Our basic data analytic approach was exploratory and content-driven [36]. The rationale for choosing an exploratory approach was that it provided an opportunity to inductively expand the knowledge base on patients’ experiences with T2DM, the challenges they face in managing the disease, and patients’ expectations of a telephone-based DHC program for managing T2DM [37]. The specific approach to analysis was phenomenology because this approach uses people’s subjective experiences to gain an in-depth understanding of a phenomenon [38]. Since we explored the challenges patients face in managing T2DM and how those lived experiences shaped their expectations surrounding a telephone-based DHC program, the phenomenological approach appeared to be a good fit for the study. We followed the data analysis steps outlined in the Creswell and Creswell [33] guidelines. We first organized and prepared the transcripts for analysis by transcribing the audio-recorded files.

After an initial immersion in the transcripts, the 2 coders independently read all transcripts multiple times and wrote analytic memos to capture their initial impressions. We conducted both the first and second cycles of coding.

First-cycle coding was conducted line-by-line using a combination of descriptive codes (summarizing segments of data) and in vivo codes (participants’ own words). Each transcript was initially coded independently by 2 researchers to identify participants’ concrete experiences with diabetes self-management and expectations for telephone-based coaching.

After coding the first set of transcripts, the coders met to compare their code lists, collapsed overlapping codes, and agreed on common code labels and definitions, which formed the first version of the codebook. This codebook was iteratively refined as additional transcripts were coded; new codes were added when data segments did not fit existing codes, while similar codes were merged or refined to improve clarity and consistency. Each codebook revision was discussed and agreed upon by both coders, and discrepancies were resolved through discussion, with a third qualitative expert consulted when consensus could not be reached.

In the second coding cycle, the team used pattern coding to group the first cycle of codes into higher-order categories or themes that reflected broader concepts (eg, combining multiple diet-related codes into a broader category reflecting diet management) [39].

These categories were then systematically compared across transcripts to identify recurring patterns and relationships. These categories were further abstracted into themes and subthemes that described patterns across participants. Candidate themes were continuously checked against the coded data and raw transcripts to ensure they were grounded in participants’ accounts. Negative or divergent cases that did not fit initial interpretations were examined. Themes were finalized only when all coders agreed that they comprehensively captured the dataset’s range of perspectives.

Two researchers independently coded all transcripts, and assistance from a third coder was sought when the first 2 coders could not reach an agreement on coding. The intercoder agreement level was calculated using the Huberman method [39], and an agreement level of >90% was observed. NVivo software version 12 Plus (Lumivero) was used to organize, code, and categorize the information obtained from the transcripts.

A substantial amount of time was spent on the analysis, and expert help was sought when necessary. Deidentified data were used throughout the analysis. An analytic perspective was maintained throughout the analysis, using deidentified data while remaining empathetic to participants’ life experiences [40]. Opinions from all participants were given equal weight when interpreting the data to avoid selective interpretation [40]. It was ensured that the data remained fully confidential, and only the researchers involved in this study had access to it [39].

The analytical team consisted of 2 UAB-based researchers, one with expertise in public health and the other in health services research with prior qualitative research experience, who collaboratively coded the transcripts. Both researchers had long-standing experience working with adults with T2DM in the Deep South, including predominantly African American, socioeconomically disadvantaged populations receiving care within the UAB Health System [32,41,42]. To ensure credible analysis, a third qualitative expert, a health service researcher with extensive experience in qualitative studies, was consulted in case of coding disagreement or interpretive uncertainty.

Diabetes self-management presents persistent, multifaceted challenges that test patients’ knowledge, resources, and resilience daily. Participants in this study expressed challenges across four key domains, including (1) managing diet, (2) physical activity, (3) medication access barriers, and (4) blood glucose monitoring (Table 2).

The most frequently reported challenges were related to dietary management. Participants described diet as a major challenge in managing diabetes in their daily lives (n=10). Many individuals struggled to resist ingrained eating habits and sweet cravings. While some mentioned that sweets were “getting around me” despite their efforts to limit intake (P12), others just compensated for it with an additional dose of insulin instead of practicing abstinence (P9). Participant (P11) reported having adequate nutritional knowledge but lacking appropriate behavioral execution, citing situational limitations on healthy options and a lack of willpower in choosing better alternatives. For some, knowledge gaps in interpreting food labels and the fear of eating incorrectly were concerns. Participant (P8) reported that the initial diagnosis of diabetes brought overwhelming confusion regarding grocery shopping, reading labels, and food selection. Additional concerns about managing diet included situational limitations such as a busy work schedule (P5) and struggles with daily meal planning (P3), which led to choosing convenient food over meeting nutritional requirements.

Physical activity was hindered by physical limitations, including low energy, fatigue, limb pain, weakness, and visual impairment, as well as by environmental factors for some participants (n=4). Diabetes-related complications, such as leg pain and weakness, impaired mobility (P7), and ankle swelling, further hindered ambulation (P1) for participants. Extreme climates also interfered with outdoor walking routines for some participants (n=3), as participant P11 mentions: “I try to walk. I haven’t been doing that lately. (Because it is cold outside).” Participants not limited by physical limitations reported engaging in some mild-to-moderate physical activity. Walking emerges as the most common and accessible form of exercise.

Medication adherence was generally good among the participants in this study; however, structural barriers disrupted treatment access for some individuals. Barriers related to medication access included the unavailability of prescribed drugs and lack of insurance coverage (n=3). These challenges resulted in delays or changes in prescribed regimens. The available and potentially alternative medicines either had high copays, were not affordable for patients, or were not covered by insurance at all. Such experiences highlight financial barriers, insurance limitations, and medication shortages, leading to disturbances in adherence, even in highly motivated individuals. In these situations, participants often relied on health care providers for alternative solutions.

Glucose monitoring also emerged as an essential component of diabetes self-management for most participants, who incorporated blood glucose checks into their daily routines using traditional glucometers or continuous glucose monitoring (CGM) systems. However, structural barriers similarly affected access to monitoring tools (n=3). Challenges included supply shortages and insurance-related limitations on obtaining glucometers or CGM devices. Some participants expressed difficulty maintaining regular monitoring due to a lack of supplies or discomfort with fingerstick testing. Participant P12 discontinued fingerstick testing after exhausting test strips and now relies on symptom recognition for hyperglycemia or hypoglycemia. Participant P4 faced insurance barriers to receiving CGM monitors and mentioned that they are in the process of resolving access issues.

Overall, the findings highlight that diabetes self-management is influenced by a complex interplay of behavioral, physical, and structural factors. Participants faced persistent difficulties in adhering to dietary recommendations, engaging in consistent physical activity, accessing and affording prescribed medications, and maintaining regular blood glucose monitoring. These challenges were compounded by long-standing behaviors, physical comorbidities, environmental constraints, and systemic barriers within health care and insurance systems. The convergence of these individual and contextual factors underscores the multifaceted nature of diabetes self-management and reveals the necessity for comprehensive, patient-centered interventions that address not only education and motivation but also accessibility, affordability, and continuity of care.

The multifaceted nature of diabetes self-management necessitated understanding patients’ needs and expectations for an efficient, personalized telephone-based DHC program. Four subthemes emerged within this theme: (1) targeted education, (2) perception of a health coach, (3) delivery modality, and (4) incentive option. A brief description of each subtheme is provided in Table 3.

Based on their experiences with current T2DM management strategies, the participants were asked whether they would like to incorporate specific knowledge into future T2DM interventions to help them monitor and manage their diabetes. Participants expressed targeted interests in diabetes education topics, with dietary guidance emerging as the most common need (n=3), particularly clarification on suitable foods for individuals with diabetes, as P7 mentions, “And then I can see about my eating habits, too, see what I can eat….what I can’t eat...I know a little of it, but I don’t know a lot of it.” Participants also emphasized the need to learn day-to-day regimens to build sustainable habits (n=2), noting the difficulty of changing long-standing lifestyles and the value of clear daily guidance for effective habit formation. Additional knowledge on other topics included specific exercises (n=1), side effects of diabetes medication (n=1), and the desire for overall refreshers (n=3). However, some indicated no need for additional information (n=4), citing self-learning, prior knowledge, or prior participation in research studies as sufficient.

Upon querying participants’ thoughts on receiving a customized booklet containing diabetes education information, accompanied by a follow-up call to explain its contents, most participants expressed interest in learning more and felt it would be a valuable tool (n=11). They mentioned that the booklet would serve as a quick, convenient guide for finding any information they needed compared to searching online on an electronic device.

The concept of a health coach was introduced to the participants as a personal cheerleader, a guide, and a partner who could help them implement lifestyle changes to improve their health. When asked, the participants were asked whether they thought they would benefit from consulting with a health coach to manage diabetes. Most participants (n=8) were open to working with a health coach and expressed interest in tailored behavioral support. A few (n=2) were unsure of the idea and were willing to try if recommended, while only a couple (n=2) declined the idea of working with a coach, citing sufficient self-control and an aversion to added stress. Most participants perceived two primary benefits of working with a health coach: (1) an accountability partner for informed decision-making and keeping them on the right track and (2) providing support on personalized meal planning.

Participants also mentioned that receiving health coaching would help them develop strategies to manage their diabetes. They perceived that communicating information about their diabetes management to a health coach would help them receive constructive feedback, which, in turn, would enable them to manage their diabetes more effectively.

Participants perceived using technology, such as a phone call or text, as part of the intervention would be vital.

Participants thought daily monitoring texts or phone calls could be helpful reminders because they had memory issues (n=10). They mentioned having difficulty remembering to take medications on time or checking their blood sugar levels. When asked about their preferred communication mode, participants preferred text over calls (n=7). Participants who favored texts did not like the idea of regular phone calls, as P9 quoted, “Probably a text because I don’t really like phone calls, so I’d rather text than take a phone call.” Participants who opted for calls (n=5) preferred quick interactions, as P10 stated, “It would be preferably a call because text messages, I’m not going to lie, they’re easy for me to ignore. So yeah.” Although most participants preferred receiving a daily text over a daily call, we also asked participants about the desired length of a daily call if they were to receive one; most specified that they wanted it to be short. The majority of participants preferred the call duration to be less than 10 minutes (<2 minutes [n=4], 5‐10 minutes [n=4], as per the needs of the participants [n=1]).

Participants were also asked about their preference for weekly contact for coaching calls. Most participants (n=11) indicated that these calls would be a helpful component of the intervention, providing accountability and reminders to support routine building. One participant preferred a reduced frequency (once or twice a month), suggesting that weekly calls might be too repetitive and frustrating (P1).

Participants identified practical incentives to enhance engagement with daily calls or texts, with monetary rewards emerging as the predominant preference (n=10). Most favored financial support through cash or gift cards, with monetary rewards ranging from US $10 to US $50, viewing it as valuable supplemental income. P8 mentions: “...it would help me a lot financially... The money would come in handy.” A minority prioritized health over rewards (n=2), with P3 stating: “I’m not really worried about that [reward]. I’m more so worried about my health,” and P10 preferring alternatives that promote healthy behavior, such as a gym membership.

Participants described their expectations surrounding a potential DHC intervention within this theme. Given the challenges they face in managing diabetes, they expected future interventions to incorporate specific information related to diet, day-to-day regimens, exercise, and medication effects to optimally manage their diabetes. Most participants identified several aspects of health coaching that would benefit them in managing diabetes. They perceived that a health coach could serve as an accountability partner, a guide for meal planning, and assistance in staying on track with diabetes management, while also providing a means to discuss aspects of diabetes beyond their PCP. Most participants found that using technology, such as daily or weekly calls and texts, would help them as reminders to take medications or exercise, except for a few who indicated that daily or weekly communications could be overwhelming. Most participants stated that a monetary incentive of US $10 to US $50 would motivate them to attend those calls, while a few noted that a financial incentive might not be as effective as other incentives, such as a gym membership.