Improving the Safety and Functionality of an Artificial Pancreas System for Use in Younger Children: Input from Parents and Physicians

Abstract

Background:

Artificial pancreas (AP) systems have initially been designed for and tested in teens and adults, but there is evidence that an AP system with additional support and safety systems could greatly benefit younger children with type 1 diabetes (T1D).

Subjects and Methods:

Five pediatric endocrinologists and 15 parents of children aged 5–8 years with T1D participated in a total of four focus groups. Focus groups investigated current diabetes technology use and acceptance, as well as possible modifications to the current adult AP system, which would allow for safe and successful use in younger children. Modifications discussed include child-specific functionality for input tasks, safety features, and monitoring capabilities.

Results:

Participant suggestions included the following: passcodes for differential access to AP features by parents, ancillary caregivers, and the child; preset early, intermediate, and advanced child access categories; maximal customization for general and alarm settings; simplified meal screens utilizing the AP’ corrective blood glucose (BG) ability; automated exercise mode; spoken and dictated messaging capabilities; emergency contacts; treatment instructions for the child and caregiver; remote monitoring website and application; animated continuous glucose monitor BG trace; gamification, such as rewarding diabetes-friendly behaviors; and comprehensive training of all individuals involved in the child's diabetes care.

Conclusion:

Parents and physicians were eager for AP applications to be available for younger children, but stressed that a modified system could better serve this group's needs for safety and improved diabetes-related communication. The diverse and emerging needs of 5–8-year olds require flexible and customizable systems for T1D management.

Introduction

Younger children with type 1 diabetes (T1D) are a population that continues to experience both hypoglycemia and hyperglycemia^1,2 and thus may benefit greatly from artificial pancreas (AP) technologies targeting blood glucose (BG) control.³ These technologies utilize existing diabetes-related tools, including continuous glucose monitors (CGMs) to follow BG levels over time and insulin pumps to deliver insulin, and connect these with sophisticated algorithms that determine the amount of insulin to deliver every few minutes (Fig. 1). Several trials have demonstrated improved glycemic control in pediatric populations when using AP technology,^4

–9 although these have utilized AP systems developed for use in adults.

FIG. 1.

Current tools for diabetes care compared to AP-based diabetes care. Current care involves the option of using some tools (such as CGM and insulin pumps) that an AP system relies on. While current care relies on direct entry by parents, children, and other care givers, for the majority of the day and night, the AP system determines insulin delivery from preprogrammed data and CGM tracings, requiring only carbohydrate entry for food. In both cases, the child may not be at a mature enough state to enter data leading to insulin delivery. AP, artificial pancreas; CGM, continuous glucose monitor.

An important goal in the development of AP technology is to provide effective treatment that is adoptable by users of all ages. The Diabetes Assistant (DiAs) was the first ambulatory AP system, developed by the University of Virginia (UVa). The DiAs has been extensively tested in outpatient settings over more than 200,000 h of use in both adults and adolescents.^10,11 However, a user interface such as DiAs, while designed to facilitate BG management tasks in adults,¹² may be too advanced and even dangerous when used by younger children. Inadequate oversight, understanding, and training, and purposeful misuse significantly contribute to reported adverse events with pediatric use of insulin pumps.^13,14 Applying this information to the adoption of a pediatric AP system, it is possible that younger children will benefit most from a system that provides them with modified access to some features but reserves access to advanced features to caregivers. For instance, the younger-child AP system may be adapted to more team-centered rather than patient-centered care by including thorough training of all caregivers and promoting involvement for optimal diabetes management and patient outcomes.^15,16

Other challenges when considering pediatric use of closed-loop technology include the following: unpredictable eating and exercise, significant hypoglycemia, low insulin doses, high insulin sensitivity, and the usability of closed-loop components^15,17; the complexity of the closed-loop components is further tempered by this population's variant levels of comfort and maturity with technology use and the intimate involvement of various caregivers (i.e., parents and school staff) in many management tasks. This study employed focus groups of pediatric endocrinologists and parents of younger children with T1D to assess factors associated with diabetes technology adoption in children aged 5–8 and discuss adaptations to the DiAs user interface to promote successful and safe use in this age range. The 5–8-year age range was chosen to capture the prepubertal, early elementary population who are less likely to have developed skills to independently manage T1D and therefore rely on diverse, potentially inconsistent aid between home and school. Safety modifications and support systems facilitating continuous and involved care would greatly benefit this age range in particular, whereas older adolescents may take more responsibility for their disease management.

Methods

Study design and analysis

This study uses a phenomenological framework to explore how the experience of caring for younger children with T1D shapes needs and desires for AP modifications. Four, 2-h-long focus groups, that is, guided discussions, were conducted at UVa. One focus group consisted of pediatric endocrinologists, who were recruited for their knowledge of current diabetes technology as well as safety risks and benefits of diabetes technology use in younger children, while the latter three focus groups consisted of parents of children aged 5–8 with T1D. The focus group technique was chosen for the following reasons: (1) the interactive dynamic effect within a group is more than the sum of independent interviews, (2) participants could provide context and qualification to their statements, and (3) the fact that the moderator has the opportunity to ask relevant follow-up questions.¹⁸ Throughout the focus groups, broad, predetermined questions (outlined in Supplementary Materials 3; Supplementary Data are available online at www.liebertpub.com/dia) were followed by specific questions tailored to topics raised by the participants.

Focus groups were audio recorded, transcribed, and coded in accordance with thematic analysis¹⁹ and analyst triangulation.²⁰ This included the following steps: (1) two researchers (R.G. and S.L.R.) coded the entire transcript into inclusive categories and (2) a third researcher and subject matter expert (M.D.D.) reviewed the coded data and synthesized the themes within the predetermined categories of the following: current technology use, AP input tasks, AP safety features, and individuals involved in diabetes management. The goal of this analysis was to include all participant viewpoints, while organizing them into meaningful categories as characterized by multiple analyzers. While it was determined that there was a high degree of similarity between the primary raters' output, analyst triangulation improved diversity of perspective on the final coded categories and contributed to a richer synthesis.

Results

Participants

Each 2-h-long focus group was attended by three to six participants. The researchers attempted to achieve data saturation by initially recruiting five participants per focus group, a number chosen to promote diversity of experience, while allowing ample opportunity for each participant to talk; five recruited parents did not attend the focus groups due to last-minute conflicts, so spouses were allowed to participate to reach the recruitment goal. Five pediatric endocrinologists and 15 parents of 11 children aged 5–8 with T1D participated in separate focus groups. Pediatric endocrinologists were recruited from around the state and had a median of 21 years' (range 2–36 years) experience in treating pediatric diabetes in practices with extensive pump and CGM use. One of the pediatric endocrinologists had been the study physician for an AP trial; the others did not have experience with the AP.

Parents were recruited by flyers placed in the waiting area of UVa's Pediatric Diabetes Clinic or referred by physicians at additional pediatric endocrinology clinics and asked to contact the Center for Diabetes Technology (CDT) by phone or e-mail. Local clinics were chosen to increase the likelihood that families could participate in the focus groups in person. In addition, families with pump and/or CGM experience were recruited for their familiarity with these AP system components. The range in duration of diabetes was 2–7 years. No parent had experience in using a research closed-loop system, but one had an older child who had participated in an AP clinical study. Additional participant characteristics are available in Table 1.

Table 1.

Subject Characteristics

Variable	n	Mean ± SD or frequency count (%)
Pediatric endocrinologist gender	5	Two female (40%), 3 male (60%)
Parent gender	15	Eleven female (73%), 4 male (27%)
Child gender	11	Three female (27%), 8 male (73%)
Child age (years)	11	7.1 ± 0.9
Child pump use	11	11 (100%)
Child current CGM use	11	7 (64%)
Child past/intermittent CGM use	11	3 (27%)
Child no CGM use	11	1 (9%)

Pediatric endocrinologists and parents of 5–8-year-old children with T1D participated in 2-h focus groups at the University of Virginia. Focus groups examined improvements to safety features and the user interface of the AP for use in young children.

AP, artificial pancreas; CGM, continuous glucose monitor; SD, standard deviation; T1D, type 1 diabetes.

Topics

Parents and physicians were given a description of the current AP system (Fig. 1). As this system relies largely on existing technologies, participants were initially asked to report their experiences with current CGM and insulin pump management. Using technology for the management of T1D in younger children depends on the patients, the parents, and ancillary caregivers; participants were asked to provide their input on necessary characteristics of persons involved in AP diabetes management. A summary of these discussions is provided in the Supplementary Materials 1 and 2, respectively. Participants were then asked about the settings and data input for the AP. Common themes across focus groups are discussed in the body, while a comprehensive list of suggested functionalities and supporting quotations are available in the corresponding Tables 2 –4.

Table 2.

Participant Ideas on Artificial Pancreas Technology Input Tasks

Participant type	Functionality	% Support	Supporting quotation or details
Insulin settings
Physician and parent	Insulin settings should only be accessible to parents.	100% (20/20)
Parent	Insulin settings should be locked with a passcode.	87% (13/15)	I absolutely want a passcode to change settings. 100%. I don't want somebody to change the carb ratio, that needs to be me or the doctor.
Parent	Insulin settings do not need to be locked with a passcode.	13% (2/15)	[My child should have access to] everything, he has access to his pump. I mean he's not going to go in [the settings] and change anything. A kid knows not to change that.
Parent	Insulin sensitivity factors, carbohydrate ratios, and basal rates should not be altered by school or other adult supervisors.	100% (15/15)	My mom I trust because she has enough knowledge if I tell her step by step how to get in and what to do. Do I trust her to make the judgment call and know when it needs to be changed? Probably not. […] Do I trust the nurse? Absolutely not. A random person? Absolutely not, because I don't feel they are educated enough to be able to do so.
Parent	Parents should be able to access and edit insulin settings remotely.	13% (2/15)
Parent	The AP should have a safety check to ensure the AP and pump insulin settings match before reverting to pump therapy.	7% (1/15)	I think, of course it would need to be blocked, but if there's a way to have it checking to make sure that the settings in the pump and in the device are the same at all times … for whatever reason […] I change it in the pump and forget to change it in the device, like we're having too many lows and I need to up the sensitivity factor or down or whatever, then I think it would be good to have something there.
Security and general settings
Parents	Settings (e.g., alerts/alarms, remote monitoring, and user interface) should be maximally customizable.	100% (15/15)	I'm not saying not to have [a certain] feature, just have it be able to be shut off because everybody is different. You want maximum flexibility for marketing and for age groups and variability, just to have that option.
Parents	The AP should have default settings from which the parent/endocrinologist/care team can choose based on the child's developmental level, autonomy in diabetes management, and preferences.	1% (1/15)	Every one of these options that can be locked or unlocked based on early, intermediate, advanced should be preset into categories, but as the caregiver you can pick and choose certain settings. Like with checkboxes you can pick some things and not others. Like an iPhone settings.
Meals
Physician	The AP does not need a meal announcement tool in this age range because of its ability to attenuate high BG.	20% (1/5)	These 5-8-year-olds might not be eating tons of grams of carbohydrates compared to an adolescent boy who's consuming a hundred, which certainly needs to be announced. One of my thoughts for this age group that may not be eating more than 40 grams at one sitting, that the system might be able to handle it unannounced. It may be different than what you offer to an older child.
Physician	The AP meal screen could include small, moderate, and large meal options and allow the AP to correct the BG afterward.	20% (1/5)	There's the fear that the child will over-bolus themselves. My idea is that if the system can handle a 20 carb meal/snack without having any interventions done, if the child only has to recognize a small, medium, or large meal meaning 20, 40, or 60 g meals and a 20 g wiggle room the system can handle then you don't have to be precise in the carb counting. The system allows some wiggle room and flexibility, it will turn itself off or turn itself on to make up for the differences.
Physician	The AP meal screen could include pictures of commonly consumed foods, selectable portion sizes, and precalculated carbohydrate amounts.	20% (1/5)	Really thinking outside the box, it would be cool if you could have pictures of foods tagged with “this number of goldfish are this grams of carbohydrates” that could help a child count carbs more accurately.
Physician	[Most] 5- to 8-year olds will be able to make meal selections on their own using a modified meal screen.	100% (5/5)	• However, schools/parents are likely to require supervision during insulin bolusing.
Physician	The AP meal screen could include preset carbohydrate averages and typical meal times, allowing the child to press a single button to inject an appropriate meal/snack bolus.	20% (1/5)	The child is going to be on his own primarily at school. The phone knows that it's lunchtime, snack time, if there was just a button to push to say “I'm eating now” and it just delivers the appropriate amount of insulin based on the [normal number of carbs and the] clock.
Physician	The AP meal screen could allow parents to preenter snack and lunch carbohydrate quantities for the day, allowing the child to simply indicate when the meal occurs.	40% (2/5)	• A 70% insulin bolus for preentered carbohydrates could be selected to protect against hypoglycemia.
Parent	Children aged 7–8 should be involved, but not independent in entering carbohydrate quantities and bolusing insulin for meals.	100% (15/15)	• Parents disagreed about whether a password should block insulin delivery capabilities. Feedback included the following:
			○ A password would protect against purposeful mis-bolusing, as one son did to experience hypoglycemia and get “treats.”
			○ A password would protect against accidental mis-bolusing by necessitating a second pair of eyes before injection.
			○ A password would make bolusing more difficult for school staff, who often struggle with current passcodes or have the children unlock their devices.
			○ Children with T1D are often more knowledgeable about their diabetes and equipment than ancillary caregivers.
Parent	The AP should issue alerts/calls to the parent to approve the dose whenever insulin is administered.	13% (2/15)	• This feature could be used for dosing outside of normal mealtimes if desired.
Physician and parent	Parents should be able to remotely enter carbohydrates and bolus insulin.	15% (3/20)	Can you adjust something that involves a change to the way the pump is working? Meaning enter food remotely?
Exercise
Physician and parent	The main concern of an exercise feature is that the child will not remember to engage/disengage it.	100% (20/20)	The only thing that I worry about is that he's not going to remember. I don't think he's going to be like, “I'm going outside to jump on the trampoline, let me push the button on here.” I think that's why I say [the exercise function needs to be used] with adult supervision.
Parent	The exercise function should be categorized into separate sports/activities with customized target BG/basal suspend profiles.	13% (2/15)	I don't know if anyone else experiences this, but depending on the exercise my child goes down or goes up. So is there a way to have different exercises?
Physician and parent	The exercise function should be automated using Bluetooth technology to connect a pedometer, smart watch, heart rate monitor, and/or accelerometer.	67% (10/15)	• UVa's CDT has performed randomized controlled trials using a heart rate monitor connected to the AP during exercise.^18,19
Parent	Children may benefit from enabling the exercise function for the entire school day to help prevent dangerous hypoglycemia, even at the risk of temporarily higher BG.	7% (1/15)
Parent	Children may benefit from using the exercise function on weekends more than weekdays, as sedentary class time can contribute to higher BG at school.	7% (1/15)
Messaging
Parent	The AP should have a messaging function.	100% (15/15)	… The parent could be making decisions and analyze the big situation and give advice to whoever happens to be near the child.
			• Messaging could substitute the need for a separate piece of technology for diabetes management-specific communication.
			• Messaging could be especially helpful for substitute nurses; during field-trips; and at friends' and families' houses.
Parent	The AP should have an audio dictation messaging system usable by preliterate children.	20% (3/15)	I'd say the spoken messages, if there was a way to really simplify it from the child's side when he's trying to send it. For a 5-year-old […] if it just says ‘talk’ and we tell them “ok, when you press this you need to talk” it may be useful as long as they don't abuse that privilege and turn it into a toy.

BG, blood glucose; CDT, Center for Diabetes Technology.

Table 3.

Participant Ideas on Artificial Pancreas Technology Safety Features

Participant type	Functionality	% Support	Supporting quotation or details
Passcodes
Physician and parent	AP passcodes settings should be flexible.	100% (20/20)	• Families should be able to decide together which features need to be initially blocked and features to add back as their child's autonomy increases.
Physician and parent	AP passcodes should operate differently than current pump/iPhone passcodes.	100% (20/20)	• Current pump/iPhone passwords do not keep children out, but can increase the difficulty of diabetes management for adult caregivers.
Physician and parent		100% (20/20)	I don't like the lock function on the [pump brand] because it's just pushing two buttons that he figured out how to push in a matter of two weeks of having it, and the nurse … I told her how to unlock it, but she gets him to unlock it.
Physician and parent	Children with diabetes generally respect their pump as a medical device, and therefore easy passcodes do not currently pose additional risk.	35% (7/20)	Kids are smart! Even when we lock their pumps they figure out how to unlock them. We don't have a lot of kids who play with their pumps and inadvertently give themselves insulin. The parents are instilling this is not a toy, you don't play with it, you don't let other people play with this.
Physician and parent	The AP's phone platform may be more enticing to children and could pose more risk than current diabetes technology.	100% (20/20)	It's the 5 yo that's mischievous and wants to play video games on the device, I have a friend who has that child
Parent	Requiring a passcode for ancillary caregivers is unnecessary and anxiety inducing for school staff; therefore, passcodes should be used to lock settings only.	66% (10/15)	I think the worst thing is if somebody freaks out because of the passcode and your child is having a condition that can't be remedied because they can't get into this thing.
			I would want the option to put on passcodes. I absolutely want a passcode to change settings, […] that needs to be me or the doctor. But anything that happens at school doesn't need a passcode.
			The nurse and the child should [have access to] the same thing. My son is teaching his nurse because he's the only child in the entire school that has it.
			Plus with your caregivers if you have someone who's really nervous about it, like the county nurses the more steps they have to do the more freak out they have.
Physician and parent	Parents should be able to remotely unlock the AP's screen in case of emergency or ancillary caregiver difficulty with the passcode.	15% (3/20)	• Remote unlocking should not be relied upon for access in case the parent is unavailable.
Physician and parent	AP passcode type should be customizable, as different formats will be preferable for different families and schools.	100% (20/20)	Suggestions for passcode format:
			• Number and letter combination
			• Dynamically changing passcode (e.g., eight-digit date)
			• Fingerprint recognition
			• Facial recognition
			Like an iPhone. Some people have the thumbprint, some people the four-digit code, some eight. Or you can turn it off. I had mine turned off, but one time I lost my phone and freaked out, so now I have it.
Emergency contacts/phone call
Physician and parent	The AP should list emergency contacts.	100% (20/20)	• Emergency contacts could be listed on a strip across the top of the home or lock screen, or as an unlocked button on the home screen.
Parent	The AP should have buttons that can directly call caregivers, the physician, and 911.	13% (2/15)
Parent	The AP should list emergency contacts, but block phone capabilities.	33% (5/15)	• Children may not have the discretion to use emergency phone call buttons appropriately.
			… If [someone] felt the need to [call], they would do it on their own phone versus a medical device.
			What would a kid consider an emergency? They might not know the difference, they might just want to call their mom and dad.
Parent	The AP should automatically call an emergency contact if the data indicated a dangerous low or prolonged high blood sugar.	20% (3/15)	• Concurrent forms of contact (e-mail, remote monitoring alert, and app alert) should also be used in case caregivers are unavailable by phone.
Parent		20% (3/15)	Say you're worried about the kid turning off the alarm, could there be a setting on there that automatically calls the parent if he resets the alarm 3 times?
Alerts and alarms
Physician and parent	Alert and alarm preferences will vary across families and schools. Customization will increase the likelihood of the technology being successfully adopted.	100% (20/20)	• Some children will respond to quieter or vibrating alarms, while others will require loud and persistent alarms. This may also vary by location and time.
			• Similarly, parents and ancillary caregivers will prefer different alert types.
			When he first got the pump and didn't know what to do about the alarms, he would go to the nurse because he didn't want to be embarrassed by that noise in class.
			1: If it's sitting right up against your skin in class you're going to be able to feel a vibration. But if you're running around outside you might not feel it.
			2: Also are you going to ignore it? My daughter does.
			It would be nice if there could be multiple contacts and types of alerts.
Physician and parent	The number and severity of high alarms should be limited.	20% (4/20)	… The number of high alarms is the reason most people stop using CGMs at school. (Physician)
Parent	A parent/supervisor passcode should be required to silence alarms and ensure timely treatment.	13% (2/15)
Parent	The AP should have default alarm settings. Alarms for different purposes should be distinct to aid in quick diagnosis.	100% (15/15)	Default alarms should include the following:
			• Noticeable low alarms that persist until acknowledged and treated
			• An alert for a caregiver to check the pump site when high BG persists (longer than 2 h) despite AP action
			• Loud alarm and accompanying caregiver alert for a sharp BG spike without an insulin bolus, indicating carbohydrate consumption without a bolus
			• Loud nighttime alarms accompanied by automated phone calls
			• Alarms with increasing intensity, including a voice readout of treatment instructions, when alarms are ignored and/or proper treatment is not initiated
			It would also be really great if you could have different auditory sounds for different things. Ping versus fire alarm because there are certain sounds- like when people call you and you've programed different ringtones for different people- so I can automatically say, “Oh, that's a low BG” and I don't even need to look at what it is. And then you can change it up because sometimes you get immune to the same sound.
			What if it did [the spoken alert] if they ignored the first alarm?
			One of the things that annoys me is that thing will beep for hours and she won't care. Is there a way to escalate the alarms? That's an aggravation that I have, and she'll sleep right through it, she can't hear it at night.
Instructions for child and caregivers
Physician and parent	Simple, written and pictorial instructions should accompany alarms to help children and ancillary caregivers with treatment decisions.	100% (20/20)	• For example, a hypoglycemia banner could read “Blood sugar is low. Treat with fast acting carbs.” and include pictures of appropriate treatment.
			• Instructions could also be verbal, similar to the new voice-activated EpiPen.
			• Straightforward directions could alleviate alarm-induced anxiety among school staff.
Physician and parent	Information about diabetes, treatment, and AP technology should be available within a button on the home screen.	25% (5/20)
Other features
Parent	Phones are easily lost, stolen, and broken. AP technology should be made with flexible materials, marked with “Medical Device,” and used with screen protectors and carriers.	40% (6/15)	I'm more worried about them losing it and keeping up with it. The fanny pack is actually an excellent idea for keeping track of it and maybe discouraging sharing it with friends.

Table 4.

Participant Ideas on Artificial Pancreas Technology Monitoring

Participant type	Functionality	% Support	Supporting quotation or details
CGM trace and history
Physician and parent	A CGM BG tracing should be available on the home screen of the AP.	60% (12/20)	• Current CGM users and physicians believe that the CGM trace is an important visual that helps younger children understand their BG.
Physician and parent		60% (12/20)	We went to the doctors and the first thing he did was pull up the last 24 hours and showed the doctor immediately, because he was proud of what it looked like. He knows that we're happy when his blood sugar is in a better range, and he feels better, so in his mind if he can see that it is doing positive. […] I don't know how he processes the information, but he knows what it is. He knows the top line, the bottom line, and he knows it's good when he's in between that.
Physician and parent	An animated CGM trace (such as the airplane and its contrail marking the BG trace shown in Fig. 3) should be an available, but not mandatory feature.	100% (20/20)	• Animation could be a great teaching tool and motivator.
			• Animation may negatively motivate some children to push their BG high or low to explore what happens with the screen.
			• Animation may inadvertently punish children when having out-of-range BG (such as during an illness or other uncontrollable circumstances).
			When my son was first diagnosed, he didn't understand any of it so he didn't care about it. […] He slowly is learning what those data points in there mean, and the more he understands it, the more pride he takes in taking care of himself. A big issue I've noticed is a lot of kids just don't care what their blood sugar is, why should they care when they don't understand the severity of what it can cause, it's just whatever it is. Visually teaching them that this is why, as a learning experience it would be really good for them. My seven-year-old would love that now just to have it to look at. It's like playing a video game, so anything that is positive that comes from [his diabetes management] would be great.
			Something like that could be a reward, where you could use points to switch the screen. Like [my daughter] would love to have sharks with scuba divers if she got to change it.
Gamification
Parent	Gamification should be an available, but not mandatory feature. Points should be awarded to children for good diabetes management. Points should not be awarded to children for circumstances out of the child's direct control.	100% (15/15)	Points could be awarded for the following:
			• responsible bolusing
			• entering finger sticks
			• calibrating the CGM
			• making healthy food choices
			• promptly acknowledging alarms
			• treating lows quickly
			Points should not be awarded for the following:
			• time in range
			• average BG
			I think it's even better for small children, because it's a learning opportunity for them. So they could, only thing I would say is that sometimes you get really anxious children or really competitive, so you'd want to be able to turn that off if it's causing problems. It's like a reward chart.
			When his blood sugar is high, he will blame someone; “It's your fault because YOU gave me a piece of pizza,” and this might help him take more responsibility in his choices. But if he's sick or getting sick, or something like that he would probably get to a point where he would feel like he's doing something wrong even if it's not his fault for going out of range.
			My son would probably respond more to the points. Something like that picture he might say “Let me put [the airplane] in the clouds to see what happens, let me take it down here to see what happens.” That's what he would do. If it's a point thing, that would be something better to gear towards.
Parent	Children could earn rewards for points earned.	20% (3/15)	• Points could be rewarded by the parents with physical objects/activities or through an online system.
			I have a few kids play this math game where you get points, coins, and then once you get so many coins you can go to the store with your piggy bank and buy stuff for your fish tank. So you can collect fish and fill it up with cool stuff. They'll be on it and be like, “Oh look at the brand-new fish I got,” and then they had another scene where it was a farm scene.
			[Our son] earns stickers for eating vegetables and then at the end of the week he can turn those in for a bigger sticker or something like that that's less likely to encourage him to need a toy in response to it. But that's about the only reward we do for that. I know as they get older they'll need to maintain it more on their own and be responsible for their decisions. So maybe it helps the parents, they could say “If you get a certain number of points then we'll give you $20” or whatever the situation would be.
Remote monitoring
Physician and parent	Remote monitoring should be available for parents and authorized ancillary caregivers.	100% (20/20)	• Remote monitoring could be available by website, phone application, and/or as a health overview on the care providers' mobile phone lock screen.
Physician and parent	Parents should be able to take remote action (i.e., bolus insulin through a remote phone application).	30% (6/20)	As for what the parents perhaps would be able to do, with the remote monitoring capability is there any way that they can actually do the bolus delivery from a remote site?

AP Technology: Input Tasks

Settings

Insulin settings

AP insulin settings, prescribed by endocrinologists and sometimes adjusted by parents between appointments, are entered into the system like an insulin pump (Table 2). All participants agreed that insulin settings should be changed only by the physician or parents, but disagreed about whether they needed to be locked. Considering the user-friendly interface of the AP's phone platform, the majority (87%) of parents believed entering settings should require a passcode.

Security and general settings

Parents recommended maximizing the ability to customize settings (e.g., alerts/alarms, remote monitoring, and user interface). Research has shown that adding more options can decrease the likelihood of a person choosing anything; however, more knowledge and organization of the options into categories diminish this effect.²¹ Participants supported default settings, which could be further customized by the parent based on the child's developmental level and autonomy in diabetes management.

Meals

Physicians expressed confidence in the AP’s closed-loop capabilities with regard to meals and suggested simplifying the meal screen using loose estimation of carbohydrates and subsequent AP compensation. Based on experiences with their own children, parents believed that 5–8-year olds should be involved in entering meal information. Most parents agreed it would be appropriate to lock insulin delivery capabilities for this age range, while other parents believed passcodes would make ancillary diabetes care more difficult. Many parents supported measures to aid parental involvement in school meals.

Exercise

Indicating exercise within the DiAs sets a temporary low BG target of 140 mg/dL (compared to the default 110 mg/dL) and may provisionally suspend basal insulin.^22,23 When used during exercise, this allows the system to prevent exercise-induced hypoglycemia; when used during rest, this feature may not allow for as tight a control as otherwise possible. Because this function is not a major safety hazard, physicians and parents were only concerned with whether a 5–8-year-old would remember to use the function. Participants suggested automatizing the exercise function or utilizing it all day to prevent dangerous lows when outside the house.

Messaging

Because the AP uses a mobile platform, text or wireless messaging could be enabled for child-caregiver communication.²⁴ Parents agreed that written and audio messaging should be enabled and would be especially helpful for substitute nurses; during field-trips; and at friends' and families' houses.

Parents' and physicians' further experience with and suggestions for current diabetes technology is shown in Supplementary Table 1. Thoughts regarding creating a successful care team for their child's diabetes is shown in Supplementary Table 2. (Supplementary tables [Supplementary Data] are available at http://online.liebertpub.com/doi/suppl/10.1089/dia.2017.0150.)

AP Technology: Safety Features

Passcodes

The focus groups discussed the viability of a password to block features that may not be safe for 5–8-year olds when unsupervised (i.e., Fig. 2 and Table 3). Participants unanimously agreed that current pump/iPhone passwords do not keep children out and increase the difficulty of diabetes management for adult caregivers. All but one parent, whose child had intentionally delivered insulin to induce need for hypoglycemia treatment, agreed that children with diabetes generally respect their pump as a medical device. However, participants perceived that the AP’s phone platform may be more enticing to children and could pose more risk than current diabetes technology. Participants wanted flexibility with setting passcodes so that families may decide together which features need to be initially blocked and which may be freely accessed as their child's autonomy increases. Customizable passcode formats to support different family and school preferences and the ability to set passcodes for certain times of day were also supported.

FIG. 2.

Prototype of AP-modified access screen, with buttons leading to further screens with specific utility. Parents posited that some screens (such as CGM tracings, emergency contact information, and message communication with parents) could have open-access, where other screens (such as insulin settings or carbohydrate information) require passcode access.

Emergency contacts/phone call

Focus groups considered emergency situations in which the child, adult caregiver, or diabetes-unfamiliar person would want to contact the parent. Participants supported listing emergency contacts either on a strip across the top of the home or lock screen, or within a button. Parents showed mixed support for enabling buttons that could directly contact caregivers, the physician, and 911, but most participants would like the AP to automatically call an emergency contact if the data indicated a dangerous low or prolonged high blood sugar.

Alerts and alarms

Alert and alarm preferences varied across families, and participants believed the ability to customize based on their needs would increase their likelihood of adopting the technology. Parents agreed on the following default alarm settings: noticeable low alarms that persist until acknowledged and treated; an alert for a caregiver to check the pump site when high BG persists (longer than 2 h) despite AP action; a loud alarm and accompanying caregiver alert for a sharp BG spike without an insulin bolus, indicating carbohydrate consumption without accompanying insulin; loud nighttime alarms accompanied by automated phone calls; a parent/supervisor passcode to silence alarms and ensure timely treatment; and alarms that increase in intensity, culminating in a voice readout of treatment instructions, when alarms are ignored and/or proper treatment is not initiated.

Instructions for child and caregivers

Participants unanimously supported simple, written, and pictorial instructions accompanying alarms to help children and other care providers with treatment decisions and alleviate alarm-induced anxiety among school staff. For hypoglycemia, a banner with, “Blood sugar is low. Treat with fast-acting carbs.” and pictures of appropriate treatment could help those unfamiliar with diabetes care. More in-depth information about diabetes, treatment, and AP technology should also be available.

Monitoring

CGM trace and history

Some parents believed that a CGM BG tracing should be accessible on the home screen of the AP, whereas others were not sure how their child would benefit from this information (Table 4). Current CGM users and physicians believed that a CGM trace is an important visual that helps younger children understand their BG. Investigators presented prototypes of an animated CGM screen; an airplane and its contrail mark the BG trace, while a blue sky indicates a safe BG range, clouds a high BG range, and trees a low BG range (Fig. 3). Parents thought that animation could be a great teaching tool and motivator. However, they also worried that children may get upset when having out-of-range BG, such as during an illness or other uncontrollable circumstances, or the animation could negatively motivate children to push their BG high or low and explore what happens to the screen. Animation should be an available, but not mandatory feature.

FIG. 3.

Prototype of animated CGM trace. A modifiable screen such as this could provide a visual cue for children to understand their current blood glucose level that is best when not too low (in the trees) and not too high (in the stormy clouds).

Gamification

Investigators introduced the possibility of AP gamification based on evidence of diverse mHealth apps for improving health outcomes, including an increased frequency of BG monitoring in adolescents with T1D.^25,26 Participants agreed that children would be motivated by earning points for good diabetes management such as responsible bolusing, entering finger sticks, calibrating the CGM, making healthy food choices, promptly acknowledging alarms, and treating lows quickly. Parents did not want rewards to be based on average BG or time in range since these are not in the child's direct control. Points could be rewarded by an online system or by the parents, and this feature could be turned off if desired.

Remote monitoring

Current CGM users reported relying on remote monitoring daily. Participants in each focus group expressed interest in both remote monitoring through the AP and proposed the ability to remotely bolus insulin. Parents wanted a corresponding website and phone application that could be accessed by care providers.

Discussion

Across focus groups, suggestions for AP modifications followed two major themes: (1) creating a modified and flexible AP system to help a child assume more responsibility for their diabetes management and (2) creating a system that can facilitate the involvement of and communication between various care providers. Prior research and results from these focus groups emphasize that both self-efficacy and a team-centered approach are positive for diabetes management in younger children.²⁷ The ability of the AP to integrate diverse diabetes tools (e.g., pump and CGM, meal and exercise assistant, text messaging, phone calls, diabetes care websites and social media, diabetes care information, and AP training) into one device simplifies and expedites their multipurpose use to involve parents while bolstering child problem-solving and self-management skills.²⁴

Involving the children in their care at this age is critical because it can set the stage for effective diabetes management for their entire lives. Self-management skills deemed appropriate for this age group based on the focus groups included the following: inputting exercise, monitoring their CGM trace, and knowing where/when to get help. Focus groups supported modifications to incentivize children to participate in positive diabetes management activities (e.g., animation of the CGM trace and a points system). Focus groups strongly endorsed flexible AP settings that allow responsibilities to be tailored to the individual child. It is important to note that (1) children go through many changes between the ages of 5–8, which necessitate adjustments in AP settings, and (2) improvements in quality of life (QoL) are dependent on having appropriate expectations for the technology based on the maturity of the patient.²⁸

Care teams must understand the capabilities of and sustained involvement necessary for successful AP use in younger children. As previous studies have shown, maintaining parent involvement at this age is critical to adherence to BG monitoring and glycemic control, and sustaining a higher QoL.^27,28 The AP also has the potential to alleviate parental/child stress related to diabetes when the adopters are willing to participate in a care team and the technology is adopted with realistic expectations. Features of the AP that could reduce parental/child stress and were well received during the focus groups include the following: passcodes to block insulin delivery (as the CDT recently tested in younger children⁹); customizable alerts to the child/caregivers to ensure appropriate treatment and discourage covert snacking; and easily accessible resources and incentives through animation and gamification.

Focus groups also stressed the importance of improving ancillary caregiver involvement in young children's T1D management. Participants cited poor training as a barrier, while good training was considered imperative to successful technology adoption. With the complex AP system, training is even more critical for all parties involved in diabetes management. Parents from the focus groups were overwhelmingly unsatisfied with their children's current diabetes management at school, consistent with previous published literature,²⁹ concluding that school nurses and primary teachers need enhanced training on T1D and how to reliably manage BG events. Psychoeducational factors could be included in training to help children, parents, and ancillary caregivers deal with the pressures of T1D management.³⁰ In addition, focus groups supported features to help ancillary caregivers including the following: text messaging, instructions paired with alarms, straightforward input screens (e.g., carbohydrate entry, exercise indicator, and treatment screen), and diverse remote monitoring capabilities.

Finally, the adoption of AP technology also needs to take into account barriers, stigmas, and preconceived notions associated with the hardware (i.e., insulin pump and CGM). A study assessing patients' future acceptance of closed-loop technology found that pump users who read a description of an AP system were highly trusting the system's accuracy and usefulness,³¹ whereas a study with CGM users were more concerned with the system's accuracy.³² These results were mirrored in our focus groups in that, the parent's attitude toward the AP system was dependent on current CGM use/expertise. Parents with CGM experience understood the benefits/limitations and the need for continued treatment involvement (i.e., fingersticks, calibrations, and interpreting BG trends/history). Parents with intermittent CGM experience conveyed greater dissatisfaction with sensor inaccuracy, consistent with data that nonregular CGM users have less valuable experiences with the technology.³³ Finally, parents with no CGM experience were the least aware of the potential value and limitations of CGM technology. Again, future AP users must be fully aware of the technology's capabilities, requirements, and limitations.

This study sought to move away from what engineers or adults find easy to use and toward what children aged 5–8, parents, and potentially diabetes-unfamiliar care providers can utilize as a team. A direction for future research would be to explore other perspectives (e.g., school staff, family members, certified diabetes educators, the children themselves, and a larger child age range) on proposed modifications to the AP system for children. One limitation of this study is the limited sample size; while the targeted population further limited generalizability, the researchers thought it was critical to recruit participants with children using AP components (i.e., insulin pump and CGM technology). A more diverse sample in the future could include children who do not use these components (i.e., use multiple daily injections and no CGM) to represent the larger diabetes population. The authors sought to quantify the focus group data by tracking percent support per functionality discussed, but this has limited accuracy as the open-ended format allowed focus groups to generate different ideas. While equal contribution was encouraged, there is always the potential for domination by one or several speakers as well as groupthink with the focus group format. Surveys could be used along with focus groups to increase generalizability. Focus groups and surveys specific to user interface and AP features could be administered to children and family members after experience with closed-loop systems in a clinical trial, as is being done with teen and adult populations.^34,35

Footnotes

Acknowledgment

Funding source: UVa internal Launchpad Grant Program.

Authors' Contributions

S.L.R., PhD in industrial engineering (with a focus on cognitive human factors) and professor at Clemson University, conducted the focus groups. This researcher has had 8+ years of experience conducting experimental studies and running focus groups, but had little bias or knowledge specific to the AP before this project. M.D. DeBoer, pediatric endocrinologist and researcher at the UVa CDT, and Daniel Cherñavvsky, researcher at the CDT and Chief Medical Officer at TypeZero Technologies, Inc., helped introduce the AP system and ask/answer questions requiring detailed knowledge of the AP. R.G., psychology BA and research technician at the CDT, was present at the focus groups and managed participant recruitment and consent, transcription, coding, and writing. M.D. DeBoer and S.L.R. participated in coding and critical revising as well. No relationship was established between the researchers and the participants before the focus groups, at which time the researchers gave short, professional introductions.

Author Disclosure Statement

D.R.C. is CMO of TypeZero Technologies, Inc., a company currently testing an inControl AP system purchased from the UVa Center for Diabetes Technology with the aim of marketing the product once approved. M.D. Breton is a founder and equity holder of TypeZero Technologies, Inc. He has received research support from Dexcom, Tandem, Roche, Senseonics, Sanofi, and Ascensia. M.D.B. has received consulting fees and honorarium from Sanofi, Ascensia, Dexcom, and Roche. All other authors have no competing financial interests.

References

Beck

, Tamborlane

, Bergenstal

, et al.: The T1D Exchange Clinic Registry. J Clin Endocrinol Metab, 2012; 97:4383–4389.

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.

Wadams

, Cherñavvsky

, Lteif

, et al.: Closed-loop control for pediatric type 1 diabetes mellitus. Diabetes Manage, 2015; 5:23–35.

Elleri

, Allen

, Tauschmann

, et al.: Feasibility of overnight closed-loop therapy in young children with type 1 diabetes aged 3–6 years: comparison between diluted and standard insulin strength. BMJ Open Diabetes Res Care, 2014; 2:e000040.

Russell

, Hillard

, Balliro

, et al.: Day and night glycaemic control with a bionic pancreas versus conventional insulin pump therapy in preadolescent children with type 1 diabetes: a randomised crossover trial. Lancet Diabetes Endocrinol, 2016; 4:233–243.

Phillip

, Battelino

, Atlas

, et al.: Nocturnal glucose control with an artificial pancreas at a diabetes camp. N Engl J Med, 2013; 368:824–833.

Hovorka

, Allen

, Elleri

, et al.: Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial. Lancet, 2010; 375:743–751.

Cherñavvsky

, DeBoer

, Keith-Hynes

, et al.: Use of an artificial pancreas among adolescents for a missed snack bolus and an underestimated meal bolus. Pediatr Diabetes, 2016; 17:28–35.

DeBoer

, Breton

, Wakeman

, et al.: Performance of an artificial pancreas system for young children with T1D. Diabetes Technol Ther, 2017; 19:293–298.

10.

Kovatchev

, Renard

, Cobelli

, et al.: Feasibility of outpatient fully integrated closed-loop control: first studies of wearable artificial pancreas. Diabetes Care, 2013; 36:1851–1858.

11.

Kovatchev

, Renard

, Cobelli

, et al.: Safety of outpatient closed-loop control: first randomized crossover trials of a wearable artificial pancreas. Diabetes Care, 2014; 37:1789–1796.

12.

Keith-Hynes

, Guerlain

, Mize

, et al.: DiAs user interface: a patient-centric interface for mobile artificial pancreas systems. J Diabetes Sci Technol, 2013; 7:1416–1426.

13.

Scaramuzza

, Iafusco

, Lombardo

, et al.: Adolescent use of insulin and patient-controlled analgesia pump technology: a 10-year food and drug administration retrospective study of adverse events. Pediatrics, 2008; 122:473–474; author reply 4.

14.

Cope

, Samuels-Reid

, Morrison

: Pediatric use of insulin pump technology: a retrospective study of adverse events in children ages 1–12 years. J Diabetes Sci Technol, 2012; 6:1053–1059.

15.

Elleri

, Allen

, Dunger

, Hovorkea

: Closed-loop in children with type 1 diabetes: specific challenges. Diabetes Res Clin Pract, 2011; 93 Suppl 1:S131–S135.

16.

Gonder-Frederick

, Shepard

, Peterson

: Closed-loop glucose control: psychological and behavioral considerations. J Diabetes Sci Technol, 2011; 5:1387–1395.

17.

DeBoer

, Breton

, Wakeman

, et al.: Performance of an artificial pancreas system for young children with type 1 diabetes. Diabetes Technol Ther, 2017; 19:293–298.

18.

Rogers

, Meyer

, Walker

, Fisk

: Functional limitations to daily living tasks in the aged: a focus group analysis. Hum Factors, 1998; 40:111–125.

19.

Braun

, Clarke

: What can “thematic analysis” offer health and wellbeing researchers?. Int J Qual Stud Health Well-being, 2014; 9:26152.

20.

Pope

, Ziebland

, Mays

: Qualitative research in health care. Analysing qualitative data. BMJ, 2000; 320:114–116.

21.

Mogilner

, Rudnick

, Iyengar

: The mere categorization effect: how the presence of categories increases choosers' perceptions of assortment variety and outcome satisfaction. J Cons Res, 2008; 35:202–215.

22.

DeBoer

, Cherñavvsky

, Topchyan

, et al.: Heart rate informed artificial pancreas system enhances glycemic control during exercise in adolescents with T1D. Pediatr Diabetes, 2016; doi: 10.1111/pedi.12454. [Epub ahead of print].

23.

Breton

, Brown

, Karvetski

, et al.: Adding heart rate signal to a control-to-range artificial pancreas system improves the protection against hypoglycemia during exercise in type 1 diabetes. Diabetes Technol Ther, 2014; 16:506–511.

24.

Vaala

, Hood

, Laffel

, et al.: Use of commonly available technologies for diabetes information and self-management among adolescents with type 1 diabetes and their parents: a web-based survey study. Interact J Med Res, 2015; 4:e24.

25.

Cafazzo

, Casselman

, Hamming

, et al.: Design of an mHealth app for the self-management of adolescent type 1 diabetes: a pilot study. J Med Internet Res, 2012; 14:e70.

26.

Brzan

, Rotman

, Pajnkihar

, Klanjsek

: Mobile applications for control and self management of diabetes: a systematic review. J Med Syst, 2016; 40:210.

27.

Mohebi

, Azadbakht

, Feizi

, et al.: Review the key role of self-efficacy in diabetes care. J Educ Health Promot, 2013; 2:36.

28.

Liberman

, Buckingham

, Phillip

: Diabetes technology and the human factor. Diabetes Technol Ther, 2014; 16 Suppl 1:S110–S118.

29.

Tolbert

: Managing type 1 diabetes at school: an integrative review. J Sch Nurs, 2009; 25:55–61.

30.

Murphy

, Rayman

, Skinner

: Psycho-educational interventions for children and young people with type 1 diabetes. Diabet Med, 2006; 23:935–943.

31.

van Bon

, Brouwer

, von Basum

, et al.: Future acceptance of an artificial pancreas in adults with type 1 diabetes. Diabetes Technol Ther, 2011; 13:731–736.

32.

Shepard

, Gonder-Frederick

, Vajda

, Kovatchev

: Patient perspectives on personalized glucose advisory systems for type 1 diabetes management. Diabetes Technol Ther, 2012; 14:858–861.

33.

Chase

, Beck

, Xing

, et al.: Continuous glucose monitoring in youth with type 1 diabetes: 12-month follow-up of the Juvenile Diabetes Research Foundation continuous glucose monitoring randomized trial. Diabetes Technol Ther, 2010; 12:507–515.

34.

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.

35.

Barnard

, Wysocki

, Ully

, et al.: Closing the loop in adults, children and adolescents with suboptimally controlled type 1 diabetes under free living conditions: a psychosocial substudy. J Diabetes Sci Technol, 2017:1932296817702656.

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

0.12 MB