secondary analysis, the analytical process was not done concurrently with data collection and instead completed once all data had been collected. 43

BOX 1 Interview guide
Tell me about the first time your provider talked to you about diabetes (or prediabetes).
What was your first impression of diabetes (or prediabetes)?
a. What was your first thought at the moment of diagnosis? Why did you think this?
What do you think caused your diabetes?
Why do you think it started when it did?
How severe is your diabetes (or prediabetes)? Do you expect the diagnosis to change?
(Probe: source of this information) What do you fear most about diabetes (or prediabetes)?
(Probe: short term versus long term fears/effects) multiple times to become immersed in the data and discuss patient perceptions and goals for diabetes care. The four researchers then identified and segmented data that reflected these constructs by analysing half of the interview transcripts to conduct a preliminary thematic analysis and identify potential themes for each research inquiry. In the second phase, CJWL sought to validate these themes by reviewing the analysis conducted on the segmented text while also conducting axial coding to define each theme's characteristics.
In the third phase, one RA then used the finalized codebook to analyse the second half of the interviews. This provided an additional opportunity to ensure rigour by further validating the themes, identify those most saturated for final presentation and confirm the thematic characteristics. Of 532 registered conference attendees, 387 attendees (72.7% response rate) responded to the survey. We excluded 65 responses

| Phase 2: Physician practice through a quantitative approach
that did not answer the questions from this section of the omnibus. Since this is a study of clinical practice, we also excluded an additional 38 responses that were current medical students or were missing data for year of medical school graduation and year of residency graduation. Therefore, 284 responses are included in analysis. See Table 2 for respondent demographics.

| Phase 1: Patient perceptions through a qualitative approach
The first inquiry identified the patient perceptions of the duration of the disease.

| Phase 2: Physician practice through a quantitative approach
In the first case, which described a patient who met the clinical cri-  Survey results showed that more than 40% of our sample would not use the word 'prediabetes' in counselling patients regarding their diagnosis. These practice habits likely contribute to the fact that nearly 90% of adults in the United States who have preDM are unaware of their diagnosis. 58 Alternately, this could reflect the lack of consensus regarding diagnostic thresholds for preDM 11,59 or the contested nature of the diagnosis of preDM itself. 60 66 and progression to T2DM and its associated complications.

| D ISCUSS I ON
Long-term, large-scale studies of intensive lifestyle intervention in people with preDM have demonstrated significant reductions in the rate of conversion to T2DM. [67][68][69] Informing patients of their preDM also represents an opportunity to motivate patient behaviours early in the continuum of glucose tolerance, when remission is more likely. 5,7 Evidence also suggests that recommending these lifestyle changes leads to changed patient behaviours and increased physical activity in patients with preDM. [70][71][72][73] Although more vague terms such as 'hyperglycemia' and 'impaired glucose tolerance' may be used to communicate with patients, specifically naming the diagnosis influences patient representations of disease. 74 We propose that using the term 'prediabetes' in the setting of motivational interviewing and risk communication is important to communicate a clear message and a path forward.

| CON CLUS ION
Most physicians here elected to not communicate remission or regression in patient care. Yet, patients describe two different potential timelines for diabetes, including a subset of patients who expect their diabetes can be 'cured'. Physicians should incorporate shared decision making to create a shared mental model of diabetes and its potential outcomes with patients.

D I SCL A I M ER
The

ACK N OWLED G EM ENTS
We thank Heather Rider, Jasmyne Womack and Angela Seehusen for assistance with recruitment and data collection. We thank Carla Fisher for her assistance in creating and refining the interview guide.
We acknowledge Dean Seehusen and Paul Crawford for their leader-

DATA AVA I L A B I L I T Y S TAT E M E N T
Transcribed data are not available for release.