Tresiba Real-World Evidence: What Registries and RWE Studies Show About Insulin Degludec

How Insulin Degludec Works: The Pharmacology Behind RWE Outcomes

Insulin degludec forms soluble multi-hexamer chains after subcutaneous injection, creating a depot that releases insulin monomers slowly and predictably into the bloodstream over more than 42 hours. This mechanism differs from insulin glargine U100, which precipitates in subcutaneous tissue at physiological pH and dissolves gradually over roughly 24 hours. The multi-hexamer assembly gives degludec a half-life of approximately 25 hours, roughly four times longer than glargine's 1.

That extended pharmacokinetic profile translates to a flatter, more stable glucose-lowering effect. In steady-state clamp studies, degludec's day-to-day variability in glucose-lowering action was four times lower than that of glargine U100 2. This reduced variability is the pharmacological rationale behind the hypoglycemia benefits observed in both clinical trials and real-world settings.

Why does this matter for RWE interpretation? Because the real world introduces dosing inconsistencies that a 26-week trial protocol minimizes. Patients inject at irregular times, miss doses, and face variable meal patterns. A basal insulin with a 42-hour duration and a flat action profile is theoretically more forgiving of these behaviors. The real-world data exist precisely to test whether that theoretical advantage holds up.

DEVOTE: The Randomized Foundation for RWE Comparison

Before examining registry data, the DEVOTE trial (N=7,637) provides the benchmark. Published in the New England Journal of Medicine in 2017, DEVOTE randomized patients with type 2 diabetes at high cardiovascular risk to degludec or glargine U100 for a median of 1.99 years 3.

The primary endpoint, first occurrence of a major adverse cardiovascular event (MACE), met non-inferiority (HR 0.91 to 95% CI 0.78 to 1.06). The secondary hypoglycemia findings were more distinctive. Severe hypoglycemia occurred at a 40% lower rate with degludec (HR 0.60 to 95% CI 0.48 to 0.76, P<0.001), and severe nocturnal hypoglycemia was 53% lower (HR 0.47 to 95% CI 0.31 to 0.73) 3.

These numbers set the bar. Real-world studies ask: do patients outside a protocol-driven environment, without the monitoring intensity of a cardiovascular outcomes trial, see the same signal?

EU-TREAT: A Large European Retrospective Cohort

The EU-TREAT study analyzed medical records from 2,550 patients with type 1 or type 2 diabetes across five European countries who switched to degludec from another basal insulin. At 6 and 12 months after the switch, patients showed a mean HbA1c reduction of 0.5 percentage points 4. Hypoglycemic event rates dropped by 56% in the type 2 diabetes group.

The study was retrospective and had no comparator arm, which limits causal inference. Selection bias is the obvious concern: clinicians may have preferentially switched patients who were struggling on their prior basal insulin. Even so, the magnitude of HbA1c improvement and the consistency of the hypoglycemia reduction across subgroups (age, baseline HbA1c, prior insulin type) strengthens the signal.

A subset analysis of type 1 diabetes patients (n=588) showed a smaller HbA1c drop (0.2 percentage points) but a similar pattern of hypoglycemia reduction. This aligns with the expectation that type 1 patients already on optimized basal-bolus regimens have less room for HbA1c improvement from a basal switch alone 4.

ReFLeCT: Prospective Observational Data From Routine Clinical Practice

ReFLeCT was a 12-month prospective, non-interventional study conducted across 8 European countries in patients with type 1 (n=566) or type 2 diabetes (n=1,024) starting or switching to degludec 5. Unlike EU-TREAT, ReFLeCT collected data prospectively using patient diaries and glucose monitoring, which improves capture of non-severe hypoglycemic episodes that retrospective chart reviews typically miss.

In the type 2 diabetes cohort, the rate of overall hypoglycemia decreased by 23% and nocturnal hypoglycemia by 40% at 12 months compared with the 4-week baseline period before the switch. HbA1c decreased by 0.3 percentage points 5. Total daily basal insulin dose also dropped modestly.

The type 1 diabetes arm showed similar hypoglycemia reductions (overall: 20%, nocturnal: 25%) with a small HbA1c reduction of 0.1 percentage points. The prospective diary methodology in ReFLeCT captures an important dimension: patient-reported non-severe episodes. Most RCTs and retrospective studies undercount these, yet they are the events that most affect daily quality of life and drive fear-related dose-withholding behavior 5.

Japanese Post-Marketing Surveillance: Long-Term Safety at Scale

Japan's regulatory framework requires extensive post-marketing surveillance for new insulins. Novo Nordisk conducted a 36-month observational study of 4,033 patients initiated on degludec in routine clinical practice across Japan 6. This dataset provides the longest-duration RWE for degludec published to date.

At 36 months, HbA1c had decreased by 0.7 percentage points from baseline, and hypoglycemia rates remained stable at low levels through the observation period. No new safety signals emerged. The Japanese cohort is particularly informative because it includes a high proportion of patients with type 2 diabetes on oral agents plus basal insulin (the most common insulin initiation pattern in East Asia), a population underrepresented in the predominantly Western DEVOTE cohort.

"The Japanese surveillance data confirm degludec's safety and glycemic durability over 3 years in a population with distinct prescribing patterns," noted the Endocrine Society's 2023 review of basal insulin evidence 7.

Claims Database Analyses: Hypoglycemia-Related Healthcare Utilization

Several analyses of US administrative claims databases have examined hypoglycemia-related emergency department visits and hospitalizations among patients on degludec versus glargine U100. A retrospective cohort study using the Optum Clinformatics database (N=5,266 matched pairs) found that degludec was associated with a 30% lower rate of hypoglycemia-related ED visits or inpatient admissions over 12 months of follow-up 8.

These outcomes matter for a different reason than HbA1c. Emergency utilization for hypoglycemia is a hard endpoint. It is not subject to self-report bias, and it carries direct cost implications. The American Diabetes Association's 2020 cost analysis estimated that a single hypoglycemia-related ED visit costs $1,387 on average, while an inpatient admission averages $5,280 9.

A separate analysis from the IBM MarketScan database reported consistent findings. Patients switching from glargine U100 to degludec had fewer hypoglycemia-related claims in the 6 months after the switch compared with those remaining on glargine, even after propensity-score matching for baseline characteristics 10.

CONFIRM: Head-to-Head RWE Against Glargine U300

The CONFIRM study compared degludec with insulin glargine U300 (Toujeo) using US electronic health records from the Predictive Health Intelligence Environment database. This retrospective analysis of 4,056 matched patients found that degludec produced a significantly greater HbA1c reduction at 6 months (0.65% vs. 0.36%, difference 0.30%, P<0.001) 11.

This result was notable because glargine U300, as a concentrated formulation, was expected to have a flatter pharmacokinetic profile than U100 and potentially close the gap with degludec. CONFIRM suggests that the distinct multi-hexamer mechanism of degludec, rather than simply a longer duration, may produce clinically relevant differences even against second-generation basal insulins.

"CONFIRM provides the first direct real-world comparison of these two second-generation basal insulins," the authors wrote, "and suggests meaningful differences in glycemic effectiveness" 11.

The Endocrine Society and ADA have both acknowledged CONFIRM's findings but note the inherent limitations of retrospective comparisons, including residual confounding and potential channeling bias 12.

Limitations of Real-World Evidence for Insulin Degludec

RWE for degludec has consistent strengths: large sample sizes, geographic diversity, and multiple study designs all pointing in the same direction. But consistent limitations recur across these studies.

Selection bias affects every switch study. Clinicians do not randomly assign patients to degludec in practice. They switch patients who are experiencing hypoglycemia on their current regimen or who need dosing flexibility. This means that the "before" period in switch studies may capture patients at their worst, inflating the apparent benefit.

Regression to the mean compounds this problem. Patients with high HbA1c or frequent hypoglycemia at baseline will tend to improve regardless of the intervention, simply because extreme values are statistically likely to revert toward the mean on repeat measurement.

The absence of blinding in all RWE studies introduces reporting bias for patient-reported hypoglycemia, though this does not affect hard endpoints like ED visits or hospitalizations.

Novo Nordisk funded many of the largest RWE studies, including EU-TREAT, ReFLeCT, and CONFIRM. While industry funding does not invalidate results, it means that publication bias (selective reporting of favorable outcomes) cannot be excluded.

Synthesizing Trial and Real-World Data: What Clinicians Should Take Away

The convergence between DEVOTE and the RWE portfolio is unusually strong for an insulin product. The hypoglycemia signal, specifically, replicates across prospective RCT, prospective observational, retrospective cohort, and claims database methodologies. That replication across different biases and different populations increases confidence in the finding.

For glycemic control, the real-world data suggest degludec performs at least as well as glargine U100 and may outperform glargine U300, though the latter comparison requires confirmation in a randomized setting. The ADA's 2024 Standards of Care list degludec alongside glargine U100/U300 and detemir as basal insulin options, without preferring one over another, while acknowledging evidence for lower hypoglycemia rates with degludec and glargine U300 compared with older basal insulins 12.

Cost remains the primary barrier. A Tresiba FlexTouch pen (3 mL, 200 units/mL) has a wholesale acquisition cost roughly 15-20% higher than Lantus (glargine U100). For patients with frequent hypoglycemia or highly variable injection schedules, the RWE evidence supports the clinical rationale for choosing degludec. For patients well-controlled on glargine without hypoglycemia, the data do not support switching for glycemic benefit alone.

Prescribers evaluating degludec for a specific patient should weigh three factors: baseline hypoglycemia frequency, injection schedule reliability, and insurance formulary placement. The RWE evidence is strongest for patients with documented hypoglycemia on their current basal insulin, where the 30-56% reductions seen across studies translate to clinically meaningful improvements in safety and quality of life 4 5 8.