Rethinking control‐IQ+ technology: Simple strategies for easy optimization

Abstract

Control‐IQ+ is an automated insulin delivery (AID) algorithm approved for people with type 1 diabetes aged 2+ years and adults aged 18+ years with type 2 diabetes. While numerous publications support improved glycaemia and quality of life for people with diabetes, this practice paper is intended to encourage uptake for healthcare professionals (HCPs) who are less familiar with AID systems. This includes recommendations for initializing settings, as well as research and practice‐based approaches to optimizing glycaemia with stronger settings (e.g., strong correction factor settings). In addition to automated insulin adjustments every 5 min, Control‐IQ+ delivers a large Autobolus up to once per hour, which can improve glycaemia in people who have challenges with routine bolusing. Simple bolus strategies and tips and tricks also highlight how to make Control‐IQ+ easy to use for HCPs and the person with diabetes.

1. INTRODUCTION

Automated insulin delivery (AID) has greatly enhanced diabetes care for people with type 1 diabetes (T1D) and insulin‐treated type 2 diabetes (T2D). An AID system consists of an insulin pump that uses an algorithm to automatically adjust insulin delivery based on glucose readings from a connected Continuous Glucose Monitoring (CGM) device. AID use has demonstrated superior glycaemic outcomes, improvements in quality of life and reductions in diabetes burden when compared to usual care (multiple daily injections [MDI] or insulin pump without automation) in people with T1D and insulin‐treated T2D.^{1, 2, 3, 4, 5, 6, 7, 8} AID systems provide improvement over previous therapies in helping people with diabetes reach internationally recognized glycaemic targets; however, future innovation will be required to consistently meet glycaemic benchmarks across the spectrum of people with diabetes. Nonetheless, multiple international organizations recommend AID as the preferred insulin delivery method for children and adults with T1D and people with T2D, which brings an opportunity for healthcare professionals (HCPs) to consider this therapy for more people with diabetes.^{9, 10, 11, 12} Currently, there are several AID systems available in different countries and areas of the world. These systems include the t:slim X2 and Tandem Mobi systems with Control‐IQ+ technology (Tandem Diabetes), 780G (MiniMed), Omnipod 5 (Insulet), CamAPS FX (CamDiab Ltd., Cambridge, UK) and Diabeloop DBLG1 (Diabeloop, France). In the United States, two additional AID systems are available: the iLet Bionic Pancreas (Beta Bionics) and twiist AID system (Sequel Med Tech). All AID systems to date are currently considered hybrid closed loop systems, meaning they are optimized to be used when a person delivers boluses for meals. Each system has different ways to adjust settings to optimize glycaemia, and knowing these strategies is important for healthcare professionals (HCPs) working with each system. ¹³

The t:slim X2 with Control‐IQ technology has been available since 2021 for people with T1D ages 6 and up. The system is backed up by data from three randomized control trials and a meta‐analysis of users aged 2–72 years reporting a significant reduction of A1c by 0.38% and a significant improvement in time in range (TIR, 70–180 mg/dL, 3.9–10 mmol/L) of +11.5% (2.8 h/day) with Control‐IQ, compared to usual care. These improvements were irrespective of age, socioeconomic status and baseline glycaemia^{14, 15, 16, 17} Large real‐world studies from multiple countries also report improvements in glycaemia over previous therapies,^{18, 19, 20, 21} and a post‐market prospective real‐world study confirms lower rates of severe hypoglycaemia, diabetic ketoacidosis (DKA) and improvements in quality of life over 1 year compared to historical registry data. ²² A recent additional clearance of the updated Control‐IQ+ technology (Control‐IQ+) has expanded labelling for children ages 2–5 years with T1D and adults with insulin‐treated T2D. ²³ An RCT of Control‐IQ+ technology in adults with T2D reported a 0.9% reduction in HbA1c after 3 months of Control‐IQ+ use (mean adjusted group difference of −0.6% compared to standard care), with benefit irrespective of age, medication use (including GLP1‐RA), and whether the person engaged in carbohydrate counting. ⁶ Moreover the improvement was irrespective of C‐peptide level ²⁴ or presence of GAD antibodies. ²⁵ With both T1D and T2D studies, improvement in glycaemia was highest in those with the lowest baseline control.^{6, 14}

While the Tandem AID system has been available for many years, perspectives from clinical practice and data analyses give new insight into optimizing use of this system. These insights are outlined here, with published recommendations referenced, and expert opinion given based on clinical practice. Concepts are further illustrated in hypothetical and anonymized clinician vignettes to highlight how these strategies may be used in real life. This is especially useful for HCPs who are less familiar with Control‐IQ+, including primary care physicians, who will require additional support and knowledge to routinely implement AID in their practice.^{26, 27} The purpose of this article, therefore, is to provide these practical strategies for a simple start, simple use and simple customization with Control‐IQ+, in an effort to equip HCPs to help people with diabetes (PWD) meet their personal glycaemic and lifestyle goals for diabetes therapy.

2. REVIEW OF CONTROL‐IQ+ TECHNOLOGY

Control‐IQ+ is an advanced hybrid closed loop that dynamically increases, decreases or suspends insulin delivery every 5 min to maximize time in range 70–180 mg/dL (3.9–10 mmol/L) and minimize time below 70 mg/dL (3.9 mmol/L). The every 5 min glucose‐responsive dosing uses the basal settings as the ‘baseline’ level of insulin and adjusts up or down in response to predicted glucose levels. The algorithm additionally features an Autobolus which occurs up to once an hour when glucose levels are predicted to be >180 mg/dL (10 mmol/L). This feature delivers a larger amount of insulin at one time (bolus), designed to prevent or mitigate hyperglycaemia, with safe limits to prevent insulin stacking. The Autobolus has been shown to deliver a substantial proportion of bolus insulin in both clinical trials and real‐world use, especially benefiting people with higher HbA1c levels^{14, 28} and people who have a tendency of missing boluses.^{29, 30}

Control‐IQ+ uses: (1) correction factor (CF) (2) basal rates and (3) insulin‐to‐carbohydrate ratios (ICR) to impact automation. Every AID system has different settings that can be used to optimize automation; hence, knowing these settings is key for healthcare professionals. ¹³ For example, while a user can adjust the ‘glucose target’ setting in systems like the MiniMed 780G or CamAPS, similar outcomes are achieved with Control‐IQ+ by adjusting CF and basal rates.^{31, 32, 33} Resources like www.Pantherprogram.org can help HCPs keep track of which settings are modifiable for each AID system. The algorithm features of Control‐IQ+ are shown in Table 1, including the differences between the original Control‐IQ technology and Control‐IQ+.

TABLE 1.

Features of the Control‐IQ+ algorithm.

Feature	Description
TDI and weight parameters	TDI range: 5–200 units (previously 10–100 units) Weight range: 20–440 lbs (previously 55–308 lbs) Optimized algorithm for basal rate of >3 units/h
Every 5 min automation	Dynamic increase (>4 times), decrease or full suspension of programmed basal rate in response to predicted glucose levels, targeting 112.5–160 mg/dL (6.25–8.9 mmol/L)
Autobolus	Up to once per hour bolus of insulin designed to prevent or mitigate hyperglycaemia. This is calculated using the 30 min predicted glucose level, attenuated to 60% based on future glucose, not current glucose
Algorithm correction target	110 mg/dL (6.1 mmol/L)
AID settings can be adjusted	Correction Factor (CF) Insulin‐to‐carbohydrate ratio (ICR) Basal rates
User‐given boluses	Bolus calculator uses ICR (for carbohydrate entry) and CF
Sleep Activity	Targets glucose levels 112.5–120 mg/dL (6.25–6.7 mmol/L) with more dynamic 5‐min automation and without Autoboluses. Can be set on a daily schedule
Advanced features (optional use)
Personal Profiles	Up to 6 different profiles of basal rates, ICR, CF for different scenarios (e.g., sick day, weekends, etc.)
Exercise Activity	Automation target 140–160 mg/dL (7.8–8.9 mmol/L). Autobolus remains active in the exercise mode to prevent hyperglycaemia.
New features with Control‐IQ+	Extended bolus: Can extend user‐given carbohydrate bolus of insulin up to 8 h when Control‐IQ+ is active Temp basal: Can temporarily change the amount of basal insulin as the basis of automation when Control‐IQ+ active Profile settings calculator: Available on later pump versions—automatically calculates initial pump settings for people coming from MDI (optional) Spanish language option in the United States

The Control‐IQ+ algorithm is available on two different pumps: the t:slim X2 and the Tandem Mobi (Figure 1). The t:slim X2 is a touch screen, rechargeable insulin pump that holds up to 300 units of insulin. The Tandem Mobi is a 200‐unit insulin pump that can be worn with an adhesive sleeve attached to the body, or clipped to clothing, and is controlled from a smartphone. ³⁴ The Tandem X2 and Mobi are currently compatible with multiple CGMs, including Dexcom and Libre options. Control‐IQ+ data are uploaded to the cloud automatically via Tandem pump mobile apps or via manual cable download from the t:slim. Insulin and glycaemic data can be viewed using Tandem Source, Glooko (Figure 2) and Tidepool.

FIGURE 1.

The t:slim X2 and Tandem Mobi insulin pumps with Control‐IQ+ technology.

FIGURE 2.

Tandem Source Overview report (left) and Glooko Summary Report (right).

3. STRATEGIES FOR SIMPLE START‐UP AND FOLLOW‐UP ADJUSTMENTS

3.1. Strategies for simple start

To start Control‐IQ+, a ‘pump total daily insulin’ (pump TDI) amount is estimated based on current insulin therapy or weight as described below. This pump TDI can be used to calculate initial basal rates, CF and ICR, which is the foundation for automation. Approaches will differ slightly based on whether the user is transitioning from MDI or a different insulin pump.

3.1.1. Starting control‐IQ+ from multiple daily injections

When transitioning from MDI, pump TDI may be the same as injection TDI or may need to be reduced 20–25%, especially if glycaemia is close to target (HbA1c <8% or TIR >60) or there is a high degree of hypoglycaemia. ³⁵ For people with higher HbA1c due to missing insulin doses, pump TDI can also be calculated using weight (~0.5–0.7 units/kg/day).

Recommended initial settings are found in Table 2, For basal rate settings, we recommend setting basal rates as 50–60% of the pump TDI.^{33, 35} For many people, a single basal rate is adequate, since Control‐IQ+ can compensate for day‐to‐day variability that occurs within a person, ³⁶ though this may be different for younger adults or children who have different circadian patterns of insulin secretion (see section 6 below). ³⁷

TABLE 2.

Simple start‐up with Control‐IQ+ technology.

Task	Considerations
Initial training	Can be done in person or via digital/virtual training22, 46 Set up device accounts prior to training if possible (e.g., Tandem Source, Glooko, Tidepool, CGM accounts, etc.)
Starting from Multiple Daily injections	Consider reducing injection TDI by 20%–25% if HbA1c <8% (or TBR >4%). May not be needed if HbA1c >8% or low TBR. Can consider averaging with weight‐based calculation (kg x 0.5) if TDI is not reliable 35 If Profile Settings Calculator available, user can enter injection TDI and have settings calculated automatically
Starting from other pump systems	Use TDI from other pump systems as a starting point Recalculate new starting settings as Control‐IQ+ works differently than other systems (Other systems do not use basal rates or correction factors to impact automation, so these are likely not optimized. ICR may also be tuned differently due to variable insulin action times or users entering fake carbohydrates)
Calculate starting doses	Starting settings can be intensified later. Guidelines may be used35, 38: Basal insulin: ~50%–60% of TDI divided by 24 h—can be set as single rate 65 CF: ≤1700/TDI for mg/dL (≤94/TDI for mmol/L) and strengthen if no hypoglycaemia (lower number, e.g., ≤1580/TDI for mg/dL or 88/TDI for mmol/L for best outcomes with stronger correction factor 33 ) ICR: 300–450/TDI and strengthen if no hypoglycaemia (lower number) These settings may be different for young children or special populations 36, 37
Turn on automation	Enter weight and TDI to start Control‐IQ+ (used for system start‐up automation but not instrumental over time)
Set up Sleep Activity	Sleep Activity will maximize overnight glycaemia with more aggressive continuous delivery without Autoboluses. Set 2 h after usual sleeping time, so the system can give 1–2 Autoboluses as needed to bring glucose to near normal level before start of Sleep Activity (which does not deliver Autoboluses). Sleep schedules can be different based on day of the week. Consider setting Sleep Activity 24/7 for people who desire tight glycaemia and bolus routinely for meals—this will maximize every 5 min insulin delivery without need for Autoboluses 44
Key education considerations	Maximize time in automation, wear CGM continually for optional glycaemia Best outcomes when bolusing for meals. Round numbers can be used to estimate small, medium and large meal carbohydrates (ex: 30, 60, 90g, see section below) If glucose >300 mg/dL (16.7 mmol/L) for 2 h, assume infusion site failure ‘When it doubt, switch it out’ 38 Treat hypoglycaemia <70 mg/dL (3.9 mmol/L) with less carbohydrate (5–10g) given that insulin delivery has already suspended. Ensure users have ketone testing equipment and a back‐up plan for how to transition back to subcutaneous injection in case of pump failure

Note: <70 mg/dL (3.9 mmol/L).

Abbreviations: CF, correction factor; ICR, insulin‐to‐carbohydrate ratio; TDI, total daily insulin; TBR, time below range.

For bolus settings, CF of ≤1700/TDI for mg/dL (≤94/TDI for mmol/L) and an ICR of 300–450/TDI can be a good starting place for Control‐IQ+, though guidance varies by organization and age of user.^{35, 38} Control‐IQ+ users likely benefit from more aggressive CF settings. ³³

Some Tandem pump software versions include a Profile Settings Calculator, which automatically uses the above guidelines to calculate initial pump settings for a person switching from MDI. To use this feature, a person enters their injection TDI and the pump populates the initial insulin doses (including the 25% reduction in TDI). These can then be adjusted by the HCP, pump trainer or pump user at any time (Figure 2).

3.1.2. Starting Control‐IQ+ from other pumps or AID systems

While a reduction in TDI may not be needed, it is recommended that initial pump settings are recalculated from the pump TDI when transitioning between AID systems, as settings that worked well with one AID system may not work well for another.^{11, 13} The initial basal rates, ICR and CF can be calculated as per above. Control‐IQ+ automation can start as soon as settings are entered, CGM is connected, and automation is turned on. Table 2 summarizes easy start up considerations.

3.1.3. The importance of initial follow‐up for optimization

Starting settings can be more conservative as long as the settings are optimized within the first 2 weeks of wear (see next section). It is suggested to review settings at day 3 (first infusion set change), day 7, and then as needed. This can be as simple as ‘spot checking’ the correction factor, ICR and basal rates based on TDI and intensifying as described below.

3.2. Control‐IQ report interpretation and follow‐up

HCPs can analyse overall glycaemia, insulin delivery and user behaviours in the Tandem Source Overview report or the Glooko Summary Report (Figure 2). Figure 3 highlights the review process for follow‐up, which includes glycaemic assessment, system use and optimizing settings. ³⁹

FIGURE 3.

Strategies for follow‐up assessing and optimizing Control‐IQ+ technology.

3.2.1. Review glycaemia

The first step is to review glycaemia for the past 2 weeks (or shorter period after system start) to determine next steps for system and settings optimization (Figure 3). Target glycaemia includes >70% TIR (70–180 mg/dL, 3.9–10 mmol/L), <4% TBR (<70 mg/dL, 3.9 mmol/L) and <25% Time Above Range (TAR >180 mg/dL, 10 mmol/L),^{40, 41} though personal glycaemic goals should be considered as well, especially in children, pregnant individuals and people at higher risk for health complications. ⁴¹ Time in a tight range (70–140 mg/dL, 3.9–7.8 mmol/L) can also be considered ⁴² in certain populations who desire to have an A1c below 6.5%.

3.2.2. Assess automation, sleep activity and user behaviour

It is most important to check that CGM and automation use are as high as possible, aiming for 90–95% time of Control‐IQ+ system use. ³⁹ Sleep Activity schedule should also be set for every night (7 nights per week) for most users. Beyond these essentials, the HCP can also use the Overview/Summary Report to check user behaviours, including frequency of user‐initiated boluses and infusion set changes. Reviewing the Ambulatory Glucose Profile (found on the preview/summary page) and daily reports will reveal bolus timing and frequency of missed boluses. Understanding how often the PWD boluses and when they bolus relative to their meal can give insight into how to best support their personal goals (see bolus section below).

3.2.3. Optimize settings

Stronger pump settings with Control‐IQ+ are associated with higher TIR. ³³ The correction factor is especially impactful for Control‐IQ+ automation, as it impacts both the dynamic continuous automation, as well as direct calculation of the Autobolus dose. While there is no magic formula, a CF of <1580/TDI for mg/dL (<88 in mmol/l) is associated in real‐world use with the highest mean TIR with low hypoglycaemia on average (lower number = stronger CF). ³³ Importantly, in predictive analysis, stronger CF did not predict an increase in clinical hypoglycaemia, which may empower HCPs to feel more confident making larger changes in CF without risking hypoglycaemia. This may result in a significant improvement in TIR without the user needing to change any behaviours.

Insulin‐to‐carbohydrate ratio can also be used to impact overall glycaemia, with ICR impacting glucose levels when the user enters carbohydrate grams into the system. Stronger ICRs (lower number) are associated with higher TIR. The same real‐world analysis reported that ICRs <330/TDI were associated with 77% TIR and 1.1% hypoglycaemia. ³³

Finally, basal insulin can also be adjusted to impact overall TIR, similar to the overall glucose target used in other AID systems. Basal rates can be adjusted by time of day or hourly, providing additional flexibility in optimization over a 24‐h period. For higher TIR or lower mean glucose, such as during pregnancy, the programmed basal insulin should be 10–25% higher than the delivered basal; if there is a significant fear or risk of hypoglycaemia, then the programmed basal rates should be 10–25% lower than the delivered basal. An HCP can look at the average total daily basal that is delivered by the system and compare it to the programmed basal rate total. ⁴³

For people who bolus routinely for meals, Sleep Activity can be used during all hours to tighten the automation target to 112.5–120 mg/dL (6.25–6.7 mmol/L) for every 5‐min automation. A database analysis of 7000 people using Sleep Activity all the time indicates an average TIR of 76% with 5.6 boluses given per day. ⁴⁴ The tight target works best with routine bolusing, as the Autobolus is not delivered during the use of this feature.

4. PRACTICAL MEALTIME STRATEGIES

Current AID systems are used optimally when users deliver boluses for carbohydrate ingestion. Control‐IQ+ allows a user to deliver boluses for meals based on entry of carbohydrates or units of insulin, and users can also deliver an extended bolus of up to 8 h for high‐fat meals or conditions such as gastroparesis. Mealtime strategies can be adjusted to meet user needs in simple ways, as AID has been shown to benefit people regardless of carbohydrate counting skills and bolusing behaviours.^{6, 28, 45} Though carbohydrate counting and pre‐meal bolus have been associated with better glycaemic outcomes with the use of AID systems, ⁴⁶ nearly 1 in 10 users of Control‐IQ have gone a week or longer without delivering a bolus. ³⁰ Missed boluses can be partially compensated for by Control‐IQ+, with a French study of 60 children reporting that users who miss 1 bolus per day still improved TIR by 15% compared to sensor augmented pump and those missing 2 or more boluses per day improving TIR 17%. ²⁹ These results are consistent with a US based real‐world pilot study in adults with T1D showing a 19% improvement in TIR for people who do not routinely bolus over the course of 1 year. ²⁸

4.1. Missed boluses

For people who miss boluses, the Control‐IQ+ Autobolus and dynamic 5‐min insulin delivery can be leveraged to maximize TIR. Since the Autobolus calculates a correction dose based on the future predicted glucose down to the system target of 110 mg/dL (6.1 mmol/L), strengthening the CF works well to maximize its function. Our expert opinion is that aggressive CF settings may be recommended for people who do not bolus (e.g., 1500/TDI for mg/dL or 83/TDI mmol/L, or even stronger/lower). Further, higher programmed basal rates may be used to compensate for missed meal boluses (e.g., >60% TDI, Table 3).

TABLE 3.

Practical strategies for management challenges and unique populations.

Considerations	Strategies
Bolusing considerations	For non‐bolusers: Increase basal rates to compensate for a large portion of TDI (e.g., >60%) To maximize the impact of the Autobolus, strengthen the correction factor as tight as necessary (e.g., 1500/TDI, for mg/dL or 83/TDD for mmol/L) to maximize euglycaemia without hypoglycaemia (lower number = more impact). For occasional/variable missed boluses: To maximize the impact of the Autobolus, tighten/strengthen the correction factor (lower number = more impact). The ICR may be relaxed to minimize risk of hypoglycaemia from the more aggressive correction factor (less carb insulin to offset the increased correction insulin) For simple bolusing solutions: Meal estimation (carbs): Consider simple bolusing strategies for meals such as entering 30 g (small meal), 60 g (medium meal), 90 g (large meal) of carbohydrate. Meal estimation (units): May alternatively consider using insulin units in similar way if user does not want to enter carbohydrates Fixed unit/custom: Define set amounts to deliver for routine meals
Exercise strategies	Exercise is individualized. Trial and error will be needed Prior to exercise: Set Exercise Activity 1–2 h before activity until end of exercise if decrease in glucose is expected Reduce mealtime bolus insulin dose by 25%–33% if within 2 h of exercise 52 Do not give carbohydrates immediately prior to exercise unless trending toward hypoglycaemia (so system does not respond with extra insulin) Consider delivering a 0.05 unit bolus of insulin prior to exercise to ‘lock out’ the Autobolus for 1 h During exercise: If trending toward hypoglycaemia, consider small carbohydrate amounts (3–20 g) to reduce the risk of system responding with extra insulin 52 Use a Temp Rate during exercise in addition to Exercise Activity (e.g., 50% basal rate) if more insulin reduction is needed If routine exercise schedule: Create Personal Profile to remove some of the user burden and time needed to prepare for exercise and adjust settings Example: A profile with 25% lower basal doses, 50% weaker CF and 25% weaker ICR
Paediatric considerations	In addition to settings considerations above: Prepubertal children will require less insulin (0.7–1.0 units/kg/day) compared to pubertal children (up to 1 and 2 units/kg/day) 53 Children may require different settings by time of day 37 For preschool aged children (ages 2–6 years) May need less basal insulin (e.g., 30%–50% of TDI) 9 Set higher basal rates 6pm–12 am and lower basal rates in early morning. 9 Can program a basal rate of 0.0 units/h when needed, to deliver insulin only if glucose levels are predicted to rise above target. Use extended boluses (up to 8 h with automation) for erratic eating patterns, grazing or when unsure how much of a meal the child will eat. Exercise Activity overnight can be useful for hypoglycaemia concern 66 For new onset AID 55 : Emphasize continuous use of CGM with system initiation to maximize algorithm performance Calculate total daily insulin: If DKA: base on insulin administered in the past 24 h. If no DKA: 0.75 units/kg/day if ketones >1 mmol/L or 0.5–0.6 units/kg/day if ketone‐negative. Keep mealtime simple: Use fixed carbohydrate assumptions (e.g., 25 g breakfast, 80 g lunch/dinner, 20 g snack) to derive initial ICR ratios. Avoid entering fictitious carbohydrate values. Ensure monitoring and early follow‐up: train in hypoglycaemia and hyperglycaemia correction. Schedule follow‐up visit 8–10 days post‐discharge to review pump data, optimize settings and reinforce training.
Type 2 diabetes considerations	Control‐IQ+ can be considered for wide range of people with type 2 diabetes not meeting therapeutic goals 6 People using a GLP‐1 RA can expect similar improvement in glycaemia to people not using a GLP‐1RA without the potential weight gain from insulin 67 If using a GLP‐1 RA at time of initiation, initial TDI should be reduced 10–20% when starting Control‐IQ+. 68 If coming from MDI, may not need to reduce pump TDI if HbA1c >8% (See Table 2). Formal carbohydrate counting training not required for success‐ Meal estimation and fixed bolus amounts produce similar glycaemia6, 45
Pregnancy considerations	Control‐IQ+ is not currently cleared for use in pregnancy. Considerations from clinical trial 63 : Update settings every 1–2 weeks Use Sleep Activity day and night during pregnancy (to lower system target to 112.5–120 mg/dL, 6.25–6.7 mmol/L) Adjust basal rates to be 20% higher than delivered basal insulin after 20 weeks gestation Intensify CF as needed (in study used <1620/TDI for mg/dL and 90/TDI for mmol/L) Bolus for meals 10–15 min pre‐meal in the first trimester, 20–30 min prior in second trimester and 30–45 min before a meal in the third trimester.

Abbreviations: AID, automated insulin delivery; CF, correction factor; CGM, continuous glucose monitoring; DKA, diabetic ketoacidosis; GLP‐1 RA, glucagon‐like peptide‐1 receptor agonist; ICR, insulin‐to‐carbohydrate ratio; MDI, multiple daily injection; TDI, total daily insulin.

4.1.1. Clinical example: Optimizing glycaemia without boluses

Description: David is a 36‐year‐old male with T1D for 26 years, who uses MDI and has an A1c above target at 8.5%. Even though David is prescribed 12 units of glargine daily and 8 units of aspart for each meal, he admits to being burned out by mealtime dosing and often misses meal doses. His HCP discusses using Control‐IQ+ for improvement in glycaemia, even if David is not intending to deliver mealtime boluses.

Therapy plan: David's HCP estimates that David is taking about 20 units a day and does not reduce this dose to calculate settings. Since David is not planning on bolusing, his HCP calculates 70% of his insulin to be delivered by basal: 14 units/day equates to a single basal rate of 0.6 units/h and a strong CF of 70 mg/dL (1400/TDI) or 3.9 mmol/L (78/TDI in mmol/L) to optimize the impact of the Autobolus. The HCP further chooses a relaxed carb ratio, knowing that David will rarely use this setting. These settings are intended to maximize glycaemia in someone who does not bolus but reduce the risk of hypoglycaemia (with the relaxed ICR) if he does.

Result: David starts Control‐IQ+ and achieves a TIR of 74%, with 3.0% time <70 mg/dL (3.9 mmol/L) and 0.9% <54 (3.0 mmol/L) mg/dL in the first 2 weeks, with an average of 0 boluses/day. He is receiving 90% of his insulin by basal and the remaining by Autobolus. Follow‐up occurs every 3–6 months to monitor for continued TIR >70%, with the HCP continuing to assess if David would like to start bolusing, which would require settings adjustments to more conservative automation.

4.2. Simple bolusing strategies

Although Control‐IQ+ is intended for use with adequate carbohydrate counting skills, multiple studies show that there are high error rates in carbohydrate counting by people with T1D, which leads to higher glycaemic variability.^{47, 48, 49, 50, 51} Moreover, carbohydrate counting is used minimally among people with T2D on insulin therapy. The RCT of Control‐IQ+ in adults with T2D reported that >50% of participants used simple bolus strategies, such as approximating carbohydrates into small/medium/large amounts (e.g., 30, 60, 90 g carbohydrate) or fixed units of insulin based on meal size (e.g., 5, 8, 10 units of insulin). ⁶ The Tandem Diabetes Care website has resources to support simple bolus strategies as described below:

4.2.1. Meal estimation bolus strategy: Carbohydrates

In this approach, the person employs basic carbohydrate awareness and education to be able to estimate the approximate carbohydrate amount in a qualitative measure. This leads to a person bolusing an amount of carbohydrates for a ‘small, medium or large’ meal.

Example: Sam has an appropriate understanding of what foods contain carbohydrates and is able to approximate carbohydrate load in his meal choices. His HCP suggests that Sam enters 30 carbs into the bolus calculator for a small meal, 60 carbs for a medium meal and 90 carbs for a large meal. The HCP can then adjust at subsequent follow‐up visits based on post‐meal glycaemic targets.

4.2.2. Meal estimation: Units

Dietitians or diabetes educators can evaluate carbohydrate contents of a person's regular breakfast, lunch and dinner. Based on estimated carbohydrate intake, recommendations can be made for fixed amounts of insulin prior to meals. Post‐prandial glycaemic profiles can be evaluated, and the fixed amount could be adjusted over time.

Example: Kat has some understanding of what foods contain carbohydrates. Her HCP suggests that Kat enter 2 units into the bolus calculator for a small meal, 4 units carbs for a medium meal and 6 units carbs for a large meal. The HCP can then adjust at subsequent follow‐up visits based on post‐meal glycaemic targets.

4.2.3. Fixed unit/custom strategy

For people who eat very similar meals each time of the day, custom food strategy could be helpful. Dietician or diabetes educators can evaluate regular breakfast, lunch and dinner meal composition and based on it decide on insulin dose.

Example: Jan eats a breakfast bowl of shredded wheat with milk 4–5 days of week (~60 g of carbs), a turkey sandwich for lunch with 30 g of carbs during work most work days and a protein with mashed potatoes 2–3 times per week for dinner. The HCP can recommend that Jan take 4 units for bowls of cereal, 2 units for turkey sandwich and 5 units for dinner. These dosages can be adjusted based on postmeal glycaemic responses. Additionally, the person can be taught to observe different meals and insulin requirements over time to add more custom food and insulin dosing.

5. SIMPLE EXERCISE STRATEGIES

Control‐IQ+ can also be used to mitigate the risks of hypoglycaemia during exercise and maximize euglycaemia. Consensus recommendations for physical activity and AID include setting a higher glucose target 1–2 h before glucose‐lowering exercise (or maintain a regular glucose target if a glucose rise is expected), decrease meal insulin doses if within 2 h prior to exercise, minimize insulin on board at the time of exercise start and small, frequent carbohydrate ingestion during exercise as needed. ⁵²

The Control‐IQ+ Exercise Activity and Temp Rate features are useful as well as behavioural strategies to optimize exercise glycaemia (Table 3). Exercise Activity is a temporary mode that allows the algorithm to target a higher range of 140–160 mg/dL (7.8–8.9 mmol/L) and can be set for a duration of 30 min to 8 h. As per the guidelines above, this should be set 1–2 h before activity to minimize hypoglycaemia. ⁵² Additionally, Temp Rate can be used to further impact insulin delivery during exercise. Temp rate increases or decreases the basal rate (starting place for automation) and can be set from 0% to 250% for a duration of 15 min to 72 h. Control‐IQ+ is the only system currently that allows Tamp Rates to be used with automation turned on.

It is important not to consume ‘pre‐emptive carbohydrates’ prior to exercise, unless trending toward hypoglycaemia, as the rise in sensor glucose can trigger Control‐IQ+ to deliver an Autobolus, even during use of Exercise Activity. In addition to not taking a pre‐exercise snack, another way the Autobolus can be mitigated is by delivering a small manual bolus dose of insulin (e.g., 0.05 units) immediately prior to exercise. Since Autoboluses only deliver insulin if a bolus has not been given in the last hour, this will temporarily disable the system's Autobolus feature for the next 60 min.

Finally, when exercise is planned in advance, or performed routinely at the same time of day, a secondary Personal Profile can be programmed for exercise. This is done by the user creating a new profile, duplicating their routine settings and then changing the settings related to the time of day for exercise. The advantage of a Personal Profile over a Temp Rate for structured exercise is that the user will not have to program any changes at the time of exercise.

5.1. Clinical example: Optimizing glycaemia with aerobic exercise

Description: Sasha is a 15‐year‐old girl with T1D for 6 years and uses Control‐IQ+. She plays high school soccer four times per week. Sasha eats a meal heavy in carbohydrates and fat 30–45 min prior to practice, and sets Exercise Activity 30–60 min prior to start. Most days, however, she experiences hyperglycaemia at the start of exercise (250–300 mg/dL, 13.9–16.7 mmol/L), followed by subsequent hypoglycaemia in the middle of practice. She feels frustrated and tired during practice, often having to sit out to treat hypoglycaemia shortly after she starts.

Therapy Plan: Sasha and her diabetes care team discuss changing the timing of her meal to 2 h before practice. It is further suggested that she eat a lower‐fat meal like a turkey sandwich and decrease her mealtime insulin bolus by 20%. This will reduce the amount of active insulin at practice start, reducing the likelihood of hypoglycaemia later in practice. Sasha is also encouraged to set Exercise Activity 1–2 h before soccer practice. Sasha can enter a 0.05 unit bolus immediately before practice to stop the Autobolus delivery for the next 60 min during her soccer practice.

Result: Within 2 weeks, Sasha notices fewer glucose spikes above 200 mg/dL (11.1 mmol/L) during soccer and she feels more energized. She also has less frequent episodes of hypoglycaemia after practice start, likely due to less active insulin circulating in her body. By adjusting her meal timing and composition and refining her bolus insulin strategy before soccer practice, Sasha experiences more satisfaction from her routine practice.

6. CHILDREN AND NEW ONSET CONSIDERATIONS

The International Society of Paediatric and Adolescent Diabetes (ISPAD) recommends AID systems like Control‐IQ+ for youth with diabetes in order to improve TIR by minimizing hypoglycaemia and hyperglycaemia, improving person‐reported outcomes and reducing burden of care, especially in the overnight period. ⁹ This is supported by RCTs of Control‐IQ+ in young children aged 2–5 years ¹⁷ and in school‐aged children. ¹⁵ Youth who start AID can follow the conventional Control‐IQ+ guidance above, though they may differ from adults in initial insulin pump settings. Children, especially young children, may require lower basal rates and a higher proportion of bolus insulin (e.g., 30–50% of insulin by basal). Further, prepubertal children will require less insulin (0.7–1.0 units/kg/day) compared to pubertal children (up to 1 and 2 units/kg/day). ⁵³ Diabetes in preschool‐aged children (ages 2–6 years) can be particularly difficult, with unpredictable activity levels and food intake challenging traditional management. Table 3 offers practical suggestions for this age.

6.1. Control‐IQ+ at new onset diabetes

ISPAD Clinical Practice guidelines and recent data support initiation of AID systems at the time of diagnosis, with evidence for improved TIR, minimized TBR, improved treatment satisfaction and quality of life, results that persist past 12 months.^{54, 55, 56} Practical recommendations for initiation of Control‐IQ+ at diagnosis include many of the same suggestions as above, with initial emphasis on consistency with the CGM and Control‐IQ+ automation use and frequent reassessment of settings as the honeymoon period progresses (Table 3). While the weight and TDI parameters in Control‐IQ+ are primarily used by the algorithm at initialization, HCPs may wish to periodically update these parameters every few months as the child's insulin needs change.

6.1.1. Clinical example: Starting Control‐IQ+ in a new onset T1D 13‐year‐old female ⁵⁷

Description: Geraldine is a 13‐year‐old girl who presents to the hospital in moderate‐to‐severe DKA, and standard DKA management led to rapid clinical recovery within 30 h. Geraldine's family is interested in starting Control‐IQ+ to normalize glycaemia as soon as possible.

Therapy plan: Geraldine's healthcare team started CGM on day 2 of the hospital admission and started Control‐IQ+ on day 3. Initial settings were based on 24‐h insulin needs (0.9 units/kg/day), with basal rates calculated as 50% of TDI. Her initial CF was calculated using 1500/TDI (83/TDI for mmol/L, based on HCP preference for stronger CF), with the ICR empirically estimated based on the assumed carbohydrate intake per meal and subsequently adjusted. The family received intensive training on how to use the device, deliver boluses, and how to make sure the CGM is always connected and Control‐IQ+ set to ‘on’. Geraldine navigates the insulin pump menu herself with support from her parents on bolusing. The HCP reviewed CGM data daily and made minor adjustments to settings as needed. The family achieved autonomy in managing both the device and glycaemic corrections within days.

Results: Geraldine was discharged on day 9 after diagnosis (6 days into using Control‐IQ+). At 2 weeks, Geraldine's TIR was 95%, with no CGM readings <54 mg/dL (3.0 mmol/L). The Glucose Management Indicator (GMI) was 6.3%, CGM use was 96.3%, and daily insulin requirements averaged 0.7 units/kg/day (48% basal, 52% bolus).

7. CONSIDERATIONS FOR TYPE 2 DIABETES

Recent guidelines endorse AID for people with T2D, ¹² which represents a shift in clinical care and will come with new implementation challenges. Many people with T2D are treated in a primary care setting, where implementation of AID is less common. Primary care providers have reported knowledge barriers to feeling comfortable with AID, including knowing who is an appropriate candidate, being able to answer questions about the device, and reviewing and interpreting data.^{26, 58} Clinical inertia also leads to suboptimal insulin initiation and titration for people with T2D. ⁵⁹ It has been suggested that initiating AID early in people with T2D requiring insulin may improve A1c faster and reduce the burden on patients and providers with insulin titration. ⁶⁰

An RCT of Control‐IQ+ in adults with T2D demonstrated participants experiencing glycaemic improvements with Control‐IQ+ irrespective of age, gender, duration of diabetes, literacy, numeracy and baseline glycaemia, though those with the highest HbA1c levels at baseline had the largest improvement after 3 months. ⁶ The study participants reflected the real world T2D population, including 39% from minority race or ethnicity, >50% using a glucagon‐like peptide‐1 (GLP‐1) receptor agonist and/or a sodium‐glucose cotransporter 2 inhibitor (SGLT2i) medication and 61% having a baseline HbA1c ≥8%. Importantly, even participants on glucose‐lowering medications had additional glycaemic benefit with Control‐IQ+, indicating that people not meeting therapeutic goals who require intensive insulin can benefit from Control‐IQ+. The study provides insight into simple strategies for use of the system. During the trial, the majority of participants used simple bolusing strategies, such as meal estimation or fixed doses, with similar glycaemic outcomes regardless of bolusing strategy, as outlined in Table 3. ⁴⁵ Resources such as Figure 3 are intended to bridge knowledge gaps for practical implementation.

7.1. Clinical example: Control‐IQ+ in a person with T2D

Description: Sandra is a 41‐year‐old female living with T2D for 8 years using Tirzepatide 15 mg weekly, and basal bolus insulin therapy: 96 units Tresiba per day and 40 units of Humalog for meals three times per day. Sandra reports missing insulin injections ‘a few times per week’, and her HbA1c is 7.8%. Together with her healthcare team, Sandra decides to start Control‐IQ+ to further improve glycaemia.

Therapy plan: Sandra's reported TDI is 216 units/day, but Sandra admits that she is likely not administering this full amount and estimates her actual daily doses are closer to 180 units/day. Because she is on a GLP1 receptor agonist, her HCP reduces the dose by 20% to 144 units/day and calculates a basal rate of 3 units/h (144/2 = 72 units/24 h = 3 units/h), a CF of 12 (1700/144 = 11.8) and ICR of 1 unit for every 3.1 carbs (450/144 = 3.1). The dietician reviews Sandra's eating habits, which are high in carbohydrates but routine from day to day, and decides that small/medium/large meal estimation would support Sandra, who has basic carbohydrate awareness. Therefore Sandra is taught to input 70, 90, 110 carbs into the bolus calculator.

Result: During Sandra's first 3 days on Control‐IQ+, TIR was 79%, but Sandra reported having hypoglycaemia after ‘a few’ of her meals, which made her reluctant to deliver a meal bolus for those meals. The HCP verified 2 episodes of glucose <70 mg/dL (3.9 mmol/L) after boluses, so adjusted her ICR from 1:3.1 to 1:5 (less aggressive). One week later, Sandra's TIR was 89%, with no episodes of hypoglycaemia. She delivered 3 boluses per day, and Control‐IQ+ was additionally delivering 4 Autoboluses per day. At 2 weeks, Sandra's mean glucose was 154 mg/dL (8.6 mmol/L), which equates to a glucose management index of 7%, representing a significant improvement over her pre‐Control‐IQ+ HbA1c of 7.8%.

8. CONSIDERATIONS IN PREGNANCY

New data are emerging on the use of AID systems in pregnancy^{61, 62, 63} A recent RCT randomized 91 pregnant women with T1D to Control‐IQ or standard care (MDI or insulin pump with CGM) and measured time in pregnancy target (63–140 mg/dL, 3.5–7.8 mmol/L) from 16 to 34 weeks of gestation. Control‐IQ+ demonstrated a mean adjusted treatment group difference of +12.5% compared to standard care group (65.4% versus 50.3% time in pregnancy target range), outcomes in line with other AID system RCT results. ⁶⁴

Control‐IQ+ is not currently cleared for diabetes in pregnancy. Practical recommendations for using Control‐IQ in pregnancy have been published in the literature,^{43, 63} and are included in Table 3.

9. CONCLUSION

Control‐IQ+ is an adaptable AID system that can be used with simple strategies to meet challenges that people with diabetes face. HCPs can feel empowered to use Control‐IQ+ in a wide variety of people with diabetes requiring insulin and can use simple guidelines and strategies to improve glycaemia and quality of life. The unique contribution of CF and the Autobolus may make Control‐IQ+ an option even for those who do not bolus regularly or present other glycaemic challenges. Future research will continue to improve upon current strategies, and new developments in automation will find novel approaches to meeting needs.

CONFLICT OF INTEREST STATEMENT

VNS's institution has received research grants from Dexcom, Alexion, Eli Lilly, Enable Bioscience, Zucara Therapeutics, Cystic Fibrosis Foundation, DEKA Research, Medtronic, the NIH and Breakthrough T1D. VNS has received sensors at no cost from Dexcom and Abbott for various investigator‐initiated research projects. Novo Nordisk provided drugs at no cost for the ADJUST T1D clinical trial. VNS has received honoraria from Dexcom, Abbott Diabetes, Ascensia Diabetes Care, Tandem Diabetes Care, Medtronic, Insulet, Sequel Med Tech, NovoNordisk, Eli Lilly, Sanofi, Biomea Fusion and T1D scout, outside of this submitted work. PC has received personal fees from Tandem, Insulet, Medtronic, Abbott, Dexcom, Roche, Vertex and CML Ltd. IH receives speaker and consulting fees from Tandem Diabetes Care, Dexcom, Abbott Diabetes Care, Insulet, Medtronic, Ypsomed, Sanofi & Novo Nordisk. IR receives personal fees from Sanofi, Theras, Menarini, MOVI and grants to her Organization from Sanofi and Medtronic. DPZ has received honoraria for speaking engagements from Ascensia Diabetes Care, Insulet Corporation, Dexcom Inc. and Medtronic. DPZ is on the advisory board for the Diabetes Research Hub and Enhance‐d. DPZ has received research support from The Leona M. and Harry B. Helmsley Charitable Trust and the Stanford Diabetes Research Center. LHM is an employee and shareholder of Tandem Diabetes Care.

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no datasets were generated or analysed during the current study.