CoChat
Sign Up
Cited by 40Diabetes Management and ResearchDiabetes Management and EducationDiabetes Treatment and Management

Shining the Spotlight on Multiple Daily Insulin Therapy: Real-World Evidence of the InPen Smart Insulin Pen

Open access
MacLeod, Janice, Im, Glen Heungyong, Smith, Madison, Vigersky, Robert A.·2024·Diabetes Technology & Therapeutics
DOI 10.1089/dia.2023.0365PMC PMC10794824
Open this paper in CoChat — free accountOpen PDF

Abstract

Objective: Connected insulin pens are creating opportunities for the millions of individuals with diabetes using multiple daily injections (MDI) therapy across the globe. Continuous glucose monitoring (CGM) data from connected insulin pens are revealing gaps and opportunities to significantly improve care for this population. In this article, we report real-world findings of the InPen™ smart insulin pen paired with CGM (InPen system), used by persons with type 1 diabetes (T1D) and type 2 diabetes (T2D). Methods: A retrospective cohort analysis was conducted with the real-world data collected from the InPen system of individuals ( N = 3793 with T1D, N = 552 with T2D, and N = 808 unidentified) who used the system from January 01, 2020, to December 31, 2021. Diabetes management (e.g., missed and mistimed insulin dosing, mismatched food intake, and correction dose delivery) and glycemic outcomes were assessed. Results: In the overall and T1D populations, a dosing frequency of ≥3 doses per day and a missed dose frequency of <20% was associated with improved glycemia. In adults with T2D, missing <20% of doses was the significant factor determining improved glycemia. Conclusion: This analysis, integrating data from a smart insulin pen and CGM, provides insights into the impact of dosing behavior on glycemic outcomes and informs counseling strategies for the diabetes care team, through technologically advanced insulin management for those using MDI therapy.

AI snapshot

Written from the paper’s full text.

TL;DR

In a retrospective real-world analysis of 5,153 InPen smart insulin pen users paired with CGM, dosing ≥3 rapid-acting insulin doses/day and missing <20% of doses were associated with better glycemia in the overall and type 1 diabetes populations, while missing <20% of doses was the significant factor in adults with type 2 diabetes. On-time dosing and fewer missed correction opportunities correlated with higher time-in-range.

Key findings

  • Overall, ≥3 doses/day plus missed dose rate <20% (N = 962) was associated with TIR 70.3 ± 19.2%, TAR 26.9 ± 19.8%, and GMI 7.0 ± 0.8, compared with <3 doses/day (N = 1909; TIR 52.0 ± 26.0%, TAR 46.3 ± 26.4%, GMI 7.9 ± 1.3) and ≥3 doses/day with missed dose rate ≥20% (N = 2282; TIR 50.1 ± 20.2%, TAR 48.2 ± 20.8%, GMI 7.8 ± 0.9).
  • In adults with type 2 diabetes, missing <20% of doses was the significant factor determining improved glycemia.
  • Cluster analysis showed Cluster 3 had the highest on-time dosing (72.9 ± 1.0%) and TIR 69.0 ± 19.8%, while Cluster 1 had the lowest on-time dosing (27.9 ± 8.5%) and TIR 37.4 ± 20.5%; mean TIR and TAR differed significantly across clusters (P < 0.0001). On-time dose frequency correlated with TIR (r = 0.570, P < 0.0001).
  • Higher dose-count groups tended to have lower TAR and GMI and higher TIR; for example, ≥6+ doses/day (N = 382) had TIR 61.0% and GMI 7.33 versus ≥1 dose/day (N = 462) with TIR 50.8% and GMI 8.01, with significant difference in mean TIR (P < 0.05). Missed correction opportunities per day showed strong negative correlation with TIR (r = −0.852, P < 0.0001).

Methods

  • Retrospective cohort analysis of real-world InPen system data (InPen smart insulin pen paired with personal CGM) from January 1, 2020, to December 31, 2021. Users were adults and pediatrics who used the system ≥30 days; the analyzed dataset included the last 14 days of bolus dosing and CGM data. Users without at least one bolus record and <70% sensor glucose data were excluded, and the first 30 days were excluded. Overall N = 5153 (3793 type 1 diabetes, 552 type 2 diabetes, 808 unidentified).
  • Dosing behaviors included average rapid-acting doses/day, correction-only versus meal/meal+correction doses, and missed/mistimed doses. Meal events were detected by sensor glucose rate-of-change; on-time bolus was 30 min before to ≤15 min after the start of glucose rise, late was >15 to ≤60 min after, and missed was no bolus from ≤30 min before to ≥60 min after. Missed correction opportunities were defined as elevated sensor glucose where dose recommendations were no less than each user’s correction threshold.
  • Statistical analysis used ANOVA with Tukey post hoc (significance 0.05), Pearson correlation, and k-means clustering (k-means++ with standardized numerical features) in Python/SciPy and BigQuery Machine Learning. Continuous variables are mean ± SD; categorical variables are n (%).

Limitations

  • The paper states that the population only included users with compatible real-time CGM alongside the smart insulin pen; a follow-up study using blood glucose monitoring would be needed for smart insulin pen users without CGM. The study period was limited to 30 days, not the entire duration of usage, and race/ethnicity data were not available.
  • The paper states that the analyses compared groups split from real-world data and do not prove direct causal relationships; for example, more frequent bolus dosing was associated with higher TIR, but other factors such as on-time, late, or missed dosing and optimized therapy settings may contribute.
  • Evident from the text: this was a retrospective observational analysis without a control group, and demographic information was self-reported and optional, with missing age, diabetes type, and gender data (e.g., 808 type not specified, 695 age not specified), which limits generalizability.

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

Integrity signals

Registry

Crossref · 10.1089/dia.2023.0365

Cites

25 of 25 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

4 of 12 cited sources checked · 1 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    All three clusters had a significant difference in mean time-in-range (P < 0.0001).

    “All three groups had a significant difference in mean TIR (P < 0.0001) and TAR (P < 0.0001).”
    Quote in the paperSubstantial claim
  • Findingown data

    In adults with T2D, missing <20% of doses was the significant factor determining improved glycemia.

    “In adults with T2D, missing <20% of doses was the significant factor determining improved glycemia.”
    Quote in the paperSubstantial claim
  • Findingown data

    In the T1D population, a dosing frequency of ≥3 doses per day and a missed dose frequency of <20% were associated with better glycemic outcomes compared to lower dosing frequency or higher missed dose rate.

    “In the overall and T1D populations, a dosing frequency of ≥3 doses per day and a missed dose frequency of <20% was associated with improved glycemia.”
    • Also stated in: § Results · plain
      “In the overall and T1D populations, a dosing frequency of 3+ doses/day and a missed dose frequency of <20% were associated with better glycemic outcomes compared to those with a lower dosing frequency or higher missed dose rate.”
    Quote in the paperSubstantial claim
  • Findingown data

    Increased correction dose frequency is associated with improved time-in-range.

    “This analysis, in fact, shows that increased correction dose frequency is associated with improved TIR.”
    Quote in the paperSubstantial claim
  • Findingown data

    All three clusters had a significant difference in mean time above range (P < 0.0001).

    “All three groups had a significant difference in mean TIR (P < 0.0001) and TAR (P < 0.0001).”
    Quote in the paperSubstantial claim
  • May combine two claims. 11
  • Classified as probably not a substantial claim. 9
  • The quote only partly states this claim. 1

  • May combine two claims.

    Premisecitation

    In a prospective, observational trial of 94 participants with T1D, the use of a connected insulin pen with continuous glucose monitoring resulted in significant improvements in glycemia and a significant decrease in missed doses.

    “In a prospective, observational trial of 94 participants with T1D, the use of a connected insulin pen with CGM resulted in significant improvements in glycemia and a significant decrease in missed doses.13”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [13] Adolfsson P, Hartvig NV, Kaas A, et al. Increased time in range and fewer missed bolus injections after introduction of a smart connected insulin pen. Diabetes Technol Ther 2020; 22: 709– 718. 32003590 PMC7591375 source

      The Methods describe a prospective, observational study, and the Results report significant improvements in glycemic measures among 94 adults with T1D using CGM. Missed bolus injections fell significantly, but that analysis included only 81 participants. The Discussion cautions that this single-arm study cannot establish that the connected pen directly caused the improvements, so “resulted in” overstates the finding. (LLM, conf=0.90)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Premisecitation

    Research from individuals using multiple daily injection therapy with continuous glucose monitoring and connected insulin pen or cap devices revealed that missed, mistimed, and mismatched insulin bolus doses are common and occur in both older and younger populations with T1D or T2D, and are associated with higher A1C.

    “Research from individuals using MDI therapy with CGM and connected insulin pen or cap devices revealed that missed, mistimed, and mismatched insulin bolus doses are common and occur in both older and younger populations with T1D or T2D, and are associated with higher A1C.13–18”
    Quote in the paperSubstantial claim2 of 6 sources checked
    • cites [13] Adolfsson P, Hartvig NV, Kaas A, et al. Increased time in range and fewer missed bolus injections after introduction of a smart connected insulin pen. Diabetes Technol Ther 2020; 22: 709– 718. 32003590 PMC7591375 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [14] Munshi MN, Slyne C, Greenberg JM, et al. Nonadherence to insulin therapy detected by bluetooth-enabled pen cap is associated with poor glycemic control. Diabetes Care 2019; 42: 1129– 1131. 30862650 source
    • cites [15] Norlander LM, Anderson S, Levy CJ, et al. Late and missed meal boluses with multiple daily insulin injections. Diabetes 2018; 67: A259. source
    • cites [16] Randlov J, Poulsen JU. How much do forgotten insulin injections matter to hemoglobin a1c in people with diabetes? A simulation study. J Diabetes Sci Technol 2008; 2: 229– 235. 19885347 PMC2771482 source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [17] Robinson S, Newson RS, Liao B, et al. Missed and mistimed insulin doses in people with diabetes: A systematic literature review. Diabetes Technol Ther 2021; 23: 844– 856. 34270324 source
    • cites [18] Toschi E, Slyne C, Greenberg JM, et al. Examining the relationship between pre- and postprandial glucose levels and insulin bolus timing using bluetooth-enabled insulin pen cap technology and continuous glucose monitoring. Diabetes Technol Ther 2020; 22: 19– 24. 31483167 source

  • May combine two claims.

    Premisecitation

    A study of pediatric individuals with T1D identified six habits associated with reduction of HbA1c, four of which are related to insulin delivery.

    “A study of pediatric individuals with T1D identified six habits associated with reduction of HbA1c.28 Four of these six habits are related to insulin delivery, demonstrating the importance of ensuring that individuals with diabetes have the support they need to manage the complexity of determining when to deliver an insulin dose and how much insulin is needed.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [28] Lee JM, Rusnak A, Garrity A, et al. Feasibility of electronic health record assessment of 6 pediatric type 1 diabetes self-management habits and their association with glycemic outcomes. JAMA Netw Open 2021; 4: e2131278. 34709387 PMC8554640 source

  • May combine two claims.

    Findingown data

    Cluster 3 had the highest adherence to on-time dosing and better glycemic outcomes.

    “Cluster analysis demonstrated that Cluster 1 (N = 1125) had the lowest adherence to on-time dosing and the suboptimal glycemic outcomes, while Cluster 3 (N = 1889) demonstrated the highest adherence to on-time dosing and had better glycemic outcomes.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Users tended to have lower time above range and lower glucose management indicator as they dosed more often with a higher correction frequency.

    “Users tended to have lower TAR and GMI, as they dosed more often with a higher correction frequency.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Cluster 1 had the lowest adherence to on-time dosing and suboptimal glycemic outcomes.

    “Cluster analysis demonstrated that Cluster 1 (N = 1125) had the lowest adherence to on-time dosing and the suboptimal glycemic outcomes, while Cluster 3 (N = 1889) demonstrated the highest adherence to on-time dosing and had better glycemic outcomes.”
    Quote in the paperSubstantial claim

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

    Study design and conductnone given

    Demographic information on race/ethnicity was not available in this real-world analysis.

    “Demographic information on race/ethnicity was not available in this real-world analysis.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductreasoning

    The analyses were designed to compare between groups split from the population and do not prove a direct causal relationship between variables.

    “Based on the real-world data, the analyses were designed to compare between groups split from the population. This may be useful to reveal the difference between the groups by the certain behaviors, which was an aim for this study. However, it does not prove the direct causal relationship between variables.”
    Quote in the paperSubstantial claim

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

    Study design and conductnone given

    The study period is limited to 30 days, not during the entire duration of usage.

    “Second, the study period is limited to 30 days, not during the entire duration of the usage.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based). The quote only partly states this claim.

    Interpretationreasoning

    Smart insulin pen systems provide precision insulin management for all individuals on insulin therapy regardless of delivery method.

    “It demonstrates that smart insulin pen systems support MDI therapy patients in daily dosing decisions and may facilitate communication with the diabetes care team, which provides precision insulin management for all individuals on insulin therapy regardless of delivery method.”
    Quote in the paperNot a substantial claim

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

    Study design and conductreasoning

    Internal Review Board approval or exemption was not necessary because the study did not involve human subject protection requirements.

    “As the present study did not involve subject recruitment, enrollment, or participation in a trial and did not fall under human subject protection requirements (per the Department of Health and Human Services, Code of Federal Regulations, Title 28, Chapter 1, Part 46), Internal Review Board approval or exemption was not necessary.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Insulin stacking is estimated to occur in 60% of insulin bolus doses, putting the individual at risk for hypoglycemia.

    “Insulin stacking (overlap of bolus doses) is estimated to occur in 60% of insulin bolus doses putting the individual at risk for hypoglycemia.22”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [22] Schmidt S, Norgaard K. Bolus calculators. J Diabetes Sci Technol 2014; 8: 1035– 1041. 24876436 PMC4455388 source

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    The first 30 days were excluded from the analysis to avoid measuring dosing behaviors or glycemic outcomes associated with onboarding of the diabetes technology.

    “The first 30 days were excluded from the analysis to avoid measuring dosing behaviors or glycemic outcomes associated with the onboarding of the diabetes technology.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Users without at least one bolus record and <70% of sensor glucose data available in the last 14 days between the first 30 to 59 days of InPen use were excluded.

    “Users without at least one bolus record and <70% of sensor glucose (SG) data available in the last 14 days between the first 30 to 59 days of InPen use were excluded.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    In an observational study of 50 participants with T1D on multiple daily injection therapy using a Bluetooth-enabled pen cap and continuous glucose monitoring, 37% of total boluses resulted in persistent 3-hour postprandial hyperglycemia, while 10% resulted in clinically significant 3-hour postprandial hypoglycemia (glucose <55 mg/dL).

    “In an observational study of 50 participants with T1D on MDI therapy using a Bluetooth-enabled pen cap and CGM, it was noted that 37% of total boluses resulted in persistent 3-h postprandial hyperglycemia, while 10% resulted in clinically significant 3-h postprandial hypoglycemia (glucose <55 mg/dL).18”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [18] Toschi E, Slyne C, Greenberg JM, et al. Examining the relationship between pre- and postprandial glucose levels and insulin bolus timing using bluetooth-enabled insulin pen cap technology and continuous glucose monitoring. Diabetes Technol Ther 2020; 22: 19– 24. 31483167 source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisereasoning

    Real-world data permit analysis of a large data set that includes both sensor glucose and the time and amount of the insulin dose and reflect the actual conditions that patients encounter in daily life.

    “Real-world data not only permit the analysis of a large data set that includes both sensor glucose and the time and amount of the insulin dose but also importantly reflect the actual conditions that patients encounter in daily life.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based). May combine two claims.

    Interpretationreasoning

    Intelligent dosing technology can support both the person on multiple daily injection therapy and their care team, further reducing the considerable burden of insulin delivery.

    “Intelligent dosing technology can support both the person on MDI therapy and their care team, further reducing the considerable burden of insulin delivery.”
    Quote in the paperNot a substantial claim
17 of 17

  • The paper relies on this without support.

    This study is the first exploration and analysis of behaviors and outcomes for multiple daily injection patients using real-world data.

    “A strength of this study is that it is the first exploration and analysis of behaviors and outcomes for MDI patients using real-world data.”
1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 2

  1. Vigersky RA Cordero TL MacLeod J et al. Artificial intelligence: The next frontier in diabetes therapy. Nature: Sponsor Feature. Available from: https://www.nature.com/articles/d42473-021-00265-6#:∼:text=The%20%E2%80%9Cinternet%20of%20things%E2%80%9D%20applied minimize%20the%20burden%20of%20diabetes [Last accessed: July31 2023].
  2. Kerr D, , Warshaw H. Smart insulin pens will address critical unmet needs for people with diabetes using insulin. Endocr Today, 2019; 17:21–22.

Cited by

Sorted by citations

Related

Authors
MacLeod, J., Im, G. H., Smith, M., & Vigersky, R. A.
Affiliations
Medtronic Diabetes, Northridge, California, USA.
Corresponding author
MacLeod, Janice
ORCID
Venue
Diabetes Technology & Therapeutics
Publisher
SAGE Publications
Published
2024-01
Volume
26
Issue
1
Pages
33-39
ISSN
  • 1520-9156 (ISSN-L)
  • 1557-8593
DOI
10.1089/dia.2023.0365
PMC
PMC10794824
OpenAlex
W4387765534
Cited by
40
References
25
Citations by year
2026: 13 · 2025: 11 · 2024: 16
FWCI
6.88
Type
journal article
Language
en
Open access
Licence
Locations
  • Diabetes Technology & Therapeutics (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • PubMed Central (submittedVersion, cc-by)
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 › Medicine › Endocrinology, Diabetes and Metabolism › Diabetes Treatment and Management
Keywords
smart insulin pens, connected insulin pens, continuous glucose monitoring, type 1 diabetes, type 2 diabetes, glycemic outcomes, insulin dosing frequency
MeSH terms
Adult, Blood Glucose, Diabetes Mellitus, Type 1, Diabetes Mellitus, Type 2, Humans, Hypoglycemic Agents, Insulin, Retrospective Studies, Blood Glucose Self-Monitoring, Insulin, Regular, Human

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