Abstract
OBJECTIVE Disparities in type 1 diabetes related to use of technologies like continuous glucose monitors (CGMs) and utilization of diabetes care are pronounced based on socioeconomic status (SES), race, and ethnicity. However, systematic reports of perspectives from patients in vulnerable communities regarding barriers are limited.
RESEARCH DESIGN AND METHODS To better understand barriers, focus groups were conducted in Florida and California with adults ≥18 years old with type 1 diabetes with selection criteria including hospitalization for diabetic ketoacidosis, HbA1c >9%, and/or receiving care at a Federally Qualified Health Center. Sixteen focus groups were conducted in English or Spanish with 86 adults (mean age 42 ± 16.2 years). Transcript themes and pre–focus group demographic survey data were analyzed. In order of frequency, barriers to diabetes technology and endocrinology care included 1) provider level (negative provider encounters), 2) system level (financial coverage), and 3) individual level (preferences).
RESULTS Over 50% of participants had not seen an endocrinologist in the past year or were only seen once including during hospital visits. In Florida, there was less technology use overall (38% used CGMs in FL and 63% in CA; 43% used pumps in FL and 69% in CA) and significant differences in pump use by SES (P = 0.02 in FL; P = 0.08 in CA) and race/ethnicity (P = 0.01 in FL; P = 0.80 in CA). In California, there were significant differences in CGM use by race/ethnicity (P = 0.05 in CA; P = 0.56 in FL) and education level (P = 0.02 in CA; P = 0.90 in FL).
CONCLUSIONS These findings provide novel insights into the experiences of vulnerable communities and demonstrate the need for multilevel interventions aimed at offsetting disparities in diabetes.
Introduction
Health outcomes in type 1 diabetes in the U.S. are profoundly shaped by socioeconomic status (SES), race, and ethnicity from childhood and throughout the life span. People living with type 1 diabetes from low SES households face elevated risks for suboptimal glycemic control, diabetic ketoacidosis (DKA), disease morbidity, and mortality (1–12). Race and ethnic minority status further compound disparate outcomes for non-Hispanic Black (NHB) and Hispanic communities with type 1 diabetes as risks for living in poverty and economic deprivation are two times higher than for White communities (13). In addition, disparate outcomes in type 1 diabetes independent of SES exist based on race and ethnic minority status for multiple comorbidities and diabetes-related death (14–18). These disparities, though long-standing, are being magnified by the devastating impact of coronavirus disease 2019 (COVID-19) for underserved communities with diabetes, and the urgency of addressing health disparities and structural racism in type 1 diabetes is dire (19–24).
Two areas related to disease management in which differences in SES, race, and ethnicity might have bearing on outcomes are utilization of diabetes-related technologies like continuous glucose monitors (CGMs) and insulin pumps as well as routine endocrinology preventative visits. There are marked benefits for using CGM and insulin pumps in type 1 diabetes including improved HbA1c and quality of life; however, these technologies are underutilized in the U.S., and this is especially true for low SES communities and NHBs (18,25–28). Moreover, national claims data in the U.S. demonstrate a significant portion of people with type 1 diabetes do not receive any care from an endocrinologist (29), and other research indicates pervasive underutilization of subspecialists by low SES and NHB individuals living with chronic diseases like type 1 diabetes (30).
Understanding barriers to technology use and to routine endocrinology care for underserved communities with type 1 diabetes is critical in developing targeted interventions to address disparate outcomes. Research on type 1 diabetes tends to rely solely on recruitment efforts through major endocrinology centers (31) and subsequently underrepresents communities with type 1 diabetes most in need of strategic outreach. As part of a larger needs assessment for spearheading a Project Extension for Community Health Outcomes (ECHO) type 1 diabetes outreach program (32,33), focus groups were conducted with individuals living with type 1 diabetes with intentional recruitment efforts aimed at including the perspectives of low SES and racially diverse adults with type 1 diabetes.
Research Design and Methods
Focus groups were conducted from fall of 2018 to spring of 2019, prior to the COVID-19 pandemic, in Florida and California with sessions offered in English and Spanish. Recruitment for focus groups included intentional efforts to have representation from adults who may not see endocrinologists on a regular basis and are often underrepresented in existing research by targeting adults with type 1 diabetes ≥18 years who 1) were hospitalized for DKA in the previous year, 2) had HbA1c >9% and “no showed” to two or more consecutive endocrinology visits, or 3) were seen for primary care needs at a Federally Qualified Health Center (FQHC).
Focus group sessions lasted 1 h, followed a standard protocol script with a semistructured format, and were audio recorded. A multidisciplinary team of diabetes professionals including endocrinologists, clinical health psychologists, epidemiologists, sociologists, community health workers, and public health professionals identified focus group domains and developed questions using existing research on disparities in health outcomes and gaps in knowledge therein (Table 1). Participants completed pre–focus group surveys and, after sessions were finished, were provided $65 cash as incentive for their time. Pre–focus group surveys developed in REDCap were completed on tablets and contained questions about overall participant demographics, technology use, routine endocrinology care use, social support systems, and other questions related to diabetes management. The focus groups were led by trained faculty members with expertise in qualitative research. Acting as the moderator, the faculty member was supported by graduate assistants who had roles that included keeping track of time spent on each prompt to ensure balance, note taking, and focus group debriefs that followed each session. To ensure saturation goals were achieved and for balance across states, we aimed to conduct a minimum of 6 and maximum of 8 focus groups in each state (for a total of 12–16) (34,35). Given the ethical responsibilities surrounding recruiting communities facing unique health risks, at the conclusion of focus groups, participants could opt to learn more about diabetes resources available to them in their local area and were given direct contact numbers for follow-up with the faculty leading the focus group.
Focus group domains, sample questions, and prompts
In total, eight sessions were held in Florida (seven in English and one in Spanish) and eight sessions in California (all in English) averaging five attendees per group. In Florida, six focus groups were filled using call lists for recruitment of people with type 1 diabetes who had been hospitalized for DKA in the past year and/or who had HbA1c values >9% and “no showed” to two or more consecutive endocrinology visits. The other two focus groups in Florida were filled through recruitment at a rural FQHC and a nonrural FQHC. In California, six focus groups were filled using recruitment at FQHCs in rural and nonrural areas, and two were filled with participants from a local conference for people with type 1 diabetes.
Basic demographic characteristics were summarized by site and stratified by key themes of interest: technology use and endocrinology care. Frequencies and percentages are presented for categorical data and means ± SDs for continuous variables. Differences in technology use (pumps and CGMs) by race/ethnicity, SES, and level of education were evaluated using Fisher exact tests to accommodate small sample sizes. Cochran-Mantel-Haenszel nonzero correlation tests were used to assess for trends of increasing frequencies of endocrinology visits by increasing levels of SES and level of education; race/ethnicity testing was not conducted due to sparsity of data. All data management and analysis were conducted using SAS 9.4 (SAS Institute, Cary, NC). Statistical significance was evaluated with a predetermined level of significance of α = 0.05.
Focus group audio recordings were transcribed by Datagain, an external qualitative analysis vendor, and a thematic analysis was conducted using two coders using a method of constant comparison associated with grounded theory (36). In addition to the external thematic analysis conducted by Datagain, subsequent content analysis (36) was conducted by two coders of all quotes from participants related to barriers to technology use and routine endocrinology care. The Social Ecological Model (37) (Fig. 1) was used as a framework for the coding scheme. The Social Ecological Model allows for identification of barriers at multiple levels, including individual, interpersonal, community, and societal/policy level (37). The Social Ecological Model provides an ideal scaffolding in disparities research, allowing barriers to be examined at different levels, and, subsequently, facilitates the development of targeted interventions accordingly (38). We adapted the Social Ecological Model in a coding schematic with categories that included: 1) individual-level barriers related to personal values, beliefs, and preferences; 2) provider-level barriers related to interactions with health care providers or clinic protocols; and 3) system/policy-level barriers related to insurance or financial coverage. The κ statistic, computed in SAS, was used to evaluate the interrater reliability for identification of the primary barrier themes related to access and use for each endocrinology and technology. Raters were evaluated on their agreement about the presence or absence of selected themes across the transcripts. κ coefficients were evaluated using guidance by Landis and Koch, in which the strength of the κ coefficients of agreement is 0.01–0.20, slight; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; and 0.81–1.00, almost perfect (39).
Social Ecological Model and multilayered barriers for underserved communities.
Results
Findings From Pre–Focus Group Surveys
A total of n = 86 adults (51 in FL and 35 in CA) with type 1 diabetes participated in focus groups (see Table 2 for participant demographics). Almost half of the participants in both states had total annual household incomes of <$35,000/year (47% in FL and 49% in CA), an overrepresentation of this demographic group, as only 38% of households fall in this income bracket in the U.S. (40). However, education levels were lower for participants in Florida (41% of FL participants had a high school degree or less; 11.5% of CA participants had a high school degree or less). In both states, >60% of participants were non-Hispanic White (NHW), and Florida had the greatest number of NHB and Hispanic participants.
Focus group demographics and technology use
Surveys completed prior to focus groups revealed low levels of technology use for Florida’s participants: 38% used CGMs in Florida compared with 63% in California and 43% used pumps in Florida versus 69% in California (Table 3). There were also significant differences in pump use by SES (P = 0.02 in Florida; P = 0.08 in California) and race/ethnicity (P = 0.01 in Florida; P = 0.80 in California). In California, there were significant differences in CGM use by race/ethnicity (P = 0.05 in California; P = 0.56 in Florida) and education level (P = 0.02 in California; P = 0.90 in Florida). Thus, while rates of technology use were higher in California, there were statistically significant differences based on race and ethnic minority status and education.
Pre–focus group surveys: visits to endocrinologists
In Florida, 16% reported never seeing an endocrinologist in the past year, and 22% had seen an endocrinologist only one time. In California, 51% had never seen an endocrinologist in the past year, and 23% only saw an endocrinologist once. Despite recruitment methods being heavily focused on those hospitalized with DKA in Florida, and this metric including seeing an endocrinologist during hospitalizations, the number of participants meeting recommendations to see an endocrinologist four times a year for routine preventive care was still very low (20% in Florida and 11% in California). In Florida, evidence suggests there was an association between SES and more visits with an endocrinologist as measured by total household income (P = 0.04). There was not a statistically significant or clear pattern in the distribution of frequency of visits and increasing education in California.
Findings From Focus Group Transcripts: Barriers to Technology Use and Endocrinology Care
The first step in thematic analysis was conducted by Datagain and identified two overall themes related to barriers for participants: 1) unclear qualifying criteria for receiving a CGM or pump and 2) endocrinologists are judgmental and lack understanding. To better differentiate these themes, a secondary content analysis (36) was subsequently conducted using all answers related to barriers for technology use and endocrinology use (a total of 28,496 words from focus group transcripts) using an adaptation of the Social Ecological Model and multilayered factors that impact health outcomes. Two coders identified if participants were referring to 1) individual-level barriers related to personal values, beliefs, and preferences, 2) provider-level barriers related to interactions with health care providers or clinic protocols, and 3) system/policy-level barriers related to insurance or financial coverage. In order of frequency, the barriers were identified as provider-level (16,209 words used in 125 separate accounts), system-level (10,460 words used in 84 separate accounts), and individual-level (1,827 words used in 16 separate accounts) with an interrater reliability score of 0.80 (95% CI 0.72, 0.89) for endocrine themes and 0.86 (95% CI 0.73, 0.99) for technology themes, indicating strong agreement in identification of major themes between coders.
The most pervasively identified barrier for technology use and endocrinology use was provider-level barriers (Table 4). For technology use, participants repeatedly listed system-level barriers related to financial coverage or insurance as preventing them from using or continuing to use technologies like CGM. Participants noted the “living hell” and “bureaucracy nightmare” of being unable to obtain coverage for desired technologies: a Florida participant shared: “When I was a child, I would be eligible for an insulin pump … but I can’t get this covered as an adult. I’ve had no such luck.” In California, another participant noted: “I’ve come into an issue with insurance and I don’t know how to get a pump. I’ve been trying to get a pump and a CGM for like 4 years.” In other instances, individual-level barriers were described, like difficulty wearing a CGM or pump in jobs in which participants sweated excessively or completed manual labor. “I am a waitress and mom, and the pump was just so much to carry that around” or “I sweat constantly as a construction worker, and it just doesn’t work.”
Quotes about provider-level technology and endocrinology barriers
Yet, the most commonly identified barriers to technology use were at the provider level and related to information exchanged in face-to-face visits. This is referred to as interpersonal barriers in the Social Ecological Model (37). Participants in both states chronicled similar narratives of seeking technologies like CGM to improve their diabetes management but being overtly blocked or discouraged by endocrinologists who conveyed a range of discouraging feedback. The reasons given to participants as to why they could not pursue CGM or pump frequently cited “poor control” (HbA1c >9%) or an implication that a technology was too difficult for the participant to handle. Participants noted the irony of needing and wanting to improve glycemic control but being blocked from a technology based on elevated HbA1c levels. Examples of quotes that demonstrate this are as follows:
“I have had my [endocrinologist] recommend that I get my A1C down below 7 before they will even attempt to let me use any kind of device [pump or CGM]. With my lifestyle and work and everything, the way it is, it’s super hard to manage the diabetes as it is, which personally I think that would be a tool that would help me out a lot.” –Florida Focus Group Participant
“I tried to get on the pump, and he [endocrinologist] told me I needed to lose weight and the only way I was going to get on a pump was to lose weight and to be in a lot of control.” –California Focus Group Participant
Provider-level barriers evoked through interpersonal encounters were also the most commonly cited barriers for utilization of routine endocrinology visits. In recalling the transition from pediatric endocrinology to adult endocrinology, participants noted a stark contrast from warm and highly engaged pediatric endocrinologists to the world of adult endocrinology in which they described “being completely on [their] own.” In both states, participants frequently described clinical encounters with adult endocrinologists in which they experienced being demeaned, “belittled,” and misunderstood as a person living with type 1 diabetes. A Florida participant noted: “I feel like sometimes I don’t even want to go to the doctor because I’m getting a lecture. I’m getting like in trouble. I feel like a little kid who’s going to the principal.” California participants described similar experiences noting “endocrinologists are more like the judges…and the visits I’m on trial,” and “basically, they kind of talk to you like…belittling you that you don’t know what you’re doing with your diabetes or something.”
Moreover, participants did not see the value in the visit with adult endocrinologists as a mechanism of improving their overall health, and, thus, many simply stopped going. Participants also noted frustrations over protocols at endocrinology offices that did not allow them to schedule appointments as needs arise or in ways that accommodate working-class communities. One California focus group participant noted:
“And usually the [endocrinology] appointments are like 10 AM, 11 AM, or 2 PM. So, I end up having to take half the day off work to go to the appointment, and then also find time for the blood work and everything. So that was one…another, like, reason why I went the 2 years without seeing my endocrinologist was there was just too much going on, and it didn’t seem like a valuable enough resource to me to make it work within my schedule.” –California Focus Group Participant
Additionally, distance was identified as a barrier to utilization of endocrinology care in the focus groups that took place in rural FQHC settings. Examples of distance and transportation barriers are as follows:
“I have neuropathy in both of my feet and occasionally, I feel it in my hands. On top of the distance, doing that drive or walking or standing for long periods of time is difficult. I also have a sleep issue that they can’t figure out what the issue is. So, long drives and being alone for long drives, because I don’t…all my friends and all my family work…it also…it can become hazardous at times.” –Florida Focus Group Participant
“I have to take three transports to come because, of course, I don’t have a car. Or count on the goodwill of someone to bring me, but it’s quite far.” –Florida Focus Group Participant
The only notable difference between focus group transcripts in Florida and California was that participants in Florida described being “kicked out” of their adult endocrinology clinics for “no showing” to a scheduled visit. The guidelines for being removed from a clinic for failing to come to a routinely scheduled visits varied, but were only described by Florida participants. A Florida focus group participant described this as follows: “I was just recently dismissed from the endocrinology practice because I had missed three appointments. Two of which I was admitted in the hospital for, so [I’m seeing] nobody at this point.” This small difference between Florida and California aside, thematic analyses revealed similar types of experiences and barriers in both states. The urgency of improving diabetes management was abundantly clear for participants, many of whom were facing dire health complications, including lower limb amputation and recurrent hospitalization for DKA.
Conclusions
Findings from focus groups conducted in Florida and California with strategic recruitment efforts designed to include representation from underserved communities with type 1 diabetes demonstrate multilayered barriers related to technology use and routine endocrinology care. Barriers at the system level related to financial costs were described as prohibiting utilization of technology, but overall, the most commonly described barriers for technology and endocrinology use were related to provider-level factors and interpersonal communications during clinical encounters. It is important to note that, in some cases, providers may simply be conveying their own perceptions about rules related to coverage by which insurance and device companies have established thresholds related to HbA1c. Regardless, these thresholds, and providers educing them in face-to-face interactions, result in systemically denying access to patients most in need of these tools.
Furthermore, the lack of empathy and understanding that patients experienced from adult endocrinologists was identified as a significant obstacle. These findings support other studies that have documented the experience of stigma for communities living with type 1 diabetes (41,42) and, additionally, demonstrate the powerful role that providers can play in evoking type 1 diabetes–related stigma. For underserved communities, the intersectionality of type 1 diabetes as a disease with other status characteristics like SES, as well as structural racism, may compound the experience of stigma. While this study focused on barriers at the individual, provider, and systems levels, future directions should include consideration of the synthesis of barriers at various levels and its unique impact on underserved communities.
To our knowledge, this study provides one of the only accounts of perspectives from underserved communities with type 1 diabetes in the U.S. in two very different geographic locations—the deep South and the Western Pacific Coast. The states also differ in that Florida did not expand Medicaid, whereas California did. Thus, the similarities in focus group themes from both states point to the importance of multilayered interventions in various contexts in the U.S. However, the study has limitations, including the small cohort size and limited generalizability. Given the limited size of the cohorts in both states, caution should be used when examining the significance testing from the pre–focus group survey data. Additionally, the differences in technology use in both states might be related to larger policy-level differences related to insurance coverage (43). Finally, focus groups did not allow for a systemic analysis of potential differences in experiences related to racial and ethnic minority status. While there was one focus group conducted in Spanish, groups were not structured in such a way to assign participation based on race or ethnicity, and questions were not directly included about racial discrimination. Discussion about racial discrimination did come up at two focus groups in Florida for NHB and Hispanic participants; however, future research will benefit from overt efforts to strategically focus on structural racism.
The urgency of addressing disparate outcomes for underserved communities with diabetes is abundantly clear in data we see emerging regarding the COVID-19 pandemic and further highlights systemic and pervasive health inequalities within the U.S. (19–22). Multilayered interventions are needed to promote health equity in diabetes. System-level barriers related to access and financial cost are widely recognized as obstacles for underserved communities, as is the need for efforts aimed at improving disease management at the individual level. Based on these data, we posit that there is also a critical need for provider-level interventions focused on reducing implicit biases and improving the experiences of people with type 1 diabetes in clinical encounters. Endocrinologists are important gatekeepers for the patients they serve and should be included in targeted interventions for reducing health disparities in diabetes.
Article Information
Acknowledgments. The authors thank focus group participants for sharing experiences and perspectives.
Funding. Funding for this research was provided by the Leona M. and Harry B. Helmsley Charitable Trust.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. A.F.W. designed the study, oversaw the conduct of the study, and wrote the manuscript. M.J.G. and S.L.F. had access to data and conducted quantitative analysis. K.K.H., N.C., M.J.H., D.N., A.A., S.W., and D.M.M. contributed input to the study design, reviewed findings, and provided critical revisions to the manuscript. C.A.-Z., X.R., G.A., K.Y., and S.L. conducted focus groups and aided with qualitative data analysis and coding. J.M. conducted qualitative analysis and coding and provided critical revisions to the manuscript. J.K., E.M., and B.F. aided with conducting focus groups, served as content experts for the study analysis, and provided critical revisions to the manuscript. A.F.W. and D.M.M. are the guarantors of this work and, as such, had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Prior Presentation. This work was presented at the 80th Scientific Sessions of the American Diabetes Association, 12–16 June 2020.
- Received November 10, 2020.
- Accepted March 31, 2021.
- © 2021 by the American Diabetes Association