ORIGINAL ARTICLE
Insulin pump therapy in children with type 1 diabetes: analysis
of data from the SWEET registry
Agnieszka Szypowska
1
| Anke Schwandt
2,3
| Jannet Svensson
4
| Shlomit Shalitin
5,6
|
Roque Cardona-Hernandez
7
|GunForsander
8,9
| F ri da Sundberg
9
| CarineDeBeaufort
10,11
|
David Maahs
12
| Claudio Maffeis
13
| Stephen M.P. O‘Riordan
14
| Iveta Dzivite Krisane
15
|
Mauro Scharf
16
|Sofi a Castro
17
| Maia Konstantinova
18
| Barbora Obermannova
19
|
Kristina Casteels
20,21
|DamlaGökşen
22
|JúliaGalhardo
23
| Christina Kanaka-Gantenbein
24
|
Birgit Rami-Merhar
25
| Laszlo Madacsy
26
| the SWEET Study Group
1
Department of Paediatrics, Medical
University of Warsaw, Warsaw, Poland
2
Institute of Epidemiology and Medical
Biometry, ZIBMT, University of Ulm, Ulm,
Germany
3
German Center for Diabetes Research (DZD),
Munich-Neuherberg, Germany
4
Pediatric Department, Copenhagen
University Hospital, Herlev, Denmark
5
The Jesse Z and Lea Shafer Institute of
Endocrinology and Diabetes, National Center
for Childhood Diabetes, Schneider Children’s
Medical Center of Israel, Petach Tikva, Israel
6
Sackler Faculty of Medicine, Tel Aviv
University, Tel Aviv, Israel
7
Division of Endocrinology and Diabetes,
Hospital Sant Joan de Déu., Barcelona, Spain
8
Institute for Clinical Sciences, Sahlgrenska
Achademy, University of Gothenburg,
Gothenburg, Sweden
9
The Queen Silvia Childrens Hospital,
Sahlgrenska University Hospital, Gothe nburg,
Sweden
10
DCCP—Clinique pédiatrique de Luxembourg,
Luxembourg, Luxembourg
11
Department of Pediatric Endocrinology,
UZBrussels, Brussels, Belgium
12
Barbara Davis Center for Childhood
Diabetes, University of Colorado Denver
Aurora, Colorado, USA
13
Pediatric Diabetes and Metabolic Disorders
Unit & Regional Center for Pediatric Diabetes,
University Hospital, University of Verona,
Verona, Italy
14
Paediatrics Diabetes & Endocrine Unit,
Department of Paediatrics & Child Health,
Cork Uni versity Hospital, University College
Cork, Cork, Ireland
Background: Intensified insulin delivery using multiple daily injections (MDI) or continuous sub-
cutaneous insulin infusion (CSII) is recommended in children with type 1 diabetes (T1D) to
achieve good metabolic control.
Objective: To examine the frequency of pump usage in T1D children treated in SWEET (Better
control in Paediatric and Adolescent diabeteS: Working to crEate CEnTers of Reference) cen-
ters and to compare metabolic control between patients treated with CSII vs MDI.
Methods: This study included 16 570 T1D children participating in the SWEET prospective,
multicenter, standardized diabetes patient registry. Datasets were aggregated over the most
recent year of treatment for each patient. Data were collected until March 2016. To assess the
organization of pump therapy a survey was carried out.
Results: Overall, 44.4% of T1D children were treated with CSII. The proportion of patients with
pump usage varied between centers and decreased with increasing age compared with children
treated with MDI. In a logistic regression analysis adjusting for age, gender and diabetes dura-
tion, the use of pump was associated with both: center size [odd ratio 1.51 (1.47-1.55),
P < .0001) and the diabetes-related expenditure per capita [odd ratio 1.55 (1.49-1.61),
P < .0001]. Linear regression analysis, adjusted for age, gender, and diabetes duration showed
that both HbA1c and daily insulin dose (U/kg/d) remained decreased in children treated with
CSII compared to MDI (P < .0001).
Conclusions: Insulin pump therapy is offered by most Sweet centers. The differences between
centers affect the frequency of use of modern technology. Despite the heterogeneity of cen-
ters, T1D children achieve relatively good metabolic control, especially those treated with insu-
lin pumps and those of younger age.
KEYWORDS
childhood diabetes, continuous subcutaneous insulin infusion, CSII, MDI, multiple daily
injections
Received: 4 May 2016 Revised: 6 June 2016 Accepted: 9 June 2016
DOI 10.1111/pedi.12416
38 © 2016 John Wiley & Sons A/S.
Published by John Wiley & Sons Ltd
wileyonlinelibrary.com/journal/pedi Pediatric Diabetes October 2016; 17 (Suppl. 23): 38–45
15
Children’s University Hospital Children’s Endocrinology Centre, Riga Stradins
University, Riga, Latvia
16
Pediatric Endocrinology, Hospital Nossa Senhora Das Graças, Brazil
17
Child and Young Department, APDP—Diabetes, Lisbon, Portugal
18
Medical University—Clinic of Endocrinology, Diabetes and Genetics, Sofia
University Pediatric Hospital, Sofia, Bulgaria
19
Department of Pediatrics, University Hospital Motol and 2nd Faculty of
Medicine, Charles University in Prague, Prague, Czech Republic
20
Department of Pediatrics, University Hospitals Leuven, Leuven,
Belgium
21
Department of Development and Regeneration, KU Leuven,
Belgium
22
Faculty of Medicine Pediatric Endocrinology and Diabetes, Ege University,
İzmir, Turkey
23
Unit of Pediatric Endocrinology and Diabetes, Hospital Dona Estefânia,
Lisbon, Portugal
24
Diabetes Center, Division of Endocrinology, Diabetes and Metabolism
First Department of Pediatrics, Medical School National and Kapodistrian
University of Athens-Greece “Aghia Sophia” Children’s Hospital, Athens,
Greece
25
Department of Pediatrics and Adolescent Medicine of Medical University of
Vienna, Vienna, Austria
26
First Department of Pediatrics, Semmelweis University, Budapest,
Hungary
Conflict of Interest: The authors declare no potential conflict of interests.
Corresponding Author: Agnieszka Szypowska Assoc. Prof., MD, PhD, Medical
University of Warsaw, Department of Paediatrics, Żwirki i Wigury 63A, 02-091
Warsaw, Poland (agnieszka.szypowska@gmail.com).
1 | INTRODUCTION
Many clinical trials have demonstrated that tight blood glucose con-
trol reduces the risk of developing microvascular diabetes complica-
tions in all patients with type 1 diabetes (T1D). The beneficial effects
of reducing cardiovascular disease, retinopathy, nephropathy, and
neuropathy are well documented in the DCCT-EDIC study.
1,2
In order
to reduce these long-term effects of hyperglycemia, glucose control
should be optimized as early as possible in the course of type
1 diabetes.
3
An intensified insulin regimen is necessary in T1D patients to
achieve near-normal glucose control. However, only a relatively small
percentage of patients achieve these glycemic targets.
4
Continuous
subcutaneous insulin infusion (CSII) is a very effective treatment
modality, which is safe and widely used in children and adolescents
with T1D. There are many benefits for CSII therapy in children and
adolescents such as: optimum blood glucose control, reduction of
recurrent/severe and disabling hypoglycemia and improved quality of
life.
5
Insulin pump therapy allows a greater flexibility in insulin dosing
and meal planning, when compared with multiple daily injections
(MDI).
6,7
Meta-analyses demonstrate a reduction in glycated hemo-
globin A1c (HbA1c) and severe hypoglycemia with CSII therapy com-
pared to MDI.
8
Following the evidence-based demonstrations of the benefits of
CSII therapy, the first international consensus statement of insulin
pump indications and practice in children was created in 2007.
9
Insulin pump therapy in the pediatric age group has markedly
increased in the last decade (2007-2016). However, use of CSII
remains limited in some European countries. The main reasons are a
lack of funding by National Healthcare Insurance Systems, low num-
ber of trained physicians to deliver insulin pump therapy and a lack of
trained diabetes educators.
10
We hypothesized that children with T1D treated in SWEET
(Better control in Paediatric and Adolescent diabeteS: Working to
crEate CEnTers of Reference) centers with CSII had a better meta-
bolic control than those using MDI. The aims of this study are: to
examine the frequency of pump usage in T1D children treated in
SWEET centers and to compare the cross-sectional metabolic control
between patients treated with CSII vs. MDI.
2 | METHODS
The analysis is based on data from SWEET, a prospective, multicen-
ter, standardized diabetes patient registry. Currently, 48 diabetes care
institutions are participating in the SWEET project. For the data col-
lection, the SWEET project users use a heterogeneous environment:
Centers may use DPV (“Diabetes-Patienten-Verlaufskodumentation,”
https://sweet.zibmt.uni-ulm.de/software.php), DIAMAX, data down-
load from of existing registries or may use own local databases to col-
lect data. All centers longitudinally record demographic and clinical
data of patients with diabetes and transfer anonymized local data to
the SWEET database twice yearly. To increase the quality of data,
inconsistent/implausible data are reported back for verification or
correction after each data upload. Patients’ data are collected in
SWEET database at the Institute of Epidemiology and Medical Biom-
etry, University of Ulm, Ulm, Germany. The current analysis involved
19 European countries (39 centers) and 7 countries outside Europe.
Until March 2016, 28 713 patients with diabetes were available
in the database. Datasets with missing age, sex or diabetes duration
were excluded. For the present analysis, patients with type 1 diabetes,
aged 0-18 y and with a diabetes duration ≥1 y were included. The
exclusion criterion was missing insulin therapy (Figure 1). Datasets
were aggregated over the most recent year of treatment for each
patient. Three SWEET centers were excluded from analysis due to
incomplete data. The final cohort available for analysis was
n = 16 570 children with type 1 diabetes treated in 46 centers.
Diabetes control was assessed by glycated hemoglobin value
(HbAlc), which was measured locally in each center. In order to adjust
for differences between laboratories, multiple of the mean (MOM)
method was used to mathematically standardize HbA1c values to the
reference range of the Diabetes Control and Complications Trial
[DCCT, 21-43 mmol/mol (4-6%)].
11
Insulin pump usage was defined
as at least one visit with pump therapy. Body mass index standard
deviation score (BMI-SDS) was measured using the World Health
Organization (WHO) charts.
12
Further analysis was performed in three subgroups according to
age: 0-<6, 6-<12, 12-18 y. Diabetes duration was grouped into <5
and ≥5 y. Center size was defined as the number of patients in each
center.
SZYPOWSKA ET AL. 39
Severe adverse events were not evaluated due to incomplete
data concerning number of severe hypoglycemia [34% (5618/16 570)
of missing data] and diabetic ketoacidosis episodes [60% (9870/
16 570) of missing data]. Moreover, no data regarding ethnicity of
the participants were available in the SWEET database.
Children were divided into two groups depending on the method
of insulin therapy: insulin pump (CSII) or MDI. We evaluated the fre-
quency of insulin pump usage in the different age groups and com-
pared outcomes of diabetes control between children using insulin
pump therapy or MDI.
In order to collect data on the organization of pump therapy in
SWEET centers, each center was invited to complete an online sur-
vey. The questions included were the following: presence of National
Pump Registry, reimbursement of insulin pump therapy, initiation and
discontinuation of pump treatment and 24-h access to a pediatric
diabetes team member. Participants were able to mark only one
answer for each question. The survey was completed by 32 centers
(67%) and the results are seen in Table 1.
2.1 | STATISTICAL ANALYSIS
Wilcoxon or χ
2
-tests were performed to compare the demographic
characteristics and clinical outcomes between patients using CSII and
MDI therapy. The results are presented as median and interquartile
range (25 and 75 percentile), or numbers and percentages. Linear
regression model adjusted for age, gender and diabetes duration was
run to compare clinical outcomes between CSII and MDI groups. In
the linear regression analysis the mean and the standard error
(SE) were used to describe the differences. The Spearman rank corre-
lation and logistic regression model adjusted for center size and age,
sex and diabetes duration was used to analyze if center size had an
impact on pump use. The link between countries’ national health care
spending per person with diabetes (expressed in US dollars, R =2)
and use of pumps in children was evaluated using data from the IDF
Diabetes Atlas 2015
13
derived from the methods described by Zhang
et al.
14
Analyzing this link, we used Spearman rank correlation and
logistic regression model adjusted for country’s diabetes-related
health care spending per person, age, sex, and diabetes duration. For
logistic regression analysis, results are presented as an odds ratio with
95% confidence intervals. Statistical analysis was performed using
Statistical Analysis Software 9.4 (
SAS, SAS Institute Inc., Cary, NC,
USA). P-values <.05 were considered statistically significant.
3 | RESULTS
The current analysis included 16 570 participants (51.5% male).
Median age was “14 (Quartile 1; Quartile 3:10.6; 16.7) y”, diabetes
duration “5.3 (3.0; 8.5) y”, HbA1c “7.8 (7.1; 8.8)% [62 mmol/mol (54;
73) mmol/mol]”, BMI-SDS “0.55 (−0.1; 1.19)” and daily insulin dose
“0.87 (0.68; 1.1) U/kg/d”. We included 12 839 (77%) participants
from European countries and 3731 (23%) children from countries
outside Europe.
Results showed that 44.4% of T1D children and adolescents
were treated with CSII (varying from 0% to 90% per center). There
was comparable percentage of pump users in European countries
(45.8%) and in countries outside Europe (39.3%).
Center size did not influence pump use in the full group
(r = 0.25, P = .088). In a logistic regression analysis adjusting for age
categories, gender and diabetes duration categories, the use of pump
was associated with center size with an estimated odd ratio 1.51
(1.47-1.55), P < .0001. This means that for every 500 patient increase
in center size, the probability of a child being on a pump increases by
51%. There was no significant correlation between proportion of CSII
users in each center and HbA1c (r = −0.26, P = .084).
For the countries represented in this cohort, the mean diabetes-
related expenditure per person with diabetes per country in 2015
ranged from 95 to 1168 USD. When using a logistic regression
adjusting for demographics, pump usage was associated with the
SWEET database
28713
The group after exclusion
of patients with missing:
age, sex, diabetes duration
and diabetes duration > age
28349
Patients with type 1
diabetes
26378
Patients 18 years of age
23095
Patients with diabetes
duration
1 year
19851
Final study population after
exclusion of patients with
missing insulin therapy
16570
FIGURE 1 Selection of study population
40 SZYPOWSKA ET AL.
diabetes-related expenditure per capita [estimated odd ratio 1.55
(1.49-1.61), P <.0001] such that for every 4000 USD increase in
spending per capita, the probability of use of pump therapy increased
by 55% (Figure 2).
Comparison made between children in different age groups, trea-
ted with MDI or CSII therapy, are depicted in Table 2. The frequency
of pump use decreased with increasing age (52% in children aged
0-<6 y, 49% in youth aged 6-<12 y, and 42% in patients 12-18 y,
χ
2
= 89.9, P = .0001]. In the entire database there was no sex-related
difference in pump use.
In our sample, pump users were younger than injections users
[CSII: 13.5 (10;16.3) y vs MDI:14.4 (11; 17) y, P < .0001] and had
longer diabetes duration [CSII: 5.9 (3.5; 8.9) vs MDI: 4.9 (2.8; 8) y,
P < .0001].
The unadjusted HbA1c levels in the CSII users “7.7 [7; 8.5]%,
[60.7 (53; 69) mmol/mol]” was significantly lower than in the MDI
users “8.0 [7.2; 9.1]%, [63.9 (55; 76) mmol/mol]”, P < .001 and
HbA1c was lower in all age groups using CSII (Figure 3). Linear
regression model analysis, adjusting for age, gender and diabetes
duration showed that HbA1c remained higher in children treated with
MDI compared to CSII (P < .0001). Similar results were shown in lin-
ear regression analysis adjusting for gender and diabetes duration in
age-related subgroups. In all age groups, HbA1c was higher in chil-
dren treated with MDI compared to CSII (P < .0001).
Children using CSII therapy used lower total daily insulin dose
compared to youth treated with MDI [CSII: 0.83 (0.66; 1.02) U/kg/d
vs MDI: 0.9 (0.7; 1.13) U/kg/d, P < .0001) and daily insulin dose was
lower in both groups of youth 6-<12 y and 12-18 y old treated with
CSII (Figure 3). In a linear regression model adjusting for age, gender
and diabetes duration, children treated with MDI had higher daily
insulin dose (U/kg/d) than subjects using CSII (P < .0001).
The unadjusted BMI-SDS was higher in MDI group “0.51 (−0.14;
1.18)” compared to CSII “0.58 [−0.05, 1.2]”, P = .000. The BMI-SDS
was higher only in children with CSII and 6->12 y of age (Figure 3).
Linear regression analysis adjusting for diabetes duration showed a
similar BMI-SDS in both treatment groups (P = 0.399).
4 | DISCUSSION
Current results showed that insulin pump therapy was used in 44.4%
children aged 0-18 y treated in the SWEET centers. The data repre-
sents only the frequency of pump therapy in the SWEET centers, and
does not necessarily reflect the management of diabetes by country.
Our survey showed that national data on pump use are present only
in 38% of countries represented in our SWEET group. Owing to lack
of national data, the analysis of an overall use of CSII in SWEET
countries was not possible.
The prevalence of CSII therapy usage has increased in many
countries, but there are still differences in CSII usage among coun-
tries.
10
The overall use of insulin pumps in patients 0-18 y was 74%
(2011) in Slovenia,
7
and 58-65% (2015) in United States.
15
The data
from three large registries of pediatric type 1 diabetes patients ana-
lyzed in 2015 showed that insulin pump was used by: 41% of chil-
dren in the German/Austrian Prospective Diabetes Follow-up
Registry (DPV), 47% of youth in the US T1D Exchange (T1DX) and
14% of children in the English/Welsh National Paediatric Diabetes
Audit (NPDA).
16
Data from the Swedish National Diabetes Register
reported that in 2013 one out of every four women and one out of
every five men used insulin pump treatment. Over half of all Swedish
children with T1D are treated with CSII.
17
The percentage of Danish
children on CSII increased to approximately 50% in 2011.
18
French
national survey performed in 2007, which represented 60–75% of
the national estimated population of T1D children aged 0-18 y,
showed that 15% of children used insulin pump.
19
TABLE 1 Survey results of the organization of pump therapy in
SWEET centers.
Presence of National registers of pump use:
1. Yes: 38%
2. No: 62%
Cost of CSII covered by:
1. Health care system: 75%
2. Shared by heath care and family: 16%
3. Fully by family: 9%
Insulin pump initiation
1. Inpatient: 31
2. Outpatient: 47
3. Depends on circumstances: 22
How soon after diagnosis a patient is started in insulin pump
1. At diagnosis (no injection at all): 6.5%
2. Within the initial hospitalization/education phase: 6.5%
3. After a minimum of 6 mo after diagnosis: 9%
4. Individualized approach: 78%
24 h access to a pediatric diabetes team member
1. Yes: 81%
2. No: 19%
Minimum age criteria for pump initiation
1. Yes: 3%
2. No: 97%
Discontinuing off pump therapy when a family fails to achieve
satisfactory control
1. Yes: 44%
2. No: 22%
3. No clear policy: 34%
Abbreviations: CSII, continuous subcutaneous insulin infusion; SWEET,
Better control in Paediatric and Adolescent diabeteS: Working to crEate
CEnTers of Reference
FIGURE 2 Correlation between proportion of patients treated with
continuous subcutaneous insulin infusion (CSII) in each center and a
country’s diabetes related health care spending per person with
diabetes expressed in US dollars (R2_DM).
SZYPOWSKA ET AL. 41
Different factors may influence frequency of pump use in a cen-
ter. Insulin pump therapy is more expensive than injections.
20
There-
fore, socioeconomic status of families influences the choice of the
method of diabetes management in T1D children, especially when
there is inadequate or no reimbursement of these devices by national
healthcare systems or insurance companies.
5,21
Our survey showed,
that in one third of the centers, costs of CSII therapy are shared by
insurance and family or fully covered by family. Moreover, we noted
a significant correlation between the proportion of pump users in
each center and the diabetes-related health care expenditure per per-
son with diabetes per country. The association between low socioec-
onomic factors and low frequency of CSII usage was confirmed by
other research.
22
In T1D therapy, systematic reviews of cost-
effectiveness showed a superiority of CSII over MDI. CSII therapy
was associated with an improvement in both global life expectancy
and quality-adjusted life expectancy. This was achieved by a decrease
in HbA1c and by a lower number of hypoglycemia episodes in this
group.
23
Our results indicate differences between centers. The correlation
did not show any impact of center size on pump usage, but after
adjustment for age, sex and diabetes duration we noted that insulin
pump therapy was more often used in larger centers. Potential rea-
sons for lower CSII usage in smaller sized center may include organi-
zation of diabetes care, insufficient number of physicians specialized
in pump therapy or low number of trained diabetes educators.
24
.
However, in France no significant variation in the rate of CSII was
observed according to the size of the centers, although the frequency
of pump use remained extremely variable, ranging from 1.3 to 53% of
T1D children.
19
Diabetes education of T1D children and their families
is an essential part of diabetes care with effects on diabetes out-
come.
25
It is well known from clinical practice that education and
training for patients using CSII therapy is more time consuming, com-
pared to education for those using MDI and this influences work
organization in diabetes centers. The results of our survey also
emphasize individual approach to pump therapy and work organiza-
tion in different centers. As seen in Table 1 most centers use an indi-
vidual approach when starting pump, and the majority of pump starts
are based in the outpatient clinic. A 24 h hotline with access to
experts in pump therapy is common, whereas guidelines to discon-
tinue insulin pump therapy in non-compliant patients was limited.
Ideally multidisciplinary teams supporting pump users, should contain
a critical mass of staff (doctors, nurses, dieticians, and diabetes edu-
cators) with appropriate ongoing education in CSII therapy.
26
Our
results should be taken with caution because we have not performed
any in depth analysis assessing differences in the organization of
work in diabetes centers.
The age-related frequency of CSII use differs across countries.
Pump therapy is commonly used in children diagnosed before 6 y of
age and is associated with better long-term metabolic control.
27
In
our group, CSII therapy was more frequently used in the youngest
children less than 6-y-old (52%), than in teenagers over 12 y (42%).
Similar results were reported in the DPV register.
16
In this study,
74% of children less than 6-y-old received insulin pump therapy,
compared with 35-40% of older participants. The highest frequency
of pump use in the youngest children was also noted in the NPDA
register.
15
Conversely, the highest frequency of pump users was
shown in teenagers in the T1D register.
16
Observed differences may
indicate that different government funding arrangements for insulin
pumps (or diabetes-related health care expenditure), may play a sig-
nificant role in those who are offered pump therapy vs those who
are not.
Our results showed no gender-related difference in pump use
but further subgroup analysis indicated statistically lower use of insu-
lin pump in boys less than 6-y-old. One may speculate that young
boys are busier than girls and they may be inclined to remove infu-
sion sets during vigorous physical activities. The parents may also
think these young boys cannot cope with a pump before they try
it. Interestingly other studies also noted that boys were treated with
a pump less often compared to girls.
16
In our study, children treated with CSII reported better metabolic
control expressed by lower HbA1c. Similar results were noted in all
three age groups. After adjustment for age, sex and diabetes dura-
tion, HbA1c was 0.5% lower in the CSII group compared with the
MDI group. Both groups of younger children (<6 and 6-<12 y), using
TABLE 2 Comparison between children in different age groups treated with MDI or CSII therapy.
1
0-<6 y 6-<12 y 12-18 y
Type of insulin
therapy
MDI
median
[Q1;Q3]
CSII
median
[Q1;Q3]
P
value
MDI
median
[Q1;Q3]
CSII
median
[Q1;Q3]
P
value
MDI
median
[Q1;Q3]
CSII
median
[Q1;Q3]
P
value
Number of patients 435 464 — 2458 2358 — 6321 4534 —
Female/male 184/251 231/233 .049 1193/1265 1165/1193 .582 3054/3267 2217/2290 .582
Diabetes duration (y) 2.1 [1.5; 3.1] 2.2 [1.5; 3.2] .582 3.6 [2.1; 5.6] 4.7 [2.9; 6.7] .0001 5.9 [3.4; 9.4] 7.3 [4.6;
10.5]
.0001
HbA1c (%; mmol/mol) 7.9 [7.2; 8.6]
63 [55; 71]
7.4 [6.8; 8.0]
57 [51; 64]
.0001 7.8 [7.0;
8.5)]
62 [53; 69]
7.4 [6.8; 8.1]
57 [51; 65]
.0001 8.2 [7.3; 9.4]
66 [56; 79]
7.8 [7.2; 8.8]
62 [55; 73]
.0001
Daily insulin dose
(U/kg/d)
0.73
[0.59; 0.90.
0.72
[0.58; 0.86.
.471 0.83
[0.66; 1.02.
0.76
[0.61; 0.91.
.0001 0.95
[0.75; 1.18.
0.89
[0.71; 1.08.
.0001
BMI-SDS 0.79
[0.21; 1.45.
0.85
[0.34; 1.51.
.143 0.45
[−0.14; 1.14.
0.54
[−0.05, 1.16.
.023 0.52
[−0.16; 1.19.
0.57
[−0.09; 1.17.
.103
Abbreviations: BMI, body mass index standard deviation score; CSII, continuous subcutaneous insulin infusion; MDI, multiple daily injections; n.s., not
significant.
1
Given are median with quartile or proportions.
42 SZYPOWSKA ET AL.
