CoChat
Sign Up
Cited by 33Diabetes Management and ResearchDiabetes Management and EducationPatient Safety and Medication Errors

Smart Pen Exposes Missed Basal Insulin Injections and Reveals the Impact on Glycemic Control in Adults With Type 1 Diabetes

Open access
Ekberg, Neda Rajamand, Hartvig, Niels Væver, Kaas, Anne, Møller, Jonas Bech, Adolfsson, Peter·2024·Journal of Diabetes Science and Technology
DOI 10.1177/19322968221104142PMC PMC10899845
Open this paper in CoChat — free account

Abstract

Background: Adherence to basal insulin injections and the effects of missed basal insulin injections in adults with type 1 diabetes (T1D) were investigated using data from continuous glucose monitoring (CGM) and smart insulin pen devices in a real-world study. Methods: This was a post hoc analysis of a prospective, real-world study conducted in Sweden. Adults with T1D who were using CGM received a smart insulin pen device (NovoPen 6) for insulin injections. Missed basal insulin doses (≥40 hours between doses) were evaluated over 14-day periods, and the probability of missing basal insulin doses was estimated. Associations between missed basal insulin doses and glycemic outcomes were also explored. Results: Thirty-two patients with 4410 acceptable CGM days (315 14-day periods) were included. The number of missed basal insulin doses ranged from 0 to 4 over 315 14-day periods. The estimated probability of missing at least one basal insulin dose over any given 14-day period was 22% (95% confidence interval: 10%-40%). Missed basal insulin doses were significantly associated with higher mean glycemic levels, higher glucose management indicator, and lower time in range (70-180 mg/dL [3.9-10.0 mmol/L]). Similar results were observed when adjusted for missed bolus insulin doses; age and sex had no statistically significant effect on any glycemic parameter. Conclusions: This is the first study, based on accurate real-world injection data, to demonstrate the challenge of adherence to basal insulin injections in patients with T1D, and document that just one missed basal injection per week can result in clinically significant changes in glycemic control.

AI snapshot

Written from the paper’s full text.

TL;DR

In a post hoc analysis of a Swedish real-world study, smart insulin pen data showed that adults with type 1 diabetes missed basal insulin doses and that these missed doses were associated with worse glycemic control. The estimated probability of missing at least one basal insulin dose in any 14-day period was 22% (95% CI: 10%-40%).

Key findings

  • Among 32 adults with T1D and 315 14-day periods, missed basal insulin doses ranged from 0 to 4; the estimated probability of missing at least one basal dose over a 14-day period was 22% (95% CI: 10%-40%).
  • Each missed basal insulin dose was associated with lower TIR (−2.63%, 95% CI: −4.41 to −0.71; P = .005), higher mean glucose (+0.44 mmol/L, 95% CI: 0.19-0.69; P < .001), higher GMI (+0.19%, 95% CI: 0.08-0.30; P < .001), and higher TAR level 2 (+2.91%, 95% CI: 0.99-4.73; P = .002).
  • Two missed basal insulin doses corresponded to a TIR change of −5.3%, described as clinically relevant; missed basal doses were not associated with higher %CV (P = .855).
  • Patients with high %CV had a higher estimated probability of missing at least one basal dose than those with low %CV (46.6% vs 7.3%; P = .007), while age and sex had no statistically significant effect on glycemic parameters.

Methods

  • Post hoc analysis of a one-arm, prospective, real-world study conducted at 12 diabetes clinics in Sweden from May 2017 to April 2019; adults with T1D using CGM received a smart insulin pen (NovoPen 6) for basal insulin degludec, with or without a smart pen for bolus insulin.
  • Included patients had at least one 14-day period with at least two basal injections and ≥70% CGM coverage; missed basal insulin dose was defined as ≥40 hours between basal injections.
  • Probabilities of missed basal doses were estimated with a cumulative linked mixed model with patients as random effects; associations with glycemic outcomes used linear mixed models, unadjusted and adjusted for missed bolus doses, age, and sex.

Limitations

  • The paper states limitations including a small study population and limited background information on included patients.
  • The paper states that data were grouped into 14-day intervals for simplicity, making the exposure-outcome model coarse and preventing inclusion of temporal covariates such as time since last missed basal dose; HCP visits could fall before, during, or after a 14-day period, so their detailed effect was not easily studied.
  • Evident from the text: this was a post hoc, one-arm observational analysis without a control group, so the reported associations are not causal; the adjusted model excluded 4 patients without a connected bolus pen, reducing N to 28.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1177/19322968221104142

Cites

25 of 25 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

3 of 8 cited sources checked · 0 supported · 1 flagged · accuracy not yet measured

  • Interpretationcitation

    This was the first study to investigate the effects of missed basal insulin injections using accurate injection data from a smart insulin pen.

    “The present study was the first to investigate the effects of missed basal insulin injections using accurate injection data from a smart insulin pen; previous studies using data from smart insulin pens have primarily focused on missed bolus insulin doses.4,8”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [4] Morris AD Boyle DI McMahon AD, et al. Adherence to insulin treatment, glycaemic control, and ketoacidosis in insulin-dependent diabetes mellitus. The DARTS/MEMO Collaboration. Diabetes Audit and Research in Tayside Scotland. Medicines Monitoring Unit. Lancet. 1997; 350: 1505- 1510. 9388398 source
    • cites [8] Peyrot M Rubin RR Kruger DF Travis LB. Correlates of insulin injection omission. Diabetes Care. 2010; 33(2): 240- 245. 20103556 PMC2809256 source

      The citation supports the sentence’s characterization of previous studies as using smart-pen data and focusing on missed bolus doses, not its claim about the citing authors’ present study. This paper instead used an Internet survey of self-reported injection omission (Research Design and Methods) and explicitly says it had no objective measure of insulin use (Conclusions, Study strengths and limitations). It therefore cannot serve as an example of a prior study using smart-pen injection data. (LLM, conf=0.88)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured
    • Rests on an assumption: Smart insulin pen (NovoPen 6) injection data accurately reflect the actual timing and dose of basal insulin injections.
  • Premisecitation

    Previous studies using data from smart insulin pens have primarily focused on missed bolus insulin doses.

    “The present study was the first to investigate the effects of missed basal insulin injections using accurate injection data from a smart insulin pen; previous studies using data from smart insulin pens have primarily focused on missed bolus insulin doses.4,8”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [4] Morris AD Boyle DI McMahon AD, et al. Adherence to insulin treatment, glycaemic control, and ketoacidosis in insulin-dependent diabetes mellitus. The DARTS/MEMO Collaboration. Diabetes Audit and Research in Tayside Scotland. Medicines Monitoring Unit. Lancet. 1997; 350: 1505- 1510. 9388398 source
    • cites [8] Peyrot M Rubin RR Kruger DF Travis LB. Correlates of insulin injection omission. Diabetes Care. 2010; 33(2): 240- 245. 20103556 PMC2809256 source

      The Research Design and Methods section describes an Internet survey, and Study strengths and limitations says the study had no objective measure of insulin use. Searches for smart insulin pens, missed bolus doses, and basal injections found no passage addressing what earlier smart-pen studies primarily examined. (LLM, conf=0.78)

      Experimental — accuracy not yet measured
  • Interpretationreasoning

    Two missed basal insulin doses during a two-week period correspond to a change in TIR of −5.3%.

    “Thus, two missed basal insulin doses correspond to a change in TIR of −5.3%; a difference of 5% is considered clinically relevant.”
    • Also stated in: § Discussion · plain
      “One missed basal insulin dose per week (or two missed basal insulin doses during a two-week period) was sufficient to correspond to a clinically relevant (a difference of ≥5%) change in TIR (−5.3%).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [23] Battelino T Danne T Bergenstal RM, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range. Diabetes Care. 2019; 42(8): 1593- 1603. 31177185 PMC6973648 source

      The Relationship Between TIR and A1C section (p. 14) discusses a 10% change in TIR, but does not link TIR to missed basal insulin doses. Searches for two missed doses, a −5.3% change over two weeks, and a clinically relevant 5% difference found no passage addressing those points. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
  • Findingown data

    Age and sex did not have a statistically significant effect on any glycemic parameter in the adjusted model.

    “age and sex had no statistically significant effect on any glycemic parameter.”
    • Also stated in: § Associations Between Missed Basal Insulin Doses and Glycemic Parameters · plain
      “Age and sex did not have a statistically significant effect on any of the glycemic parameters (Supplemental Table S2).”
    Quote in the paperNot a substantial claim
  • Findingreasoning

    Missed bolus insulin doses were significantly associated with TIR (P = .008).

    “In contrast to this, %CV was significantly associated with missed bolus doses (P < .001) and, as such, missed bolus doses were also associated with TIR (P = .008) and TAR L2 (P = .008).”
    Quote in the paperNot a substantial claim
  • The quote only partly states this claim. 4
  • May combine two claims. 7
  • Classified as probably not a substantial claim. 6

  • The quote only partly states this claim.

    Findingown data

    Each missed basal insulin dose was associated with an increase in mean glucose of 0.44 mmol/L (95% CI 0.19 to 0.69; P < .001) in the adjusted model.

    “Each missed basal dose corresponds to a reduction of TIR of −2.6% and an increase in mean glucose of 0.44 mmol/L.”
    • Also stated in: § Associations Between Missed Basal Insulin Doses and Glycemic Parameters · plain
      “Each missed basal insulin dose was associated with an increase in the mean glycemic level (P < .001), a reduced TIR (P = .005), an increased TAR L2 (P = .002), and higher GMI (P < .001) (Table 3).”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Findingown data

    Each missed basal insulin dose was associated with a reduction in time in range (TIR) of −2.63% (95% CI −4.41 to −0.71; P = .005) in the adjusted model.

    “Each missed basal insulin dose was associated with an increase in the mean glycemic level (P < .001), a reduced TIR (P = .005), an increased TAR L2 (P = .002), and higher GMI (P < .001) (Table 3).”
    • Also stated in: § Associations Between Missed Basal Insulin Doses and Glycemic Parameters · plain
      “Each missed basal dose corresponds to a reduction of TIR of −2.6% and an increase in mean glucose of 0.44 mmol/L.”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Findingown data

    Each missed basal insulin dose was associated with an increase in time above range level 2 (TAR L2) of 2.91% (95% CI 0.99 to 4.73; P = .002) in the adjusted model.

    “Each missed basal insulin dose was associated with an increase in the mean glycemic level (P < .001), a reduced TIR (P = .005), an increased TAR L2 (P = .002), and higher GMI (P < .001) (Table 3).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    No statistically significant differences were observed for the probability of missing basal insulin injections between patients with high versus low TIR, or between younger versus older patients.

    “No statistically significant differences were observed for the probability of missing basal insulin injections in patients with high versus low TIR, or for younger versus older patients.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Findingown data

    Each missed basal insulin dose was associated with a higher glucose management indicator (GMI) of 0.19% (95% CI 0.08 to 0.30; P < .001) in the adjusted model.

    “Each missed basal insulin dose was associated with an increase in the mean glycemic level (P < .001), a reduced TIR (P = .005), an increased TAR L2 (P = .002), and higher GMI (P < .001) (Table 3).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The results of the unadjusted model were consistent with the results of the adjusted model.

    “The results of the unadjusted model (Supplemental Table S1) were consistent with the results of the adjusted model (Table 3).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    TIR and age were not associated with the probability of missing a basal insulin dose.

    “TIR and age were not associated with the probability of missing a basal insulin dose, indicating that younger patients did not forget more basal insulin doses than older patients.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    All patients included in this analysis received basal insulin degludec; 28 of the 32 patients also administered their bolus insulin injections using a smart insulin pen during the study.

    “All patients included in this analysis received basal insulin degludec; 28 of the 32 patients also administered their bolus insulin injections using a smart insulin pen during the study.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The post hoc analysis included 32 adults with type 1 diabetes, contributing 4410 acceptable CGM days (315 14-day periods).

    “Thirty-two patients with 4410 acceptable CGM days (315 14-day periods) were included.”
    • Also stated in: § Patient Characteristics · plain
      “Overall, 32 patients (mean age: 35.6 years) were included in this analysis, with 4410 acceptable CGM days (315 14-day periods).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Patients with high glycemic variability tended to forget more basal insulin doses, but high glycemic variability was not a direct result of the missed basal insulin doses.

    “Collectively, these data demonstrate that patients with high glycemic variability tended to forget more basal insulin doses, but that high glycemic variability was not a direct result of the missed basal insulin doses.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Study design and conductreasoning

    A detailed study of the effect of the HCP visit with the smart pen and changes over time in basal insulin adherence is not easily included in this framework.

    “Also, since an HCP visit may fall before, during, or after a 14-day period, a detailed study of the effect of the HCP visit with the smart pen and changes over time in basal insulin adherence is not easily included in this framework.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Missed bolus insulin doses correlate with higher HbA1c levels in patients with T1D.

    “This is consistent with other studies of patients with T1D that have demonstrated that missed bolus insulin doses correlate with higher HbA1c levels.24,25”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [24] Norlander LM Anderson S Levy CJ, et al. Late and missed meal boluses with multiple daily insulin injections. Diabetes. 2018; 67(suppl 1):992-P. source
    • cites [25] Burdick J Chase HP Slover RH, et al. Missed insulin meal boluses and elevated hemoglobin A1c levels in children receiving insulin pump therapy. Pediatrics. 2004; 113(3, pt 1): e221-e224. 14993580 source

  • May combine two claims.

    Study design and conductreasoning

    The model of exposure and outcome is coarse, and temporal covariates such as time since last missed basal insulin dose cannot be included in this analysis.

    “Consequently, the model of exposure and outcome is coarse, and we cannot include temporal covariates, such as time since last missed basal insulin dose in this analysis.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Using smart insulin pen devices in conjunction with CGM could enable the timely detection of missed basal insulin injections.

    “Furthermore, using smart insulin pen devices in conjunction with CGM could enable the timely detection of missed basal insulin injections, improving the quality of discussions between HCPs and patients, thereby increasing patient awareness, empowering patients to improve their diabetes management, and providing ongoing diabetes management education and support.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    The small study population and limited background information on included patients are limitations of this analysis.

    “However, there are limitations, such as the small study population and limited background information on the included patients, which should be considered when interpreting these results.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation

    Meals were identified from the CGM signal using the clinically validated GRID algorithm.

    “Meals were identified from the CGM signal using the clinically validated GRID algorithm. 22”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [22] Harvey RA Dassau E Zisser H Seborg DE Doyle FJ 3rd. Design of the glucose rate increase detector: a meal detection module for the health monitoring system. J Diabetes Sci Technol. 2014; 8(2): 307- 320. 24876583 PMC4455414 source
16 of 16

  • The paper relies on this without support.

    Smart insulin pen (NovoPen 6) injection data accurately reflect the actual timing and dose of basal insulin injections.

    “This is the first study, based on accurate real-world injection data, to demonstrate the challenge of adherence to basal insulin injections in patients with T1D”
    “A key strength of this analysis was that the use of smart insulin pens enabled the collection of accurate injection timing data under real-world conditions”

  • The paper relies on this without support.

    The 32 adults with T1D included in this analysis are representative of adults with T1D generally.

    “This is the first study, based on accurate real-world injection data, to demonstrate the challenge of adherence to basal insulin injections in patients with T1D”
    “This is the first study to demonstrate, based on real-world injection data collected from smart insulin pens, that adherence to basal insulin injections can be a real challenge for patients with T1D and that missed basal insulin injections have a considerable impact on glycemic control.”
2 of 2

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Conflicting confidence intervals for missed basal doses

    Minor

    The abstract reports a 95% confidence interval of 10%-40% for the probability of missing at least one basal dose, but Table 2 gives 10.5%-89.5% for the same estimate. The upper limits cannot be reconciled by rounding.

  • Missed-dose threshold differs between methods and figure

    Minor

    The methods count an interval of exactly 40 hours as a missed dose, whereas the Figure 1 caption excludes it by specifying more than 40 hours.

2 of 2

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Ekberg, N. R., Hartvig, N. V., Kaas, A., Møller, J. B., & Adolfsson, P.
Affiliations
  • Karolinska University Hospital (SE)
  • Novo Nordisk (Denmark) (DK)
  • Hallands sjukhus Kungsbacka (SE)
Corresponding author
Ekberg, Neda Rajamand
ORCID
Venue
Journal of Diabetes Science and Technology
Publisher
SAGE Publications
Published
2022-07-01
Volume
18
Issue
1
Pages
66-73
ISSN
1932-2968 (ISSN-L)
DOI
10.1177/19322968221104142
PMC
PMC10899845
OpenAlex
W4283779445
Cited by
33
References
25
Citations by year
2026: 7 · 2025: 12 · 2024: 8 · 2023: 6
FWCI
3.94
Type
journal article
Language
en
Open access
Licence
Locations
  • Journal of Diabetes Science and Technology (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by)
Funders
Novo Nordisk
Topics
  • Health Sciences › Medicine › Endocrinology, Diabetes and Metabolism › Diabetes Management and Research
  • Health Sciences › Medicine › Endocrinology, Diabetes and Metabolism › Diabetes Management and Education
  • Health Sciences › Health Professions › Emergency Medical Services › Patient Safety and Medication Errors
Keywords
type 1 diabetes, basal insulin, insulin adherence, smart insulin pens, continuous glucose monitoring, glycemic control, glucose management indicator, time in range, adults
MeSH terms
Glycemic Control, Adult, Blood Glucose, Diabetes Mellitus, Type 1, Humans, Hypoglycemic Agents, Insulin, Prospective Studies, Blood Glucose Self-Monitoring

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review