523882
research-article2014
DSTXXX10.1177/1932296814523882Journal of Diabetes Science and TechnologyHeinemann et al
Editorial
We Need More Research and Better
Designs for Insulin Infusion Sets
Lutz Heinemann, PhD1, John Walsh, PA, CDTC2,
and Ruth Roberts, MA3
Keywords
insulin therapy, infusion set, insulin pumps, CSII, unexplained hyperglycemia, pump occlusions
An insulin infusion set (IIS) is an indispensable conduit in
insulin pump systems that transfers insulin from the reser-
voir to the subcutaneous tissue. Some form of IIS is required
for all pumps, including patch pumps. Transfer of insulin
across the skin occurs through a stainless steel needle or
through a Teflon cannula once its steel needle has been
removed during the insertion process. With their ease of
insertion, steel needles are inserted manually while Teflon
catheters can be inserted manually, with an auto-insertion
device, or in some cases either.
An insulin pump delivers only rapid-acting insulin, so any
failure of the IIS can become a critical health issue. At best,
IIS failure increases health care costs, while at the worst it can
lead to elevated HbA1c levels, diabetic ketoacidosis (DKA),
or death. Even though IIS designs have improved and fewer
adverse events occur today, pump plumbing (IIS) continues
to retain many similarities to home plumbing in the era of
galvanized pipes fraught with a variety of ways to fail.
When an IIS issue of unexplained hyperglycemia or
occlusion arises, several factors may be involved, including
the physical design of the IIS, material composition of the
catheter, tendency for dislodgement, presence or absence of
a safety anchor or anchoring tape, type of insulin, strength of
the adhesive used to fix the set to the skin, and the lubricant
used to coat the exterior of a Teflon or steel cannula. Any of
these may impact glucose outcomes, duration of use, and
comfort.
Variations in the manufacturing process over time can
also impact IIS reliability. Small changes or adjustments in
IIS materials or the manufacturing process will sometimes
have a considerable impact on IIS properties in daily life. For
example, the thickness of the Teflon wall of a cannula is
directly related to its risk of kinking during insertion, while
its comfort is inversely related to a lesser degree. Fixation of
IIS systems to the skin with adhesive also has a major impact
on glucose outcomes, yet this remains a relatively ignored
area for clinical research.
IIS reliability seems to vary by user and set. In pump
blogs and in clinic and support group discussions, many
users report that they have experienced acute metabolic dete-
riorations with one IIS or another that vanish once they
Downloaded from dst.sagepub.com at GEORGIAN COURT UNIV on March 6, 2015
Journal of Diabetes Science and Technology
2014, Vol. 8(2) 199­–202
© 2014 Diabetes Technology Society
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1932296814523882
dst.sagepub.com
switch to a different one, particularly when switching from a
Teflon IIS to steel. Although anecdotal reports have to be
taken with care, the number and consistency of these reports
suggest that the type of IIS an individual chooses has a real
impact on glucose control.
Patient Experience and Data About IIS
Failure
Despite their critical role in insulin delivery, the number of
research studies on IIS remains relatively small even though
IIS failure is not rare.1 One research study compared 45
pump wearers’ standard infusion set (44 Teflon and 1 steel)
worn over a 1-month period to the Accu-Chek FlexLink infu-
sion set studied during its third month of use in the trial.
Insertion devices were used by 80% of participants with their
standard IIS and by 93% with the IIS (FlexLink) being tested.
Set failure, defined as immediate or unexplained hyperglyce-
mia within the next 6 hours that could not be corrected,
occurred during 8.9% of insertions of the standard sets (45 of
507 insertions). On the FlexLink, the initial failure rate fell to
3.2% (15 of 488 insertions). This study strongly suggests that
infusion set design can reduce initial IIS failure rates and
with further design modifications might eventually be able to
reduce overall failure rates.2
Initial failures were surprisingly common and are usually
associated with use of an auto-insertion device. They
occurred on the standard set once every 22.5 days with 2 day
wear, and once every 33.7 days with 3 day wear. Subsequent
to these initial failures, an additional 8% of sets failed in both
groups. Just over 40% of participants had to replace an infu-
sion set early at least once during this 1-month study in each
IIS test group. One participant experienced DKA and 6 had
1
Science & Co, Düsseldorf, Germany
2
Advanced Metabolic Care and Research, Escondido, CA, USA
3
Diabetes Services, Inc., San Diego, CA, USA
Corresponding Author:
Lutz Heinemann, PhD, Science & Co, Kehler Str 24, 40468 Düsseldorf,
Germany.
Email: l.heinemann@science-co.com
200 Journal of Diabetes Science and Technology 8(2)
measurable levels of ketones with hyperglycemia during the
60 days of study.
During an extended wear study of QuickSet Teflon and
Sure-T steel IISs for 7 days or until failure, Quick-Sets were
found to have a 15% initial failure rate. After the end of 1
week, the failure rate for steel and Teflon gradually equaled
out at 64% for each IIS. During the entire 77 weeks of wear,
30% of sets failed because of hyperglycemia following a
failed correction dose, 13% were removed for pain, 10%
were pulled out by accident, 10% had erythema and/or indu-
ration of >10 mm, 5% fell out because of loss of adhesion,
and 4% were removed for infection.3
After 7 days, 36% of the infusion sets were still operating
and 32.5% were used for a full week (25 out of the total 77
weeks) with no increase in hyperglycemia or daily insulin
requirements. The main predictor for length of wear in this
study was the individual subject. The ability to use an IIS for
an extended period of time varies for people on pumps,
although it is not clear whether early failure is innate to indi-
viduals, or may be influenced by better IIS training.
A poster presented at the ADA and the DTM 2013 docu-
mented significant, intermittent, and unexplained elevations
in blood glucose levels in pump wearers.4 By chance, pres-
sure sensors had been installed in the infusion lines and
researchers at Becton Dickinson were able to connect ele-
vated pressure levels within the infusion line prior to the glu-
cose elevations. Because an occlusion alarm was never
triggered and the hyperglycemia was able to be corrected
over time, these episodes were described as “hidden occlu-
sions” and may explain some instances of unexplained
hyperglycemia. If confirmed in other studies, an IIS redesign
may be required to correct this issue.
In unpublished data from the Actual Pump Practices
Study by 2 of the authors and Timothy Bailey, MD, the pres-
ence of an occlusion alarm was significantly associated with
higher average glucose levels in 364 pump users during an
average of 73 days of wear (P < .0001).5 In the low glucose
tertile with an average glucose of 144 mg/dl, the average
number of occlusions was 1.36 occlusions a month, while the
middle and high tertiles, with average glucoses of 181 and
227 mg/dl respectively, averaged over 3.1 occlusion alarms a
month. Among 12 other variables studied, only the frequency
of hypoglycemia was more strongly associated with the
average glucose level.
Data from a recent survey in Germany confirmed that
success of the IIS is highly relevant for daily success with
pump therapy.6 The metabolic control of patients using steel
IIS (7.7%) was not different from that of those using Teflon
IIS (7.8%); however, a smaller number of patients observed
unexplained increases in glycemia with steel IIS (in total
405: yes n = 210 [52%] versus no n = 195 [48%]) than with
Teflon (535: yes n = 313 [59%] versus no 222 [41%]; P < .05
[chi-square test]).
In reviewing continuous glucose monitoring (CGM)
records, those who wear pumps often do not have awareness
Downloaded from dst.sagepub.com at GEORGIAN COURT UNIV on March 6, 2015
or memory for likely IIS failures and often use terms that
may signify IIS failure, like “scarring,” “site failure,” or
“poor absorption.” Many pump wearers are unable to iden-
tify the frequency of IIS failure or the impact it has on their
overall control. However annoying or dangerous at the time
they occur, IIS failure events can be easily forgotten amid
daily life events and be difficult to distinguish from possible
absorbtion issues. However, a simple test trial with an alter-
nate reliable IIS will usually resolve this uncertainty. Any
delay in correcting unexplained hyperglycemia leads to frus-
tration and has been identified as a common source for dis-
continuation of insulin pump therapy.1
Differences in Usage of IIS Between
Countries
In view of the differences between patients with respect to
age, sex, BMI, physical activities, and insulin pump varia-
tions, a clear need for a variety of IIS types exists. However,
many pump wearers remain on the same IIS that is shipped
with their pump or that the diabetes nurse supplied during
their introduction to insulin pump therapy.
Most IIS are manufactured by a single company in
Denmark. There is a scarcity of published data about how
many pump wearers in different countries use steel versus
Teflon IIS, and about which brands or styles are used in vari-
ous locations. Survey data suggest that steel sets are more
commonly used in Germany (~40% of sets) and Europe
(~25%) than in the United States (~10%), and this may cor-
relate with the market share for various brands of insulin
pumps used in different countries.6
Steel Versus Teflon
Prior to starting on an insulin pump, a person may believe
that steel needles will be less comfortable or carry more
risk during physical work or sports activities than Teflon.
While it is understandable that having a thin sharp needle in
the skin might impose such risk, it appears that this is not
the case. Patients generally report that steel IIS are simpler
and easier to insert and as comfortable as Teflon. Although
an occasional user will switch from steel to Teflon due to
issues with bleeding or occlusion, the majority of Teflon
users who try a steel set stay with it. Steel IIS are also less
expensive at about $8 per set in the United States compared
to $12 per set for Teflon. In a 2004 report, consumables
made up the largest annual expense for those wearing an
insulin pump, and IIS costs made up 84.2% to 89.5% of the
annual expense.7
Compared to steel IIS where an insertion device is not
needed, auto-insertion devices are commonly used with
Teflon IIS. Given the few individuals who have a real fear of
needles or are unable to manually insert an IIS, the wide-
spread use of these devices is puzzling. Do patients really
prefer Teflon IIS for good reasons or is there a marketing or
Heinemann et al 201
training issue here? Is this preference for auto-inserters the
main factor that sways pump wearers to use Teflon more than
steel? And why do many pump manufacturers prefer Teflon
instead of steel IIS? Variations in which styles and brands of
IIS are used in different countries may be driven by the mar-
keting activities and market share of pump manufacturers.
Commissions for pump sales representatives are often par-
tially determined by the number of sales they produce for a
particular IIS.
Although a “soft” cannula may be easier to promote to
those on pumps, there may also be economic advantages for
insulin pump companies that make Teflon IIS sets more
attractive. This may not be a good argument for patients if
significant health risks are associated with certain IIS or with
the insertion device used to install them. For most dextrous
individuals, manually inserting a steel IIS provides better
control with elimination of kinking and a lower rate of inser-
tion failure compared to use of an insertion device.
Length of use recommendations for IIS is generally 2
days for steel sets and 3 days for Teflon sets, although we are
unable to find any data supporting these recommendations
and at least one study found no difference.8 For example, a
study of Tender and Comfort IIS (both Teflon and identical
other than the name) found that glucose levels gradually rose
with each day of use.9 As a result of the study, the authors
recommended that this Teflon set that is inserted at about a
30 degree angle be used for only 2 days.
This angled set is designed and commonly used without
the infusion line being anchored with tape, so any tug or
movement of the infusion line is directly transferred to its
Teflon cannula under the skin. In our experience, this can
lead to skin irritation, redness, and on occasion leakage of
insulin along the Teflon cannula back to the skin. When the
infusion line is simply anchored with tape, these problems
largely disappear, and the same set can be worn for extended
periods of time.
Discussion and Summary
The lack of reliable data about IIS performance is not reas-
suring. We have outlined some areas where IIS issues should
be addressed, such as whether significant clinical differences
exist between steel and Teflon IIS, whether auto-insertion
devices create increased health risks, and the need to improve
adhesion of sets to the skin. We encourage analysis of data-
bases on pump-related medical events to confirm the extent
of IIS issues and identify characteristics of IIS that minimize
unwanted outcomes. IIS manufacturers and insulin pump
companies whose customers use them, and diabetes advo-
cacy groups like the ADA, EASD, and AACE, would ideally
fund surveys and studies to evaluate IIS performance. To
ensure unbiased results, both the design and data analysis of
surveys and studies should be evaluated and published by
independent clinicians to guarantee a balanced and unbiased
presentation of the data.
Downloaded from dst.sagepub.com at GEORGIAN COURT UNIV on March 6, 2015
CGM records provide a convenient method to identify and
correct individual IIS problems. These records could also be
used to identify variations in failure rates among different IIS
brands and styles, between steel and Teflon, and with the use
of an auto-insertion device. Larger data studies with more
data may be needed to determine if clinically relevant differ-
ences exist between particular varieties of steel or Teflon IIS.
Manufacturers, who are required by law to monitor this,
should be able to analyze the number of patient complaints
received by their call centers that may relate to IIS failure.
In view of the impact that the IIS can have on the success
of pump therapy, wearers should be encouraged to try vari-
ous IIS during their introduction to pump therapy by both
clinicians and pump manufactures. Given the potential sav-
ing in cost of product and user frustration, as well as poten-
tial reductions in complaint calls and unnecessary pump and
IIS replacements, pump manufacturers and insurance com-
panies may find it advantageous to cover the minimal cost
for this essential training. It would also be important to pro-
mote safe and successful use of IIS via social networks and
encourage relevant groups to place more attention on this
topic to increase IIS reliability.
Regulatory agencies require glucose meters to meet per-
formance standards. Why aren’t infusion sets that are much
more directly related to health outcomes also required to
meet such standards?
In summary, IIS failure has a major impact on glucose
control and is not uncommon for many pump wearers. It
deserves more attention. We believe that the selection
between steel or Teflon IIS, the use of auto-insertion devices,
and methods that reduce tugging and pulling between the
infusion line and cannula have a larger impact on the glyce-
mic success of insulin pump therapy than is widely recog-
nized. Patients who experience episodes of unexplained
hyperglycemia or poor absorption with rapid swings in gly-
cemia or development of unexplained ketoacidosis should
vary the type of IIS they use to determine whether such acute
deteriorations are related to their infusion set. In critical situ-
ations, such as young children on small insulin doses or a
pregnant woman who requires consistent insulin delivery,
patients should be advised about the reliability of steel and
certain Teflon IIS once these are properly anchored by design
or with tape. Infusion sets and sites need to carefully moni-
tored. The goal for anyone who wears a pump would be to
have less than one IIS failure per month and preferably none.
Given the considerable price of each IIS, the frequency and
variety of failures, and the negative impact on glycemia, a
major redesign of IIS and auto-insertion devices is needed.
This commentary will hopefully initiate an open discussion
about the type of IIS that is used and its relevance for insulin
pump therapy. We encourage clinical studies that systemati-
cally evaluate the impact of different IIS in patients with diabe-
tes and prove or disprove our beliefs. It is important that both
types of IIS remain on the market as we believe that steel IIS
produce a clear clinical advantage for many pump wearers.
202 Journal of Diabetes Science and Technology 8(2)
Abbreviations
AACE, American Association of Clinical Endocrinologists; ADA,
American Diabetes Association; BMI, body mass index; CGM,
continuous glucose monitoring; DKA, diabetic ketoacidosis; DTM,
Diabetes Technology Meeting; EASD, European Association for
the Study of Diabetes; HbA1c, glycosylated hemoglobin; IIS, insu-
lin infusion set.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest
with respect to the research, authorship, and/or publication of this
article: LH holds shares in the Profil Institute for Metabolic Research,
Neuss, Germany, and the Profil Institute for Clinical Research, San
Diego, USA. JW is a consultant for a number of companies develop-
ing insulin pumps and infusion sets (including Roche Diagnostics).
He is employed by Advanced Metabolic Care and Research, which
conducts numerous studies in diabetes including insulin pumps,
infusion sets, and insulins. RR has no disclosures. LH and JW are
also consultants for a range of companies that develop new diagnos-
tic and therapeutic options for the treatment of diabetes.
Funding
The author(s) received no financial support for the research, author-
ship, and/or publication of this article.
Reference List
1.  Heinemann L, Krinelke L. Insulin infusion set: the Achilles heel
of continuous subcutaneous insulin infusion. J Diabetes Sci
Technol. 2012;6:954-964.
Downloaded from dst.sagepub.com at GEORGIAN COURT UNIV on March 6, 2015
2.  Renard E, Guerci B, Leguerrier AM, Boizel R. Lower rate of
initial failures and reduced occurrence of adverse events with a
new catheter model for continuous subcutaneous insulin infu-
sion: prospective, two-period, observational, multicenter study.
Diabetes Technol Ther. 2010;12:769-773.
3.  Patel PJ, Benasi K, Ferrari G, et al. Randomized trial of infu-
sion set function: steel versus teflon. Diabetes Technol Ther.
2014;16:15-19.
4.  Keith S, McVey E, Pettis RJ. High sensitivity occlusion detec-
tion using fluid pressure monitoring during basal insulin infu-
sion. Diabetes. 2013;62(suppl 1):A249-A250.
5.  Walsh J, Roberts R, Bailey T. Guidelines for insulin dosing in
continuous subcutaneous insulin infusion using new formulas
from a retrospective study of individuals with optimal glucose
levels. J Diabetes Sci Technol. 2010;4:1174-1181.
6. Reichert D, Weber D, Kaltheuner M, Scheper N, Faber-
Heinemann G, Heinemann L. Realität der Insulinpumpentherapie
in Diabetesschwerpunktpraxen: Daten von 1142 Patienten aus
40 diabetologischen Schwerpunktpraxen. Diabetes, Stoffwechsel
und Herz. 2013;22:367-375.
7. Colquitt JL, Green C, Sidhu MK, Hartwell D, Waugh N.
Clinical and cost-effectiveness of continuous subcutane-
ous insulin infusion for diabetes. Health Technol Assess.
2004;8:iii, 1-iii, 171.
8.  Liebner T, Holl R, Heidtmann B, et al. Insulinpumpenkatheter
Komplikationen im Kindes- und Jugendalter. Diabetologie und
Stoffwechsel. 2011;6:S52.
9.  Schmid V, Hohberg C, Borchert M, Forst T, Pfutzner A. Pilot
study for assessment of optimal frequency for changing cath-
eters in insulin pump therapy-trouble starts on day 3. J Diabetes
Sci Technol. 2010;4:976-982.
534798
research-article2014
DSTXXX10.1177/1932296814534798Journal of Diabetes Science and TechnologyAuthor Name

Journal of Diabetes Science and Technology

Corrigendum
2014, Vol. 8(3) NP1­
© 2014 Diabetes Technology Society
Reprints and permissions:
sagepub.com/journalsPermissions.nav
DOI: 10.1177/1932296814534798
dst.sagepub.com

We Need More Research and Better Designs for Insulin Infusion Sets. 2014;8(2);199-202. DOI: 10.1177/1932296814523882

In the editorial cited above, an error was made in the identification of an insulin infusion set (IIS). The IIS identified as Accu-
Chek® FlexLink Plus on page 199 was actually the Accu-Chek® FlexLink IIS. One sentence has been added on page 199:
“This study strongly suggests that infusion set design can reduce initial IIS failure rates and might be able to reduce overall
failure rates.” and one sentence has been removed from page 200: “Despite its promising results in this study, the FlexLink
Plus IIS was recalled in February 2011 due to excessive kinking and bending of the cannula.” These corrections are now avail-
able in the pdf version as of 11 April, 2014.