Omnipod 5 (SmartAdjust AID)

Omnipod 5 is a tubeless automated insulin delivery (AID) system: the on-body Pod contains insulin and runs SmartAdjust. The US product supports Dexcom G6/G7 systems and FreeStyle Libre 2 Plus/3 Plus with the appropriate Pod and control device. These are specific combinations, not interchangeable sensors for every Pod. Local options differ: the Australian manufacturer page lists Dexcom G6, G7 and Libre 2 Plus.⁹⁸¹⁴

Components

The system combines a tubeless Pod, a compatible CGM, and the Omnipod 5 app on a supported control device. SmartAdjust runs inside the Pod: it predicts glucose up to 60 minutes ahead and increases, decreases or pauses insulin every 5 minutes. In the US, Dexcom combinations support compatible iPhone/Android models; Libre 3 Plus supports the Controller or compatible Android devices, while Libre 2 Plus requires the Controller. Libre 3 Plus is not currently supported on the iPhone app.²⁰²¹

The enhanced algorithm — now shipped

The FDA cleared the enhanced Omnipod 5 algorithm on 3 December 2025,¹⁸ and Insulet began the US over-the-air rollout on 3 June 2026, delivered through the Controller or smartphone app. The enhanced algorithm requires both the app update and compatible Pods; an app update alone does not upgrade older Pods.⁸³⁴ Two changes matter most: an optional 100 mg/dL (5.6 mmol/L) glucose target, lowering the selectable range to six settings from 100 to 150 mg/dL (5.6–8.3 mmol/L); and fewer interruptions — a new advisory alarm lets the system keep auto-correcting every 5 minutes through prolonged high-glucose events instead of forcing a switch to Manual Mode.³⁸ The same rollout made FreeStyle Libre 3 Plus live as a sensor option in the US, with LibreLinkUp caregiver following.⁴

The FDA summary describes in-silico validation (computer simulation) for the added 100 mg/dL target, rather than a prospective clinical trial of that setting. The older pivotal study results below should not be presented as measured results of the enhanced algorithm.¹⁸

Insulet reports an observational analysis of 403 people with type 1 diabetes aged 2+ who used both a 150 and a 110 mg/dL target for consecutive periods of 14–90 days each. Median TIR increased 11.8 percentage points and time below 70 mg/dL increased 0.23 points. This does not isolate a causal target effect and did not test the new 100 mg/dL setting.⁴

Recalls you should know about (2026)

Omnipod Pods — the same Pods this system runs on — were the subject of two separate Class I FDA recalls in 2026, the FDA's most serious category: a tear in the Pod's internal tubing and a tear in the cannula just above the skin. FDA communications confirmed Class I status on 29 April and 2 July 2026, respectively.²¹ They are different defects, and both cause insulin under-delivery. The part that matters most for day-to-day safety: in both cases the under-delivery may happen without triggering any alarm.²¹ As of the FDA notices, Insulet had reported 29 serious injuries for the tubing defect and 24 for the cannula defect, with no deaths.²¹

Both recalls are lot-specific. FDA instructions say to stop using affected Pods immediately, replace an active affected Pod with an unaffected one, and contact the healthcare team about alternative insulin delivery if supplies run out. Check lot numbers with Insulet and request replacement Pods; do not rely solely on alarms. Follow the system instructions and your clinical sick-day plan for unexplained high glucose.²¹ Full detail is on the Omnipod 5 pump record.

Does the sensor change the results?

A real-world cohort included 353 adults, with paired CGM data for 268 and paired HbA1c for 193. Dexcom G6 versus Libre 2 Plus use was not associated with the likelihood of improvement, but this was not a randomized sensor-equivalence trial and does not establish identical results for all compatible sensors.¹¹

Glycemic outcomes (achieved)

The 3-month, single-arm type 1 pivotal trial analyzed 112 children aged 6–13.9 and 128 participants aged 14–70. The trial enrolled 129 in the older group; one participant did not enter the automated-delivery phase and was not included in that analysis. Mean TIR (70–180 mg/dL / 3.9–10.0 mmol/L) during AID was 68.0% and 73.9%, respectively; HbA1c fell by 0.71 and 0.38 percentage points. In the older group, median time below 70 mg/dL fell from 2.00% to 1.09%. Three severe hypoglycemia events and one DKA event occurred during the pivotal study; the DKA followed an infusion-site failure.¹⁰

A July 2026 meta-analysis of four randomized trials of tubeless hybrid closed-loop systems (514 participants) estimated a TIR difference of +15.31 percentage points (95% CI 7.67–22.95) and an HbA1c difference of −0.46 points (95% CI −0.79 to −0.14) versus standard care. Between-study heterogeneity was high (I² 95% for TIR and 91% for HbA1c). These are pooled differences from comparator treatment, not achieved glucose levels or evidence that Omnipod 5 is superior to another AID brand.⁶

An optional extension included 224 participants, with a median 22.3 months of AID use. Mean TIR during the extension was 65.9% in children and 72.9% in adolescents/adults; median Automated Mode use was 96.1% and 96.3%. HbA1c follow-up was available for up to 15 months and CGM follow-up for up to two years. Seven severe hypoglycemia events and one DKA event occurred during the extension.¹²

Hypoglycemia, including impaired awareness

A single-centre observational study published online in July 2026 compared outcomes before and 12 months after Omnipod 5 initiation in people with impaired hypoglycemia awareness. Its abstract reports 68 individuals overall and baseline comparisons involving 32 with impaired awareness and 32 with normal awareness. In the impaired-awareness group, time below 70 mg/dL (3.9 mmol/L) fell from 2.8% to 1.7%, and weekly sensor-detected episodes below that threshold fell from 5.4 to 4.0. Residual hypoglycemia remained higher than in the normal-awareness group. This observational finding is not proof that the system eliminates severe hypoglycemia or restores awareness.⁵

Automation level

This is a hybrid closed loop: SmartAdjust adjusts insulin automatically, while users enter carbohydrate amounts and initiate meal boluses. Omnipod 6 STRIVE results also include user boluses; Insulet’s fully closed-loop development is a separate programme.²⁰²²

Exercise

The Activity feature uses a 150 mg/dL (8.3 mmol/L) target for a chosen duration up to 24 hours. In a 38-participant randomized crossover trial, enabling it 30 or 60 minutes before a 70-minute treadmill session reduced insulin delivery and glucose decline. Hypoglycemia occurred in 42% during standard Automated Mode, 29% with 60-minute advance activation and 24% with 30-minute activation; neither comparison was statistically significant. The authors recommend considering additional strategies such as earlier activation and/or carbohydrate intake.²³¹³

Form factor

Omnipod 5’s defining practical advantage is its tubeless on-body Pod: no tubing to catch on clothing and no separate insulin pump to clip or disconnect. The Pod is rated IP28 for immersion to 7.6 metres (25 feet) for up to 60 minutes. The Controller is not waterproof; the chosen CGM has its own water limits.¹⁴ This supports showering and swimming within the component instructions, rather than a claim that the entire system is waterproof.

Ages, indications and access

In the US, Omnipod 5 is indicated for type 1 diabetes from age 2 and for type 2 diabetes from age 18. SmartAdjust should not be used by people needing less than 5 units of insulin daily. US supply uses the pharmacy channel; individual coverage still requires checking. The system is also marketed in the UK, Australia and selected European countries. Local labeling, sensor availability and funding are separate questions; a US phone or sensor option should not be assumed available elsewhere.^820161415

The ADA Standards of Care in Diabetes—2026 recommend AID as the preferred insulin-delivery method for type 1 diabetes over multiple daily injections, non-automated pumps and sensor-augmented pumps (7.25a, grade A). Recommendation 7.8 supports considering early device initiation, including at diagnosis; 7.8a says C-peptide, autoantibody status or a minimum insulin-treatment duration should not be prerequisites for pump/AID initiation. These are guideline recommendations, not guaranteed payer coverage or a preference for Omnipod over another brand.⁷

What closed-loop users say they want (cross-system survey, not Omnipod-specific)

The SELECT HCL study — "Priorities, Expectations, Perceptions of Time in Tight Range and Support Needs Relating to Closed-Loop Therapy Among Adults With Type 1 Diabetes" — surveyed "166 participants" with type 1 diabetes who were using, awaiting initiation of, or considering closed-loop therapy of any brand (median age 47 years; HbA1c 52 mmol/mol [6.9%]): "Sensor accuracy (39.3%), glycaemic performance (37.6%) and patch pump without tubing (27.5%) were the most commonly prioritised factors when selecting sensors, algorithms and insulin pumps, respectively."²⁴ On day-to-day ambition, the survey also probed "Perceptions of Time in Tight Range": most respondents saw time-in-tight-range at or above 50% as beneficial for health (63.8%) while far fewer saw it as achievable (36.1%) — a gap between what users hope a loop could do and what they believe it will do for them.²⁴ And on getting started, "therapy improved QoL (78.6%), reduced hyperglycaemia (75.0%) and hypoglycaemia (72.3%), and met expectations (81.1%)," with "adequate support during closed-loop initiation was associated with therapy meeting participants' expectations (OR 16.6, 95% CI 5.0-52.4, p < 0.001)" — a cross-sectional association, not proof that support causes satisfaction.²⁴