Cost,Hassle,and On-Body Experience: Barriers to Diabetes Device Use in Adolescents and Potential Intervention Targets

Abstract

Background:

Adolescents with diabetes have the highest A1cs of all age groups. Diabetes devices (insulin pumps and continuous glucose monitors [CGM]) can improve glycemic outcomes, and although the uptake of devices has increased, they remain underutilized in this population. This study characterizes adolescent-reported barriers to diabetes device use to determine targets for clinician intervention.

Methods:

We surveyed 411 adolescents with type 1 diabetes (mean age 16.30 ± 2.25 years) on barriers to diabetes device use, technology use attitudes (general and diabetes specific), benefits and burdens of CGM, self-efficacy for diabetes care, diabetes distress, family conflict, and depression. We characterize barriers to device uptake; assess demographic and psychosocial differences in device users, discontinuers, and nonusers; and determine differences in device use by gender and age.

Results:

The majority of adolescents used an insulin pump (n = 307, 75%) and more than half used CGM (n = 225, 55%). Cost/insurance-related concerns were the most commonly endorsed barrier category (61%) followed by wear-related issues (58.6%), which include the hassle of wearing the device (38%) and dislike of device on the body (33%). Adolescents who endorsed more barriers also reported more diabetes distress (P = 0.003), family conflict (P = 0.003), and depressive symptoms (P = 0.014). Pump and CGM discontinuers both endorsed more barriers and more negative perceptions of technology than current users, but reported no difference from device users in diabetes distress, family conflict, or depression. Gender was not related to the perceptions of devices.

Conclusions:

Clinicians can proactively assess attitudes toward diabetes technology and perceptions of benefits/burdens to encourage device uptake and potentially prevent device discontinuation among adolescents.

Introduction

In the past decade, self-management of type 1 diabetes has changed dramatically due to diabetes technologies beyond blood glucose meters and insulin injections. Insulin pump use now exceeds multiple daily injection use in the United States, with recent data indicating that 63% of type 1 diabetes exchange (T1Dx) registry participants use an insulin pump.¹ Adolescents have the highest utilization rate of insulin pumps at 68%. Continuous glucose monitors (CGM) are used by a growing percentage of individuals with diabetes. CGM use in adolescents has increased 13-fold from only 3% in 2011 to 38% in 2018.¹ Diabetes technology use is associated with lower A1c values compared with those who do not use insulin pumps or CGM.¹

To promote diabetes technology utilization in individuals with diabetes, it is essential to understand barriers to device uptake. Previous studies identified nonmodifiable barriers to CGM use, including cost and inadequate insurance coverage,^2

–5 and modifiable barriers, including concerns with accuracy,² annoyances from alerts/alarms,^5
–7 body issues related to device wear,^5,7 and interruption from daily life.^8,9 While some of these data are related to barriers in adolescents and children, they are either qualitative in nature,^3,6 or report on outdated technologies.^7,8 The most recent and extensive description of device barriers was reported in 1503 adults with type 1 diabetes who identified cost and/or insurance coverage as the most significant barrier to all diabetes device use (61% of individuals), followed by the hassle of wearing devices (47%) and not liking the device on the body (35%).⁵

A fundamental gap in the extant literature is the lack of a comprehensive, large-scale inquiry into diabetes device barriers in adolescents with diabetes. This is an important population to characterize because adolescents have the highest HbA1c values of any age group across the life span, peaking at the age of 19 years.^1,10 They further have unique developmental and psychosocial characteristics that make their diabetes care experience different from adults.^11,12 The period of adolescence is marked by significant changes to emotional and cognitive awareness,^13,14 increased reliance on peer group identity,¹⁴ changes to family dynamics,^14,15 and increased ability to take responsibility for diabetes care tasks.¹⁵ Furthermore, unlike adults, the decision to wear devices is often shared between the adolescent and the parent/caregiver, and adolescents may be asked to wear devices even if they are personally ambivalent to the technology. By understanding the barriers to diabetes devices for adolescents, clinicians may better determine realistic strategies to promote device use and overall self-management practices in this vulnerable population.

Therefore, the purpose of this study is to characterize barriers to diabetes device use in adolescents with type 1 diabetes. The three aims were to (1) describe barriers endorsed by adolescents about diabetes devices; (2) explore demographic and psychosocial differences in individuals who use devices, have discontinued devices, and have never used devices; and (3) determine associations between gender and age with diabetes device uptake.

Methods

Study design

Data were collected from adolescents with type 1 diabetes, ages 12–19 years, in a cross-sectional online survey. Information about the study was e-mailed to participants in the T1Dx registry in spring 2018 (13,152 e-mails sent to parents and adults ages 18–19 from 64 T1Dx registry sites), although registry participation was not a requirement to complete surveys. Parents provided consent for their children younger than 18 years to participate, and adolescents 18–19 years old provided consent for themselves. Survey data were collected via REDCap database,¹⁶ and surveys took 20–30 min to complete. The study was approved by the Stanford Institutional Review Board. Deidentified data analyses were also approved by the Colorado Multiple Institutional Review Board.

Measures

Demographic and diabetes management information

Study participants reported their date of birth, gender, and race/ethnicity. They were also asked whether they currently or previously used an insulin pump and/or CGM. For individuals who were participants in the T1Dx clinic registry, data were linked to registry information, including duration of diabetes and recent HbA1c values (when available).

Checklist of device barriers

Participants were given a list of 19 barriers to diabetes device use, derived from a previous study,⁵ and asked to endorse the specific barriers that applied to them. Three of the barriers were nonmodifiable: cost of the device, cost of supplies, and insurance coverage. The other 16 barriers were modifiable, person-focused barriers, including not liking the device on the body, feeling nervous the device might not work, hassle of wearing devices all the time, and worries about what other people will think. Participants could select between 0 and 19 barriers.

Benefits of CGM and burdens of CGM surveys

The benefits of CGM scale (abbreviated BenCGM) and burden of CGM scale (abbreviated BurCGM) were used to assess specific perceptions about CGM use.¹⁷ Each scale comprised eight items related to BenCGM (e.g., “CGM makes taking care of diabetes easier”) and BurCGM (e.g., “CGM takes too much effort to use”). Individual items are scored on a Likert scale of 1–5 (1 = strongly disagree, 5 = strongly agree), and the total scale score is the mean of all items. The BenCGM and BurCGM scales were originally validated in this sample, with Cronbach's alphas being 0.89 (BenCGM) and 0.87 (BurCGM), indicating good reliability.

Technology attitudes: general and diabetes specific

Attitudes toward technology were measured with a 6-item general scale (Cronbach's alpha = 0.95) and a 5-item diabetes-specific scale (Cronbach's alpha = 0.96) used in a previous study.⁵ Participants rated their agreement on a 1–5 Likert scale (1 = strongly disagree, 5 = strongly agree) with statements such as “Technology has made my life better,” and “Diabetes technology has made my life easier.” Scores were sum totals.

Diabetes distress

Diabetes distress was measured with the 20-item Problem Areas in Diabetes scale, Pediatric version (PAID-Peds) (Cronbach's alpha = 0.94).¹⁸ Participants rated statements about burden of diabetes in the past month on a 5-point Likert scale (0 = agree, 4 = disagree) denoting how strongly they agree with each statement. The total score is the mean of all items, normalized to a 100-point scale by multiplying by 25, with higher scores indicating higher distress.

Self-efficacy

Self-efficacy was measured by the Self-Efficacy for Diabetes Management scale (Cronbach's alpha = 0.91), a 10-item questionnaire with a single total mean score.¹⁹ The stem question asks “How sure are you that you can do each of the following, almost all the time?” and then lists diabetes tasks. Participants rated each item on a 1–10 scale (1 = not at all, 10 = completely sure).

Family conflict

Family conflict was assessed using the 19-item Revised Diabetes Family Conflict scale—Child Version (Cronbach's alpha = 0.93).²⁰ The questionnaire stem states, “During the past month, I have argued with my parents about…,” followed by 19 items related to diabetes tasks that participants rate on a 3-point scale (1 = almost never, 2 = sometimes, 3 = almost always). Scale score is the sum (possible range 19–57), with higher numbers indicating more conflict.

Depressive symptoms

Depressive symptoms were assessed with the Patient Health Questionnaire-8 (PHQ-8), which comprised the same items from the PHQ-9,^21
–23 with the suicide question omitted (Cronbach's alpha = 0.92). The questionnaire includes eight items related to how the participant has felt in the last 2 weeks (e.g., “Little interest or pleasure in doing things”). Participant responses were scored 0 (not at all) to 3 (nearly every day), providing a final total score with a possible range from 0 to 24.

Analysis

Demographic data are reported as counts and proportions. Since HbA1c was only available for a small portion of the total sample (n = 73 out of 411, 18%), t-test and χ ² statistics were used to determine differences between those who had HbA1c data and those who did not. Duration of diabetes had 22% missing data that were missing at random (due to T1Dx data on file) and multiple imputation was used to fill in missing values.

For Aim 1 (characterize barriers), percentages of adolescents who endorsed each barrier were tabulated and ranked; Spearman correlations were used to determine correlation between the number of barriers and demographic and psychosocial measures. Differences in number of barriers by gender and age (divided into younger ages 12–15 years and older ages 16–19 years) were assessed by the Mann–Whitney U rank test. In addition, the barriers were condensed into five main categories for further analysis: cost (cost of devices, supplies, and insurance), wear related (hassle, not liking device on body, not liking look of diabetes devices on body, worries what others will think, people will notice and ask questions), technology anxiety (nervous device won't work, nervous to rely on technology, don't want more information), time (too busy to use, not wanting to take more time), and lack of support (diabetes care team will not write prescription, diabetes care team not talked about technology options, not enough support from diabetes care team, not enough support from family, family does not think devices are important).

For Aim 2, one-way ANOVA was used to determine differences between current pump users, pump discontinuers, and pump never-users in number of barriers to device use, age, duration of diabetes, and psychosocial measures. Since the samples were different sizes, post hoc Gabriel tests were used to determine significant differences between groups. We did not test for differences in demographic or psychosocial variables between those who used different combinations of technology (e.g., pump/CGM, nonpump/CGM) as the sample size for nonpump, CGM users was too small (n = 20).

For Aim 3, two-way ANOVA was used to determine if there was an interaction between age and gender on device use, number of barriers, and psychosocial variables. Significant differences between males and females were assessed by t-test and χ ².

Significance was set at the alpha <0.05 level. The Benjamini–Hochberg procedure was applied to each aim to control for the false discovery rate. All analyses were performed using SPSS version 26 (IBM Corporation, Chicago, IL).

Results

Sample characteristics

In total, 411 adolescents, ages 12–19 years, with type 1 diabetes completed the online questionnaires. Mean age of respondents was 16.30 ± 2.25 years, with 9.54 ± 3.53 years duration of diabetes, and 52.8% females (n = 217). Eighty-three percent of respondents identified as white, non-Hispanic; 11.4% as Hispanic/Latino and 5.6% as other race/ethnicity. For device use, 307 (74.7%) participants used insulin pump, and 225 (54.7%) participants used CGM (6 participants omitted an answer about either pump or CGM). Of the CGM users, 99% used real-time CGM and 1% (n = 4) used intermittently scanned glucose monitoring. Mean HbA1c was 8.6% ± 1.44% (70 mmol/mol) (n = 73). There was no difference in age, duration of diabetes, proportion of insulin pump or CGM users between participants who had HbA1c data and those who did not.

Aim 1: barriers to device use

The most frequently cited barriers were nonmodifiable factors: cost of devices, cost of supplies (e.g., infusion sets, CGM sensors), and insurance coverage (Table 1). The most frequently endorsed modifiable barriers include the hassle of wearing devices (37.7%), disliking the device on the body (33.1%), and disliking how the device looks on the body (29.2%). One-quarter of participants (24.6%) were nervous that the device might not work, and approximately one in five did not want to take more time to manage diabetes or were worried what others would think of them. The least commonly cited barriers (<2% of sample) included not having support from family or diabetes providers, not wanting information about their diabetes, and family not thinking diabetes devices were important. There was no difference in the percentage of males and females who endorsed each barrier, nor were there differences between young adolescents (ages 12–15 years) and older adolescents (ages 16–19 years).

Table 1.

Most Frequently Cited Barriers to Device Use Reported by Study Participants (n = 411)

Barriers	n	%
Nonmodifiable
Cost of device	193	47.0
Cost of supplies	188	45.7
Insurance coverage	171	41.6
Modifiable
Hassle of wearing devices all of the time	155	37.7
Do not like having diabetes devices on my body	136	33.1
Do not like how diabetes devices look on my body	120	29.2
Nervous that the device might not work	101	24.6
Do not want to take more time from my day to manage diabetes	84	20.4
Worried about what others will think about me	82	20.0
Nervous to rely on technology	71	17.3
Do not like diabetes devices because people notice them and ask questions about them	70	17.0
Too busy to learn how to use a new technology or device	29	7.1
Not able to get diabetes care team to write a prescription	14	3.4
My diabetes care team has never talked with me about diabetes technology options	12	2.9
Do not understand what to do with the information or features of the devices	8	1.9
Not enough support from my family	6	1.5
Not enough support from my diabetes care team in using devices	6	1.5
Do not want to have more information about my diabetes	6	1.5
My family does not think diabetes devices are important for taking care of my diabetes	1	0.2

Participants reported a mean of 3.54 ± 0.59 barriers and a median of 3.00 (IQR 3) barriers to diabetes devices (range 0–16 out of 19 total barriers). A total of 38 participants (9.2%) endorsed no barriers to diabetes devices. The number of barriers was positively skewed, and thus, Spearman's rho (r _s) was used to report correlation. More barriers to device use were correlated with lower diabetes technology attitudes (r _s = −0.133, P = 0.014) and lower self-efficacy (r _s = −0.165, P = 0.002) (Table 2). Endorsing more barriers was also moderately correlated with higher family conflict (r _s = 0.213, P = 0.002), higher diabetes distress (r _s = 0.4, P = 0.002), and higher depression scores (r _s = 0.272, P = 0.014). The number of barriers was not significantly associated with any other demographic or psychosocial variable.

Table 2.

Correlation of Number of Endorsed Barriers with Demographic and Psychosocial Metrics (n = 411)

	Spearman's rho (r _s)	P^*
Age	0.091	0.087
Duration of diabetes	0.008	0.880
Technology attitudes	−0.092	0.087
Diabetes technology attitudes	−0.133	0.014^a
Self-efficacy	−0.165	0.003^a
Family conflict	0.213	0.003^a
Diabetes distress	0.400	0.003^a
Depression	0.272	0.014^a

P value adjusted for multiple comparisons.

Significant at the P < 0.05 level.

When barriers were condensed into five broad categories, 60.8% (n = 250) endorsed cost-related concerns, 58.6% (n = 241) wear-related concerns, 31.9% (n = 131) technology anxiety-related concerns, 24.1% (n = 99) endorsed time concerns, and 6.3% endorsed support-related concerns (Supplementary Table S1). Participants endorsed a mean of 1.82 ± 1.28 barrier categories out of five. Nearly one-third (31.9%) of the sample endorsed both cost- and wear-related issues, and nearly one-quarter (24.3%) endorsed both anxiety- and wear-related issues (Supplementary Table S1).

Aim 2: differences between device users, discontinuers, and nonusers

The majority of participants used an insulin pump (307 out of 406, 75.6%), whereas 31 participants discontinued using an insulin pump (7.6%), and 68 participants never used a pump (16.8%) (Table 3). Pump discontinuers reported the largest number of barriers to diabetes devices (5.15 ± 2.88) compared with current pump users (3.36 ± 2.50) and pump nonusers (3.62 ± 2.72, F = 6.97, P = 0.002). Insulin pump discontinuers were older (mean age 17.19 ± 2.13 years) than pump users and nonusers (F = 3.73, P = 0.002), but were not different in duration of diabetes (F = 2.29, P = 0.141). Current insulin pump users reported the highest perceived BenCGM, the lowest perceived BurCGM, and the most positive technology and diabetes technology attitudes. There were no differences between pump groups in diabetes distress, self-efficacy, family conflict, or depression scores. In post hoc analyses, the only difference between pump discontinuers and those who had never used insulin pump was that pump discontinuers cited more barriers (P = 0.015).

Table 3.

Differences Between Pump Users, Discontinuers, and Nonpump Users (N = 406)

		Pump users n = 307 (75.6%)	Pump discontinuers n = 31 (7.6%)	Pump nonusers n = 68 (16.8%)	F	P^*
No. of barriers reported	0–19, lower is less barriers	3.36 ± 2.50	5.15 ± 2.88	3.62 ± 2.72	6.97	0.002^a,b
Age		16.16 ± 2.31	17.19 ± 2.13	16.63 ± 1.89	3.73	0.046^a
Diabetes duration		9.74 ± 3.60	9.20 ± 3.84	8.74 ± 3.40	2.29	0.141
BenCGM score	1–5, higher is more perceived benefit	4.13 ± 0.66	3.81 ± 0.71	3.78 ± 0.69	9.58	0.002^a,c
BurCGM score	1–5, higher is more perceived burden	2.12 ± 0.69	2.48 ± 0.68	2.45 ± 0.70	9.00	0.002^a,c
Technology attitudes: general	5–30, higher is more positive attitude	24.84 ± 5.44	22.36 ± 6.31	21.70 ± 6.31	10.12	0.002^a,c
Technology attitudes: specific	5–25, higher is more positive attitude	21.60 ± 4.07	18.33 ± 6.17	19.44 ± 4.59	10.99	0.002^a,c
Diabetes distress	0–100, higher is more distress	52.80 ± 22.84	62.70 ± 4.46	51.84 ± 25.42	2.69	0.108
Self-efficacy	1–10, higher is more self-efficacy	7.00 ± 1.80	6.76 ± 1.70	7.05 ± 1.85	0.31	0.740
Family conflict	19–57, higher is more conflict	27.53 ± 7.71	30.42 ± 10.84	29.27 ± 9.21	2.62	0.740
Depression	0–24, higher is more depression	4.94 ± 5.55	5.97 ± 6.38	5.84 ± 6.19	1.00	0.455

P value adjusted for multiple comparisons.

Gabriel post hoc test indicated significant difference between users and discontinuers.

Gabriel post hoc test indicated significant difference between discontinuers and nonusers.

Gabriel post hoc test indicated significant difference between users and nonusers.

BenCGM, benefits of continuous glucose monitors; BurCGM, burdens of CGM.

CGM was used by 227 (55.4%) participants in the sample, whereas 55 participants (13.4%) discontinued CGM, and 128 participants (31.2%) never used CGM (Table 4). CGM discontinuers reported the largest number of barriers to device use (4.31 ± 2.85, F = 3.67, P = 0.004), significantly more than CGM users (P = 0.017), but not different from those who had never used CGM. CGM discontinuers also had the longest duration of diabetes (F = 3.61, P = 0.044), although there was no difference in age between the groups (F = 0.95, P = 0.327). Current CGM users reported the highest perceived BenCGM, lowest perceived BurCGM, highest technology attitudes, highest diabetes technology attitudes, and highest self-efficacy. There were no differences in diabetes distress, family conflict, or depression among the three groups.

Table 4.

Differences Between Continuous Glucose Monitor Users, Discontinuers, and Noncontinuous Glucose Monitor Users (n = 410)

		CGM users, n = 227 (55.4%)	CGM discontinuers, n = 55 (13.4%)	CGM nonusers, n = 128 (31.2%)	F	ANOVA P ^*
No. of barriers reported	0–19, lower is less barriers	3.29 ± 2.52	4.31 ± 2.85	3.66 ± 2.54	3.67	0.044^a
Age		16.20 ± 2.33	16.55 ± 2.09	16.37 ± 2.14	0.95	0.327
Diabetes duration		9.76 ± 3.53	10.42 ± 3.69	8.81 ± 3.57	4.82	0.017^a,b
BenCGM score	1–5, higher is more perceived benefit	4.37 ± 0.48	3.53 ± 0.71	3.65 ± 0.65	89.27	0.003^a,b
BurCGM score	1–5, higher is more perceived burden	1.89 ± 0.57	2.69 ± 0.66	2.56 ± 0.65	68.81	0.003^a,b
Technology attitudes: general	5–30, higher is more positive attitude	25.32 ± 5.51	23.22 ± 6.13	22.46 ± 5.65	11.35	0.003^a,b
Technology attitudes: specific	5–25, higher is more positive attitude	22.00 ± 4.09	20.11 ± 5.04	19.71 ± 4.41	12.47	0.003^a,b
Diabetes distress	0–100, higher is more distress	51.40 ± 23.68	55.96 ± 21.29	55.71 ± 23.74	1.78	0.209
Self-efficacy	1–10, higher is more self-efficacy	7.25 ± 1.72	6.72 ± 1.75	6.70 ± 1.89	4.82	0.017^b
Family conflict	19–57, higher is more conflict	27.12 ± 7.93	29.89 ± 8.21	28.77 ± 8.71	3.32	0.051
Depression	0–24, higher is more depression	4.96 ± 5.46	4.54 ± 5.74	5.75 ± 6.11	1.12	0.327

P value adjusted for multiple comparisons.

Gabriel post hoc test indicated significant difference between users and discontinuers.

Gabriel post hoc test indicated significant difference between users and nonusers.

Aim 3: device use by age and gender

Two-way ANOVA determined there were no significant interactions between gender and age for use of insulin pump, CGM, or any of the psychosocial instruments. After adjusting for multiple comparisons, males and females did not differ in number of barriers, perceived BenCGM and BurCGM, technology attitudes, diabetes technology attitudes, diabetes distress, self-efficacy, family conflict, or depression. Older age was significantly correlated with more depressive symptoms (r = 0.197, adjusted P = 0.009).

Conclusions

This report indicates that adolescents with type 1 diabetes have substantial and specific barriers to diabetes device use, and is the most comprehensive report of adolescent perceptions on device use to date. Cost- and wear-related issues are experienced by the majority of adolescents. By examining barriers in relation to actual device use and correlating with other patient-reported outcomes, a fuller understanding of the adolescent experience emerges. Previous reports on barriers to diabetes devices reflect adult perceptions,^4,5 relate to outdated technologies,^4,7,8 or are qualitative in nature^3,6 making this adolescent-specific report timely and pertinent to understanding diabetes technology uptake. Despite the pronounced barriers, >75% of the current sample used an insulin pump, 55% used a CGM, and 51% used both technologies, which is higher than previously reported in the T1Dx registry data,¹ and may reflect the ongoing trend of increasing diabetes technology use.

The most common barriers to diabetes device use are cost of devices, supplies, and insurance coverage (cost-related issues), followed by the hassle of devices and dislike of devices on the body (wear-related issues). These barriers are similar to the barriers reported in adults with type 1 diabetes,⁵ and the top eight barriers are nearly the same (cost of supplies and cost of devices are inverse). Although cost issues are often considered “nonmodifiable” barriers, they can be tangentially addressed through device optimization strategies. Pump optimization can include clinicians aiding adolescents in selecting infusion sets that will last the longest (tentative evidence suggests oblique insertion sets are associated with the least amount of failure²⁴), educating on proper insertion techniques, and suggesting that adolescents fill reservoirs with only the amount of insulin needed for 2–3 days of wear. To reduce CGM supply costs due to failure, clinicians can educate adolescents on proper site selection and skin preparation,²⁵ as well as proper calibration techniques. While clinicians may have limited ability to influence costs for individuals, they can use their aggregate influence to advocate for better reimbursement from insurance and government entities for diabetes devices for the adolescent population.

Hassle of device wear has been noted with older technologies,^3,7,26 and continues to be a central theme as diabetes technologies evolve. Current diabetes devices require a moderate to large amount of work from the user to wear: inserting and replacing devices, maintaining adhesion of the device to the skin, configuring device placement around clothes, and managing device failures. In addition to wear-related hassles, devices also engender general hassles such as responding to alerts, suspending insulin delivery, calibrating sensors, and troubleshooting problems. As device technology improves, the burden of user interaction will hopefully diminish. Even in the past 4 years, some CGM systems have eliminated the need for routine calibrations and confirmation via blood glucose for insulin dosing.²⁷ Future automated insulin delivery systems will further reduce hassle by automating more insulin dosing, requiring less input from the user.^28,29 Until all systems significantly reduce the burden of diabetes, clinicians can help adolescents set appropriate expectations of diabetes device use, including workload and hassle of wearing devices.³⁰ These conversations can be important both for device selection and also for daily device use.

In addition to hassle, adolescents reported other specific wear-related issues, including not liking the diabetes device on the body and disliking how it looks on the body, which are consistent with previous studies.^3,5
–7 These concerns are likely due to both physical discomfort and body image or self-image concerns wearing on-body devices. Furthermore, one in five adolescents endorsed being “worried about what others will think of me,” which was twice as frequently endorsed by adolescents than by adults with diabetes (10%).⁵ This is an important distinction between diabetes device-using adolescents and adults—adolescents have significantly higher concern around what others may think of them. Diabetes devices are a visible sign of being different than peers, which may be distressing at a time when peer relationships are of paramount psychological importance.¹² Possible strategies to address on-body concerns include clinicians proactively discussing discreet placement of infusion sites and CGM sensors with adolescents, and promoting strategies to preserve skin integrity.²⁵ Clinicians should be open to the possibly of adolescents benefiting from occasional “pump breaks” or “CGM breaks” to feel less tethered to diabetes devices for special occasions. Strategies to increase the social norm of wearing diabetes devices, or increasing adolescents' comfort with disclosing their diabetes,¹⁵ may help reduce the worries around what others will think of the device wearer.

Barriers related to lack of support were endorsed by <10% of the adolescent sample, which may indicate that adolescents generally feel supported in using diabetes devices both from family and from the diabetes care team, or at least do not cite lack of support as a significant barrier to device use. This pattern is encouraging considering that many clinicians perceive a large number of barriers to diabetes device use,³¹ although perhaps this is less in the case with clinicians who work with adolescents.

While there were several significant correlations between the number of device barriers and psychosocial measures, actual device use is most significantly associated with the number of barriers and cognitive perceptions of diabetes devices (technology/diabetes technology attitudes and BenCGM and BurCGM). Cognitive perceptions of adolescents are likely the most proximal measure to understanding behavior, and can easily be assessed in routine diabetes care. Underlying diabetes distress, family conflict, and depressive symptoms are associated with the number of barriers endorsed, and can also be considered pertinent to the adolescent experience, although not significantly different in device users and discontinuers.

Insulin pump and CGM discontinuers perceive more barriers, lower technology attitudes, less benefit, and more burden to devices than current device users. Furthermore, those who had discontinued insulin pumps were older than current pump users, indicating additional likelihood of pump discontinuation in older adolescent years. Clinicians should include routine assessment of perceived barriers to pump therapy in all adolescents, placing more emphasis on addressing barriers as adolescents age. For CGM users, self-efficacy for diabetes management was significantly lower in discontinuers than in CGM users. This is consistent with self-efficacy literature that reports an association between self-efficacy and diabetes self-management.¹⁹ For adolescents who do not have strong self-efficacy for managing diabetes, the CGM may feel overwhelming and difficult to manage. It is not clear, however, in this cross-sectional study if higher self-efficacy predated CGM use, or if those who use CGM successfully feel self-efficacious to do so, and it is further unclear if this is also true for intermittently scanned CGM. Future prospective work could help clarify this relationship.

This study elucidates a number of important clinical intervention targets for adolescents using diabetes devices, and highlights the heightened importance adolescents place on worrying about what others will think of them compared with adults who use devices. The strengths of this study include the large sample size of adolescents with diabetes, and the use of comprehensive measures of cognitive, emotional, and psychological aspects of diabetes care.

Limitations to this study include the cross-sectional study design, and potential underrepresentation of adolescents not using any diabetes technology (i.e., using insulin injections with meter and finger sticks), or using intermittently scanned CGM. Considering the sample is only a small fraction of eligible adolescents in the T1Dx registry, data may not be representative of all youth with T1D using devices, and may hold bias toward individuals who are more engaged with diabetes technology. Different recruitment methods may yield additional information about diabetes technology trends in this age group. Furthermore, all of the psychological and diabetes technology measures are self-report, and more objective measures (e.g., % time wearing CGM, % time in range) may have uncovered additional relationships of interest. The limited availability of laboratory-acquired HbA1c data is also a limitation. Nonetheless, the current findings provide timely insight into the unique adolescent experience with diabetes devices for self-management of diabetes.

As diabetes technology continues to evolve, user interaction with diabetes devices will change. In the meantime, understanding the adolescent experience with current devices provides meaningful opportunities to reduce barriers, reinforce benefits, and partner with adolescents in the care of their diabetes.

Footnotes

Acknowledgments

The authors thank the Leona M. and Harry B. Helmsley Charitable Trust for funding this research, and all the participants who contributed to this research.

Author Disclosure Statement

L.H.M. is a certified trainer for Medtronic Diabetes, has received consulting fees from Tandem Diabetes Care, DexCom, Inc., Capillary Biomedical, and Clinical Sensors, and her group receives research funding from Medtronic, Tandem Diabetes Care, DexCom, Insulet, and Lilly. P.F.C. is a statistical consultant for Academic Impressions, Inc. K.K.H. has research support from Dexcom for an investigator-initiated protocol and consultant fees from Med-IQ and LifeScan Diabetes Institute. M.L.T., J.J.W., S.J.H., K.A.D., have no disclosures.

Funding Information

This work was funded by the Leona M. and Harry B. Helmsley Charitable Trust.

Supplementary Material

Supplementary Table S1

References

Foster

, Beck

, Miller

, et al.: State of type 1 diabetes management and outcomes from the T1D exchange in 2016–2018. Diabetes Technol Ther, 2019; 21:66–72.

Hirsch

, Armstrong

, Bergenstal

, et al.: Clinical application of emerging sensor technologies in diabetes management: consensus guidelines for continuous glucose monitoring (CGM). Diabetes Technol Ther, 2008; 10:232–244; quiz 245–236.

Pickup

, Ford Holloway

, Samsi

: Real-time continuous glucose monitoring in type 1 diabetes: a qualitative framework analysis of patient narratives. Diabetes Care, 2015; 38:544–550.

Polonsky

, Hessler

: What are the quality of life-related benefits and losses associated with real-time continuous glucose monitoring? A survey of current users. Diabetes Technol Ther, 2013; 15:295–301.

Tanenbaum

, Hanes

, Miller

, et al.: Diabetes device use in adults with type 1 diabetes: barriers to uptake and potential intervention targets. Diabetes Care, 2017; 40:181–187.

Messer

, Johnson

, Driscoll

, Jones

: Best friend or spy: a qualitative meta-synthesis on the impact of continuous glucose monitoring on life with Type 1 diabetes. Diabet Med, 2018; 35:409–418.

Tansey

, Laffel

, Cheng

, et al.: Satisfaction with continuous glucose monitoring in adults and youths with Type 1 diabetes. Diabet Med, 2011; 28:1118–1122.

Cemeroglu

, Stone

, Kleis

, et al.: Use of a real-time continuous glucose monitoring system in children and young adults on insulin pump therapy: patients' and caregivers' perception of benefit. Pediatr Diabetes, 2010; 11:182–187.

Rubin

, Peyrot

: Treatment satisfaction and quality of life for an integrated continuous glucose monitoring/insulin pump system compared to self-monitoring plus an insulin pump. J Diabetes Sci Technol, 2009; 3:1402–1410.

10.

Miller

, Foster

, Beck

, et al.: Current state of type 1 diabetes treatment in the U.S.: updated data from the T1D Exchange clinic registry. Diabetes Care, 2015; 38:971–978.

11.

American Diabetes Association: Children and Adolescents: Standards of Medical Care in Diabetes 2019. Diabetes Care, 2019; 42(Suppl 1):S148–S164.

12.

Spear

: The adolescent brain and age-related behavioral manifestations. Neurosci Biobehav Rev, 2000; 24:417–463.

13.

Casey

, Getz

, Galvan

: The adolescent brain. Dev Rev, 2008; 28:62–77.

14.

Sanders

: Adolescent psychosocial, social, and cognitive development. Pediatr Rev, 2013; 34:354–358; quiz 358–359.

15.

Chiang

, Maahs

, Garvey

, et al.: Type 1 diabetes in children and adolescents: a position statement by the American Diabetes Association. Diabetes Care, 2018; 41:2026–2044.

16.

Harris

, Taylor

, Thielke

, et al.: Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform, 2009; 42:377–381.

17.

Messer

, Cook

, Tanenbaum

, et al.: CGM benefits and burdens: two brief measures of continuous glucose monitoring. J Diabetes Sci Technol, 2019; 13:1135–1141.

18.

Markowitz

, Volkening

, Butler

, Laffel

: Youth-perceived burden of type 1 diabetes: problem areas in diabetes survey-pediatric version (PAID-Peds). J Diabetes Sci Technol, 2015; 9:1080–1085.

19.

Iannotti

, Schneider

, Nansel

, et al.: Self-efficacy, outcome expectations, and diabetes self-management in adolescents with type 1 diabetes. J Dev Behav Pediatr, 2006; 27:98–105.

20.

Hood

, Butler

, Anderson

, Laffel

: Updated and revised diabetes family conflict scale. Diabetes Care, 2007; 30:1764–1769.

21.

Kroenke

, Spitzer

: The PHQ-9: a new depression and diagnostic severity measure. Psychiatric Annals, 2002; 32:509–521.

22.

Smarr

, Keefer

: Measures of depression and depressive symptoms: Beck Depression Inventory-II (BDI-II), Center for Epidemiologic Studies Depression Scale (CES-D), Geriatric Depression Scale (GDS), Hospital Anxiety and Depression Scale (HADS), and Patient Health Questionnaire-9 (PHQ-9). Arthritis Care Res (Hoboken), 2011; 63 Suppl 11:S454–S466.

23.

Richardson

, McCauley

, Grossman

, et al.: Evaluation of the Patient Health Questionnaire-9 Item for detecting major depression among adolescents. Pediatrics, 2010; 126:1117–1123.

24.

Rabbone

, Minuto

, Toni

, et al.: Insulin pump breakdown and infusion set failure in Italian children with type 1 diabetes: 1-year prospective observational study with suggestions to minimize clinical impact. Diabetes Obes Metab, 2018; 20:2551–2556.

25.

Messer

, Berget

, Beatson

, et al.: Preserving skin integrity with chronic device use in diabetes. Diabetes Technol Ther, 2018; 20(S2):S254–S264.

26.

Iturralde

, Tanenbaum

, Hanes

, et al.: Expectations and attitudes of individuals with type 1 diabetes after using a hybrid closed loop system. Diabetes Educ, 2017; 43:223–232.

27.

U.S. Food and Drug Administration: FDA expands indication for continuous glucose monitoring system, first to replace fingerstick testing for diabetes treatment decisions. www.fda.gov/newsevents/newsroom/pressannouncements/ucm534056.htm (Accessed December 28, 2018 ).

28.

Forlenza

, Cameron

, Ly

, et al.: Fully closed-loop multiple model probabilistic predictive controller artificial pancreas performance in adolescents and adults in a supervised hotel setting. Diabetes Technol Ther, 2018; 20:335–343.

29.

Ramkissoon

, Herrero

, Bondia

, Vehi

: Unannounced meals in the artificial pancreas: detection using continuous glucose monitoring. Sensors (Basel) 2018;18 pii:E884 doi:10.3390/s18030884.

30.

Messer

: Why expectations will determine the future of artificial pancreas. Diabetes Technol Ther, 2018; 20(S2):S265–S268.

31.

Tanenbaum

, Adams

, Hanes

, et al.: Optimal use of diabetes devices: clinician perspectives on barriers and adherence to device use. J Diabetes Sci Technol, 2017; 11:484–492.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB