A Silent Burden: What patients taught us about CGM alerts—and how better design can ease the load

Insights on Continuous Glucose Monitoring (CGM) alert systems from our UX research study.

A Silent Burden: What patients taught us about CGM alerts—and how better design can ease the load

Insights on Continuous Glucose Monitoring (CGM) alert systems from our UX research study

CGMs have one purpose: to help people maintain healthy blood glucose levels. They achieve their aim when they provide:

Accurate, continuous blood-glucose measurement
Reliable, useful reporting and alerts

From our research, many of the most significant pain points are in the alert experience, with substantial impacts on patient quality of life.

Our UX research study on CGMs shed light on the real-world design gaps, how they negatively impact patients and their families, and how they might be addressed.

CGMs have one purpose: to help people maintain healthy blood glucose levels. They achieve their aim when they provide:

Accurate, continuous blood-glucose measurement
Reliable, useful reporting and alerts

From our research, many of the most significant pain points are in the alert experience, with substantial impacts on patient quality of life.

Our UX research study on CGMs shed light on the real-world design gaps, how they negatively impact patients and their families, and how they might be addressed.

CGMs have one purpose: to help people maintain healthy blood glucose levels. They achieve their aim when they provide:

Accurate, continuous blood-glucose measurement
Reliable, useful reporting and alerts

From our research, many of the most significant pain points are in the alert experience, with substantial impacts on patient quality of life.

Our UX research study on CGMs shed light on the real-world design gaps, how they negatively impact patients and their families, and how they might be addressed.

Backstory: How Lived Experience Became Research

This project began with a personal spark: two of our Radian team members have children with type 1 diabetes. We’ve experienced the power and frustrations of today’s CGMs up close.

That led us to a broader study, including 12 in-depth interviews with patients spanning preteens to people in their 80s. Participants told us about life with diabetes and showed us exactly how they use their CGM apps (and related follow, watch, pump, and 3rd-party apps).

Covering Dexcom, Libre, Medtronic, and Eversense systems, these conversations gave us rich insights into the onboarding, alerts, reporting, and integrations experience of people using today’s CGMs.

The CGM Value Journey

To be successful, CGM systems must deliver on their value promise at each step:

Measure glucose levels accurately
Transmit the data seamlessly to where it’s needed
Transform it into clear insight for the patient (and caregivers)
Capture attention through an optimal sensory experience
Communicate the insight to support appropriate action

A failure at any step can lead to frustration and increase the risk of severe glycemic events.

Real Life Stories

Let’s take a look at a few snapshots taken directly from our research (names changed and stock images used for patient privacy):

Maria: The False Alarm
Adult
libre
Maria is woken at 3 AM by an urgent low alert. She does a fingerstick and discovers her glucose is within range. Annoyed, she snoozes all alarms for the next six hours, choosing uninterrupted sleep over a system she can't trust.
“The problem is, I was sleeping on my sensor [causing a compression low]. But nobody told me not to do that. I followed the instructions. I had to learn about things like that by going onto Reddit.”
Blake: The Call for Autonomy
child
dexcom
Blake has a smartphone with the CGM app, but its nighttime alerts don’t always wake him up. And the companion follow app his parents use to monitor his glucose can’t be set up to only send critical alerts. This reduces Blake’s ability to manage his diabetes independently and creates more sleep disruption for his parents.
“We need a 'find my phone'-like persistent alert for my child at night, as alerts often don't wake him."
John: The Smartphone Catch-22
child
dexcom
For John, it’s easy to forget to bolus at lunchtime at school. But when John’s CGM picks up his high glucose level, that alert only goes to his pump, because, like many kids, he’s not ready for a smartphone. The pump’s alerts don’t capture his attention sufficiently, and Mom can’t receive the alerts because the CGM follow app requires John to use a smartphone.
"You have to give your kid a smartphone, with all the trouble that brings, or you get no remote monitoring and alerts."
Alex: The Irritating Buzz
adult
Eversense
Alex is in an important meeting at work. A sharp, constant vibration starts buzzing against his arm—an alert from his on-body transmitter. The vibration is intensely annoying and distracting. He wishes he could change it to be softer for non-urgent alerts, but there's no way to customize the intensity or pattern.
"It's like a cat scratching on your arm... and there's no way to mute it, alter it, or tone it down. You can't turn it up when you're asleep and turn it down when you're awake. There's no way to modify it."
Sierra: The DIY Insights
adult
Dexcom
Sierra gets an alert from her CGM app on her smartphone with a single arrow pointing down. She ignores it and instead glances at two apps on her smartwatch which tell her she is "decreasing slightly" and that her glucose is falling at a rate of "-2 mg/dL per minute." This rich, quantitative data, which she had to route through multiple third-party apps, gives her the confidence to watch and wait, rather than over-treating a minor dip.
“I prefer to get my alerts through Night Scout or Sweet Dreams. They say things like 'slightly decreasing' and give you the actual number of how much. The [CGM] app just gives you arrows. It's just not as helpful.”
Casey: The Painful Interruption
Parent
dexcom
Casey is on her smartphone at work when an alert blares in her ear. Her follow app doesn’t know she’s on the phone. What can she do? Turn alerts down or off at work, with the risk of missing something important.
"Alerts screaming loudly when on a phone call are painful."

Maria: The False Alarm
Adult
libre
Maria is woken at 3 AM by an urgent low alert. She does a fingerstick and discovers her glucose is within range. Annoyed, she snoozes all alarms for the next six hours, choosing uninterrupted sleep over a system she can't trust.
“The problem is, I was sleeping on my sensor [causing a compression low]. But nobody told me not to do that. I followed the instructions. I had to learn about things like that by going onto Reddit.”
Blake: The Call for Autonomy
child
dexcom
Blake has a smartphone with the CGM app, but its nighttime alerts don’t always wake him up. And the companion follow app his parents use to monitor his glucose can’t be set up to only send critical alerts. This reduces Blake’s ability to manage his diabetes independently and creates more sleep disruption for his parents.
“We need a 'find my phone'-like persistent alert for my child at night, as alerts often don't wake him."
John: The Smartphone Catch-22
child
dexcom
For John, it’s easy to forget to bolus at lunchtime at school. But when John’s CGM picks up his high glucose level, that alert only goes to his pump, because, like many kids, he’s not ready for a smartphone. The pump’s alerts don’t capture his attention sufficiently, and Mom can’t receive the alerts because the CGM follow app requires John to use a smartphone.
"You have to give your kid a smartphone, with all the trouble that brings, or you get no remote monitoring and alerts."
Alex: The Irritating Buzz
adult
Eversense
Alex is in an important meeting at work. A sharp, constant vibration starts buzzing against his arm—an alert from his on-body transmitter. The vibration is intensely annoying and distracting. He wishes he could change it to be softer for non-urgent alerts, but there's no way to customize the intensity or pattern.
"It's like a cat scratching on your arm... and there's no way to mute it, alter it, or tone it down. You can't turn it up when you're asleep and turn it down when you're awake. There's no way to modify it."
Sierra: The DIY Insights
adult
Dexcom
Sierra gets an alert from her CGM app on her smartphone with a single arrow pointing down. She ignores it and instead glances at two apps on her smartwatch which tell her she is "decreasing slightly" and that her glucose is falling at a rate of "-2 mg/dL per minute." This rich, quantitative data, which she had to route through multiple third-party apps, gives her the confidence to watch and wait, rather than over-treating a minor dip.
“I prefer to get my alerts through Night Scout or Sweet Dreams. They say things like 'slightly decreasing' and give you the actual number of how much. The [CGM] app just gives you arrows. It's just not as helpful.”
Casey: The Painful Interruption
Parent
dexcom
Casey is on her smartphone at work when an alert blares in her ear. Her follow app doesn’t know she’s on the phone. What can she do? Turn alerts down or off at work, with the risk of missing something important.
"Alerts screaming loudly when on a phone call are painful."

Real Life Stories

Let’s take a look at a few snapshots taken directly from our research (names changed and stock images used for patient privacy):

Maria: The False Alarm
Adult
libre
Maria is woken at 3 AM by an urgent low alert. She does a fingerstick and discovers her glucose is within range. Annoyed, she snoozes all alarms for the next six hours, choosing uninterrupted sleep over a system she can't trust.
“The problem is, I was sleeping on my sensor [causing a compression low]. But nobody told me not to do that. I followed the instructions. I had to learn about things like that by going onto Reddit.”
Blake: The Call for Autonomy
child
dexcom
Blake has a smartphone with the CGM app, but its nighttime alerts don’t always wake him up. And the companion follow app his parents use to monitor his glucose can’t be set up to only send critical alerts. This reduces Blake’s ability to manage his diabetes independently and creates more sleep disruption for his parents.
“We need a 'find my phone'-like persistent alert for my child at night, as alerts often don't wake him."
John: The Smartphone Catch-22
child
dexcom
For John, it’s easy to forget to bolus at lunchtime at school. But when John’s CGM picks up his high glucose level, that alert only goes to his pump, because, like many kids, he’s not ready for a smartphone. The pump’s alerts don’t capture his attention sufficiently, and Mom can’t receive the alerts because the CGM follow app requires John to use a smartphone.
"You have to give your kid a smartphone, with all the trouble that brings, or you get no remote monitoring and alerts."
Alex: The Irritating Buzz
adult
Eversense
Alex is in an important meeting at work. A sharp, constant vibration starts buzzing against his arm—an alert from his on-body transmitter. The vibration is intensely annoying and distracting. He wishes he could change it to be softer for non-urgent alerts, but there's no way to customize the intensity or pattern.
"It's like a cat scratching on your arm... and there's no way to mute it, alter it, or tone it down. You can't turn it up when you're asleep and turn it down when you're awake. There's no way to modify it."
Sierra: The DIY Insights
adult
Dexcom
Sierra gets an alert from her CGM app on her smartphone with a single arrow pointing down. She ignores it and instead glances at two apps on her smartwatch which tell her she is "decreasing slightly" and that her glucose is falling at a rate of "-2 mg/dL per minute." This rich, quantitative data, which she had to route through multiple third-party apps, gives her the confidence to watch and wait, rather than over-treating a minor dip.
“I prefer to get my alerts through Night Scout or Sweet Dreams. They say things like 'slightly decreasing' and give you the actual number of how much. The [CGM] app just gives you arrows. It's just not as helpful.”
Casey: The Painful Interruption
Parent
dexcom
Casey is on her smartphone at work when an alert blares in her ear. Her follow app doesn’t know she’s on the phone. What can she do? Turn alerts down or off at work, with the risk of missing something important.
"Alerts screaming loudly when on a phone call are painful."

Real Life Stories

Let’s take a look at a few snapshots taken directly from our research (names changed and stock images used for patient privacy):

Maria: The False Alarm
Adult
libre
Maria is woken at 3 AM by an urgent low alert. She does a fingerstick and discovers her glucose is within range. Annoyed, she snoozes all alarms for the next six hours, choosing uninterrupted sleep over a system she can't trust.
“The problem is, I was sleeping on my sensor [causing a compression low]. But nobody told me not to do that. I followed the instructions. I had to learn about things like that by going onto Reddit.”
Blake: The Call for Autonomy
child
dexcom
Blake has a smartphone with the CGM app, but its nighttime alerts don’t always wake him up. And the companion follow app his parents use to monitor his glucose can’t be set up to only send critical alerts. This reduces Blake’s ability to manage his diabetes independently and creates more sleep disruption for his parents.
“We need a 'find my phone'-like persistent alert for my child at night, as alerts often don't wake him."
John: The Smartphone Catch-22
child
dexcom
For John, it’s easy to forget to bolus at lunchtime at school. But when John’s CGM picks up his high glucose level, that alert only goes to his pump, because, like many kids, he’s not ready for a smartphone. The pump’s alerts don’t capture his attention sufficiently, and Mom can’t receive the alerts because the CGM follow app requires John to use a smartphone.
"You have to give your kid a smartphone, with all the trouble that brings, or you get no remote monitoring and alerts."
Alex: The Irritating Buzz
adult
Eversense
Alex is in an important meeting at work. A sharp, constant vibration starts buzzing against his arm—an alert from his on-body transmitter. The vibration is intensely annoying and distracting. He wishes he could change it to be softer for non-urgent alerts, but there's no way to customize the intensity or pattern.
"It's like a cat scratching on your arm... and there's no way to mute it, alter it, or tone it down. You can't turn it up when you're asleep and turn it down when you're awake. There's no way to modify it."
Sierra: The DIY Insights
adult
Dexcom
Sierra gets an alert from her CGM app on her smartphone with a single arrow pointing down. She ignores it and instead glances at two apps on her smartwatch which tell her she is "decreasing slightly" and that her glucose is falling at a rate of "-2 mg/dL per minute." This rich, quantitative data, which she had to route through multiple third-party apps, gives her the confidence to watch and wait, rather than over-treating a minor dip.
“I prefer to get my alerts through Night Scout or Sweet Dreams. They say things like 'slightly decreasing' and give you the actual number of how much. The [CGM] app just gives you arrows. It's just not as helpful.”
Casey: The Painful Interruption
Parent
dexcom
Casey is on her smartphone at work when an alert blares in her ear. Her follow app doesn’t know she’s on the phone. What can she do? Turn alerts down or off at work, with the risk of missing something important.
"Alerts screaming loudly when on a phone call are painful."

A Hierarchy of Needs

As we talked to patients and looked at the patterns of their experience and feedback, a clear hierarchy of needs emerged for CGM alert systems:

Reliable and Accurate. Consistent delivery of accurate reporting where it's needed.
Customizable. Granular control of alert sounds and haptics, context-specific profiles, and robust data portability.
Intelligent. Smart, predictive alerts and context-aware experience, handling edge cases with sophistication.
Empowering. Advanced pattern recognition with highly personalized insights, and adaptive, mastery-based guidance.

Few people we talked to were getting their needs at levels 1-2 met comprehensively. So, they often gave limited consideration to levels 3-4 and CGMs' newer features that attempt to serve these needs. They are stuck dealing with more foundational diabetes management concerns.

A Closer Look at the Pain Points

Let’s explore the two most common alert problem areas from our research.

Data Transmission

Problem: People often can’t get alerts where and when they really need them.

A smartwatch isn’t compatible at all, transmission to it isn’t reliable, or connectivity requires the presence of their smartphone as well.
A patient must have a smartphone to transmit data to a follow app, and even then, alerts aren’t in parity with the CGM app.
A third-party app with a preferred alert experience can’t access the data easily, timely, or at all.
A connectivity (wifi or cell reception) dependence for both devices for caregivers to receive alerts, and when connectivity isn’t available, “no data” alerts are unhelpful and lack good snoozing options.
Aggressive battery-life management can hurt alert reliability.

Impact: These challenges make it harder for people to get the alerts they need, when and where they need them. They leave patients uninformed in critical moments, create anxiety for caregivers, and restrict paths to improve insight and attention capture. A failure in transmission can break the value chain in multiple places.

Sensory Experience

Problem: Different people, in different contexts, need different sensory experiences for effective CGM alerts.

CGMs have limited or no context awareness. Alert settings must be manually changed for varying situations.
Audible alert settings are often the most configurable, but aren’t appropriate for every situation.
Haptics usually have little or no configurability (intensity and pattern in general and per alert type), undermining the potential to be the preferred sensory experience.
Escalation isn’t possible (a multi-step alert experience that demands more attention if ignored), often forcing a choice between sensory options that are too discrete or too disruptive.
Snoozing options are limited or nonexistent, leaving you with alert fatigue or missing reminders.
There can be complexity and confusion between app settings and device-wide settings (e.g. DnD).

Impact: These challenges put a stressful burden on patients caught between the fear of missing critical alerts and being disrupted by irritating, embarrassing, or painful alerts. The results include sleep disruption, anxiety, emotional and cognitive fatigue, less time in range, and ultimately, a greater risk of dangerous glycemic events.

Experience Mapping for Parent-Teen Team

Managing a Complex Alert Ecosystem

For most CGM users, the companion app for their CGM sensor serves as the central hub of their alert ecosystem. They depend on it to reliably:

Collect blood-glucose data from the sensor
Communicate that data in useful ways
Make that data available to other devices, apps, and people in their alert ecosystem

But not all CGM users have a smartphone. And many want or need to receive at least some of their alerts on other devices and apps. They often build complex alert ecosystems in an effort to overcome the shortcomings and limitations of the first-party CGM experience.

This places a significant burden on patients and their families, leading to increased cognitive load and emotional burden and risks from dependence on fail-prone workarounds.

The Bigger Picture

Even though our research uncovered many pain points with the alert systems in today’s CGMs, the bigger picture is more optimistic.

Three out of every four participants in our study indicated they would be “very disappointed” if they had to switch to another CGM system, and indicated they were “happy” or “very happy” with their CGM overall. These systems are life-changing, even with their many shortcomings.

But there’s an opportunity to significantly improve patient quality of life and grow market share through more rapid advancements in user experience.

Improving the Alert Experience

What does our research point to as some of the best opportunities to improve the CGM alert experience?

We’ve used the value journey to help surface and understand user experience problems. Now our hierarchy of needs provides a framework to inform our product roadmap.

Level 1: Reliability

Data portability – secure, efficient data access through open protocols with sufficient performance to support real-time alert experiences
Smartwatch support – first-class connections between sensors and all major smartwatch platforms
Smartwatch transmission – data portability via connected smartwatches
Inaccuracy handling – detect known factors reducing measurement accuracy (e.g. miscalibration, end of sensor life, compression) and respond proactively with correction/recalibration or situational education to users
OS handling – anticipate system-level alert impediments (e.g. power management, focus modes) and respond with contextual education to users
“No data” handling – when sensor data isn’t available to transmit, distinguish between sensor malfunction/failure and a lack of connectivity and send context where appropriate (e.g. last known location)
Satellite connectivity – as satellite connectivity emerges for smartphones, lobby for and utilize access for data transmission

Level 2: Customization

Context management – support creation of context profiles with unique alert preferences and thresholds with automatic application based on things like schedule, location, and app use (e.g. on a phone call).
Haptics – support creation of alert haptics with custom intensity and patterns
Escalations – support creation of multi-step alert escalations with configuration of sound, haptics, device(s), and timing
Data richness – provide a few options for how alert insights are presented for people who value richer data in written (“slightly decreasing”) or numeric form (rate of change per unit of time).
Recurrence – support easy snoozing (for X minutes) and repeat (every X minutes if unresolved) options for non-critical alerts
Follow app parity – provide the same alert customization experience in the follow app as in the primary CGM app
Status updates to followers – to keep followers informed, allow the CGM user to indicate that action is being taken on an alert, and to include a message if appropriate

Level 3: Intelligence

Predictive alerts – provide predictive alerts that are personalized by a progressive understanding of their correlated impact on blood glucose overall and by context
Progressive UI – as users master basic capabilities, make them aware of more advanced capabilities that may be helpful little by little
Pattern matching – use advanced data analysis combined with research-informed algorithms to surface correlations between blood glucose and time of day, day of the week, location, input events, demographics, and other factors, providing guidance to better anticipate and understand blood glucose levels and manage insulin dosing, glucose intake, meal choices, activity, and rest

Level 4: Empowerment

Adaptive, Contextual Education – proactive education relevant to the immediate situation (related to inputs and outputs) and what the user is doing in the app
Enable Mastery – provide a proactive education and skill-building path personalized to user technical skill, health literacy, and desired mastery level, timing, and pacing
Feedback Loop – solicit feedback on the education and guidance users receive and further personalize their experience based on this feedback and improve the education for all users

From Devices to a System of Care

The next leap forward in CGM may not be in the sensor, but in the silence. Patients and families are already building fragile ecosystems of apps, devices, and human backups to compensate for alert shortcomings—clear signals of unmet needs.

For product and design leaders, the opportunity is to move from designing devices and apps to orchestrating systems of care: alerts that are predictably reliable, adaptive to context, and measured not only in time-in-range but in trust, sleep, and peace of mind.

Alerts aren’t a side feature–they are the experience of CGMs. Done right, they can fade into the background, freeing patients and families to focus not on their diabetes, but on their lives.

Stay In Touch

We’ll soon publish our next article from this research. Would you like a brief email then, so you don’t miss it? (We’ll treat your inbox and time with care.)

Managing a Complex Alert Ecosystem

For most CGM users, the companion app for their CGM sensor serves as the central hub of their alert ecosystem. They depend on it to reliably:

Collect blood-glucose data from the sensor
Communicate that data in useful ways
Make that data available to other devices, apps, and people in their alert ecosystem

This places a significant burden on patients and their families, leading to increased cognitive load and emotional burden and risks from dependence on fail-prone workarounds.

The Bigger Picture

Even though our research uncovered many pain points with the alert systems in today’s CGMs, the bigger picture is more optimistic.

But there’s an opportunity to significantly improve patient quality of life and grow market share through more rapid advancements in user experience.

Improving the Alert Experience

What does our research point to as some of the best opportunities to improve the CGM alert experience?

We’ve used the value journey to help surface and understand user experience problems. Now our hierarchy of needs provides a framework to inform our product roadmap.

Level 1: Reliability

Data portability – secure, efficient data access through open protocols with sufficient performance to support real-time alert experiences
Smartwatch support – first-class connections between sensors and all major smartwatch platforms
Smartwatch transmission – data portability via connected smartwatches
Inaccuracy handling – detect known factors reducing measurement accuracy (e.g. miscalibration, end of sensor life, compression) and respond proactively with correction/recalibration or situational education to users
OS handling – anticipate system-level alert impediments (e.g. power management, focus modes) and respond with contextual education to users
“No data” handling – when sensor data isn’t available to transmit, distinguish between sensor malfunction/failure and a lack of connectivity and send context where appropriate (e.g. last known location)
Satellite connectivity – as satellite connectivity emerges for smartphones, lobby for and utilize access for data transmission

Level 2: Customization

Context management – support creation of context profiles with unique alert preferences and thresholds with automatic application based on things like schedule, location, and app use (e.g. on a phone call).
Haptics – support creation of alert haptics with custom intensity and patterns
Escalations – support creation of multi-step alert escalations with configuration of sound, haptics, device(s), and timing
Data richness – provide a few options for how alert insights are presented for people who value richer data in written (“slightly decreasing”) or numeric form (rate of change per unit of time).
Recurrence – support easy snoozing (for X minutes) and repeat (every X minutes if unresolved) options for non-critical alerts
Follow app parity – provide the same alert customization experience in the follow app as in the primary CGM app
Status updates to followers – to keep followers informed, allow the CGM user to indicate that action is being taken on an alert, and to include a message if appropriate

Level 3: Intelligence

Predictive alerts – provide predictive alerts that are personalized by a progressive understanding of their correlated impact on blood glucose overall and by context
Progressive UI – as users master basic capabilities, make them aware of more advanced capabilities that may be helpful little by little
Pattern matching – use advanced data analysis combined with research-informed algorithms to surface correlations between blood glucose and time of day, day of the week, location, input events, demographics, and other factors, providing guidance to better anticipate and understand blood glucose levels and manage insulin dosing, glucose intake, meal choices, activity, and rest

Level 4: Empowerment

Adaptive, Contextual Education – proactive education relevant to the immediate situation (related to inputs and outputs) and what the user is doing in the app
Enable Mastery – provide a proactive education and skill-building path personalized to user technical skill, health literacy, and desired mastery level, timing, and pacing
Feedback Loop – solicit feedback on the education and guidance users receive and further personalize their experience based on this feedback and improve the education for all users

From Devices to a System of Care

Alerts aren’t a side feature–they are the experience of CGMs. Done right, they can fade into the background, freeing patients and families to focus not on their diabetes, but on their lives.

Stay In Touch

We’ll soon publish our next article from this research. Would you like a brief email then, so you don’t miss it? (We’ll treat your inbox and time with care.) You can also follow us on LinkedIn.

Subscribe to the Radian Podcast for the latest in digital health product development.

Youtube

Spotify