Patient Markers of Successful Diabetes Management

This study was a secondary analysis of semistructured interviews that were conducted to compare the comprehensiveness and efficiency of semistructured interviews and group concept mapping for eliciting patient-important outcomes for diabetes care as part of a Patient Centered Outcomes Research Institute–funded methodology study. The full methods and findings of the primary study have been described elsewhere (19). Only interview data are included in this secondary analysis.

Participants were recruited from an urban academic medical center. Research staff enrolled participants during an ED visit (acute care) within 7 days post–hospital discharge (post–acute care) and at scheduled primary care visits (primary care). Interviews were continued until thematic saturation was achieved for each recruitment venue.

This study used a convenience sample of English-speaking adults who were ≥18 years of age, had type 1 or type 2 diabetes, and were able to provide written informed consent. Participants enrolled in the acute care or post–acute care setting were required to be receiving treatment for a diabetes-related problem. Those enrolled in the primary care setting were required to have had at least two A1C measurements >7.5% in the previous year. Sampling was conducted in three settings to allow for capture of differences in goals and priorities related to diabetes that individuals with diabetes may have when coping with varying severity of disease and care challenges.

Potential participants were excluded if they had a new diagnosis of diabetes during the recruitment visit; had a significant diabetes-related complication such as end-stage renal disease, amputation, or blindness; were undergoing medical clearance for a detox center or any involuntary court or magistrate order; were in police custody or incarcerated; or had major communication barriers that would compromise providing written informed consent.

ED participants were screened using the ED’s electronic medical record (EMR) during scheduled shifts; research staff identified and approached potential participants for interviews. For post-hospitalization and primary care settings, research staff screened regularly generated EMR patient lists to identify potential participants. Post-hospitalization participants were contacted, gave verbal consent, and were interviewed by telephone. Primary care participants were contacted by phone before a scheduled visit to assess their interest in study participation. Participants met with the interviewer before or after their primary care visit, provided written consent, and were interviewed in person. Participants were compensated $25. Research approval was obtained from Thomas Jefferson University’s institutional review board.

Participants completed demographics forms, and their last recorded A1C and BMI values were obtained from the EMR. The authors used an open-ended, semistructured interview guide to discuss outcomes most important to participants when making decisions regarding the management of their diabetes. The guide was developed, tested, and refined in collaboration with our Patient and Key Stakeholder Advisory Board (PAKSAB). PAKSAB members included individuals with diabetes, caregivers for family members with diabetes and other chronic conditions, a diabetes educator, and a primary care physician. Interview questions were designed to elicit participants’ beliefs about the causes of their diabetes, their worries and concerns, how well-managed they considered their diabetes to be, their definition of successful management, their care goals and challenges, and anything else they wished to discuss about their diabetes care. See Supplementary Materials for the entire interview guide.

Interviews were one on one, lasted ∼30 minutes each, and were audio-recorded. Four team members served as interviewers: two Master’s-prepared research coordinators with training in clinical research and two PAKSAB members (a patient advocate and a nurse practitioner). All interviews were transcribed professionally with identifying information removed. Transcripts were checked by a team member for accuracy.

Interviews were analyzed using NVivo, v. 11.0, software (20). The team decided a priori to develop a “goals” node (concept) to capture all ideas related to participant goals or treatment priorities because this concept was the primary focus of the overall research project. The goal types and definitions were developed de novo from the interview content. Other nodes represented concepts that emerged during analysis using a conventional content analysis approach (21). Interviews were coded by three team members, including two PAKSAB members. A fourth researcher performed double-coding of a random subsample of interviews to ensure coding consistency. The coders individually read the transcripts, identified codes and subcodes that emerged, and coded independently. They then met to review coding, resolve discrepancies, and refine the codebook in an iterative process. The coding team routinely reviewed the κ coefficient and percentage of agreement and met to discuss discrepancies. The minimum acceptable κ coefficient was set at 0.60, and the target was 0.80. The average coefficient was 0.70.

For the overall study findings, attention to study credibility, confirmability and dependability, and transferability ensured the rigor of the qualitative research process (22). Credibility was addressed by data triangulation, which consisted of interviewing participants from differing care settings (ED, primary care, and post–hospital discharge), using multiple coders, and assessing intercoder reliability. Two PAKSAB members served as coders, and the full PAKSAB reviewed and discussed study findings to provide member-checking. A summary of findings was also disseminated to interview participants. Confirmability was supported through detailed records of all steps of the study process and the use of the Consolidated Criteria for Reporting Qualitative Research checklist (23) for reporting the study procedures and findings. Transferability was supported through description of the participants and use of quotes to support themes.

For the analysis reported in this article, we analyzed interview text related to patients’ perceptions of successful management that was coded to the node “measuring management success.” A matrix coding query was completed in NVivo to identify the intersections between the “measuring management success” node and other nodes. Two coders independently reviewed the node content and the matrix to identify themes and met to resolve discrepancies. Differences in themes between care settings were also examined. SPSS Statistics, v. 26 (24) was used to calculate descriptive statistics for participants.

A total of 126 individuals were approached for interviews. Thirty-one declined, primarily because of a lack of interest in participating or not feeling well enough to participate; 95 were enrolled; and 89 completed interviews. Participant demographics are shown in Table 1. The mean participant age was 55 years. Sixty-eight percent of participants were Black, and the same percentage were high school graduates. Most individuals had type 2 diabetes (95.5%), and the mean A1C was 10.2 ± 3.3%. Participants’ markers of successful diabetes management focused on several themes: 1) A1C, blood glucose, and numbers; 2) engagement in medical care; 3) taking medication and medication types; 4) symptoms; 5) diet, exercise, and weight; and 6) stress management and social support.

All participants cited A1C measured by their health care professional (HCP) and/or self-monitoring of blood glucose as a marker. Nearly all cited meeting A1C targets as an important measure of successful management. As one participant said, “In June, … I went to get another blood test done, and [my A1C] came down from 12 to 8.3, and I go this month [to get] another blood work done, and I'm hoping that I got it down to 7 something, whatever. My goal is to get it down to 6, even below 6” (ID 217). Although most cited A1C as an important target, participants expressed varying levels of understanding of A1C. As one put it, “Well, my AC1 level is I think 12, and he said that was pretty high and, to me, I don’t even know nothing of AC1 level or 1C or whatever you call the level. I don’t know. But I’m learning” (ID 105).

Participants referred to blood glucose using a variety of terms, including “blood sugar,” “sugars,” or “my numbers.” They mentioned both tracking of blood glucose using logs and specific numbers as their markers for success. One said, “If it’s in that 120 range, [I] want it to stay in that 120 range. So, … you don’t want it to go up—that’s to the 200s. Like I say, 1s are fine, 100. Yeah. And 180s and stuff like that, that’s good” (ID 208). Another said, “If you wake up feeling good, okay, fine. You get on the machine, and if it says 120—between 120 and 80— you know you're in good shape” (ID 101). Some cited the concurrent use of both A1C and blood glucose values, as in this comment: “You need an A1C count to know whether it’s well managed or not. It’s also the glucose meter, as well” (ID 109).

Although most participants reported additional metrics when assessing their diabetes management success, they usually stressed the importance of A1C and blood glucose levels as the primary way in which their HCPs monitored management success. OF participants who acknowledged A1C as a marker, some also expressed frustration when these measures did not reflect their self-management efforts or said that successful A1C management was ultimately unattainable. As one put it, “My AC1 had fallen to 8.6. I thought it was good. No, it wasn’t. So, my reward from my AC1 falling to 8.6 is now I’m on this Bydureon. I have been able to lose a little bit of weight. I’m losing weight too slowly. So, it just seems like it’s not a winnable situation” (ID 124).

In addition to A1C and blood glucose targets, participants commonly referred to the completion of medical care processes as measures of successful diabetes management; these included completion of A1C and blood glucose testing by HCPs. One participant said, “You get that uplift when you go see your doctor, and she tells you what your blood level’s been for them 3 months” (ID 123). Another mentioned attendance at medical appointments: “I think by seeing him every 3 months and having that A1C test done is an accountability” (ID 114). Completion of additional screening and monitoring tests was often mentioned. As one participant said, “When I come in for—like I am today, he is going to give me a complete examination. He's going to look for different things—look at my feet to see if there is any change that I haven’t noticed. And through, I guess, his questions to see what I've been doing—my weight and that kind of stuff” (ID 113).

Participants also highlighted the importance of constructive patient-provider discussions and adherence to physician recommendations as important measures of management success. Although participants understood the importance of routine monitoring, some wished for more collaborative treatment plan discussions that extended beyond current clinical markers to include additional metrics. These included referrals to additional diabetes management resources to address psychosocial well-being, as well as clarification of management goals. One participant said, “I need to have a more definitive definition of what management means … but it’s just nothing concrete here when you’re talking about managing what can be a life-threatening disease” (ID 130). Another shared:

Closely linked with engagement in medical care were factors related to diabetes medication. For some, this meant taking medications as an intermediate goal to achieve an ultimate blood glucose target. One participant put it this way: “It’s managed pretty good because I check it and I take a pill. And, once I’m taking the pill like I’m supposed to, it stays in check” (ID 203). Others viewed taking prescribed medications as an independent goal and measure of successful management. One participant said, “… because your medication should be taken on time, instead of skipping. And, some days I used to go maybe all day long and whatnot without taking my medication…. [And] since it’s been up, I really try to take it … on time, around the same times, because I take a dose in the morning and a dose in the evening. I think that’s important” (ID 206). Others noted a more fluid approach to medication-taking based on physical feelings and their schedule. As one said, “Sometimes I skip the metformin. The insulin, I don’t skip” (ID 227). However, among those who viewed medication-taking patterns as a marker of successful management, some acknowledged periods of reduced medication efficacy as judged by other clinical targets.

Some participants saw reducing or eliminating their medications as management success. One said, “My goal is to cut down on my medication because I believe that once I went on insulin—because I’m on oral medication and insulin—I believe that once I went on insulin, I wouldn’t have to take them oral medications, but I did” (ID 119). One individual reported achieving this goal:

Many participants also reported the presence or absence of physical symptoms as a marker of successful diabetes management. The symptoms were varied but primarily related to fatigue, dizziness, and sleepiness, which are often symptoms of hyperglycemia:

Blurred vision and frequent urination, also potential indicators of hyperglycemia, were often discussed. One participant noted that, “The only way I can tell is if I don’t urinate a lot, then I know it’s all right. So, I urinate a lot, then I know it’s high, and if I got blurry vision in my eyes, then I know it’s high again, too” (ID 202). Finally, the presence or absence of neuropathy symptoms was cited as an indicator of successful management. One participant said, “Sometimes I get neuropathy in my feet because I … wasn’t taking care of it at first. So, I had to get myself back into the swing of things mentally” (ID 309).

A number of participants described adherence to diet, exercise, and weight goals as important markers of successful diabetes management. Of these three, discussions of diet predominated. A small portion of participants reported successful dietary practices. One said, “The fact that I don’t eat the things that I shouldn’t eat helps me to believe that I’m doing the right thing” (ID 310). Descriptions of healthy eating included frequent vegetable and water intake, as well as avoidance of sugar. However, the majority of participants expressed challenges with healthy eating and a need for further nutritional support. As one put it, “Exactly how I’m eating … [it] basically comes from my diet…. If I know it’s not successful—by what I’m eating, [the] same habits I had that I’m trying to break—is hard. And I’m still continuing. I mean, some days I have my good days. Some days I have my bad days. But it’s hard to concentrate on how I’m eating” (ID 307).

Although less frequently mentioned, some participants strove for exercise and weight loss as management goals, driven either by HCP recommendations or their own concerns. AS one said, “Without the exercise, I noticed if I don’t exercise for a week, my sugars go up crazy. If I exercise and stay on my regimen and eat the proper things, it stays in check” (ID 216). Another said, “If I drop 50 lb, this will—because I have type 2 diabetes—all this might just go away” (ID 301).

The final theme identified was stress management and relationships with family and friends. Some participants indicated reliance on family and friends for practical and emotional support when discussing successful diabetes management. As one put it, “It’s weird, because it’s one of those things you have to manage on your own, but it’s kind of hard to manage it on your own. It’s a little easier now because I have my daughter living with me, and she knows. So, she knows, ‘Uh, oh. He’s doing this. Come here,’ [and tell me to ] take this or … eat this or whatever the case may be. Or, she’ll be like, ‘Oh, you’re going to the bathroom a lot. Did you test your blood sugar?’” (ID 201). Others considered family and friends a stressor, particularly with regards to their role as caregivers. Accordingly, stress level was also cited as a marker of successful diabetes management. One participant said, “I don’t know about other people, but my diabetes came when I was older, and worrying makes my sugar go up and down” (ID 310).

A number of participants saw either avoiding or emulating family members’ experiences with diabetes as an indicator of their success. One said of her brother, who had died of diabetes-related kidney failure, “And then, when he got older out on his own, he went through the amputation, and he [wound] up on kidney dialysis. And then, he passed away at the age of 33 because of kidney failure. So, I use him as an example for myself because I don’t want to go through that” (ID 119). Others saw being a role model to family as a sign of successful management. One person said, “I just want to be an inspiration for someone who’s going through the same thing that I’m going through. So, I can make it out of this. And if I can do it, they can do it” (ID 221).

In this study, we interviewed 89 individuals with a history of elevated A1C, diabetes-related ED visits, or diabetes-related hospitalizations from the primary care, acute, and post–acute care settings to understand their perceptions of successful diabetes management. This was a primarily Black population, with recent diabetes-related ED visits or hospital admissions or elevated A1C who sought care in an urban academic health system. Although most participants relied predominantly on clinical assessment of A1C or blood glucose to assess management success, we identified several other patient markers of successful management, including engagement in medical care; taking medication and the types taken; presence or absence of symptoms; diet, exercise, and weight management; and stress management and family. Patients often relied on multiple markers of diabetes management to gauge their personal success.

Patients’ knowledge of A1C values, as well as other clinical measures, plays an important role in achieving A1C targets and supporting the clinical management of diabetes. However, it is widely reported that an understanding of A1C is not sufficient for meeting A1C targets and that the language of “poor control” can lead to feelings of stigma and shame for individuals who do not meet these targets (15,16). To date, few studies have explored patient-identified indicators of successful diabetes management. Our findings provide additional support for Gopalan et al.’s markers of successful diabetes management, which included A1C and blood glucose monitoring, type and number of medications taken, presence or absence of symptoms, and engagement in the self-care behaviors of diet, exercise, weight management, and seeking medical care (18). Our study expands on this work by engaging a larger sample size from a range of care settings and finding evidence of support among patients for the incorporation of additional measures that extend beyond A1C into clinical discussions and treatment planning.

We identified a new subtheme within engagement in medical care—specifically, the setting of explicit diabetes management goals based on shared decision-making with HCPs. A large body of evidence has identified the importance of shared decision-making in diabetes care, particularly given the complexity of treatment regimens and the frequent presence of comorbidities (25,26). We also found a new theme of stress management and family as a patient-identified marker of successful diabetes management. Within this theme, family and friends were viewed as sources of either stress or support, and prior experiences of family members and friends with diabetes helped shape participants’ views of successful management. This finding aligns with our previous research on the influence of family members’ diabetes experiences (27) and additional research on the positive and negative influences of family members on diabetes self-management (28). Both themes expand the concept of diabetes management beyond the individual patient and illustrate the key role of relationships, both patient-provider and personal, in diabetes management success.

Our study is one of the first to ask patients directly how they assess whether their diabetes management is successful. Other study strengths include its large sample across multiple care settings. Additionally, our focus on a primarily Black, lower-income population highlights patient measures of successful management in a population disproportionately affected by type 2 diabetes.

Our study does have limitations. Our sample only included individuals with elevated A1C and/or diabetes-related ED visits or hospitalizations; therefore, individuals meeting ADA-recommended glycemic goals were not included in our sample. Additionally, although we sampled a particularly vulnerable patient population, this strategy may have limited the generalizability of our findings to other patient groups. Although we randomly selected individuals to approach from among those eligible, some declined to participate, which may have contributed to selection bias. Future studies should explore patients’ perceptions of successful diabetes management in a wider range of demographic groups and geographic settings and explore how to incorporate patient-important measures of successful management into diabetes clinical care, education, and quality measurement.

Individuals with diabetes have multiple markers of successful management beyond A1C and have a desire for management discussions that incorporate these additional markers. This fact has multiple implications for clinicians and other members of the multidisciplinary diabetes care team. First, our findings indicate that a number of individuals may benefit from discussions focused on clarifying measurements of A1C and blood glucose, which may help to improve clinical outcomes. Second, our list of patient-identified indicators of successful diabetes management can spur conversations with patients about their own management goals that can be incorporated into their care plans along with clinical targets, which may improve quality of life. Follow-up conversations about these goals can use a strengths-based approach to provide positive reinforcement and identify additional supports such as family members that patients can draw on for help in meeting their goals. Depending on the conversation topic, clinicians may wish to draw on the growing body of diabetes decision aids (25,26,29). Clinicians may also wish to track patient-reported management measures, as is becoming more common (30).

Diabetes educators will likely note that the patient-identified indicators of successful diabetes management are consistent with the American Association of Diabetes Educators’ AADE7 Self-Care Behaviors (healthy eating, being active, monitoring, taking medication, problem-solving, reducing risks, and healthy coping) (31) that form the core of diabetes self-management education (DSME). Despite strong evidence of the efficacy of DSME for improving knowledge, self-management behaviors, psychosocial well-being, and A1C and the ADA’s recommendation that all individuals with diabetes receive DSME, it remains woefully underused (32). Our study findings present an impetus for clinicians to create new referral and delivery processes with diabetes educators to enhance patient access to DSME, which can facilitate meeting patient-centered diabetes management goals.

The authors thank the Patient and Key Stakeholders Advisory Board members for their assistance in designing and conducting this study. They also thank Lori Latimer for her assistance with data organization and analysis.

Research reported in this article was funded through a Patient-Centered Outcomes Research Institute (PCORI) Award (ME-1503-28476). The statements presented in this article are solely the responsibility of the authors and do not necessarily represent the views of PCORI or its Board of Governors or Methodology Committee.

No potential conflicts of interest relevant to this article were reported.

A.T.C. and P.A. drafted the manuscript, and all authors contributed substantially to its revision. A.T.G. and A.M.B.D. performed data collection and primary data analysis, with all other team members contributing to data interpretation. A.T.C., A.T.G., A.M.B.D. managed the data. G.D.M., M.D.L., B.G.C., J.E.H., and K.L.R. designed the study. M.D.L. and K.L.R. conceived the study and obtained research funding. K.L.R. supervised the conduct of the trial and data collection. All authors read and approved the final manuscript. A.T.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.