Frequency and Detection of Insulin Infusion Site Failure in the Type 1 Diabetes Exchange Online Community

Abstract

Insulin infusion site (IIS) failures are a weakness in insulin pump therapy. We examined experience with IIS failures among U.S. individuals with diabetes on insulin pump through survey distributed to the T1D Exchange Online Community. Demographic factors, IIS characteristics, and diabetes-related perceptions were assessed by logistic regression to determine odds of higher (≥1 per month) or lower (<1 per month) reported IIS failure frequency. IIS failures were common; 41.4% reported ≥1 per month. IIS failure is usually detected through development of hyperglycemia rather than pump alarm. No assessed demographic factor or IIS characteristic was predictive; however, higher odds of ≥1 failure per month were associated with feelings of burnout (odds ratios [OR] 1.489 [1.024, 2.165]) and considering pump discontinuation (OR 2.233 [1.455, 3.427]). IIS failures are frequent and unpredictable, typically require hyperglycemia for detection, and are associated with negative perceptions. More should be done toward preventing IIS failures and/or detecting them sooner.

Introduction

Insulin pumps represent a major therapeutic advancement for patients with diabetes, and adoption has increased rapidly in recent years.¹ Subcutaneous insulin delivery depends on the integrity of many pump components, but perhaps the most tenuous is the insulin infusion site (IIS).² The IIS includes a cannula intended to deliver insulin into the subcutaneous space. Available options can differ in cannula angle, material, length, and insertion mechanism.³ The majority are approved for only 2–3-day use,^2,4 although an extended wear (up to 7 days) option was recently approved.⁵

IIS failure is most often attributed to cannula-associated issues (kinking, occlusion, and dislodgement) or insertion site inflammation/infection.^2,6,7 Without recognition and replacement, IIS failures lead to inadequate insulin delivery, hyperglycemia, and increased risk for diabetes-related ketoacidosis (DKA).^2,5,8,9 Correspondingly, IIS failures are the primary contributor to DKA development and the leading cause of DKA-related hospitalizations among insulin pump users.^9

–12 In severe instances, insulin pump delivery failures, including from IIS insertion errors, have resulted in fatalities.⁸

Development of technologies to reduce incidence and improve detection of IIS failures is ongoing.^5,6,11,13,14 We present survey data from the T1D Exchange Online Community to explore real-world experience with IIS failure.

Research Design and Methods

Study design

A voluntary English-only online survey hosted on SurveyMonkey^® was available from September 3, 2020, to November 19, 2020. It was distributed through two emails to members of the T1D Exchange Online Community and published on T1D Exchange social media accounts, including Twitter (four posts), Instagram (four posts), Facebook (four posts), and LinkedIn (one post). Eligible respondents were ≥18 years old, residing in the United States, and diagnosed with or caring for an individual with type 1 or type 2 diabetes actively using an insulin pump. The survey was initially conducted for market research; however, waiver for analysis and publication was obtained retrospectively from the Children's Hospital Los Angeles Institutional Review Board.

Survey

Participants were asked to self-report demographics as well as most recent hemoglobin A_1c (HbA_1c), estimated average total daily dose (TDD) of insulin, type and brand of diabetes technologies (i.e., continuous glucose monitor [CGM], pump, and infusion set), duration of insulin pump use, and estimated average IIS wear time. Participants were also asked about IIS failures, defined as:

…when your infusion set/pod does not correctly infuse a bolus or basal rate of insulin from your pump, potentially resulting in hyperglycemia or a loss of blood sugar control. These failures can be caused by numerous issues, including: inflammation at the infusion site (and other skin complications), occlusions, leakages, kinking of the cannula, adhesion problems, and dislodgement from the infusion site.

From this definition, respondents reported their estimated frequency of IIS failure, how they are first alerted of IIS failure, and what actions they first take after an IIS failure. Current perceptions, including sense of diabetes control, feeling of burnout from diabetes, and whether there was any recent consideration for insulin pump discontinuation were also queried (full survey available in Supplementary Data).

Statistical analysis

There are no established guidelines for the target limit of IIS failure frequency, but it is generally understood that fewer failures are better since even a single failure can affect glycemic control. In our study, frequent IIS failure was defined as occurring at least once per month, which was clearly defined by the questionnaire. Although less frequent failures could impact patients, the survey did not discern data on frequencies between at least once every 6 months and once per month.

Logistic regression (LR) was used to calculate odds ratios (ORs) for variables hypothesized to contribute to IIS failure frequency ≥1 per month. More frequent IIS failures were assessed for association with perceptions of diabetes control, feelings of burnout, and consideration for discontinuing pump. A single multivariable LR was performed incorporating all IIS characteristic variables, except presence of inserter due to multicollinearity. All analyses were performed using R code and packages.¹⁵

Results

In total, 1079 responses were received, of whom 339 participants did not complete the survey and were excluded from the final analysis, and 33 duplicates were also excluded after identification by manual review and email confirmation. Final analysis included 707 responses, comprising 631 (89.3%) adults with diabetes on insulin pump and 76 (10.7%) caregivers. Incompatible IIS characteristics (e.g., a reported infusion site type that does not offer the reported cannula length) were encountered on 102 (14%) surveys. In each case, all IIS characteristics were removed from analysis, but the remainder of survey responses were included.

The resulting cohort (Table 1) predominantly comprised adults (90.8%). Individuals were also mostly of female gender and white race. A majority (63.6%) reported most recent HbA_1c meeting American Diabetes Association goal <7%. Mean HbA_1c was 6.8% ± 1.0%. Median time on insulin pump was 14 years [interquartile range; IQR 7.1, 20.1], with >95% of the cohort on CGM and nearly half reporting automated insulin delivery (AID) use. IIS characteristics (n = 603) tended to have a 90° insertion angle (61.2%), short (72.8%), and Teflon (89.4%) cannulas, and a built-in or optional inserter (89.2%) (see Supplementary Table S1 for classification of characteristics). A majority (>70%) reported exchanging IIS every ≤3 days on average. Of note, 24.5% used Omnipod, which does not allow extension of IIS wear beyond 3 days.

Table 1.

Participant Characteristics

Characteristic	N = 707
Age^a	47.7 ± 18.5
Gender
Female	489 (69.2)
Male	209 (29.6)
Transgender	1 (0.1%)
Genderqueer (not exclusively male or female)	4 (0.6%)
Prefer not to answer	4 (0.6%)
Race/ethnicity^b
American Indian or Alaska Native	1 (0.1%)
Asian or Asian American	3 (0.4%)
Black or African American	7 (1.0%)
Hispanic or Latino	10 (1.4%)
Middle Eastern or North African	1 (0.1%)
Mixed/biracial	19 (2.7%)
White or Caucasian	655 (92.6%)
Prefer not to answer	11 (1.6%)
HbA_1c ^c (%)	6.8 ± 1.0
TDD of insulin^d (U/day)	38 [28–50.9]
CGM use	673 (95.2%)
Insulin pump brand^e
Insulet	173 (24.5%)
Medtronic	247 (34.9%)
Roche	2 (0.3%)
Tandem	284 (40.2%)
Ypsomed	1 (0.1%)
Time on insulin pump (years)	14.0 [7.1–20.1]
Use of AID system^f	312 (44.1%)
Time on AID system (years)	1.2 [0.5–7.2]
Average IIS exchange frequency
≤3 Days	518 (73.3%)
>3 Days^g	189 (26.7%)
Time since last diabetes education
<6 Months	204 (28.9%)
6–12 Months	146 (20.7%)
1–2 Years	104 (14.7%)
2–4 Years	107 (15.1%)
>4 Years	146 (20.7%)

Data are reported as N (%) for binary categorical variables, mean ± SD for continuous normally distributed variables, or median [IQR] for continuous non-normally distributed variables.

There were 65 individuals (9.2%) with age <18 years.

Participants were able to select multiple options.

HbA_1c was not reported by 15 (2.1%) participants.

TDD of insulin was not reported by 48 (6.8%) participants.

Pump use was part of criteria for analysis; pump types included Insulet OmniPod Eros, Insulet OmniPod DASH, Medtronic Paradigm series, Medtronic 630G, Medtronic 670G, Tandem t:flex, Tandem t:slim, Roche Accu-Check Spirit, and Ypsomed YpsoPump.

Reported AID systems included Medtronic 670G, Tandem Control IQ, and Open Source utilizing Omnipod or Medtronic pump.

Includes 4 (0.6%) participants who reported use >7 days.

AID, automated insulin delivery; CGM, continuous glucose monitor; HbA_1c, hemoglobin A_1c; IIS, insulin infusion site; IQR, interquartile range; SD, standard deviation; TDD, total daily dose.

Experience with IIS failure was common (Fig. 1) with 683 respondents (96.6%) having ≥1 IIS failure every 12 months. Nearly half (41.4%) indicated encountering ≥1 failure per month.

FIG. 1.

Participants' self-reported frequencies of IIS failure. Depicts survey results from self-estimated IIS failure frequencies. The most selected answer was “Rarely (Once every 2–6 months)”; however, a frequency of IIS failure ≥1 per month was selected by 41.4% of the cohort. IIS, insulin infusion site.

Of participants reporting any history of IIS failure, 56.2%, 25.0%, 9.9%, 5.7%, and 3.2% reported that their most common method of first being alerted to an IIS failure was by experiencing hyperglycemia on glucometer or CGM, a pump alarm, hyperglycemia symptoms, a site-specific issue (see/smell/feel insulin leak, inflammation, and swelling), or unknown/not alerted, respectively. The most reported initial action taken after an IIS failure was discovered, giving a pump bolus without IIS replacement (47.7%), followed by replacing the IIS and then giving a pump bolus (43.5%).

Additional reported initial actions included giving an alternative source of fast-acting insulin (syringe/pen/inhaled insulin) and replacing IIS (4.8%), giving an alternative source of fast-acting insulin (syringe/pen/inhaled insulin) and not replacing IIS (2.0%), other/nonspecific strategies (e.g., waiting additional time), and nothing (0.7%).

None of the assessed participant characteristics, including age, household income, HbA_1c, insulin TDD, duration of insulin pump use, and estimated average IIS wear time >3 days was found to have significant odds associated with higher (≥1 per month) or lower (<1 per month) reported IIS failure frequency. In addition, no individual IIS attributes including angled, steel, and long cannula types as well as maximum wear time hard stop of 3 days (i.e., Omnipod) were predictive of ≥1 IIS failure per month. ORs for feelings of good diabetes control were not significant.

Feeling burnout was associated with higher odds of IIS failure ≥1 per month (OR 1.489, 95% confidence interval [CI] 1.024–2.165), whereas those who reported that they were considering ending pump use had the highest odds of having at least monthly IIS failures (OR 2.233, 95% CI 1.455–3.427). Full set of ORs with forest plot is available in Supplementary Table S2.

Conclusions

This study characterizes the frequency, method of detection, diabetes-related perceptions, and potential predictors associated with IIS failure in a large real-world cohort of individuals with diabetes. A broad characterization of IIS failure was used to include any incident in which inadequate insulin delivery is recognized, regardless of whether it meets previously specified criteria, which often differ considerably in studies.^{3,5,7,11,13,14,16} With this definition, the universality of experience with IIS failure is readily apparent: More than 95% of participants reported at least one occurrence in the prior year.

Greater burden of reported IIS failure was not predicted by age, household income, HbA_1c, insulin TDD, duration of pump use, or average IIS wear time. In contrast, a recent posthoc analysis by Kanapka et al. examining 22,741 infusion site wears among 263 participants across two clinical trials found that participants aged <18 years or with higher baseline HbA_1c had greater odds of IIS failure using a definition of prolonged (≥2 h) hyperglycemia before IIS exchange.¹⁶ In addition, longer IIS wear duration >3 days has been associated with higher number of catheter-related adverse events, including hyperglycemia (blood glucose >250 mg/dL), site reactions, and “silent occlusions.”^7,12,17

No specific IIS characteristic was found to predict the higher IIS failure rate. Previous studies have shown steel cannulas have better durability of favorable IIS properties and greater resistance to kinking than Teflon.¹⁸ In their analysis, Kanapka et al. found higher rates of IIS failure in association with 90° Teflon cannulas, but a difference between steel and angled Teflon cannulas was not apparent.¹⁶ However, in an open-label crossover study comparing 90° Teflon with steel, Patel et al. found that the strongest predictor of IIS failure was by individual rather than IIS type.³

Monthly occurrences of IIS failure were associated with stronger feelings of burnout and more commonly considering pump discontinuation. This aligns with previous findings that insulin pump discontinuation is commonly due to IIS failure.¹⁹

Only a quarter of participants reported first being notified of IIS failure by their pump (with no association to pump type or AID use). This is unsurprising, as occlusion alarms are often triggered after delays >2 h or not at all.²⁰ Most commonly, the first indication of IIS failure was after hyperglycemia already developed. Therefore, detecting IIS failure earlier, before hyperglycemia develops, represents an important gap.

This retrospective survey has several limitations, including sample bias from using a voluntary online English-only survey and potential for misestimations of event frequencies or glycemic parameters from reliance on self-reported data, contingent upon participant understanding and recall. The results showing no correlation between insulin pump failure rates and income or education are limited by the homogeneity of the cohort, which consisted of mostly white English-proficient individuals, all of whom were using insulin pumps per inclusion criteria.

The cohort does not reflect the T1D Exchange population, in which insulin pump use was 60%, 66%, and 62% and mean HbA_1cs were 8.9%, 7.8%, and 7.7% among adults aged 18–25, 29–49, and ≥50 years old, respectively.¹ Importantly, the definition of IIS failure is also subjective, and the broader definition used does not include a time component, which may affect individual recognition variability. Although some may consider events occurring beyond 3 days as misuse rather than true failures, we feel the definition appropriately reflects the well-known fact that patients commonly use their IIS beyond 3 days in real-world settings.^16,17

The median duration of pump use of 14 years with IQR of 7 years suggests respondents were very experienced users and are likely to be more attuned to IIS failure than those who are new to pump therapy or have shorter duration of use. Even still, self-estimated IIS failure frequency has a strong potential for general underestimation, and there is additional potential for desirability bias, in which participants may feel pressure/embarrassment/self-conscious, and under-report IIS wear time. Finally, those who elect to use a specific IIS type may have done so previously to address a prior higher proclivity toward IIS failure, diminishing an otherwise perceivable difference in effect.

Data to clarify these nuances and several additional variables of potential interest, such as body mass index, insulin type, frequency and duration of insulin suspensions, and incidence of DKA were not collected as part of the survey. Ultimately, detection of true causative associations between IIS failure rates and IIS characteristics will require prospective randomization, likely using each individual as their own control in light of the findings by Patel et al.³ The full survey is included in the supplement, but some responses were not included in the article due to inadequate responses or inadequate definition of terms in the questions.

In summary, these data provide valuable information on the frequency, unpredictability, and burdens of IIS failure, which often go undetected until hyperglycemia develops. This is associated with burnout and discontinuation of pump use. Therefore, there is a critical need for more discerning methods to detect IIS failure earlier, as well as further developments in the IIS technology to minimize the incidence and associated burdens of these failures.

Footnotes

Authors' Contributions

M.S.H. researched data, conducted statistical analysis, and wrote/edited the article. J.L.D. researched data and wrote/edited the article. M.L.B. conducted statistical analysis. L.E.B., J.C.G., J.H.W., and R.A.L. researched data and reviewed/edited the article. G.Z. and J.E. managed IRB approval and reviewed/edited the article.

Funding Information

Creation and distribution of the survey were funded by Diatech Diabetes, Inc. The T1D Exchange is supported through the Leona M. and Harry B. Helmsley Charitable Trust. M.S.H. is supported by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) of the National Institutes of Health (NIH) under Award Number 5K12 DK122550. R.A.L. is supported by a Diabetes, Endocrinology and Metabolism Career Development Grant (1K23 DK122017, P30 DK116074) from NIDDK and has additional research support from JDRF.

Author Disclosure Statement

M.S.H. has consulted for Dexcom, Inc. R.A.L. has consulted for Abbott Diabetes Care, Biolinq, Capillary Biomedical, Deep Valley Labs, Morgan Stanley, Gluroo, and Tidepool. L.E.B., J.C.G., and J.H.W. are employees at Diatech Diabetes, Inc. M.L.-B. was an employee/intern at Diatech Diabetes, Inc. J.L.D., G.Z., and J.E. have no relevant disclosures.

Supplementary Material

Supplementary Data

Supplementary Table S1

Supplementary Table S2

References

Foster

, Beck

, Miller

, et al. State of type 1 diabetes management and outcomes from the T1D Exchange in 2016–2018. Diabetes Technol Ther, 2019; 21(2):66–72.

Heinemann

, Krinelke

. Insulin infusion set: The Achilles heel of continuous subcutaneous insulin infusion. J Diabetes Sci Technol, 2012; 6(4):954–964.

Patel

, Benasi

, Ferrari

, et al. Randomized trial of infusion set function: Steel versus teflon. Diabetes Technol Ther, 2014; 16(1):15–19.

Medtronic Secures FDA Clearance for First Ever Infusion Set That Can Be Worn Up To 7 Days By People With Diabetes. Dublin, Ireland: Press Release: Medtronic PLC; 2021.

Lal

, Hsu

, Zhang

, et al. Longevity of the novel ConvaTec infusion set with Lantern technology. Diabetes Obes Metab, 2021; 23(8):1973–1977.

Schmid

, Hohberg

, Borchert

, et al. Pilot study for assessment of optimal frequency for changing catheters in insulin pump therapy-trouble starts on day 3. J Diabetes Sci Technol, 2010; 4(4):976–982.

Pfützner

, Sachsenheimer

, Grenningloh

, et al. Using insulin infusion sets in CSII for longer than the recommended usage time leads to a high risk for adverse events: Results from a Prospective Randomized Crossover Study. J Diabetes Sci Technol, 2015; 9(6):1292–1298.

Ziegler

, Williams

, Yarid

, et al. Fatalities due to failure of continuous subcutaneous insulin infusion devices: A report of six cases. J Forensic Sci, 2019; 64(1):275–280.

Dogan

ADA

, Jørgensen

, Gjessing

. Diabetic ketoacidosis among patients treated with continuous subcutaneous insulin infusion. J Diabetes Sci Technol, 2017; 11(3):631–632.

10.

Flores

, Amir

, Ahmed

, et al. Causes of diabetic ketoacidosis among adults with type 1 diabetes mellitus: Insulin pump users and non-users. BMJ Open Diabetes Res Care, 2020; 8(2):e001329.

11.

Cescon

, DeSalvo

, Ly

, et al. Early detection of infusion set failure during insulin pump therapy in type 1 diabetes. J Diabetes Sci Technol, 2016; 10(6):1268–1276.

12.

Thethi

, Rao

, Kawji

, et al. Consequences of delayed pump infusion line change in patients with type 1 diabetes mellitus treated with continuous subcutaneous insulin infusion. J Diabetes Complications, 2010; 24(2):73–78.

13.

Gibney

, Xue

, Swinney

, et al. Reduced silent occlusions with a Novel Catheter Infusion Set (BD FlowSmart): Results from Two Open-Label Comparative Studies. Diabetes Technol Ther, 2016; 18(3):136–143.

14.

Howsmon

, Baysal

, Buckingham

, et al. Real-time detection of infusion site failures in a closed-loop artificial pancreas. J Diabetes Sci Technol, 2018; 12(3):599–607.

15.

R Core Team. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing: Vienna, Austria; 2021. Available from: https://www.R-project.org/ [Last accessed: November 10, 2022].

16.

Kanapka

, Lum

, Beck

. Insulin pump infusion set failures associated with prolonged hyperglycemia: Frequency and relationship to age and type of infusion set during 22,741 infusion set wears. Diabetes Technol Ther, 2022; 24(6):396–402.

17.

Sampson Perrin

, Guzzetta

, Miller

, et al. A web-based study of the relationship of duration of insulin pump infusion set use and fasting blood glucose level in adults with type 1 diabetes. Diabetes Technol Ther, 2015; 17(5):307–310.

18.

Højbjerre

, Skov-Jensen

, Kaastrup

, et al. Effect of steel and teflon infusion catheters on subcutaneous adipose tissue blood flow and infusion counter pressure in humans. Diabetes Technol Ther, 2009; 11(5):301–306.

19.

Wood

, Moreland

, Volkening

, et al. Durability of insulin pump use in pediatric patients with type 1 diabetes. Diabetes Care, 2006; 29(11):2355–2360.

20.

Klonoff

, Freckmann

, Heinemann

. Insulin pump occlusions: For patients who have been around the (infusion) block. J Diabetes Sci Technol, 2017; 11(3):451–454.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.02 MB

2.09 MB