O R I G I N A L A R T I C L E
Comparing continuous subcutaneous insulin infusion and
multiple daily injections in children with Type 1 diabetes in
Sweden from 2011 to 2016
—A longitudinal study from the
Swedish National Quality Register (SWEDIABKIDS)
Anna-Lena Fureman
1|
Mikael Lilja
2|
Torbjörn Lind
3|
Stefan Särnblad
4|
Marie Bladh
5|
Ulf Samuelsson
61
Department of Clinical Sciences, Pediatrics, Östersund Hospital, Umeå University, Umeå, Sweden
2
Department of Public Health and Clinical Medicine, Unit of Research, Education and Development, Östersund Hospital, Umeå University, Umeå, Sweden
3
Department of Clinical Sciences, Pediatrics, Umeå University, Umeå, Sweden
4
Department of Pediatrics, Örebro University, Hospital and School of Medical Sciences, Örebro University, Örebro, Sweden
5
Department of Obstetrics and Gynecology and Biomedical and Clinical Sciences, Linköping University Hospital, Linköping University, Linköping, Sweden
6
Division of Pediatrics, Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden Correspondence
Anna-Lena Fureman, Department of Clinical Sciences, Pediatrics, Östersund Hospital, Umeå University, Umeå, Sweden.
Email: annalena.fureman@regionjh.se Funding information
Department or Research and Development Region Jämtland Härjedalen; Stiftelsen Samariten; Thuringstiftelsen
Abstract
Objective: This study aimed to compare metabolic control measured as hemoglobin
A1c (HbA1c), the risk of severe hypoglycemia, and body composition measured as
body mass index standard deviation scores (BMI-SDS) in a nationwide sample of
chil-dren and adolescents with Type 1 diabetes with continuous subcutaneous insulin
infusion (CSII) and multiple daily injections (MDI), respectively.
Research Design and Methods: Longitudinal data from 2011 to 2016 were extracted
from the Swedish National Quality Register (SWEDIABKIDS) with both
cross-sectional (6 years) and longitudinal (4 years) comparisons. Main end points were
changes in HbA1c, BMI-SDS, and incidence of severe hypoglycemia.
Results: Data were available from 35,624 patient-years (54% boys). In general,
HbA1c decreased approximately 0.5% (2
–5 mmol/mol) from 2011 to 2016
(ptrend
< 0.001) and the use of CSII increased in both sexes and all age groups. Mean
HbA1c was 0.1% (0.7
–1.5 mmol/mol) lower in the CSII treated group. Teenagers,
especially girls, using CSII tended to have higher BMI-SDS. There was no difference
in the number of hypoglycemias between CSII and MDI over the years 2011
–2016.
Conclusions: There was a small decrease in HbA1c with CSII treatment but of little
clinical relevance. Overall, mean HbA1c decreased in both sexes and all age groups
without increasing the episodes of severe hypoglycemia, indicating that other factors
than insulin method contributed to a better metabolic control.
K E Y W O R D S
continuous subcutaneous insulin infusion, metabolic control, multiple daily injection, HbA1c, hypoglycemia
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2021 The Authors. Pediatric Diabetes published by John Wiley & Sons Ltd.
increases the risk of severe complications later in life. Even people with well controlled Type 1 diabetes have a doubled risk of death from any cause as well as from cardiovascular causes compared to the general population, and the risks are several times higher among patients with poor glycemic control.4
Traditionally, Type 1 diabetes has been treated with subcutane-ous insulin injections 5–6 times daily. More than 40 years ago contin-uous subcutaneous insulin infusion (CSII) was introduced as a way to achieve a better metabolic control.5,6In randomized controlled trials7 CSII among adolescent and adults with Type 1 diabetes has been demonstrated to improve metabolic control by lowering hemoglobin A1c (HbA1c) and decrease complications.8 Also, in children CSII is
considered the most physiological way to administer insulin today.9In young children, CSII reduces the incidence of severe hypoglycemic events10 which is an important feature of its use in this age group.11,12
However, in terms of efficacy, studies show varying results regarding the ability of CSII to sustain lower HbA1c levels after the first 6–12 months of use.13-15Smaller studies previously carried out in Sweden have given somewhat different results concerning the effi-cacy of CSII. In 2008, Skogsberg et al. were not able to show a better metabolic control with CSII compared to multiple daily injections (MDI).16In another follow-up of patients who started CSII treatment in the years 2005–2009 compared to matched controls, CSII treat-ment improved the metabolic control during the first year but not thereafter.17The group with CSII treatment had fewer incidents of
severe hypoglycemia during follow-up but the number of ketoacidosis events increased.
A better glucose metabolic control may contribute to a normal body mass index (BMI), but there is no evidence from previous studies that CSII treatment can accomplish this.13,18,19
The aim of this retrospective study using real life data was to compare CSII treatment with MDI treatment in terms of metabolic control, BMI, and prevalence of severe hypoglycemia in a large national cohort of children and adolescents with Type 1 diabetes.
2
|
M E T H O D
The study population consisted of all children and adolescents from 0 to 17 years of age with Type 1 diabetes, as reported in the Swedish National Diabetes Register for Pediatric Diabetes in Sweden, SWEDIABKIDS, hereinafter referred to as the Register. The period 2011–2016 was selected because of the general decrease in mean HbA1c that occurred during these years while the frequency of CSII
ter is continuously validated against the Swedish Prescribed Drug Register for incident cases.1 The Prescribed Drug Register is
maintained by the National Board of Health and Welfare and covers all redeemed medical prescriptions since July 2005.20
Since 2008, the Register has been web-based and clinical data are recorded from every patient visit to the out-patient pediatric clinics, which is usually three to four times per year per patient. In 2016, data from 7310 unique patients up to 17 years of age were recorded in the Register from 27,186 visits21and on average 6915 patients per year were recorded from 2011 to 2016. The target for HbA1c for children and adolescents with diabetes Type 1 was <7% (<53 mmol/mol) during 2011–2016 but was revised to <6.5% (<48 mmol/mol) in 2017. The indications for CSII treatment during this period was difficulties achieving a good metabolic control with MDI treatment, eating disorders, frequent episodes with severe hypoglycemia, dawn phenomenon in teenagers and young age with difficulties administering low doses of insulin with MDI. The start of CSII treatment was often done with the patient and family staying at the hospital for 2–3 days. Some diabetic teams invited a representative from the CSII-industry to help out with the educa-tion to the patients and families while other teams educated the families themselves. During this period CSII treatment to small chil-dren became more frequent and the method to count carbohy-drates was nationally introduced around 2011.
2.2
|
Data measurement
At the different hospitals using the Register, analyses of HbA1c are performed according to different laboratory methods but all are qual-ity assured through Equalis (External Qualqual-ity Assurance in Laboratory Medicine in Sweden),22which makes it possible to compare HbA1c
values across different clinics. HbA1c is presented as International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) units (mmol/mol) and National Glycohemoglobin Standardization Program (NGSP) units. In comparison to NGSP, an HbA1c value of 53 mmol/ mol corresponds to 7.0% and 75 mmol/mol to 9.0%.23The conversion of HbA1c from IFCC (mmol/mol) to NGSP (%) was done using the tool provided by the National Glycohemoglobin Standardization pro-gram.24 Severe hypoglycemia was defined as blood glucose of
<70 mg/dl (3.9 mmol/L) together with an event involving seizure or unconsciousness. Anthropometrical data (weight and height) were col-lected during each clinical visit. BMI standard deviation scores (BMI-SDS) were calculated during each visit using the BMI reference values for Swedish children.25
2.3
|
Cross-sectional cohort
Anonymized individual-level data was extracted from the Register from January 1, 2011 to December 31, 2016. For each year and for each patient, data on age, sex, HbA1c, diabetes duration, insulin deliv-ery method, BMI-SDS, and severe hypoglycemic events were col-lected. The cumulative number of severe hypoglycemic events was collected from the last visit for that year for every patient, so that each number corresponds to one individual having at least one or more severe hypoglycemic events during that year. Patients that had changed insulin delivery method during a year were excluded for that year; this group consisted of approximately 10% of the patients. The recorded number of ketoacidosis events were very few and since it was not routine at all clinics to register them during these years, we deemed that it was not reliable to use this data.
2.4
|
Longitudinal cohort
A cohort was also created consisting of all patients from 0 to 15 years of age and with a diabetes duration of >6 months in 2013, based on data extracted from the Register from January 1, 2013 to December 31, 2016. The patients needed to have had at least one registered HbA1c value for each of the 4 years from 2013 to 2016 and they had to have remained on the same insulin delivery method, that is, CSII or MDI, and not become >18 years of age throughout the follow-up period. An initial group of 6674 unique patients was identified but 1411 patients were excluded due to missing HbA1c values. Another 868 patients chan-ged their insulin delivery method, and 467 patients were also excluded as data were missing on their insulin delivery method during this period. This left 3928 patients in the final cohort. For each year and for each patient, data on age, sex, A1c, diabetes duration, insulin delivery method, BMI-SDS, and severe hypoglycemic events were collected.
2.5
|
Drop-out analyses
2.5.1
|
The group missing HbA1c values in the
cohort: (data not shown)
There were no differences in sex or BMI-SDS distribution in the group missing HbA1c values compared to the group with HbA1c values in the cohort. In the group, 74% received MDI treatment, they had a shorter diabetes duration and there were very few patients with severe hypoglycemia. No systematic skewness was found in the group with missing HbA1c values. However, younger patients were over-represented thus having a shorter diabetes duration.
2.6
|
Statistical analysis
The annual data was stratified according to insulin treatment (CSII vs. MDI), age groups (0–6, 7–12, 13–17 years old), and HbA1c groups (<6.5 [<47.9 mmol/mol], 6.5–7.3 [48.0–55.9 mmol/mol], 7.3–8 [56.0– 63.9 mmol/mol], 8–8.7 [64.0–71.9 mmol/mol], and > 8.7 [>72 mmol/mol]%). The age groups were based on major clinical differ-ences in developmental psychology, ability to self-care, and bodily maturation and hormonal impact on the glucose metabolic control. As girls tend to have a higher HbA1c than boys,26a sex stratified analysis
was performed. The longitudinal data were stratified according to insulin treatment (CSII vs. MDI) and sex. All statistical analyses were performed using IBM SPSS version 24 with independent two-sample t-tests for normally distributed data and one-way ANOVA for calcu-lating p-trends. The multivariate analyses included: (a) multiple logistic regression in which the dependent variable was set to insulin delivery method and the independent variables were HbA1c, BMI-SDS, dura-tion of diabetes, and sex, and (b) multivariable linear regressions with two models, one where the dependent variable was set to HbA1c and F I G U R E 1 Mean HbA1c with 95% confidence intervals and frequency of device use for the years 2011–2016 for multiple daily injections (MDI) and continuous subcutaneous insulin infusion (CSII) among boys and girls, stratified for age groups: 0–6, 7–12, and 13–17 years old. *p < 0.05, **p < 0.01, ***p < 0.001
the independent variables were sex, BMI-SDS, and duration of diabe-tes and the other model where the dependent variable was set to BMI-SDS and the independent variables were HbA1c, sex, and dura-tion of diabetes. Statistical significance was defined as p < 0.05. A lin-ear mixed model was used to study HbA1c, BMI-SDS, and severe hypoglycemic events of the cohort.
3
|
R E S U L T S
Baseline data are presented in Table S1. In total, data from 35,624 patient-years (54% boys) of patients from 0 to 17 patient-years of age were available.
3.1
|
Cross-sectional data
3.1.1
|
Choice of insulin delivery method
Between 2011 and 2016, the frequency of CSII treatment increased from 41% to 60% in girls and from 35% to 56% in boys (ptrendfor
both <0.001) (Figure 1, Table S1). Patients using CSII were generally younger (Figure 2(B), Table S2) but had a longer diabetes duration compared with those treated with MDI, except for the HbA1c group >8% (64 mmol/mol) where there was no difference in patient age between the different insulin delivery methods (Figure 2(B), Table S3).
F I G U R E 2 (A) Mean BMI-SDS with 95% confidence intervals and frequency of sex for 2011–2016 for multiple daily injections (MDI) and continuous subcutaneous insulin infusion (CSII), stratified for HbA1c groups: <6.5, 6.5–8, and > 8%. BMI, body mass index (kg/m2); SDS, standard
deviation score; *p < 0.05, **p < 0.01, ***p < 0.001. (B) Mean age with 95% confidence intervals and frequency of sex for 2011–2016 for multiple daily injections (MDI) and continuous subcutaneous insulin infusion (CSII) stratified for HbA1c groups: <6.5, 6.5–8, and > 8%. *p < 0.05,
**p < 0.01, ***p < 0.001. (C) Mean diabetes duration with 95% confidence intervals and frequency of sex for 2011–2016 for multiple daily injections (MDI) and continuous subcutaneous insulin infusion (CSII) stratified for HbA1c groups: <6.5, 6.5–8, and > 8%. *p < 0.05, **p < 0.01, ***p < 0.001. (D) Frequency of hypoglycemias and frequency of sex for 2011–2016 for multiple daily injections (MDI) and continuous subcutaneous insulin infusion (CSII) stratified for HbA1c groups: <6.5, 6.5–8, and >8%. *p < 0.05, **p < 0.01, ***p < 0.001
3.1.2
|
Metabolic control
An overall decrease in HbA1c of around 0.5% (2–5 mmol/mol) was seen in the different age and sex groups from 2011 to 2016 (Figure 1, Table S1). There was no difference in mean HbA1c between the patients using CSII or MDI except for the group with HbA1c >8.7% (72 mmol/mol) in which the group using CSII had a lower HbA1c com-pared to the MDI group; (CSII: 9.2% [77.4 mmol/mol] vs. MDI: 9.4% [79.8 mmol/mol], p < 0.01) (Table S3) and in boys 13–17 years where the group using CSII had a higher HbA1c; (CSII: 7.7% [60.3 mmol/mol] vs. MDI: 7.5% [58.2 mmol/mol], p < 0.001) (Figure 1, Table S2).
3.1.3
|
BMI-SDS
There were no changes in BMI-SDS over the period in any of the age groups. Patients using CSII tended to have a higher BMI-SDS and, for example, in teenage girls, the mean BMI-SDS with CSII treatment was 0.96 during the period 2011–2016 (Figure 2(A), Table S2).
3.1.4
|
Prevalence of hypoglycemia
There was no difference in the recorded number of hypoglycemic events between the treatment groups from 2011 to 2016 despite the general decrease in HbA1c (Figure 3, Table S2).
3.2
|
Longitudinal data
3.2.1
|
Choice of insulin delivery method
Patients using CSII were younger than the patients using MDI, p < 0.001 (Table 1).
3.2.2
|
Metabolic control
In general, there was an age-dependent increase in mean HbA1c simi-lar for both insulin delivery methods (i.e., CSII or MDI). There was no
difference in mean HbA1c between the insulin delivery methods (Table 1).
3.2.3
|
BMI-SDS
There was no difference in the BMI-SDS between the insulin delivery methods in either of the sexes (Table 1).
3.2.4
|
Prevalence of hypoglycemia
There was no difference in the number of severe hypoglycemic events between the patients using CSII or MDI (Table 1).
3.2.5
|
Change in insulin delivery method
More girls (54%) than boys changed their insulin delivery method during the years 2013–2016 but the mean age (11 years) was the same. Most changes were from MDI to CSII treatment. The mean HbA1c for those who changed insulin delivery method increased for every year from 2013 to 2016, from 7.7% (60.3 mmol/mol) in 2013 to 7.9% (63.0 mmol/mol) in 2016, p < 0.001 and the increase was similar to the group not changing insulin delivery method. There was no difference in the BMI-SDS or in the number of severe hypoglycemic events between those changing and not changing insulin delivery method, (further data not shown).
Table 2 presents the multivariate logistic regression analysis with insulin delivery method as the outcome. Overall, the use of CSII was associated with female sex, higher BMI-SDS and long diabetes duration. In 2011, the likelihood of CSII was 30% higher in girls compared to boys (OR 1.30 [CI 1.16–1.45], p = 0.001), which decreased to 22% by 2016 (OR 1.22 [CI 1.10–1.36], p= 0.001).
Tables 3 and 4 presents the multivariable linear regression analy-sis with HbA1c and BMI-SDS as the outcomes, each outcome modeled separately and stratified by year. The multivariable regres-sion of HbA1c indicated that a high HbA1c was associated with higher BMI-SDS, female sex, and long diabetes duration. In 2011, HbA1c was approximately 0.1% lower with CSII treatment compared to MDI F I G U R E 3 Mean HbA1c with 95% confidence intervals and prevalence of severe hypoglycemia for the years 2011–2016 for multiple daily injections (MDI) and continuous subcutaneous insulin infusion (CSII), stratified for age groups: 0–6, 7–12, and 13–17 years old. Severe hypoglycemia defined as events with unconsciousness or seizures. *p < 0.05, **p < 0.01, ***p < 0.001
TAB L E 1 The coh ort gr oup 201 3– 201 6 st ratified for sex and insul in meth od Longitudinal data in the cohortgroup 2013 2014 2015 2016 MDI CSII MDI CSII MDI CSII MDI Boys n (%) 1130 (54.9) * 930 (45.1) 1130 (54.9) 930 (45.1) 1130 (54.9) 930 (45.1) 1130 (54.9) Age (years [range]) 11.3 (2.2;15.8) 10.6 (1.5;15.8) *** 12.5 (3.4;16.8) 11.6 (2.4;16.8) *** 13.6 (4.7;17.8) 12.6 (3.4;17.7) *** 14.8 (5.5;18.0) HbA1c (%[CI]) 7.3 (7.3;7.4) 7.5 (7.4;7.6) *** 7.4 (7.3;7.5) 7.4 (7.4;7.5) 7.5 (7.4;7.6) 7.6 (7.5;7.6) 7.6 (7.5;7.7) HbA1c (mmol/mol[CI]) 56.5 (55.8;57.2) 58.5 (57.8;59.1) *** 57.3 (56.5;58.0) 57.9 (57.4;58.5) 58.6 (57.9;59.4) 58.7 (58.1;59.2) 59.7 (58.9;60.4) BMI SDS (CI) 0.54 (0.47;0.60) 0.61 (0.54;0.68) 0.56 (0.49;0.64) 0.62 (0.56;0.69) 0.57 (0.49;0.65) 0.60 (0.54;0.66) 0.60 (0.52;0.69) Hypolycemic events n (%) 38 (3.4) 36 (3.9) 25 (2.2) 35 (3.8) 21 (1.8) 39 (4.2) *** 15 (1.3) Girls n (%) 1013 (54.2) 855 (45.8) 1013 (54.2) 855 (45.8) 1013 (54.2) 855 (45.8) 1013 (54.2) Age (years [range]) 11.0 (2.5;15.8) 10.6 (0.8;15.7) 12.2 (3.2;16.7) 11.6 (1.8;16.7) *** 13.4 (4.7;17.7) 12.4 (2.6;17.8) *** 14.6 (5.3;18.0) HbA1c (% [CI]) 7.4 (7.3;7.5) 7.5 (7.4;7.6) 7.5 (7.4;7.6) 7.5 (7.4;7.5) 7.7 (7.6;7.8) 7.5 (7.5;7.6) ** 7.8 (7.7;7.9) HbA1c (mmol/mol [CI]) 57.4 (56.6;58.1) 58.5 (57.8;59.2) 58.5 (57.6;59.4) 58.2 (57.6;58.8) 60.9 (59.9;61.9) 58.9 (58.3;59.5) ** 61.6 (60.6;62.6) BMI SDS (CI) 0.58 (0.51;0.66) 0.69 (0.63;0.75) 0.69 (0.61;0.77) 0.76 (0.7;0.82) 0.79 (0.71;0.87) 0.82 (0.76;0.88) 0.84 (0.75;0.92) Hypoglycemic events n (%) 27 (2.7) 32 (3.7) 30 (3.0) 28 (3.3) 16 (1.6) 31 (3.6) ** 14 (1.4) Note: Values presented are numbers, percentages, means with CI. Hypoglycemic events were defined as unconsciousness or seizures. Abbreviations: BMI, body mass index (kg/m 2); CI, confidence intervals; CSII, continuous subcutaneous insulin infusion; MDI, multiple daily injections; SDS, standard deviation score. *p < 0.05. ** p < 0.01. *** p < 0.001.
treatment. Furthermore, the multiple linear regression of BMI showed that a high BMI was associated with CSII treatment, higher HbA1c, and female sex. In 2011, BMI was 0.1 SDS higher in CSII treated patients compared to MDI treated patients. There was no interaction effect between sex and insulin method. The model was also made with calendar year with the same result. This finding was due to the association between diabetes duration and insulin delivery method. The diabetes duration among CSII treated subjects decreased with a year from 2011 to 2016 and during the same time reduced their HbA1c by 0.6% (4–7 mmol/mol) while diabetes duration and HbA1c have remained constant over the years among those using MDI.
4
|
D I S C U S S I O N
Our main findings were a minute decrease in HbA1c in CSII treated patients and a general, severalfold larger reduction in the mean HbA1c irrespective of treatment from 2011 to 2016, without any increase in severe hypoglycemia. The difference in HbA1c between CSII and MDI treatment was only 0.1% (0.7–1.5 mmol/mol) and of lit-tle clinical relevance. During the same period the use of CSII treat-ment increased by more than 50%. This indicates that the CSII regimen per se was no better in improving the metabolic glucose con-trol than MDI at the group level. A meta-analysis of randomized controlled studies of pediatric cohorts using CSII have shown a mod-est HbA1c reduction27but most of these studies had a duration less
than 3 months. Our study is based on real-life data which might explain the different results compared to RCT:s. Our finding corrobo-rates some previous studies,16,17,28which demonstrated no difference in glucose metabolic control between CSII and MDI. An Israeli study29
following children and young adults for 6 years after CSII initiation found a decrease in HbA1c similar to ours and also with no increase in the rate of hypoglycemia. Recently, episodes of severe hypoglycemia have been demonstrated to be associated with a progressive increase in HbA1c in children and adolescents with Type 1 diabetes.30It is thus important to avoid these events in order to avoid long-term diabetes complications.
However, our findings contradict a Danish study31 where they
followed patients for more than 5 years to show that children and adolescents in all age groups achieved an improved metabolic control on CSII treatment. There was a parallel rise in HbA1c in both MDI and CSII treated patients during follow-up,31similar to other studies17,32
which showed a significant improvement in HbA1c only in the first
year after starting CSII treatment but not thereafter. Possible explana-tions to our findings compared to other studies could be different indications for CSII treatment and different approaches to education in the multidisciplinary teams which can lead to different education of patients and families. As we show, the use of CSII has increased dur-ing recent years and at a later follow-up in 2018, 65% of children and adolescents <18 years of age were treated with CSII.1The usage of CSII varies considerably between different counties and also between pediatric clinics with frequencies ranging from 19% to 91%.1These differences remain largely unexplained and there is no correlation between mean HbA1c and frequency of usage of CSII in the pediatric clinics.1The costs for insulin pumps in Sweden are covered by the
National Healthcare system and the patient and their families do not have any costs for neither insulin nor the technology.
There is no standardized curriculum in Sweden for how to start CSII treatment in a patient, so this procedure varies between different pediatric clinics. National guidelines based upon ISPAD guidelines 2014 has been developed with indications for CSII treatment. A com-parable reduction in mean HbA1c between the two insulin delivery methods indicates that other factors must be of importance for the improvement of glucose metabolic control.
A second result was the higher BMI-SDS in teenage girls using the CSII regimen. This is consistent with the findings of Ibfelt et al.33 and Birkebaek et al.34One explanation could be that girls with a high
BMI-SDS might have an insulin resistance, which can lead to a higher HbA1c prompting healthcare practitioners to provide them with CSII treatment in an attempt to improve their metabolic control. If this is the case, those involved in health care delivery to children with Type 1 diabetes should explore ways to prevent avoidable increases in BMI through support to patients and families, for example healthy food recommendations and encouragement to do physical activity. Since Samuelsson et al.26showed that teenage girls with Type 1 diabetes
have an increased risk of complications in adulthood this emphasizes the importance of helping this group achieve a better glucose metabolic control without negatively affecting other aspects of their health, such as increasing the risk of obesity. A somewhat lower HbA1c in the group with the highest HbA1c in patients using CSII may indicate treatment benefits, but these results need to be interpreted with caution. We also found that HbA1c was higher in teenage boys using CSII regimen. One explanation could be that boys with difficulties achieving a satisfying gly-cemic control resulting in high HbA1c values are provided with CSII treatment in an assessment to improve their metabolic control. If so, this indicates that the multidisciplinary teams need to find other ways to help T A B L E 2 OR for CSII treatment 2011–2016
2011 2012 2013 2014 2015 2016 OR 95 CI OR 95 CI OR 95 CI OR 95 CI OR 95 CI OR 95 CI HbA1c 0.99 0.99–0.99 0.99 0.99–1.00 0.99 0.99–1.00 0.99 0.99–1.00 0.99 0.98–0.99 0.99 0.98–0.99 BMI-SDS 1.13 1.07–1.19 1.11 1.05–1.16 1.09 1.04–1.15 1.07 1.02–1.12 1.05 1.00–1.10 1.06 1.01–1.11 Duration 1.16 1.14–1.18 1.14 1.12–1.16 1.13 1.11–1.15 1.11 1.10–1.13 1.12 1.10–1.14 1.12 1.10–1.14 Female sex 1.30 1.16–1.45 1.21 1.09–1.35 1.15 1.04–1.28 1.16 1.05–1.29 1.17 1.06–1.30 1.22 1.10–1.36 Abbreviations: BMI, body mass index (kg/m2); CI, confidence intervals; CSII, continuous subcutaneous insulin infusion; SDS, standard deviation score.
TAB L E 3 Coef ficients from linear regres sion of Hb A1c strati fied by year, 2011 – 2016 2011 2012 2013 2014 2015 B 95% CI B 95% CI B 95% CI B 95% CI B 95% CI Method 0.093 0.151; 0.035 0.065 0.121; 0.009 0.064 0.120; 0.008 0.068 0.123; 0.014 0.129 0.179; 0.079 BMI-SDS 0.033 0.008;0.058 0.027 0.002;0.053 0.048 0.023;0.073 0.033 0.009;0.058 0.050 0.027;0.072 Duration (years) 0.110 0.103;0.118 0.114 0.107;0.122 0.121 0.114;0.128 0.109 0.102;0.116 0.089 0.082;0.095 Female sex 0.106 0.051;0.161 0.091 0.037;0.145 0.060 0.005;0.114 0.088 0.035;0.142 0.101 0.052;0.150 Note: B, coefficient; Method, continuous subcutaneous insulin infusion (CSII). Abbreviations: BMI-SDS, body mass index (kg/m 2) standard deviation score; CI, confidence interval. TAB L E 4 Coefficients from linear regr ession of BMI strati fied by year, 2011 – 2016 2011 2012 2013 2014 2015 B 95% CI B 95% CI B 95% CI B 95% CI B 95% CI Method 0.135 0.076;0.194 0.122 0.065;0.180 0.110 0.054;0.167 0.089 0.033;0.145 0.070 0.014;0.126 HbA1c (%) 0.034 0.008;0.060 0.029 0.003;0.055 0.049 0.023;0.075 0.036 0.020;0.062 0.062 0.034;0.090 Duration (years) 0.008 0.000;0.016 0.006 0.002;0.014 0.006 0.001;0.014 0.010 0.003;0.018 0.014 0.007;0.022 Female sex 0.190 0.135;0.246 0.194 0.138;0.249 0.181 0.126;0.236 0.158 0.103;0.213 0.165 0.110;0.220 Note: B, coefficient; Method, continuous subcutaneous insulin infusion (CSII). Abbreviations: BMI-SDS, body mass index (kg/m 2) standard deviation score; CI, confidence interval.
these patients and families. This could, for example, include contact with a psychologist or screening for neuropsychiatric disorders such as atten-tion deficit hyperactivity disorder (ADHD) which if present complicates everyday life for children and adolescents with diabetes Type 1.35
One of the factors that could be of importance for the improvement of the glucose metabolic control could be the use of carbohydrate counting,36which became an established method at all pediatric clinics in Sweden during the study. Also, the use of real time continuous glucose monitoring (rtCGM) and intermittent scanning continuous glucose moni-toring (isCGM) became more prevalent and may have improved the chances of the patients achieving a more stable glucose balance with more “time in range” (70–180 mg/dl [3.9–10 mmol/L]) and a lower risk of hypo-glycemia. Data on the use of CGM was not registered in SWEDIABKIDS until 2016. Since then, there has been an increase in CGM usage and in 2018 a majority (93%) of all children and adolescents with diabetes Type 1 in Sweden used CGM.1 Previous studies have demonstrated an
increased treatment satisfaction and quality of life with CSII treatment16,28 but at present we have no such data in SWEDIABKIDS. Further studies are required to determine the contribution of these or other factors con-tributed most on the lowering of HbA1c in Swedish children with Type 1 diabetes from 2011 to 2016.
One major strength of this study is the large number of patients with real life data. Since the National Diabetes Register SWEDIABKIDS has a coverage of nearly 98% of prevalent cases, the study gives a good repre-sentation of the pediatric population with Type 1 diabetes in Sweden. A weakness of this study is that we do not know the indication for patients starting CSII treatment. It could be due to poor glucose metabolic control, young age, patient preference, or a combination of all these factors. Another weakness is the number of patients that changed insulin delivery method each year as this group is quite large (10%) and again, we do not know the reason behind the change. In the longitudinal cohort, the group missing HbA1c values is also quite large (21%) which may be explained by a short diabetes duration and young patients having extra visits to the clinic without HbA1c testing being done. The frequency of severe hypo-glycemic events increased in the cohort group toward the end of the period, which can be explained by increasing age.
Even though we have been able to show only a minute benefit of CSII treatment in improving the glucose metabolic control compared to MDI, we find that the investment may nevertheless be worthwhile based on treatment benefits which we have not studied, for example, improved treatment satisfaction,16,17improved quality of life,28 facili-tating carbohydrate counting,36 and the coming combination with
CGM to create closed-loop systems.37
To conclude, we found, based on real-life data a minute decrease in HbA1c with CSII treatment compared to MDI treatment with little clini-cal relevance. There was a general reduction in HbA1c irrespective of insulin delivery method over the years 2011–2016 and this without increasing the incidences of severe hypoglycemia.
A C K N O W L E D G M E N T S
None of the authors have any conflict of interest and the contribu-tions are described in the following text
A U T H O R C O N T R I B U T I O N A N D C O N F L I C T O F I N T E R E S T Financial support was provided by the Department of Research and Development Region Jämtland-Härjedalen, Stiftelsen Samariten, and Thuringstiftelsen. We have no potential conflicts of interest to report. A.F. contributed to the design of the study, collected and ana-lyzed the data, wrote the first draft of the manuscript, and reviewed and edited the manuscript. M.L. contributed to the design of the study, collected and analyzed the data, contributed to the discussion, and reviewed and edited the manuscript. U.S. contributed to the design of the study, collected, and analyzed the data, contributed to the discus-sion, and reviewed and edited the manuscript. T.L. and S.S. contributed to the design of the study, contributed to the discussion, and reviewed and edited the manuscript. M.B. assisted with the statistical analysis and reviewed and edited the manuscript. U.S. is the guarantor of this work and as such, had all the access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Parts of this study were presented in abstract form at the annual meeting of the International Society for Pediatric and Adoles-cent Diabetes, Boston, USA, on October 30, to November 2, 2019.
P E E R R E V I E W
The peer review history for this article is available at https://publons. com/publon/10.1111/pedi.13217.
D A T A A V A I L A B I L I T Y S T A T E M E N T Data can be provided upon request to the authors.
O R C I D
Anna-Lena Fureman https://orcid.org/0000-0002-3390-778X
Stefan Särnblad https://orcid.org/0000-0002-1440-9961
R E F E R E N C E S
1. Swediabkids-Swedish national quality registry for diabetes in children and adolescents. http://www.ndr.nu/pdfs/Yearreport_Swediabkids_ 2018_Eng.pdf. Accessed 1/3/2021.
2. The Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med. 1993;329:977-986.
3. The Diabetes Control and Complications Trial Research Group. Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: diabetes control and complications trial. J Pediatr. 1994;125:177-188. 4. Lind M, Svensson A-M, Kosiborod M, Gudbjörnsdottir S. Glycemic
control and excess mortality in type 1 diabetes. N Engl J Med. 2014; 371:1972-1982.
5. Pickup JC, Keen H, Parsons JA, Alberti KG. Continuous subcutaneous insulin infusion: an approach to achieving normoglycaemia. Br Med J. 1978;1:204-207.
6. Tamborlane WV, Sherwin RS, Genel M, Felig P. Reduction to normal of plasma glucose in juvenile diabetes by subcutaneous administration of insulin with a portable infusion pump. N Engl J Med. 1979;300:573-578. 7. Pickup J, Mattock M, Kerry S. Glycaemic control with continuous
sub-cutaneous insulin infusion compared with intensive insulin injections in patients with type 1 diabetes: meta-analysis of randomised con-trolled trials. BMJ. 2002;324:705.
Technol Ther. 2005;7:876-884.
12. Berghaeuser MA, Kapellen T, Heidtmann B, Haberland H, Klinkert C, Holl RW. Continuous subcutaneous insulin infusion in toddlers starting at diagnosis of type 1 diabetes mellitus. A multicenter analy-sis of 104 patients from 63 centres in Germany and Austria. Pediatr Diabetes. 2008;9:590-595.
13. Jakisch BI, Wagner VM, Heidtmann B, et al. Comparison of continu-ous subcutanecontinu-ous insulin infusion (CSII) and multiple daily injections (MDI) in paediatric type 1 diabetes: a multicentre matched-pair cohort analysis over 3 years. Diabet Med. 2008;25:80-85.
14. Ahern JA, Boland EA, Doane R, et al. Insulin pump therapy in paediat-rics: a therapeutic alternative to safely lower HbA1c levels across all age groups. Pediatr Diabetes. 2002;3:10-15.
15. Plotnick LP, Clark LM, Brancati FL, Erlinger T. Safety and effective-ness of insulin pump therapy in children and adolescents with type 1 diabetes. Diabetes Care. 2003;26:1142-1146.
16. Skogsberg L, Fors H, Hanas R, Chaplin JE. Improved treatment satis-faction but no difference in metabolic control when using continuous subcutaneous insulin infusion vs. multiple daily injection in children at onset of type 1 diabetes mellitus. Pediatr Diabetes. 2008;9:472-479. 17. Brorsson AL, Viklund G, Örtqvist E, Olinder AL. Does treatment with
insulin pump improve glycaemic control in children and adolescents with type 1 diabetes? A retrospective case-control study. Pediatr Dia-betes. 2015;16:546-553.
18. Weinzimer SA, Ahern JH, Doyle EA, et al. Persistence of benefits of con-tinuous subcutaneous insulin infusion in very young children with type 1 diabetes: a follow-up report. Paediatrics. 2004;114:1601-1605. 19. Kordonouri O, Hartmann R, Lauterborn R, Barnekow C, Hoeffe J,
Deiss D. Age-specific advantages of continuous subcutaneous insulin infusion as compared with multiple daily injections in paediatric patients: one-year follow-up comparison by matched-pair analysis. Diabetes Care. 2006;29:133-134.
20. Wettermark B, Hammar N, MichaelFored C, et al. The new Swedish Prescibed drug register-opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf. 2007;16:726-735.
21. The Swedish Pediatric Diabetes Quality Registry SWEDIABKIDS, 2016. https://swediabkids.ndr.nu/DocumentsAnnualReport.aspx. Accessed 1/3/2021.
22. Lindblad B, Nordin G. External quality assessment of HbA1c and its effect on comparison between Swedish pediatric diabetes clinics. Experiences from the Swedish pediatric diabetes quality register (Swediabkids) and equalis. Clin Chem Lab Med. 2013;51(10):2045-2052.
23. Hoelzel W, Weykamp C, Jeppson JO, et al. IFCC reference system for measurement of hemoglobin A1c in human blood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study. Clin Chem. 2004;50:166-174.
24. NGSP converter tool. https://ngsp.org/convert1.asp. Accessed 1/3/2021. 25. Karlberg J, Luo ZC, Albertsson-Wikland K. Body mass index reference values (mean and SD) for Swedish children. Acta Paediatr. 2002;91(3):362.
young people at diagnosis of type 1 diabetes: pragmatic randomised controlled trial and economic evaluation. Br Med J. 2019;365:I1226. 29. Shalitin S, Gil M, Nimri R, De Vries L, Gavan MY, Phillip M. Predictors of
glycaemic control in patients with type 1 diabetes commencing continu-ous subcutanecontinu-ous insulin infusion therapy. Diabet Med. 2010;27:339-347. 30. Pilgaard KA, Breinegaard N, Johannesen J, et al. Episodes of severe hypoglycemia is associated with a progressive increase in hemoglobin A1c in children and adolescents with type 1 diabetes. Pediatr Diabe-tes. 2020;21:808-813.
31. Olsen B, Johannesen J, Fredheim S, Svensson J. Insulin pump treat-ment; increasing prevalence, and predictors for better metabolic out-come in Danish children and adolescents with type 1 diabetes. Pediatr Diabetes. 2015;16:256-262.
32. Mameli C, Scaramuzza AE, Ho J, Cardona-Hernandez R, Suarez-Ortega L, Zucotti GV. A 7-year follow-up retrospective, international, multicenter study of insulin pump therapy in children and adolescents with type 1 diabetes. Acta Diabetol. 2014;51(2):205-210.
33. Ibfelt EH, Vistisen D, Falberg Rønn P, et al. Association between glycaemic outcome and BMI in Danish children with type 1 diabetes in 2000-2018: a nationwide population-based study. Diabet Med. 2020;00:e14401. 34. Birkebaek NH, Kahlert J, Bjarnason R, et al. Body mass index standard
deviation score and obesity in children with type 1 diabetes in the Nordic countries. HbA1c and other predictors of increasing BMISDS. Pediatr Diabetes. 2018;19:1198-1205.
35. Hilgard D, Konrad K, Meusers M, et al. Comorbidity of attention defi-cit hyperactivity disorders and type 1 diabetes in children and adoles-cents: analysis based on the multicentre DPV registry. Pediatr Diabetes. 2017;18:706-713.
36. Tascini G, Berioli MG, Cerquiglini L, Santi E. Carbohydrate counting in children and adolescents with type 1 diabetes. Nutrients. 2018; 10(1):109.
37. DeBoer MD, Buckingham BA, Chernavvsky D, Wadwa RP. iDCL trial research group. A randomized trial of closed-loop control in children with type 1 diabetes. N Engl J Med. 2020;383(9):836-845.
S U P P O R T I N G I N F O R M A T I O N
Additional supporting information may be found online in the Supporting Information section at the end of this article.
How to cite this article: Fureman A-L, Lilja M, Lind T, Särnblad S, Bladh M, Samuelsson U. Comparing continuous subcutaneous insulin infusion and multiple daily injections in children with Type 1 diabetes in Sweden from 2011 to 2016— A longitudinal study from the Swedish National Quality Register (SWEDIABKIDS). Pediatr Diabetes. 2021;1–10.
